Diagnosis and treatment of community-acquired lower ... · improvements in therapeutic efﬁ cacy,...

Diagnosis and treatment of

community-acquired lower

respiratory tract infections:

Strategies for effi cient management

Diagnosis and treatment of

community-acquired lower

respiratory tract infections:

Strategies for effi cient management

Diagnostiek en behandeling van buiten het ziekenhuis opgelopen lagere luchtweginfecties.

Methoden voor een effi ciënte benadering

(met een samenvatting in het Nederlands)

Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de Rector Magnifi cus, Prof. dr.

W.H. Gispen, ingevolge het besluit van het College voor Promoties in het openbaar te verdedigen op vrijdag 15 april 2005 ‘s ochtends

om 10:30 uur

Door Jan Jelrik Oosterheert geboren op 25 juni 1976 te Zwolle

Promotoren: Prof. dr. I.M. Hoepelman 1,2

Prof. dr. M.J.M. Bonten 1,2,3

Co-promotor: Dr. M.M.E. Schneider 1

1. Universitair Medisch Centrum Utrecht. Divisie Interne Geneeskunde en Dermatologie. Afdeling Interne Geneeskunde en Infectieziekten

2. Eijkman Winkler Instituut voor Microbiologie, Infectieziekten en Ontsteking

3. Julius Centrum voor Gezondheidswetenschappen en Eerstelijns Geneeskunde

ISBN: 90-9019329-4 Layout: Febodruk BV, EnschedeDruk: Febodruk BV, Enschede

Dit proefschrift werd mede mogelijk gemaakt door de fi nanciële steun van Uniprom Diagnostics en Binax USA, Wyeth Farmaceuticals, Novo Nordisk Farma B.V. en Pfi zer BV.

Contents

Chapter 1. General introduction and outline of the thesis 7

Chapter 2. Trends and causes in pneumonia-related morbidity and mortality in adults in the Netherlands Ned Tijdschr Geneeskd. 2003; 147 (9): 381-6 21

Chapter 3. Severe community-acquired pneumonia: What’s in a name… Curr Opin Infect Dis. 2003; 16 (2): 153-9 37

Chapter 4. Risk factors for early clinical failure in patients with severe community-acquired pneumonia Submitted 57

Chapter 5. How good is the evidence for the recommended empiric antimicrobial treatment of patients hospitalised because of community-acquired pneumonia? J Antimicrob Chemother. 2003; 52 (4): 555-63 75

Chapter 6. Introducing British or North American Guidelines for Community-Acquired Pneumonia in The Netherlands, predicted effects on antibiotic use and adequacy of treatment Submitted 95

Chapter 7. An algorithm to determine cost-savings of targeting antimicrobial therapy based on the results of rapid diagnostic testing J Clin Microbiol. 2003: 41 (10): 4708-13 115

Chapter 8. Impact of rapid viral diagnosis by real-time PCR in patients with lower respiratory tract infections Submitted 133

Chapter 9. Costs and effects of early IV to oral switch in severe CAP: a multi center randomized trial. Submitted 151

Chapter 10. Summary and general discussion 171

Samenvatting in het Nederlands 177

Dankwoord 183

Curriculum Vitae 187

Chapter 1

1General introduction

and outline of the thesis

8

Chapter 1

Introduction to pneumonia

Community-acquired pneumonia and other lower respiratory tract infections are among the most common infectious diseases world wide. Despite advances in health care and the development of new antibiotics, respiratory tract infections still account for numerous hospitalizations and deaths each year. For example, in the Netherlands, 20.000 patients need hospitalization and 6.500 patients die each year because of pneumonia and these fi gures are still increasing. Treatment of these episodes is associated with high costs for healthcare and considerable antibiotic consumption. Estimated annual costs for treating CAP were > 1 billion US$ in the UK in 1997 and 9.7 billion US$ in the USA in 2001 (1-4). Overuse of antibiotics occurs frequently, especially in case of viral infections. (5) Unnecessary antibiotic use is regarded as an important factor in the global rise of antibiotic resistance. Further increase in antibiotic resistance leads to treatment diffi culties, increased treatment costs and possible adverse patient outcomes. Therefore, in addition to improvements in therapeutic effi cacy, strategies to decrease costs and prevent unnecessary antibiotic use have become an important issue in the management of serious infections. In this thesis, we evaluate several strategies for effi cient diagnosis and treatment for severe community-acquired lower respiratory tract infections, with the emphasis on community-acquired pneumonia.

Pneumonia is the infl ammation and consolidation of lung tissue due to an infectious agent (6) The clinical criteria for the diagnosis include chest pain, cough, or auscultatory fi ndings such as rales or evidence of pulmonary consolidation, fever or leucocytosis. In addition, there must be radiographic evidence, such as the presence of new infi ltrates on chest radiograph, and laboratory evidence that supports the diagnosis (7). Because of differences in pathogenesis and causative micro-organisms, pneumonia is often divided in hospital acquired and community-acquired pneumonia. Pneumonia developing outside the hospital is referred to as community-acquired pneumonia (CAP). CAP represents a broad spectrum of severity, ranging from mild pneumonia that can be managed by general practitioners outside the

Chapter 1

9

Chapter 1

hospital to severe pneumonia with septic shock needing treatment in intensive care unit. Depending on severity of illness, about 20% of patients with pneumonia need hospitalization and approximately 1% of all CAP patients require treatment in ICU (8;9)

Elderly persons and those with underlying conditions, such as cerebro- and cardiovascular diseases, chronic obstructive pulmonary disease (COPD) and alcoholism, are at increased risk for developing lower respiratory tract infections and complicated courses of infection. (10;11)Micro-organisms that can cause CAP in immunocompetent persons are numerous. Throughout the world, Streptococcus pneumoniae is by far the most frequently isolated pathogen. Other frequently isolated bacteria are Haemophilus Infl uenzae and Staphylococcus aureus. (12-18) Pseudomonas aeruginosa should be considered as a possible causative micro-organism in patients with structural damage of the respiratory tract, for example in patients with bronchiectasis or COPD. (19) Atypical pathogens as Mycoplasma pneumoniae, Chlamydia pneumoniae and Legionella pneumophila are also causes of community-acquired pneumonia, although their contribution to the etiology of CAP varies widely. (12;13;15;18) Most frequent viral causes of community-acquired pneumonia include infl uenza virus, para-infl uenzavirus and coronavirus. (18;20) However, even in studysettings with extensive diagnostic testing, approximately 50% of episodes of CAP remain of unknown etiology (12-18) Some clinical features are associated with causative micro-organisms and may, therefore, guide initial therapy, such as fl ucloxacillin treatment for CAP preceded by infl uenza because of a high risk of S. aureus infection, and cephalosporins with antipseudomonas activity in patients with bronchiectasis or other structural damage to the lungs, who are at risk for pseudomonas colonization. (see table 3) In general however, the microbial cause of CAP cannot be predicted upon clinical, radiological and laboratory features. (21-24)

Routine diagnostic procedures to identify these pathogens include Gram-staining of expectorated sputum, culturing of blood and sputum and serologic testing of acute and reconvalescent blood samples. Recently, diagnostic tests that can provide results within minutes to

10

Chapter 1Chapter 1

hours have become available. These include rapid urinary antigen testing for Legionella pneumophila serogroup 1 and S. pneumoniae and real-time PCR-tests of respiratory samples to detect respiratory viruses. (25-27) Whether these diagnostic techniques enhance the etiologic yield in CAP, lead to cost-savings or more targeted antimicrobial treatments, however, is unknown.

The course of pneumonia can be complicated by development of progressive pneumonia, despite appropriate antimicrobial treatment, development of pleural empyema, uncontrolled sepsis or death. (28) The clinical response of patients within the fi rst 2-3 days of treatment appears to be important for heath-outcome of patients with CAP. Non-response is associated with high morbidity and mortality, and patients that are clinically stable at day 3 infrequently deteriorate thereafter. Early prediction of whom is at risk for a complicated course could help physicians in improving effi ciency of pneumonia management. For the prediction of mortality, several risk-classifi cations exist which include combinations of underlying illnesses, age and clinical features. Whether these classifi cations can also be used to guide clinical care is controversial. (29-32) The use of critical pathways or guidelines may improve quality of care and may reduce length of hospital stay (33;34).

Treatment of CAP aims to reduce the bacterial load in lung tissue and should, therefore, be directed towards the causative agent. As development of antibiotic resistance is linearly related to the quantity of antibiotics prescribed, ideally, antibiotics should only be used in bacterial infections and be withheld in non-bacterial or self-limiting infections. Choosing an adequate treatment is simplifi ed if results of diagnostic procedures are available. However, at the time treatment has to be initiated, causative micro-organisms are usually unknown. Therefore, initial treatment in respiratory infections is mostly empirical, covering several suspected causative micro-organisms, taking into consideration epidemiological features, host-factors, local resistance patterns of micro-organisms and farmacokinetic and farmacodynamic features of antibiotics.

11

Chapter 1R

esp

irato

ry

infe

ctio

nM

ost

lik

ely

p

ath

og

en

sU

SA

UK

Neth

erl

an

ds

CAP,

outp

atie

nts

S.

pneu

monia

e, M

. pneu

monia

e, C

. pneu

monia

eM

acro

lide

or

dox

ycyc

linFl

uoro

quin

olo

ne

Am

oxic

illin

Ben

zylp

enic

illin

CO

PD:

amox

icill

in +

cla

vula

nic

ac

idIn

fl uen

za:

Flucl

oxac

illin

Dia

bet

es /

alc

oholis

m:

2nd

gen

erat

ion c

ephal

osp

orin

Susp

icio

n o

f at

ypic

al:

add

mac

rolid

e

CAP,

inpat

ients

S,

pneu

monia

e,

H.

Infl uen

zae,

M

. pneu

monia

e,

C.

pneu

monia

e,

Asp

irat

ion,

Legio

nel

la.

exte

nded

spec

trum

ce

phal

osp

orin p

lus

a m

acro

lide,

β-

lact

am /

β-l

actm

ase

inhib

itor

plu

s a

mac

rolid

e, o

r a

fl uoro

quin

olo

ne

alone

(not

seve

re,

adm

itte

d

for

oth

er r

easo

ns)

am

oxic

illin

(n

ot

seve

re)

amox

icill

in

+ m

acro

lide

Ben

zylp

enic

illin

CO

PD:

amox

icill

in +

cla

vula

nic

ac

idIn

fl uen

za:

Flucl

oxac

illin

Dia

bet

es /

alc

oholis

m:

2nd

gen

erat

ion c

ephal

osp

orin

Asp

irat

ion:

ben

zylp

enic

illin

/

amox

icill

inSusp

icio

n o

f at

ypic

al:

add

mac

rolid

e

CAP,

ICU

S.

pneu

monia

e,

Legio

nel

la s

pp,

H.

infl uen

zae,

ente

ric

gra

m n

egat

ive

bac

illi,

S.

aure

us,

asp

irat

ion

exte

nded

spec

trum

ce

phal

osp

orin p

lus

mac

rolid

e,

β-la

ctam

/ β

-lac

tam

ase

inhib

itor

plu

s ei

ther

fl u

oro

quin

olo

ne

or

mac

rolid

e.

In s

truct

ura

l lu

ng d

isea

se:

antipse

udom

onal

agen

ts.

Asp

irat

ion:

fl uoro

quin

olo

ne

with o

r w

ithout

clin

dam

ycin

, m

etro

nid

azole

or

a β-

lact

am /

β-

lact

amas

e in

hib

itor.

β-la

ctam

+ -

ase

inhib

itor

or

cephal

osp

orin (

2nd /

3rd

gen

.) p

lus

mac

rolid

e ±

rifa

mpic

in

β-la

ctam

+ -

ase

inhib

itor

or

cephal

osp

orin (

2nd /

3rd g

en.)

plu

s m

acro

lide

Tab

le 1

Most

lik

ely

pat

hogen

s an

d p

refe

rred

em

piric

al t

her

apy

in low

er r

espirat

ory

tra

ct infe

ctio

ns.

Bas

ed o

n t

he

reco

mm

endat

ions

of ATS,

BTS a

nd I

DSA a

nd S

WAB

guid

elin

es for

com

munity-

acquired

pneu

monia

(30;3

2;3

5;3

6)

12

Chapter 1Chapter 1

Based on etiologic data, in the Netherlands, the recommended choice for initial therapy for patients needing hospitalisation is a β-lactam agent such as penicillin G or amoxicillin covering infection with S. pneumoniae. Only when a strong suspicion exists for pneumonia caused by C. pneumoniae, M. pneumoniae or L. pneumophila, or when treatment in an ICU is necessary, a macrolide antibiotic such as erythromycin or azithromycin should be added. (35) Based on expected larger roles of atypical pathogens and possible anti-infl ammatory effects of macrolides, North-American and British guidelines now recommend treatment of beta-lactams plus macrolides or treatment with fl uoroquinolones for all patients admitted to hospital because of CAP. (30;32;36;37)

Micro-organism Preferred targeted therapy, duration

S. pneumoniaePenicillin susceptiblePenicillin resistant (MIC< 2μg/mL)

Penicillin G, AmoxicillinAgents based on in vitro susceptibility testes, including cefotaxime and ceftriaxone, fl uoroquinolone, vancomycin

H. infl uenzae Cephalosporin (2nd or 3rd generation) doxycyclin, β-lactam + β-lactamase inhibitor, azithromycin, TMP-SMX

M. pneumoniae Doxycyclin, macrolide

C. pneumoniae Doxycyclin, macrolide

L. pneumophila Macrolide ± rifampin; fl uoroquinolone

S. aureusMethicillin susceptible

Methicillin resistant

Nafcillin / oxacillin ± rifampin or gentamicin; (BTS) Flucloxacillin ± rifampicinVancomycin ± rifampin or gentamycin

M. catharralis Cephalosprin (2/3) TMP/SMZ, macrolide

Anaerobes β-lactam+β-lactamase inhibitor, clindamycin

P. aeruginosa Aminoglycoside + antipseudomonal β-lactam (eg. Piperacillin) or carbapenem

C. psittacci Doxycyclin

Coxiella burnetii Tetracyclin

Infl uenzavirus Amantidine or rimantidine, zanamavir or oseltamivir

Table 2

Preferred antimicrobial therapy when causative micro-organisms are known. Based on the

recommendations of ATS, BTS and IDSA guidelines for community-acquired pneumonia (30;32;36)

and the ATS guidelines for hospital acquired pneumonia (41)

13

Chapter 1

As soon as results from microbiological investigations become available, antimicrobial therapy should be streamlined to the causative micro-organism or withheld in viral infections, if possible. (see table 1 and table 2)

Feature Possible pathogens

Alcoholism S. pneumoniae, anaerobes, gram-negative bacilli, tbc

COPD/smoker S. pneumoniae, H. infl uenzae, M. catharralis, Legionella

Nursing home residencyS. pneumoniae, gram-negative bacilli, H. infl uenzae, S. aureus, anaerobes, C. pneumoniae, tbc

Poor dental hygiene Anaerobes

Exposure to birds C. psittaci, C. neoformans, H. capsulatum

Exposure to rabbits Francisella tularensis

Exposure to farm animals or parturient cats Coxiella burnetii

Infl uenza activity in community Infl uenza, S. pneumoniae, S. aureus, H. infl uenzae

Aspiration Anaerobes

Structural disease of the lung p. aeruginosa, S. aureus

Injection drug use S. aureus, anaerobes, tbc, Pneumocystis carinii

Recent antibiotic therapy Drug resistant pneumococci, P. aeruginosa

Cooling towers, air conditioning etc. Legionella pneumophila

shelter for homeless / jail S. pneumoniae, tbc

Diabetic ketoacidosis S. pneumoniae, S. aureus

Solid organ transplantation S. pneumoniae, H. infl uenzae, Legionella spp, P. carinii, Strongyloides stercoralis

Sickle cell S. pneumoniae

HIV cd4<200/μL P carinii, S. pneumoniae, H. infl uenzae, C. neoformans, M. tbc, Rhodococcus equi

Granulocytopenia Aerobic gram negative rod-like bacteria such as E. coli or K. pneumoniae

Table 3

Epidemiologic conditions or host factors related to specifi c pathogens in patients with community-

acquired pneumonia. Based on the recommendations of the ATS and the CTS for the management

of community-acquired pneumonia. (32;37)

14

Chapter 1Chapter 1

Controlled trials that specifi cally have addressed the question

how long pulmonary infections should be treated are lacking. The length of treatment is usually based on the pathogen, response to treatment, comorbid illness, and complications. Current practice is to treat pneumonia caused by S. pneumoniae until the patient has

been afebrile for 72 h. (36) Pneumoniae caused by bacteria that can cause necrosis of pulmonary parenchyma (e.g., S. aureus, P. aeruginosa, Klebsiella, and anaerobes) should probably be treated for about 2 weeks. (36) Pneumonia caused by M. pneumoniae or C. pneumoniae should probably be treated for at least 2 weeks, as should

legionnaires’ disease in immunocompetent individuals. Azithromycin may be used for shorter courses of treatment because of its very long

half-life in tissues (36) although longer courses are probably needed for Legionella infections (38).

Treatment with oral antibiotics is mostly suitable for milder infections. Parenteral treatment is indicated if oral drugs are ineffi ciently absorbed or large doses of drugs are required. As oral absorption can be decreased in severely ill patients, initial treatment for patients hospitalized because of pulmonary infections is mostly intravenously to warrant optimal serum- and pulmonary levels of antibiotics. In conventional treatment approaches, intravenous therapy is continued until defi nite clinical cure has been achieved. However, already during the phase of clinical improvement, replacing intravenous by oral antibiotics may be adequate. For most intravenous antibiotics, oral equivalents with good bio-availability are available. Therefore, the concept of early transition from intravenous to oral antibiotics for the treatment of serious infections like CAP, also referred to as “switch therapy”, “sequential therapy” or “step-down therapy”, has been evaluated. (39) Possible advantages are reductions in drug costs, early discharge, reduction in complications associated with the use of intravenous catheters such as thrombofl ebitis, decrease in hospital waste and improved freedom of movement for patients. In contrast, possible disadvantages are treatment failures when patients do not take medication adequately or when absorption of oral medication is limited. These factors could lead to reinfection, possibly resulting in readmissions or death. As the duration of intravenous therapy is the primary factor affecting length of hospital stay for patients with CAP,

15

Chapter 1

(40) early switch from intravenous to oral equivalents may, therefore, reduce costs for treatment and length of hospital stay of community-acquired pneumonia in a cost-effective way.

Outline of the thesis

The scope of this thesis is to evaluate tools for effective management of lower respiratory tract infections, with the emphasis on reducing healthcare costs and preventing unnecessary antibiotic use. In the fi rst chapter we describe the magnitude of pneumonia related problems in the Netherlands over the past 10 years. In the two following chapters, we evaluate the use of the relevance of identifying high-risk patients by reviewing frequently used defi nitions for severe CAP and by determining risk factors associated with early clinical failure. Next, the evidence for the treatment recommendations for CAP in recently formulated foreign clinical guidelines is critically reviewed and their presumed effects on antibiotic use, costs and adequacy of treatment in the Netherlands are evaluated. Subsequently, potential effects of rapid diagnostic tests of lower respiratory tract infections in reducing cost and antibiotic use are assessed in two clinical studies. Finally, the cost-effectiveness of an early conversion from intravenous to oral therapy in hospitalized patients with severe CAP is evaluated in a randomized multicenter trial.

16

Chapter 1Chapter 1

Reference List

(1) Pinner RW, Teutsch SM, Simonsen L, Klug LA, Graber JM, Clarke MJ et al. Trends in infectious diseases mortality in the United States. JAMA 1996; 275(3):189-193.

(2) Guest JF, Morris A. Community-acquired pneumonia: the annual cost to the National Health Service in the UK. Eur Respir J 1997; 10(7):1530-1534.

(3) Marston BJ, Plouffe JF, File TM, Jr., Hackman BA, Salstrom SJ, Lipman HB et al. Incidence of community-acquired pneumonia requiring hospitalization. Results of a population-based active surveillance Study in Ohio. The Community-Based Pneumonia Incidence Study Group. Arch Intern Med 1997; 157(15):1709-1718.

(4) Lave JR, Lin CJ, Fine MJ, Hughes-Cromwick P. The cost of treating patients with community-acquired pneumonia. Seminars in respiratory and critical care medicine 1999; 20(3):189-197.

(5) Ball P, Baquero F, Cars O, File T, Garau J, Klugman K et al. Antibiotic therapy of community respiratory tract infections: strategies for optimal outcomes and minimized resistance emergence. J Antimicrob Chemother 2002; 49(1):31-40.

(6) Marrie TJ. Community-acquired pneumonia. Clin Infect Dis 1994; 18(4):501-513.

(7) Chow AW, Hall CB, Klein JO, Kammer RB, Meyer RD, Remington JS. General guidelines for the evaluation of new anti-infective drugs for the treatment of respiratory tract infections. Clinical Infectious Diseases 1992; 15(suppl 1):s62-s88.

(8) Torres A, Serra-Batlles J, Ferrer A, Jimenez P, Celis R, Cobo E et al. Severe community-acquired pneumonia. Epidemiology and prognostic factors. Am Rev Respir Dis 1991; 144(2):312-318.

(9) Moine P, Vercken JB, Chevret S, Chastang C, Gajdos P. Severe community-acquired pneumonia. Etiology, epidemiology, and prognosis factors. French Study Group for Community-Acquired Pneumonia in the Intensive Care Unit. Chest 1994; 105(5):1487-1495.

(10) Koivula I, Sten M, Makela PH. Risk factors for pneumonia in the elderly. Am J Med 1994; 96(4):313-320.

(11) Lipsky BA, Boyko EJ, Inui TS, Koepsell TD. Risk factors for acquiring pneumococcal infections. Arch Intern Med 1986; 146(11):2179-2185.

17

Chapter 1

(12) Ruiz M, Ewig S, Marcos MA, Martinez JA, Arancibia F, Mensa J et al. Etiology of community acquired pneumonia: impact of age, comorbidity and severity. Am J Resp Crit Care Med 1999; 160:397-405.

(13) Lim WS, Macfarlane JT, Boswell TCJ, Harrison TG, Rose D, Leinonen M et al. Study of community acquired pneumonia aetiology (SCAPA) in adults admitted to hospital: implications for management guidelines. Thorax 2001; 56:296-301.

(14) Park DR, Sherbin VL, Goodman MS, Pacifi co AD, Rubenfeld GD, Polissar NL et al. The etiology of community-acquired pneumonia at an urban public hospital: infl uence of human immunodefi ciency virus infection and initial severity of illness. J Infect Dis 2001; 184(3):268-277.

(15) Lieberman D, Schlaeffer F, Boldur I, Lieberman D, Horowitz S, Friedman MG et al. Multiple pathogens in adult patients admitted with community-acquired pneumonia: a one year prospective study of 346 consecutive patients. Thorax 1996; 51:179-184.

(16) Luna CM, Famiglietti A, Absi R, Videla AJ, Nogueira FJ, Fuenzalida AD et al. Community-acquired pneumonia: etiology, epidemiology, and outcome at a teaching hospital in Argentina. Chest 2000; 118(5):1344-1354.

(17) Finch R, Schurmann D, Collins O, Kubin R, McGivern J, Bobbaers H et al. Randomized controlled trial of sequential intravenous (i.v.) and oral moxifl oxacin compared with sequential i.v. and oral co-amoxiclav with or without clarithromycin in patients with community-acquired pneumonia requiring initial parenteral treatment. Antimicrob Agents Chemother 2002; 46(6):1746-1754.

(18) Bohte R, Furth R van, Broek PJ van den. Aetiology of community-acquired pneumonia; a prospective study among adults requiring admission to hospital. Thorax 1995; 50:543-547.

(19) Vegelin AL, Bissumbhar P, Joore JCA, Lammers JWJ, Hoepelman IM. Guidelines for severe community-acquired pneumonia in the western world. The Netherlands Journal of Medicine 1999; 55:110-117.

(20) van Elden LJ, van Loon AM, van Alphen F, Hendriksen KA, Hoepelman AI, van Kraaij MG et al. Frequent detection of human coronaviruses in clinical specimens from patients with respiratory tract infection by use of a novel real-time reverse-transcriptase polymerase chain reaction. J Infect Dis 2004; 189(4):652-657.

(21) Bohte R, Hermans J, van den Broek PJ. Early recognition of Streptococcus pneumoniae in patients with community-acquired pneumonia. Eur J Clin Microbiol Infect Dis 1996; 15(3):201-205.

18

Chapter 1Chapter 1

(22) Macfarlane JT, Miller AC, Roderick Smith WH, Morris AH, Rose DH. Comparative radiographic features of community-acquired legionnaires’ disease, pneumococcal pneumonia, mycoplasma pneumonia and psittacosis. Thorax 1984; 39:28-33.

(23) Farr BM, Kaiser DL, Harrison BDW, Connolly CK. Prediction of microbial aetiology at admission to hospitall for pneumonia from the presenting clinical features. Thorax 1989; 44:1031-1035.

(24) Woodhead MA, Macfarlane JT, American Thoracic Society. Comparative clinical and laboratory features of legionella with pneumococcal and mycoplasma pneumonias. Br J Dis Chest 1987; 81:133-139.

(25) Birtles RJ, Harrison TG, Samuel D, Taylor AG. Evaluation of urinary antigen ELISA for diagnosing Legionella pneumophila serogroup 1 infection. J Clin Pathol 1990; 43(8):685-690.

(26) Murdoch DR, Laing RT, Mills GD, Karalus NC, Town GI, Mirrett S et al. Evaluation of a rapid immunochromatographic test for detection of Streptococcus pneumoniae antigen in urine samples from adults with community-acquired pneumonia. Journal of Clinical Microbiology 2001; 39(10):3495-3498.

(27) Elden LJ van, Nijhuis M, Schipper P, Schuurman R, Loon AM van. Simultaneous detection of infl uenza viruses A and B using real-time quantitative PCR. Clin Microbiol 2001; 39(1):196-200.

(28) Roson B, Carratala J, Fernandez-Sabe N, Tubau F, Manresa F, Gudiol F. Causes and factors associated with early failure in hospitalized patients with community-acquired pneumonia. Arch Intern Med 2004; 164(5):502-508.

(29) Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify low-risk patients with community acquired pneumonia. N Engl J Med 1997; 336:243-250.

(30) BTS Guidelines for the Management of Community Acquired Pneumonia in Adults. Thorax 2001; 56 Suppl 4:IV1-64.

(31) Lim WS, van der Eerden MM, Laing R, Boersma WG, Karalus N, Town GI et al. Defi ning community acquired pneumonia severity on presentation to hospital: an international derivation and validation study. Thorax 2003; 58(5):377-382.

(32) American Thoracic Society. Guidelines for the management of adults with community acquired pneumonia. Am J Crit Care Med 2001; 163:1730-1754.

(33) Marrie TJ, Lau CY, Wheeler SL, Wong CJ, Vandervoort MK, Feagan BG. A controlled trial of a critical pathway for treatment of community- acquired pneumonia. CAPITAL Study Investigators. Community-

19

Chapter 1

Acquired Pneumonia Intervention Trial Assessing Levofl oxacin. JAMA 2000; 283(6):749-755.

(34) Feagan BG. A controlled trial of a critical pathway for treating community-acquired pneumonia: the CAPITAL study. Community-Acquired Pneumonia Intervention Trial Assessing Levofl oxacin. Pharmacotherapy 2001; 21(7 Pt 2):89S-94S.

(35) Kasteren MEE van, Wijnands WJ, Stobbering EE, Janknegt R, Meer JW van der. Optimization of the antibiotics policy in the Netherlands. II. SWAB guidelines for the antimicrobial therapy of pneumonia in patients at home and as nosocomial infections. The Netherlands Antibiotic Policy Foundation. Ned Tijdschr Geneeskd 1998; 142(17):952-956.

(36) Bartlett JG, Dowell SF, Mandell LA, File Jr TM, Musher DM, Fine MJ. Practice Guidelines for the management of community-acquired pneumonia in adults. Infectious Diseases Society of America. Clin Infect Dis 2000; 31(2):347-382.

(37) Mandell LA, Marrie TJ, Grossman RF, Chow AW, Hyland RH, and the Canadian community-acquired pneumonia working group. Canadian guidelines for the initial management of community-acquired pneumonia: an evidence-based update by the Canadian Infectious Diseases Society and the Canadian Thoracic Society. Clin Infect Dis 2000; 31:383-421.

(38) Matute AJ, Schurink CA, Hoepelman IM. Is a 5 day course of azithromycin enough for infections caused by Legionella pneumophila? J Antimicrob Chemother 2000; 45(6):930-931.

(39) Rhew DC, Tu GS, Ofman J, Henning JM, Richards MS, Weingarten SR. Early switch and early discharge strategies in patients with community- acquired pneumonia: a meta-analysis. Arch Intern Med 2001; 161(5):722-727.

(40) Ramirez JA, Vargas S, Ritter GW, Brier ME, Wright A, Smith S et al. Early switch from intravenous to oral antibiotics and early hospital discharge: a prospective observational study of 200 consecutive patients with community-acquired pneumonia. Arch Intern Med 1999; 159(20):2449-2454.

(41) Hospital-acquired pneumonia in adults: diagnosis, assessment of severity, initial antimicrobial therapy, and preventive strategies. A consensus statement, American Thoracic Society, November 1995. Am J Respir Crit Care Med 1996; 153(5):1711-1725.

20

Chapter 1Chapter 1

2Trends and causes in

pneumonia-related morbidity

and mortality in adults in the

NetherlandsNed Tijdschr Geneeskd. 2003; 147 (9): 381-6

JJ Oosterheert, MJM Bonten, E Hak, JWJ Lammers, MME Schneider, IM Hoepelman

22

Chapter 1Chapter 2

Abstract

Morbidity and mortality due to pneumonia-related disease has increased in the Netherlands in the past 10 years, as shown in several national registrations. There are several possible explanations. An increase in the elderly population is the most likely explanation. An increase in underlying conditions such as chronic obstructive pulmonary disease and diabetes mellitus are also possible explanations. Antibiotic resistance, inadequate treatment or a shift in causative micro-organisms probably don’t play a role. Therefore, specifi cally focusing on pneumonia may not be suffi cient to reduce the burden of pneumonia related disease and mortality.

23

Chapter 2

Introduction

Despite the use of antibiotics and associated decreases in mortality due to respiratory infections, lower respiratory tract infections such as pneumonia remain an important cause for mortality in the Netherlands. Pneumonia ranks 4th in the most frequent causes of death in our country (source: CBS mortality statistics, table 1) and although sepsis and AIDS are associated with high mortality rates, in the western world, respiratory infection is the most important infectious cause of death. (1)

Disease Absolute mortality

Coronary heart disease 17.443

Cerebrovascular disease 12.275

Lung carcinoma 8.559

Pneumonia 6.960

COPD 6.634

Decompensatio Cordis 6.458

Dementia 5.343

Colon carcinoma 4.300

Breast carcinoma 3.425

Table 1

Important causes of death in the Netherlands in 2000 (Source: CBS doodsoorzakenstatistiek, RIVM

kompas voor de volksgezondheid)

Usually, pneumonia is categorized in hospital-acquired pneumonia (noscomial pneumonia) or community-acquired pneumonia. (2) Especially the elderly and persons with underlying conditions such as cerebro- and cardiovascular diseases, COPD and alcoholism are at risk for developing lower respiratory tract infections and a complicated course of the infection. (3;4)The mortality-risk of pneumonia is dependent of combinations of underlying illnesses, age and clinical features (5;6) In general, monotherapy with a beta-lactam antibiotic is the indicated initial therapy. Combinations of beta-lactam antibiotics and macrolides are only advised when a strong suspicion of pneumonia caused by M. pneumoniae, C. pneumoniae or L. pneumophila exists, or in severe pneumonia needing treatment in an intensive care unit. (2;7)In recent years, absolute mortality because of pneumonia in the

24

Chapter 1Chapter 2

Netherlands has risen and a similar increase in mortality was found in the United States. (8-10) In this chapter, we describe this increase and explore possible explanations for the Dutch situation.

Increase in registered pneumonia-related mortality

The number of pneumonia-related deaths has risen from 3487 in 1990 to 6533 in 2000. (source: CBS mortality statistics). Mortality rates per 100.000 inhabitants increased from 31.3 to 54.3 and mortality rates standardized for the population of 1990 increased from 31.9 to 48.9. (table 2, fi gure1. Source: CBS mortality statistics). This increase is most obvious in the higher age groups. Interestingly, in 2001 and 2002 standardized and age-specifi c mortality rates for pneumonia seem do decrease. (fi gure 2). According to registrations of the National Medical Registration (LMR), the number of adults that died from pneumonia in hospitals increased from 3734 in 1991 to 6104 in 2002 (source: LMR)Not only pneumonia related mortality has increased in recent years, in the same period, the number of patients diagnosed with pneumonia by general practitioners (source: Continuous Morbidity Registration) and the number of hospitalisations because of pneumonia in adults has also risen from 16.174 hospitalisations in 1991 to 26.711 in 2002. (Source: LMR, see table 2)

Pneumonia realted morbitity and mortality 1990 2000 Difference(%)

Absolute mortality in adults 3487 6533 87

Mortality per 100.000 adults per year 31,3 54,3 73

Mortality standardized for age and sex per 100.000 adults per year (Source: CBS) 31,9 48,9 53

Absolute hospital mortality Age and sex standardized mortality per 100.000 adults per year

19913734

32,9

20026104

44,9

63

36

Absolute number of hospitalisationsHospitalisations number standardized for age and sex per 100.000 adults per year (Source: LMR)

16.174

142,2

26.711

225,0

65

58

Table 2

Mortality and Morbidity related to pneumonia in adults (20 years and older) in 1991/1991 and

2000/2002.

25

Chapter 2

25

30

35

40

45

50

55

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Figure 1

Standardized mortality (for age and sex, population of 1990 per 100.000 adults) Source: CBS

Explanations

The observed increase in pneumonia related mortality can be explained in several ways. First, a seeming increase due to registration methods and possible misclassifi cation has to be ruled out. Subsequently, etiological explanations as altered age patterns in the population and increases in incidences of underlying diseases, changes in environmental factors or changes in causative micro-organisms of lower respiratory tract infections can have contributed to the increase in pneumonia-mortality. These items will be addressed in the rest of this chapter.

Registration Methods

First, it is important to evaluate whether the registered pneumonia-related mortality refl ects a true increase. In the Netherlands, causes of death are registered by the treating physician, the municipal coroner or an appointed physician. Causes of death are categorized according to the International Classifi cation of Diseases version 9 and from 1996 onwards version 10. (ICD-9 and ICD-10). These data are managed by the Central Statistics Bureau of the Netherlands.

26

Chapter 1Chapter 2

Registration for hospital admissions is accounted for by the National Medical Registration (LMR). At discharge, the diagnosis, classifi ed according to ICD-9 Clinical Modifi cation (ICD-9-CM) that was the main cause for hospitalization is registered. This registration method has been unchanged in the period 1985-2002. The described methods used for registration have their drawbacks. For example, the use of standardized disease codes like the International Classifi cation of Diseases (ICD-codes) for scientifi c research is debated. An American study showed a 58.3% sensitivity the most frequently used ICD-codes for the diagnosis community-acquired pneumococcal pneumonia. In addition, sensitivity, positive predictive value and negative predictive value were unsatisfactory and misclassifi cation occurred. (11) Furthermore, sensitivity could have been infl uenced by increased attention for the disease, possibly as a result of attention for antibiotic resistance or infl uenza epidemics.

0

500

1000

1500

2000

1985 1990 1995 2000

65-69

70-74

74-79

80-84

85+

Figure 2

Age specifi c pneumonia mortality (per 100.000) (Source: CBS)

In registering and classifying causes of death by CBS, in 1996 a switch was made from the ICD-9 to the ICD-10 classifi cation. This could have infl uenced registered mortality. However, for pneumonia, both ICD-codes are only 70% similar, and use of ICD-9 codes would

27

Chapter 2

result in a rise in pneumonia diagnoses. Based on this change rather a decrease than an increase in pneumonia related mortality can be expected. (source: http://www3.who.int/ whosis/ bethesda/ documents/Bethesda.77.doc)Another registration system, the LMR, retrospectively, i.e. after hospital stay, makes the diagnosis. A drawback of this system is that hospital-acquired pneumonias that were not the primary reason for hospital admission also are registered as pneumonia. It remains therefore unclear whether the diagnosis is based on a community-acquired or a nosocomial pneumonia. However, it is unlikely that the drawbacks in registration systems are an explanation for the observed increase in pneumonia related mortality. First, an acute change in registration procedures would lead to a fast and short lasting increase in mortality and this would remain at a constant level thereafter. Mortality would not gradually increase over a 10-year period. Secondly, an increase in incidence, hospitalisations and mortality of pneumonia as observed in three independent systems (LMR, CBS, CMR) contradicts a systematic registration bias in one of these systems. Third, in other countries, with other registration methods, a increase in pneumonia-related mortality has also been observed. (12;13) Therefore, despite the possibility of misclassifi cation, the increase in pneumonia-related mortality seems real.

Age patterns, co-morbidity and environmental factors

Because pneumonia is most prevalent in older age groups, it is attractive to explain the increase in pneumonia-related mortality by the ageing of the population. However, the mortality rates standardized for age and sex also show an increase in pneumonia-related mortality (fi gure 1). In this standardization, mortalitly rates are calculated for a standard population, in this case the adult population of 1990. The increase in mortality is present in all age categories, but age specifi c mortality rates show that pneumonia-related mortality is especially increased in the higher age groups. (fi gure 2). It is therefore likely that other factors also have contributed to the observed increase in mortality. It is possible that the number of patients with underlying diseases that affect specifi c and non-specifi c defense against micro-

28

Chapter 1Chapter 2

organisms has increased. The increase in prevalence of diabetes mellitus and COPD as observed in recent years (source: RIVM compass for population health) could have contributed to the increase in the number of deaths because of lower respiratory tract infections.In patients with diabetes the increase in incidence was most outspoken in the age-group 70-79 year. (source: rapport CMR peilstations 2001, AIM Bartelds) (14) Although a considerable part of pneumonia and infl uenza related mortality is contributed to diabetes, there is no unequivocal relation between diabetes mellitus and the risk of death from infections. (15;16) In addition, in the Netherlands, this risk-group is annually vaccinated against infl uenza with good success and a vaccination level of about 75%. (17) (bron: RIVM kompas voor de volksgezondheid). Infl uenza related pneumonia mortality was therefore probably not contributory to the observed increase in pneumonia related mortality. The number of new HIV-infections in the Netherlands also has increased in the period 1990-2000. In 1990, 250 patients were diagnosed with new HIV infections and 375 in 2000, increasing the number of HIV infected patients to 8496 in 2002. (source: RIVM rapport 441100018). A number of facts speaks against a large role of this patient group in the increase in pneumonia-related mortality. First, the patient group of HIV infected patients has a lower mean age than the patient group with the most outspoken increase in pneumonia-related mortality. Second, a European study showed that HIV-infected patients with CD4–positive cells of > 200 / mm3 without Pneumocystis Carinii Pneumonia prophylaxis in 18 months had only 3,5% developed pneumonia. A large contribution to the pneumonia-related mortality, also because of the successful treatment strategies for this group, can therefore be excluded.Interestingly, the annual number of hospitalisations per 100.000 inhabitants because of pneumonia is higher in urban areas compared to rural areas. (source: RIVM nationaal kompas voor de volksgezondheid). It can therefore be argumented that pneumonia is a disease of densely populated areas, where infectious diseases spread more easily or that the increase in air pollution in the 90´s has contributed to the observed mortality rates. (18)An increase in ageing and the number of patients with chronic

29

Chapter 2

illnesses facilitating an infection of the lower respiratory tract are therefore only in part explanatory for the increase in pneumonia-related morbidity and mortality in recent years.

Causative micro-organisms

The increase in mortality and hospitalisations in 1990-1999 and the decline in 2000-2002, possibly indicates temporal changes in certain causative micro-organisms. Despite S. pneumoniae being the most prevalent causative micro-organism in nearly every study, in recent years, the so-called atypical micro-organisms as Mycplasma pneumoniae, Chlamydia pneumoniae and Legionella pneumophila have received much attention. (19-26)Especially infections with L. pneumophila is associated with high mortality and infections with C. pneumoniae can also lead to severe pneumonia.. (27-31) The diagnosis of C. pneumoniae infection is however diffi cult. (32) In addition, the proportion of atypical infections is variable over the years. (33;34) Whether the amount of atypical infections fl uctuates in the Netherlands also, is not clear. There is only 1 prospective study evaluating the etiology of pneumonia in the Netherlands in 1990-2000. (19) Reported infections with Legionella pneumophila have hardly or not at all risen in 1990-1999 and plays a minor role in the etiology of pneumonia, at least, if no outbreak situation, as the Bovenkarspel epidemic of 1999, is present. (35;36) Undertreatment of these atypical infections is therefore no reasonable explanation for the increase in pneumonia related mortality in the Netherlands. (37) In about 50% of pneumonia a causative micro-organism cannot be found. (19-21;24;26;38-40) Because routine investigations in the Netherlands do not include viral diagnostics, part of pneumonias may have a primary viral cause. Recently, a number of new viruses have been discovered that can cause severe lower respiratory tract infections. In children, the in 2001 discovered metapneumovirus is probably an important cause for severe respiratory tract infections. It can also play a etiological role in adults. (41;42) However, it is not probable that an increase in discovered viruses has led to an increase in pneumonia related morbidity and mortality. During epidemics, the infl uenza virus can lead to serious complications

30

Chapter 1Chapter 2

with fatal outcome, particularly in high risk patients and the elderly In 1985/1986, 1989/1990, 1993/1994, 1995/1996 and 1999/2000 infl uenza epidemics have occurred with high mortality rates. During these epidemics, there were 2.000 to 4.000 more deaths than usually in winter periods. (source: CBS Monthly statistics of the population, RIVM kompas voor de volksgezondheid) Whether all of these deaths are classifi ed as pneumonia related mortality is not clear. Only in the 1993-1994 season this infl uenza epidemic seems to have led to a peak in pneumonia related mortality. (fi gure 1). On the other hand, in mild infl uenza seasons as 2000-2001 and 2001-2002, mortality seems to have decreased. One explanation could be that the weakest population in the severe infl uenza season 1999-2000 died, leading to a decrease in mortality in the succeeding years. Concluding, clear indications of a changing etiologic spectrum contributing to the increase in mortality as a result of lower respiratory tract infections are lacking.

Antibiotic resistance

Resistance of causative micro-organisms for pneumonia has remained low: in the Netherlands, ± 1% of S. pneumoniae is less susceptible for penicillin and ± 7% of erythromycin. (source: http://www.earss.rivm.nl) An increase of less susceptible or resistant S. pneumoniae can therefore not have played a role in the Netherlands and is no explanation for the increase in mortality. Whether in vitro resistance infl uences clinical outcomes is much debated. (43;44)

Conclusion

According to several national and international registration systems, in recent years the incidence, the number of hospitalisations and mortality of pneumonia has increased. A simple explanation is absent, but above resistance and changes in causative micro-organisms, the increase of the elderly population and an increase in co-morbidities seem the most important explanations. To reduce the mortality and morbidity burden, an integrated approach of elderly patients and patients with co-morbidity by specialists in geriatrics, internal medicine, pulmonarly medicine and infectious diseases prevails over

31

Chapter 2

specifi cally dealing with solely the problem “pneumonia” by one-sided development of vaccination strategies or newer treatment strategies.

Acknowledgements

Mrs. drs. A. van der Meulen (Centraal Bureau voor de Statistiek) and mr. W.F. Hoogen Stoevenbeld (Prismant) provided important data.

32

Chapter 1Chapter 2

Reference List



(3) Koivula I, Sten M, Makela PH. Risk factors for pneumonia in the elderly. Am J Med 1994; 96(4):313-320.

(4) Lipsky BA, Boyko EJ, Inui TS, Koepsell TD. Risk factors for acquiring pneumococcal infections. Arch Intern Med 1986; 146(11):2179-2185.




(8) Thompson WW, Shay DK, Weintraub E, Brammer L, Cox N, Anderson LJ et al. Mortality associated with infl uenza and respiratory syncytial virus in the United States. JAMA 2003; 289(2):179-186.

(9) Pneumonia and infl uenza death rates--United States, 1979-1994. MMWR Morb Mortal Wkly Rep 1995; 44(28):535-537.


(11) Guevara RE, Butler JC, Marston BJ, Plouffe JF, File TM, Jr., Breiman RF. Accuracy of ICD-9-CM codes in detecting community-acquired pneumococcal pneumonia for incidence and vaccine effi cacy studies. Am J Epidemiol 1999; 149(3):282-289.


(13) Pneumonia and infl uenza death rates--United States, 1979-1994.

33

Chapter 2

MMWR Morb Mortal Wkly Rep 1995; 44(28):535-537.

(14) Ubink-Veltmaat LJ, Bilo HJ, Groenier KH, Houweling ST, Rischen RO, Meyboom-de Jong B. Prevalence, incidence and mortality of type 2 diabetes mellitus revisited: a prospective population-based study in The Netherlands (ZODIAC-1). Eur J Epidemiol 2003; 18(8):793-800.

(15) Valdez R, Narayan KM, Geiss LS, Engelgau MM. Impact of diabetes mellitus on mortality associated with pneumonia and infl uenza among non-Hispanic black and white US adults. Am J Public Health 1999; 89(11):1715-1721.

(16) King H, Aubert RE, Herman WH. Global burden of diabetes, 1995-2025: prevalence, numerical estimates, and projections. Diabetes Care 1998; 21(9):1414-1431.

(17) Hak E, Hermens RP, Hoes AW, Verheij TJ, Kuyvenhoven MM, van Essen GA. Effectiveness of a co-ordinated nation-wide programme to improve infl uenza immunisation rates in The Netherlands. Scand J Prim Health Care 2000; 18(4):237-241.

(18) Fischer P, Hoek G, Brunekreef B, Verhoeff A, van Wijnen J. Air pollution and mortality in The Netherlands: are the elderly more at risk? Eur Respir J Suppl 2003; 40:34s-38s.:34s-38s.

(19) Bohte R, Furth R van, Broek PJ van den. Aetiology of community-acquired pneumonia; a prospective study among adults requiring admission to hospital. Thorax 1995; 50:543-547.

(20) Guthrie R. Community-acquired lower respiratory tract infections, etiology and treatment. Chest 2001; 120:2021-2034.


(22) Macfarlane JT, Colville A, Guion A, Macfarlane RM, Rose DH. Prospective study of aetiology and outcome of adult lower-respiratory- tract infections in the community. Lancet 1993; 341(8844):511-514.

(23) Mandell LA. Community-acquired pneumonia. Etiology, epidemiology, and treatment. Chest 1995; 108(2 Suppl):35S-42S.

(24) Pachon J, Prados MD, Capote F, et al. Severe community-acquired pneumonia: etiology, prognosis and treatment. Am Rev Respir Dis 1990; 142:369-373.

(25) Roson B, Carratala J, Dorca J, Casanova A, Manresa F, Gudiol F. Etiology, reasons for hospitalization, risk classes, and outcomes of

34

Chapter 1Chapter 2

community-acquired pneumonia in patients hospitalized on the basis of conventional admission criteria. Clin Infect Dis 2001; 33(2):158-165.


(27) Gacouin A, Le Tulzo Y, Lavoue S, Camus C, Hoff J, Bassen R et al. Severe pneumonia due to Legionella pneumophila: prognostic factors, impact of delayed appropriate antimicrobial therapy. Intensive Care Med 2002; 28(6):686-691.

(28) Balis E, Boufas A, Iliopoulos I, Legakis NJ, Zerva L. Severe community-acquired pneumonia with acute hypoxemic respiratory failure due to primary infection with Chlamydia pneumoniae in a previously healthy adult. Clin Infect Dis 2003; 36(12):e155-e157.

(29) Ewig S, Torres A. Is Chlamydia pneumoniae an important pathogen in patients with community-acquired pneumonia? Eur Respir J 2003; 21(5):741-742.

(30) Marrie TJ, Peeling RW, Reid T, De Carolis E. Chlamydia species as a cause of community-acquired pneumonia in Canada. Eur Respir J 2003; 21(5):779-784.

(31) Miyashita N, Fukano H, Okimoto N, Hara H, Yoshida K, Niki Y et al. Clinical Presentation of Community-Acquired Chlamydia pneumoniae Pneumonia in Adults(*). Chest 2002; 121(6):1776-1781.

(32) Tuuminen T, Palomaki P, Paavonen J. The use of serologic tests for the diagnosis of chlamydial infections. J Microbiol Methods 2000; 42(3):265-279.

(33) Houck PM, MacLehose RF, Niederman MS, Lowery JK. Empiric antibiotic therapy and mortality among medicare pneumonia inpatients in 10 western states. Chest 2001; 119:1420-1426.

(34) Yu VL, Vergis EN. New macrolides or new quinolones as monotherapy for patients with community-aquired pneumonia: our cup runneth over? Chest 1998; 113:1158-1159.

(35) Hoepelman IM. Legionella epidemie in Nederland. Ned Tijdschr Geneeskd 1999; 143(34):1757-1758.

(36) den Boer JW, Friesema IH, Hooi JD. [Reported cases of Legionella pneumonia in the Netherlands, 1987-2000]. Ned Tijdschr Geneeskd 2002; 146(7):315-320.

(37) Oosterheert JJ, Bonten MJ, Schneider MM, Hoepelman IM. [Community

35

Chapter 2

acquired pneumonia; no reason to revise current Dutch antibiotic guidelines]. Ned Tijdschr Geneeskd 2003; 147(9):381-386.

(38) Macfarlane JT, Colville A, Guion A, Macfarlane RM, Rose DH. Prospective study of aetiology and outcome of adult lower-respiratory- tract infections in the community. Lancet 1993; 341(8844):511-514.

(39) Mandell LA. Community-acquired pneumonia. Etiology, epidemiology, and treatment. Chest 1995; 108(2 Suppl):35S-42S.

(40) Roson B, Carratala J, Dorca J, Casanova A, Manresa F, Gudiol F. Etiology, reasons for hospitalization, risk classes, and outcomes of community-acquired pneumonia in patients hospitalized on the basis of conventional admission criteria. Clin Infect Dis 2001; 33(2):158-165.

(41) Kahn JS. Human metapneumovirus: a newly emerging respiratory pathogen. Curr Opin Infect Dis 2003; 16(3):255-258.

(42) van den Hoogen BG, de Jong JC, Groen J, Kuiken T, de Groot R, Fouchier RA et al. A newly discovered human pneumovirus isolated from young children with respiratory tract disease. Nat Med 2001; 7(6):719-724.

(43) Garau J. Treatment of drug-resistant pneumococcal pneumonia. Lancet Infect Dis 2002; 2(7):404-415.

(44) Lonks JR, Garau J, Gomez L, Xercavins M, Ochoa dE, Gareen IF et al. Failure of macrolide antibiotic treatment in patients with bacteremia due to erythromycin-resistant Streptococcus pneumoniae. Clin Infect Dis 2002; 35(5):556-564.

36

Chapter 1Chapter 2

3Severe community-acquired

pneumonia: What’s in a

name…

Curr Opin Infect Dis. 2003; 16 (2): 153-9

JJ Oosterheert, MJM Bonten, E Hak,

MME Schneider, IM Hoepelman

38

Chapter 1Chapter 3

Abstract

Purpose: Formerly, patients with community-acquired pneumonia (CAP) admitted to an intensive care unit (ICU) were considered as having severe CAP (SCAP). Recently, guidelines have distinguished severe and non-severe community-acquired pneumonia based on clinical defi nitions. In this review, we describe the different defi nitions of SCAP, and whether a differentiation based on these defi nitions refl ects differences in etiology, risk factors, diagnostic approaches and treatment.Recent fi ndings: New defi nitions do not seem to identify patients with high risk of mortality accurately; patients not admitted to ICU could also be diagnosed as having SCAP. Host-factors, such as genetic factors and underlying diseases can infl uence severity of presentation of CAP. Distribution of pathogens in SCAP and non-severe CAP is comparable and initial antibiotic therapy in patients with severe disease should provide coverage of S. pneumoniae and L. pneumophila, as delay is associated with worse outcomes. However, recent studies also suggested additional benefi t of atypical coverage in non-severe CAP. As a result, initial therapy with a β-lactam plus a macrolide or an anti-pneumococcal fl uoroquinolone is recommended for all patients with CAP. Furthermore, the value of vaccination against pneumococci to prevent episodes of SCAP is yet unknown.Summary: As current guidelines do not adequately identify patients with high risk of mortality and ICU admittance, clinical judgment remains important. Based on distribution of pathogens, investigational procedures and therapy recommended in recent guidelines, differentiation between SCAP and non-severe CAP does not seem useful. Whether atypical coverage indeed has additional value in non-severe or pneumococcal CAP, however, remains to be determined. In addition, the preventive benefi t of infl uenza and pneumococcal vaccination for development of SCAP awaits further evidence.

39

Chapter 3

Introduction

Community-acquired pneumonia (CAP) is a common and potentially fatal disease leading to high health care costs. Estimated annual costs for treating CAP were over 1 billion dollar in the UK in 1997 and 9,7 billion dollar in the USA in 2001.(1;2) Most cases of CAP are successfully managed in primary care. However, depending on the severity of disease, about 20 percent of these patients need hospitalization and approximately 1% of all CAP patients require treatment in an intensive care unit (ICU) (3;4). These ICU patients are often referred to as having severe community-acquired pneumonia (SCAP). Recent guidelines have tried to defi ne the diagnosis of SCAP more precisely, but controversy remains on the golden standard defi nition of SCAP. (5) According to the American Thoracic Society (ATS), the Infectious Diseases Society of America (IDSA) and the British Thoracic Society (BTS), SCAP can be considered a syndrome with distinct etiology, diagnostic criteria and treatment options. (6;7) In this review we will discuss whether SCAP indeed should be approached as a distinct entity, considering current diagnostic defi nitions, etiologic factors, preventive opportunities, prognosis and therapeutic strategies.

Diagnosis of SCAP

Defi nitions

There is no golden standard defi nition to diagnose SCAP. In many trials, SCAP is pragmatically defi ned by the need of intensive care unit (ICU) admittance. However, this defi nition does not include objective measurements and is dependent on specifi c policies for ICU admittance that may vary considerably between centers with an ICU. (8) The American Thoracic Society (ATS) proposed to defi ne SCAP on the presence of a certain set of minor and major clinical signs or symptoms (Table 1) (6). According to this defi nition, an episode of CAP with two or more minor criteria or one major criterion should be considered as SCAP. The British Thoracic Society defi ned SCAP using a more or less similar set of “core”, “additional” and “pre-existing” adverse prognostic features. (Table 1) (6). Patients with two “core”

40

Chapter 1Chapter 3

adverse prognostic features should be considered to have SCAP. If only one such a feature is present, CAP could still be considered SCAP based on clinical judgment, which can be assisted by considering “pre-existing” or “additional” adverse prognostic features. Furthermore, the Pneumonia Severity Index (PSI) can be used to predict mortality and thus severity of illness in patients admitted with CAP. In the original study by Fine and colleagues, mortality in the highest risk class (V, PSI >130) was 30%. How do these defi nitions for SCAP correlate with the pragmatic defi nition of need of ICU admission? Surprisingly, in Fine’s study, only 17% of patients in risk class V had been admitted to ICU. Therefore the PSI score might have limited value as to set an objective diagnosis of SCAP. (9) Similarly, the former ATS (1993) criteria were shown to be highly sensitive but insuffi cient specifi c in predicting the need for ICU-admittance. (10) BTS and revised ATS diagnostic criteria for SCAP were evaluated in patients who did (n=170) and did not (n=1,169) receive treatment in ICU in three U.S. and one Canadian hospital. (11) ATS-criteria were the best discriminator for ICU admittance (receiver operating characteristic area under the curve (AUC) of 0.68;95% confi dence interval (CI): 0.64-0.73) and high PSI (class IV or V) discrimated best in predicting mortality (AUC of 0.75; 95% CI: 0.71-0.78). However, these results refl ect far from ideal discriminator characteristics. Therefore, current defi nitions of SCAP do not appear to be robust enough to guide clinical care. This can be illustrated with a simple clinical example: most physicians will agree that a younger patient (e.g. female, 35 years of age) admitted with pneumonia and a respiratory rate of 29/min, P02 of 62 mmHg, a systolic blood pressure of 95 mmHg (and not responding to fl uid replacement therapy) and a pulse rate of 121/min suffers from severe disease. However, this patient does neither qualify for the diagnosis SCAP according to the ATS and BTS guidelines nor for the highest risk class (V) according to the PSI. This shows that clinical judgment, which is diffi cult to describe in objective terms, remains a prerequisite in the decision to decide to hospitalize or admit a patient to an ICU. This is refl ected in the new BTS-guidelines that include subjective clinical interpretation of physicians. [6]

41

Chapter 3

ATS-guidelines 1993

Any-one of the following

Respiratory rate > 30/minSevere respiratory failure (PaO2/FiO2<250)Bilateral involvement in chest radiographInvolvement of >2 lobes in chest radiograph (mulitilobar involvement)Systolic blood pressure < 90 mm HgDiastolic blood pressure < 60 mm HgRequirement for mechanical ventilationIncrease of the size of infi ltrates by >50% in the presence of clinical nonresponse to treatment or deterioration (progressive infi ltrates)Requirements of vasopressors >4h (septic shock)Serum creatinine > 2mg/dl or acute renal failure requiring dialysis (renal failure)

ATS-guidelines (6)

Presence of two minor criteria or presence of one major criterion

Minor CriteriaRespitratory rate > 30/minPaO2 / FiO2 < 250Bilateral pneumonia or multilobar pneumoniaSystolic BP < 90 mm Hg

Major CriteriaNeed for mechanical ventilationIncrease of size of infi ltrates >50% within 48 hSeptic shock or the need for pressors for > 4 h Acute renal failure (urine output < 80 ml in 4 h or serum creatinine > 2 mg/dl in the absence of chronic renal failure)

BTS guidelines [7]

Two or more core adverse prognostic features or one core adverse prognostic feature based on clinical judgement, considering additional adverse prognostic features and pre-existing adverse prognostic features

Core adverse prognostic featuresNew mental confusionUrea > 7 mmol / l Respiratory rate 30/min or moreSystolic BP < 90 mmHg or diastolic BP < 60 mmHg

Additional adverse prognostic featuresPaO2 < 8 kPa / SaO2 < 92%CXR: bilateral / multilobar shadowing

Pre-existing adverse prognostic featuresAge 50 years and overPresence of coexisting disease

Table 1

3 Defi nitions of SCAP

42

Chapter 1Chapter 3

Initial investigational procedures in SCAP

In the diagnostic work-up, initial investigational procedures can be useful in identifying causative micro-organisms early in the stage of the disease which may be helpful in directing the most effective antimicrobial therapy. Microbiological investigations usually include cultures from tracheal aspirates if available, blood cultures and serological tests. More invasive procedures such as cultures of pleural and broncho-alveolar fl uids can be added. In one study, these invasive techniques were applied with limited adverse events and pathogens were detected in 32-48% of cases. (12) Although implications for the treatment were not measured in this study, the authors do suggest to use invasive techniques to differentiate between upper airway colonisation and active pneumonia. Investigational procedures like cultures of blood and sputum or serological tests provide results after a number of days to weeks and are, therefore, not suitable to guide initial therapy. To rapidly identify S. pneumoniae, sputum gram stain and urinary pneumococcal antigen tests, which detect the C polisaccharide wall antigen common to all S. pneumoniae strains (13), can be used. Both detection methods have their limitations with regard to test characteristics and impact on treatment decisions and, therefore, on the cost-effectiveness of the diagnostic procedure (14). A urinary antigen test can be used to identify L. pneumophila type I, and is considered to be useful in the initial management of SCAP by both ATS and BTS guidelines. [6;7](6) The accuracy of these urinary antigen test appears to be higher in the more severe episodes of disease (15)However, studies evaluating the effects of these additional and invasive diagnostic procedures on clinical practice are needed.

Does SCAP have a distinct etiology and prognosis?

Pathogens

Streptococcus pneumoniae is the most frequent causative micro-organism which can be detected in about 20% of cases of SCAP.

43

Chapter 3

Other frequently identifi ed pathogens are Haemophilus infl uenzae, Gram negative enteric bacilli (GNEB), Legionella pneumophila and Staphylococcus aureus. (12;16-19) Atypical pathogens like Mycoplasma pneumoniae may sporadically cause severe CAP. (16) Pseudomonas aeruginosa should be considered in patients with structural damage to the respiratory tract, for example in patients with bronchiectasis or chronic obstructive pulmonary disease (COPD). (17) According to Chen et al, Acinetobacter baumanii, which must be treated with a 3rd generation cephalosporin and aminoglycosides, should be considered as a possible etiologic agent when patients with a fulminant course present during the warmer and more humid months of the year or when patients are younger alcoholics. (20) Ruiz et al. showed that severe pneumonia (requiring ICU admission) was independently associated with Gram negative enteric bacilli (GNEB) and P. Aeruginosa. (16) Apart form these pathogens, all other pathogens associated with SCAP are also frequently isolated from patients with non-severe CAP. Furthermore, high levels of mixed infections have been reported in CAP as well as in SCAP in some studies. (12) (21) Importantly, up to 60% of episodes remain of unknown etiology in both types of CAP. (3) In another study comparing etiologies of CAP and SCAP, there were no differences in the incidence of L. pneumophila infections. (22) Based on clinical severity of CAP it seems therefore impossible to predict the distribution of pathogens. (table 2) Viral pneumonias due to infection with infl uenza, respiratory syncitial virus and parainfl uenza virus can be life threatening in elderly and immunocompromised patients. Infl uenza pneumonia may be complicated by direct involvement of the lung tissue or by secondary bacterial infections caused by S. aureus, S. pneumoniae, H. infl uenzae or other Gram-negative pathogens. (23) However, distribution of viral pneumonias and mixed infections is equally in SCAP and CAP. Therefore it seems unlikely that these secondary infections follow a more severe course.

Hosts

Age, genetic factors and underlying high-risk conditions that infl uence host immune system and pulmonary reserves play a role

44

Chapter 1Chapter 3

in the etiology of CAP. Although age is considered a prognostic factor for mortality, (24) some studies showed that age by itself is not of prognostic importance, but rather a marker for adverse features. (25;26) Conditions affecting aspecifi c innate immunity can increase the risk of developing SCAP. For example, patients with mutations in the gene encoding mannose binding lectin (MBL) could be at substantially increased risk of developing CAP and subsequent invasive pneumococcal disease. MBL is a key mediator of innate host immunity that activates the complement pathway and directly opsonises some infectious pathogens. Patients with MBL-codon mutations were at a 2.6 fold risk of developing invasive pneumococcal disease. Five percent of north Europeans and north Americans are homozygotes for MBL codon variants. (27) Also, TNF-α hypersecretor gene polymorphisms are associated with the development of septic shock in patients with CAP. (28) Patients with HIV-infection and CD4 < 200 cells/μl, affecting specifi c cellular immune response, are at high risk of SCAP due to Pneumocystis carinii and other opportunistic infections, whereas severe bacterial pneumonia is more prevalent in HIV-patients with CD4>200 cells/μl. (22) In patients with underlying respiratory conditions, like COPD, infection development and respiratory failure may occur more easily in case of infection. Limited infectious foci causing even small increases of ventilation-perfusion inequalities may lead to severe acute respiratory failure. (8) This patient group is prone for infection with P. aeruginosa, (17) which has been associated with a high risk of mortality. (4)

Initial Antimicrobial Treatment

Initial therapy of CAP is often empiric because etiology cannot be predicted from clinical, laboratory or radiological fi ndings, (29-31) and should be prompt as rapid antibiotic delivery is associated with better outcomes. (32;33) For example, fl uoroquinolone administration within 8 hours of ICU admission was associated with lower mortality rates (OR 0,18 95% CI: 0.04 - 0.71) in severe L. pneumophila pneumonia. (34) Because S. pneumoniae is the most frequent pathogen in SCAP and L. pneumophila is feared for the potential severity of infection, (24) coverage of both pathogens has become the standard of care in SCAP. North-American and European (except

45

Chapter 3

Dutch) guidelines (6)(7;35) recommend initial therapy with the combination of a macrolide and a broad-spectrum β-lactam antibiotic or a fl uoroquinolone with antipneumococcal activity in all patients admitted to hospital with CAP. A third generation cephalosporin with activity against P. aeruginosa should only be considered in patients with structural damage of the respiratory tract. These guidelines, therefore, do not differentiate between SCAP and non-severe CAP with regard to empirical treatment.Results from retrospective studies suggest that combinations of β-lactam antibiotics plus macrolides or monotherapy with fl uoroquinolones as initial therapy for non-severe CAP reduce length of stay and mortality, even when S. pneumoniae is the causative micro-organism. (36-41) Possible explanations for these favourable results are a large and mainly unrecognised role of atypical pathogens in the etiology of CAP, anti-infl ammatory effects of macrolides or resistance to β-lactams of the most important pathogens. However, there are inconsistencies in reported outcomes and confounding bias, (i.e. groups receiving the combination therapy as compared to the mono-therapy are not comparable with regard to average prognosis) may have infl uenced these results. For example the use of macrolides in SCAP was a marker for less severe disease in one study. (19) Therefore, adequately powered studies using a randomised controlled trial design that focus on the incremental value of macrolides to standard antimicrobial therapy are warranted. (42;43)The newer fl uoroquinolones with antipneumococcal properties can be used to treat SCAP. (44) Finch et al. showed moxifl oxacin to have better clinical and bacteriological success when compared to co-amoxiclav with or without a macrolide in the treatment of patients admitted with CAP and SCAP. (45) However, clinical failures of the newer fl uoroquinolones due to development of resistance have already been reported (46) . With regard to penicillin resistant S. pneumoniae (MIC<4μg/ml): episodes of CAP caused by such pathogens can still be adequately treated with β-lactams at the right dosage. (47)

Prevention

Vaccination strategies against pneumococci and infl uenza may prevent episodes of SCAP. Currently, a 23-valent PPV, covering the 23 most

46

Chapter 1Chapter 3

prevalent serotypes, is recommended in most Western countries for all persons at high risk for developing pneumonia. However, the available data to support these recommendations are far from consistent. Some non-experimental retrospective studies have shown that PPV might be effective and might even be cost-saving in preventing invasive pneumococcal disease. (48) However, confounding by indication might have played a considerable role in these studies affecting the validity of the results. A recent systematic review yielded confl icting results with regard to the prevention of non-invasive pneumonia and even in high-risk patients who were previously admitted to hospital with pneumonia, the health benefi ts of pneumococcal vaccination remained unclear. (49-55). Moreover, the incremental cost-effectiveness of PPV programs as compared to infl uenza vaccination programs that have been successfully implemented in most Western countries is unknown. Therefore, the real value of vaccination strategies in terms of effects on non-invasive pneumonia and costs needs to be further explored, preferably in randomised controlled trials.(56;57)Infl uenza vaccination is recommended for more or less the same group of patients that are recommended for PPV . These patients have a high risk for complications of infl uenza infection, like severe viral pneumonia or severe secondary bacterial infection. Predicting the occurrence of such complications is possible on the basis of some very simple host characteristics including age, gender, co-morbidity and external factors. (58) Especially elderly patients may benefi t from infl uenza vaccination (54). In meta-analysis vaccine effectiveness was 50% for preventing hospitalisation and 68% for preventing death. (59) Data on clinical effectiveness of the vaccine in reducing post-infl uenza complications among high-risk persons of working age are limited and indicate no or at most limited benefi ts from vaccination (58;60) (55)

Conclusion

CAP is a disease with a broad spectrum of presentations and SCAP represents the end of its severity-spectrum. New defi nitions of SCAP may identify patients with a low disease severity, contrary to the former pragmatic defi nition of needing ICU admittance. We reviewed whether distinguishing SCAP from non-severe CAP has implications

47

Chapter 3

for patient management. Importantly, the new defi nitions of SCAP have not been shown to accurately identify patients with high risk of mortality. Moreover, distribution of pathogens causing CAP and SCAP appear to be comparable, and extra diagnostic efforts to identify pathogens in SCAP, have not shown to improve adequacy of therapy. Nevertheless, in SCAP empirical therapy tends to be broad-spectrum, covering most frequently associated pathogens such as S. pneumoniae and L. pneumophila, as delay in initiating adequate therapy is associated with worse outcomes. Furthermore, coverage of atypical pathogens in all cases of CAP may improve outcome as suggested by some retrospective studies. These studies and the new defi nitions for SCAP favour the use of broad spectrum antibiotics in a large and less severely ill population. This can lead to unnecessary antibiotic use, induction of resistance, and an increase of costs and adverse events. Therefore, prospective trials should determine whether a broad-spectrum approach has real benefi ts. Similarly, some reports have suggested that vaccination prevents severe pneumococcal disease. The real benefi ts of vaccination however remain to be determined, preferably in randomised controlled trials.

48

Chapter 1Chapter 3

SC

AP

Veg

eli

n 9

9(1

7)

Ru

iz 9

9(1

6)

Ru

iz 9

9(1

8)

Lu

na 0

0(2

3)

Park

01

(no

nh

iv)

(22

)

Park

01

(hiv

) El S

olh

01

(12

)R

ello

02

(1

9)

Avera

ge

(ran

ge)

Defi

nitio

n S

CAP

ICU

ad

mitta

nce

<

48 h

Adm

itte

d

to I

CU

Adm

itte

d

to I

CU

Adm

itte

d

to I

CU

Adm

itte

d t

o

ICU

Adm

itte

d t

o

ICU

Adm

itte

d t

o

ICU

, re

quirin

g

mec

han

ical

ve

ntila

tion

Adm

itte

d

to I

CU

Patien

t ch

arac

terist

ics

No k

now

n

HIV

infe

ctio

nKnow

n H

IV

infe

ctio

nAge

> 7

5

S.

pneu

monia

e35

24

20

15

18

514

15

18 (

5-3

5)

H.

infl uen

zae

11

65

36

07

35 (

3-1

1)

GN

EB

11

65

35

11

7 (

3-1

1)

S.

aure

us

72

25

30

74 (

2-7

)

Legio

nel

la5

22

56

59

65 (

2-9

)

Myc

opla

sma

03

35

35

3 (

2-5

)

Pseu

dom

onas

45

43

015

22

4 (

2-1

5)

unkn

ow

n34

47

40

41

33

25

47

43

39 (

25-4

7)

mix

ed18

18

13

11

30

18

18 (

11-3

0)

PCP

030

15 (

0-3

0)

49

Chapter 3

CA

PLie

berm

an

9

6 (

61

)R

uiz

99

(18

)Lim

01

(21

)Lu

na 0

0(2

3)

Park

0

1(n

on

-HIV

)(2

2)

Park

01

HIV

Fin

ch 0

2(4

5)

Avera

ge

(ran

ge)

Patien

t popula

tion

Adm

itte

d w

ith

CAP,

SCAP

not

excl

uded

Adm

itte

d w

ith

CAP,

SCAP

not

excl

uded

Adm

itte

d w

ith

CAP,

SCAP

not

excl

uded

Adm

itte

d w

ith

Clin

ical

dia

gnosi

s CAP,

cat

egory

m

ild a

nd

moder

ate

Adm

itte

d C

AP,

no k

now

n

HIV

infe

ctio

n,

SCAP

excl

uded

Adm

itte

d w

ith

CAP,

know

n

HIV

infe

ctio

n,

SCAP

excl

uded

Adm

itte

d

with C

AP.

SCAP

not

excl

uded

S.

pneu

monia

e43

21

48

910

833

25 (

8-4

8)

H.

infl uen

zae

66

75

52

12

6 (

2-1

2)

GN

EB

54

01

02 (

0-5

)

S.

aure

us

2#

34

12

42 (

1-4

)

Legio

nel

la16

83

14

48*

6 (

1-1

6)

Myc

opla

sma

29

53

68

88*

9 (

3-2

9)

Chla

myd

ia18

513

35

28*

7 (

2-1

8)

Pseu

dom

onas

2#

41**

20

01 (

0-4

)

PCP

00

32

11 (

0-3

2)

unkn

ow

n19

54

25

49

49

37

39 (

19-5

4)

mix

ed38

41

47***

58

16

26 (

5-4

7)

Tab

le 2

Com

par

ison o

f et

iolo

gy

of

com

munity-

acquired

pneu

monia

and s

ever

e co

mm

unity-

acquired

pneu

monia

in r

ecen

t st

udie

s.

* t

oget

her

acc

ounte

d for

8.2

% o

f in

fect

ions,

with m

ycopla

sma

bei

ng t

he

most

pre

vale

nt

** w

as iden

tifi ed

, but

pro

bab

ly n

ot

the

etio

logic

agen

t (c

olo

niz

atio

n)

*** incl

udin

g v

iral

agen

ts

# S

. au

reus,

P.

aeru

gin

osa

and A

cinet

obac

ter

toget

her

acc

ounte

d for

2%

50

Chapter 1Chapter 3

Reference List



(3) Moine P, Vercken JB, Chevret S, Chastang C, Gajdos P. Severe community-acquired pneumonia. Etiology, epidemiology, and prognosis factors. French Study Group for Community-Acquired Pneumonia in the Intensive Care Unit. Chest 1994; 105(5):1487-1495.

(4) Torres A, Serra-Batlles J, Ferrer A, Jimenez P, Celis R, Cobo E et al. Severe community-acquired pneumonia. Epidemiology and prognostic factors. Am Rev Respir Dis 1991; 144(2):312-318.

(5) Baudouin SV. The pulmonary physician in critical care . 3: critical care management of community acquired pneumonia. Thorax 2002; 57(3):267-271.



(8) Neuhaus T, Ewig S. Defi ning severe community-acquired pneumonia. Med Clin North Am 2001; 85(6):1413-1425.


(10) Ewig S, Ruiz M, Mensa J, Marcos MA, Martinez JA, Arancibia F et al. Severe community-acquired pneumonia. Assessment of severity criteria. Am J Respir Crit Care Med 1998; 158(4):1102-1108.

(11) Angus DC, Marrie TJ, Obrosky DS, Clermont G, Dremsizov TT, Coley C et al. Severe Community-acquired Pneumonia: Use of Intensive Care Services and Evaluation of American and British Thoracic Society Diagnostic Criteria. Am J Respir Crit Care Med 2002; 166(5):717-

51

Chapter 3

723.

(12) El Solh AA, Sikka P, Ramadan F, Davies J. Etiology of severe pneumonia in the very elderly. Am J Respir Crit Care Med 2001; 163(3 Pt 1):645-651.


(14) Ewig S, Schlochtermeier M, Goke N, Niederman MS. Applying sputum as a diagnostic tool in pneumonia(*) : limited yield, minimal impact on treatment decisions. Chest 2002; 121(5):1486-1492.

(15) Yzerman EP, den Boer JW, Lettinga KD, Schellekens J, Dankert J, Peeters M. Sensitivity of three urinary antigen tests associated with clinical severity in a large outbreak of Legionnaires’ disease in The Netherlands. Journal of Clinical Microbiology 2002; 40(9):3232-3236.

(16) Ruiz M, Ewig S, Torres A, Arancibia F, Marco F, Mensa J et al. Severe community-acquired pneumonia. Risk factors and follow-up epidemiology. Am J Respir Crit Care Med 1999; 160(3):923-929.



(19) Rello J, Catalan M, Diaz E, Bodi M, Alvarez B. Associations between empirical antimicrobial therapy at the hospital and mortality in patients with severe community-acquired pneumonia. Intensive Care Med 2002; 28(8):1030-1035.

(20) Chen MZ, Hsueh PR, Lee LN, Yu CJ, Yang PC, Luh KT. Severe community-acquired pneumonia due to Acinetobacter baumannii. Chest 2001; 120(4):1072-1077.


(22) Park DR, Sherbin VL, Goodman MS, Pacifi co AD, Rubenfeld GD, Polissar NL et al. The etiology of community-acquired pneumonia at

52

Chapter 1Chapter 3

an urban public hospital: infl uence of human immunodefi ciency virus infection and initial severity of illness. J Infect Dis 2001; 184(3):268-277.

(23) Luna CM, Famiglietti A, Absi R, Videla AJ, Nogueira FJ, Fuenzalida AD et al. Community-acquired pneumonia: etiology, epidemiology, and outcome at a teaching hospital in Argentina. Chest 2000; 118(5):1344-1354.

(24) Fine MJ, Smith MA, Carson CA, Mutha SS, Sankey SS, Weissfeld LA et al. Prognosis and outcomes of patients with community-acquired pneumonia. A meta-analysis. JAMA 1996; 275(2):134-141.

(25) Leroy O, Devos P, Guery B, Georges H, Vandenbussche C, Coffi nier C et al. Simplifi ed prediction rule for prognosis of patients with severe community-acquired pneumonia in ICUs. Chest 1999; 116(1):157-165.

(26) Riquelme R, Torres A, El Ebiary M, de la Bellacasa JP, Estruch R, Mensa J et al. Community-acquired pneumonia in the elderly: A multivariate analysis of risk and prognostic factors. Am J Respir Crit Care Med 1996; 154(5):1450-1455.

(27) Roy S, Knox K, Segal S, Griffi ths D, Moore CE, Welsh KI et al. MBL genotype and risk of invasive pneumococcal disease: a case-control study. Lancet 2002; 359(9317):1569-1573.

(28) Waterer GW, Quasney MW, Cantor RM, Wunderink RG. Septic shock and respiratory failure in community-acquired pneumonia have different TNF polymorphism associations. Am J Respir Crit Care Med 2001; 163(7):1599-1604.




(32) Battleman DS, Callahan M, Thaler HT. Rapid antibiotic delivery and appropiate antibiotic selection reduce length of hospital stay of patients with community-acquired pneumonia. Arch Intern Med 2002; 162:682-688.

(33) Meehan TP, Fine MJ, Krumholz HM, Scinto JD, Galusha DH, Mockalis

53

Chapter 3

JT et al. Quality of care, process, and outcomes in elderly patients with pneumonia. JAMA 1997; 278(23):2080-2084.

(34) Gacouin A, Le Tulzo Y, Lavoue S, Camus C, Hoff J, Bassen R et al. Severe pneumonia due to Legionella pneumophila: prognostic factors, impact of delayed appropriate antimicrobial therapy. Intensive Care Med 2002; 28(6):686-691.


(36) Waterer GW, Somes GW, Wunderink RG. Monotherapy may be suboptimal for severe pneumococcal pneumonia. Arch Intern Med 2001; 161:1837-1842.

(37) Gleason PP, Meehan TP, Fine JM, Galusha DH, Fine MJ. Associations between initial antimicrobial therapy and medical outcomes for hospitalized elderly patients with pneumonia. Arch Intern Med 1999; 159:2562-2572.

(38) Stahl JE, Barza M, DesJardin J, Martin R, Eckman MH. Effect of macrolides as part of initial empiric therapy on length of stay in patients hospitalized with community-acquired pnuemonia. Arch Intern Med 1999; 159:2576-2580.


(40) Mufson MA, Stanek RJ. Bacteriemic pneumococcal pneumonia in one American city: a 20-year longitudinal study, 1978-1997. Am J Med 1999; 107(1A):34S-43S.

(41) Dudas V, Hopefl A, Jacobs R, Guglielmo BJ. Antimicrobial selection for hospitalized patients with presumed community-acquired pneumonia: a survey of nonteaching US community hospitals. Ann Pharmacother 2000; 34:446-452.

(42) Macfarlane J. Severe pneumonia and a second antibiotic. Lancet 2002; 359(9313):1170-1172.

(43) Dowell SF. The best treatment for pneumonia. New clues but no defi nitive answers. Arch Intern Med 1999; 159:2511-2512.

(44) Guthrie R. Community-acquired lower respiratory tract infections, etiology and treatment. Chest 2001; 120:2021-2034.

(45) Finch R, Schurmann D, Collins O, Kubin R, McGivern J, Bobbaers H

54

Chapter 1Chapter 3

et al. Randomized controlled trial of sequential intravenous (i.v.) and oral moxifl oxacin compared with sequential i.v. and oral co-amoxiclav with or without clarithromycin in patients with community-acquired pneumonia requiring initial parenteral treatment. Antimicrob Agents Chemother 2002; 46(6):1746-1754.

(46) Davidson R, Cavalcanti R, Brunton J, Bast DJ, de Azavedo JCS, Kibsey P et al. Resistance to levofl oxacin and failure of treatment of pneumococcal pneumonia. N Engl J Med 2002; 346:747-749.


(48) Sisk JE, Moskowitz AJ, Whang W, Lin JD, Fedson DS, McBean AM et al. Cost-effectiveness of vaccination against pneumococcal bacteremia among elderly people. JAMA 1997; 278(16):1333-1339.

(49) Appelbaum PC. Resistance among Streptococcus pneumoniae: Implications for drug selection. Clin Infect Dis 2002; 34(12):1613-1620.

(50) Gross PA. Vaccines for pneumonia and new antiviral therapies. Med Clin North Am 2001; 85(6):1531-1544.

(51) Hak E, Grobbee DE, van Essen GA, Buskens E, Verheij TJ. Pneumococcal vaccination of the elderly: do we need another trial? Arch Intern Med 2000; 160(11):1698-1699.

(52) Ortqvist A, Hedlund J, Burman LA, Elbel E, Hofer M, Leinonen M et al. Randomised trial of 23-valent pneumococcal capsular polysaccharide vaccine in prevention of pneumonia in middle-aged and elderly people. Swedish Pneumococcal Vaccination Study Group. Lancet 1998; 351(9100):399-403.

(53) Nichol KL, Baken L, Wuorenma J, Nelson A. The health and economic benefi ts associated with pneumococcal vaccination of elderly persons with chronic lung disease. Arch Intern Med 1999; 159(20):2437-2442.

(54) Christenson B, Lundbergh P, Hedlund J, Ortqvist A. Effects of a large-scale intervention with infl uenza and 23-valent pneumococcal vaccines in adults aged 65 years or older: a prospective study. Lancet 2001; 357(9261):1008-1011.

(55) Ahmed F, Singleton JA, Franks AL. Clinical practice. Infl uenza vaccination for healthy young adults. N Engl J Med 2001; 345(21):1543-1547.

(56) Ament A, Fedson DS, Christie P. Pneumococcal vaccination and pneumonia: even a low level of clinical effectiveness is highly cost-effective. Clinical Infectious Diseases 2001; 33(2078):2079.

55

Chapter 3

(57) Ament A, Baltussen R, Duru G, Rigaud-Bully C, de Graeve D, Örtquist A et al. Cost-effectiveness of pneumococcal vaccination of older people: a study in 5 western European countries. Clinical Infectious Diseases 2001; 31:444-450.

(58) Hak E, Hoes A, Verheij TJ. Infl uenza vaccination. Who needs them and When? Drugs 2002; 62(17):2413-2420.

(59) Gross PA, Hermogenes AW, Sacks HS, Lau J, Levandowski RA. The effi cacy of infl uenza vaccine in elderly persons. A meta-analysis and review of the literature. Ann Intern Med 1995; 123(7):518-527.

(60) Demicheli V, Jefferson T, Rivetti D, Deeks J. Prevention and early treatment of infl uenza in healthy adults. Vaccine 2000; 18(11-12):957-1030.


56

Chapter 1Chapter 3

4Risk factors for early clinical

failure in patients with

severe community-acquired

pneumoniaSubmitted

M Hoogewerf, JJ Oosterheert, E Hak,

IM Hoepelman, MJM Bonten

58

Chapter 1Chapter 4

Abstract

Background: There is lack of data on predictors of early clinical outcome in patients treated for severe community acquired pneumonia (SCAP). Since careful risk assessments can improve the effi ciency of pneumonia treatment, we conducted a prospective cohort study to assess clinical, biochemical and microbiological predictors for early failure.Methods: We used data from 250 non-immune compromised adult patients hospitalized with SCAP. Clinical failure was defi ned as a respiratory rate >25 in, PaO2 <55mmHg, O2 saturation <90%, haemodynamic instability, body temperature >38oC, deteriorated mental state or inability to take oral medication at day 3. Multivariate logistic regression analysis was performed to assess independent asociations of possible predictors of clinical failure at the day of admission. Results: Of 250 patients, 70 (28%) failed at day 3. Failure was mainly associated with a respiratory rate >25/min (n=34), oxygen saturation >90% (n=28) and fever (n=19). In multivariate analysis, failure was indepentently associated with arterial pH >7,35 mm Hg, altered mental state and infection with gram negative bacteria. Heart failure was negatively associated with failure. The AUC of the ROC curve of the best predictive model was 0.78. Patients who failed showed a non statistically signifi cant higer 28-day mortality rate than non-failures. Conclusions: Routine clinical, biochemical and microbiologicalinformation can be used to determine early clinical failure in SCAP. As early failure may be associated with higher mortality rate, close monitoring of high-risk patients is essential. Conversely, patients at lowrisk for failure might be switched early from IV to oral antibiotics.This could infl uence length of hospital stay.

59

Chapter 4

Introduction

Community-acquired pneumonia (CAP) is a common disease that is associated with serious complications such as respiratory insuffi ciency, sepsis and death. CAP requires hospitalization in 15-50% of patients and 5% to 10% of them are managed in an intensive care unit (ICU) (1-7). Despite advances in antimicrobial therapy, mortality among hospitalized patients is still high ranging from 2% to 30% (8). Clinical response in the fi rst 2-3 days of treatment appears to be predictive of health outcome (9-11). Non-response is associated with increased morbidity and mortality, but once clinical stability has been achieved, substantial clinical deterioration due to pneumonia is rare (12). Risk assessments could help physicians in improving effi ciency of pneumonia treatment: optimal monitoring of high-risk patients may prevent unnecessary complications, ICU admissions or deaths whereas patients at low risk for treatment failure may be switched from parenteral to oral antibiotics early in the treatment process, thereby possibly reducing length of hospital stay leading to more effi cient use of available health care resources. However, at this time, there is lack of data for response or failure to initiated therapy in patients treated for severe CAP. Current defi nitions, like the APACHE II and Pneumonia Severity Index (PSI) score only predict mortality and do not seem to be accurate enough to guide clinical care. Moreover, these algorithms include over 15 variables and are, therefore, impractical in use (13-17). The specifi c aim of the present study was to assess clinical, biochemical and microbiological predictors of early clinical failure in patients hospitalized with severe CAP.

Methods

Patients and setting

This is an analysis of patients included in a multi-center, prospective randomized trial on the cost-effectiveness of early switch from IV to oral therapy of severe CAP. The study was performed in 2 university medical centers and 5 teaching hospitals in the Netherlands between July 2000 and June 2003. The study was approved by the medical

60

Chapter 1Chapter 4

ethics committees of all participating hospitals and all patients provided written informed consent to participate. Adult patients (age 18 or above) admitted with severe CAP were eligible for inclusion into the study. Pneumonia was defi ned as a new or progressive infi ltrate on a chest X-ray plus at least two of the following criteria: cough, sputum production, rectal temperature > 38oC or < 36.1oC, auscultatory fi ndings consistent with pneumonia, leucocytosis (>10.000/mm3, or >15% bands), C-reactive protein > 3 times the upper limit of normal, positive blood culture or positive culture of pleural fl uid (18). Severe pneumonia was defi ned as PSI > 90, (Fine class IV or V) or fulfi lling the ATS-criteria for severe community-acquired pneumonia (18,19). Patients with cystic fi brosis, colonization with Gram negative bacteria due to structural damage to the respiratory tract, life-expectancy less than 1 month due to underlying diseases, infections other than pneumonia needing antibiotic treatment, immunosuppression (neutropenia (<0,5x109 / l) or a CD4 count < 200 / mm3) and those that were admitted directly in ICU were excluded. All patients received intravenous antimicrobial treatment for at least three days.

Microbiological analysis

Sputum samples and blood samples were collected, cultured and evaluated following standard procedures. Micro-organisms cultured in blood or sputum were recorded. In addition, Binax NOW-tests (Binax inc., Portland, Maine) were used to detect urinary antigen for Legionella pneumophila and S. pneumoniae. Acute and convalescent serology samples were collected and evaluated for Mycoplasma pneumoniae, L. pneumophila and Chlamydia pneumonia. Any non-contaminating micro-organism cultured from a blood or sputum sample or detected by urinary antigen testing was considered a cause for the episode of pneumonia. For Mycoplasma pneumoniae, a fourfold or greater increase in titer in paired sera or a single titer of greater than or equal to 1:40 was considered indicative of infection (Immune fl uorescence agglutination, Serodia-MycoII ®, Fujirebio, inc., (20). For Legionella pneumophila, a fourfold increase in the antibody titer to 1:128 or greater, or single titers of 1:256 or more were considered suggestive of Legionella pneumonia (21). For Chlamydia pneumoniae, detection of IgM above established values, seroconversion of IgG between

61

Chapter 4

acute and convalescence samples, high amounts of IgG in single titers or a combination of these factors were considered serological evidence of infection, according to the manufacturers instructions. (ELISA, Savyon Diagnostics Ltd)

Defi nitions

Early clinical failure

Early clinical failure was assessed after 3 days and was defi ned as death, need for mechanical ventilation, respiratory rate > 25/minute, oxygen-saturation < 90 %, P02 < 55 mm Hg, hemodynamic instability, altered mental state, or fever (<1 °C decline in temperature if >38.5 °C on admission) (22). If none of these features were present, patients were considered responders. Mental state change was defi ned as an acute altering in mental state of the patient observed by family or the treating physician.

Discordant therapy

Discordant therapy was evaluated for cases with a causative micro-organism identifi ed. Discordant therapy was defi ned as treatment with antibiotics for which identifi ed pathogens were not susceptible, e.g. β-lactam monotherapy for atypical pathogens as C. pneumoniae, M. pneumoniae and L. pneumophila

Clinical characteristics and course

On admission, clinical history and physical examination were performed. Demographic criteria and medical history were recorded. PSI and APACHE II were scored (13,15), arterial blood gas analysis, routine chemical and hematological laboratory tests were taken and a chest X-ray was performed. Antibiotic treatment as instituted by the treating physician was recorded. On the third day of hospitalization, response to treatment was evaluated by the principal investigator and dedicated study research nurses. Early failure was assessed using vital parameters, arterial blood gas analysis, routine laboratory tests and physical examination. During hospital stay, in-hospital clinical

62

Chapter 1Chapter 4

data such as temperature, blood pressure, O2-saturation, heart rate and respiratory rate were measured daily. Patients were followed for a maximum of 28 days.

Characteristic Non-failureN=180

FailureN=80 OR 95%CI P

Age 67.9 ± 16.0 69.15 ± 16.8 1.01 1.00-1.02 0.57

Fine score 108.3 ± 23.9 123.9 ± 26.2 1.03 1.01-1.04 <0.001

Apache score 13.3 ± 4.4 14.8 ± 5.1 1.07 1.01-1.13 0.03

Glasgow Coma Score 14.7 ± 1.1 14.4 ± 1.1 0.80 0.62-1.02 0.07

Female Sex 57 (32%) 23 (29%) 0.87 0.49-1.55 0.69

Nursing Home 5 (3%) 4 (5%) 1.80 0.48-7.05 0.37

Clinical Features

PO2 66.9 ± 19.5 69.3 ± 27.9 1.01 0.99-1.02 0.43

pH 7.45 ± 0.06 7.43 ± 0.08 0.03 0.00-0.18 0.006

Systolic BP 134.3 ± 24.5 138.5 ± 30.8 1.01 1.00-1.02 0.24

Temperature 38.6 ± 1.2 38.3 ± 1.2 0.85 0.68-1.06 0.15

Respiratory Rate 26.3 ± 8.1 28.1 ± 9.8 1.02 0.99-1.06 0.16

Heart rate 105.5 ± 22.0 106.7 ± 24.5 1.00 0.99-1.01 0.71

Temp. <35 or > 40 ºC 24 (13%) 11 (14%) 1.00 0.95-1.06 0.93

Systolic RR <90 mm Hg 4 (2%) 3 (4%) 1.03 0.95-1.11 0.48

Heart rate >125/min 39 (22%) 21 (26%) 1.03 0.97-1.09 0.41

Respiratory rate >30/min 55 (31%) 31 (39%) 1.03 0.99-1.05 0.20

Art. PH < 7.35 mm Hg 7 (4%) 12 (15%) 4.36 1.65-11.55 0.003

Art. P02 <60 mm Hg 56 (31%) 35 (44%) 1.72 1.00-2.96 0.05

Pleural effusion X-ray 30 (17%) 17 (21%) 1.03 0.96-1.10 0.38

Mental State Change 39 (22%) 36 (45%) 2.96 1.68-5.21 <0.001

Medical History

Chronic Heart Failure 27 (15%) 4 (5%) 0.30 0.10-0.88 0.03

Neoplasm 44 (24%) 18 (23%) 1.00 0.98-1.02 0.73

CVA 13 (7%) 7 (9%) 1.02 0.93-1.12 0.67

Kidney Disease 17 (9%) 6 (8%) 0.98 0.89-1.08 0.61

Table 1

Patient characteristics

63

Chapter 4

Statistical analysis

To identify individual risk factors for early clinical failure, descriptive statistics as proportions and means (SD) were calculated to describe baseline characteristics in the two outcome groups (early responders and early failure) using SPSS for Windows, version 11.5 (SPSS inc.,Chicago, Illinois,USA). The construction of the prognostic model started with univariate assessment of the associations of each characteristic and early failure by estimation of odds ratio’s (OR) and their corresponding 95% confi dence interval (95% CI) for the total study population. In the next stage multivariate logistic regression modelling was applied to select those variables that were associated with early failure with a p-value <0.10 as criterion for entry (23). Forward selection was additionally performed to verify whether any previously deleted potentially relevant characteristic was incorrectly eliminated from the model. For each patient we calculated the individual probability of the outcome from the fi nal model (predicted probability) (24).

Etiology Non-Failure N=180

FailuresN=80 OR 95%CI P

S. pneumoniae 49 (27%) 20 (25%) 0.89 0.49-1.63 0.71

H. infl uenzae 11 (6%) 3 (4%) 0.60 0.16-2.21 0.44

Gram-negative bacteriaE. ColiEnterobacter spp.Klebsiella spp.Pseudomonas spp.Other

17 (9%)32506

15 (19%) 62125

2.21 1.04-4.69 0.04

Chlamydia 10 (6%) 8 (10%) 1.89 0.72-4.98 0.20

Legionella 7 (4%) 3 (4%) 0.96 0.24-3.82 0.96

Mycoplasma 4 (2%) 4 (5%) 2.32 0.56-9.50 0.24

Multiple pathogens 20 (11%) 9 (11%) 1.01 0.44-2.34 0.97

Other pathogens 12 (7%) 5 (6%) 0.88 0.32-2.44 0.81

Unknown 90 (50%) 35 (44%) 0.85 0.49-1.50 0.56

Table 2

Causative micro-organisms

64

Chapter 1Chapter 4

Therapy Non-failuresN=180

FailuresN=80 OR 95%CI P

β-lactam mono 109 (61%) 44 (55%) 0.80 0.47-1.36 0.40

Macrolide mono 3 (1%) 1 (1%) 0.75 0.08-7.29 0.80

Cephalosporin mono 33 (18%) 13 (16%) 0.86 0.43-1.75 0.69

Other 15 (8%) 3 (4%) 0.43 0.12-1.52 0.19

β-lactam / macrolide combi 12 (7%) 12 (15%) 2.47 1.06-5.77 0.04

Cephalosporin / macrolide combi 6 (3%) 5 (6%) 1.93 0.57-6.53 0.29

Macrolide / other combi 3 (2%) 2 (3%) 1.51 0.25-9.23 0.65

Discordant treatment* 17(19%) 11 (24%) 1.14 0.48-2.73 0.76

Table 3

Initial therapy

* In 135 patients a micro-organism was detected.

Model evaluation

The reliability of the multivariate logistic regression model was determined by the Hosmer-Lemeshow goodness-of-fi t statistic (23). The area under the receiver-operating-curve (ROC) was used to assess the model’s discriminative ability (24,25). The area under the ROC can be explained as the probability that the logistic regression model will assign a higher probability of the outcome to a randomly chosen patient with an outcome than to a randomly chosen patient without outcome. An area under the curve (AUC) estimate of 0.5 indicates no discrimination whereas an estimate of 1.0 indicates perfect discrimination. Models with AUC values between 0.70 and 0.79 are generally considered as having moderate and ≥0.80 as good discriminative properties.

Development and applicability of the prediction rule The regression coeffi cients of the derived multivariate model were used to construct the prediction rule. The absence or presence of specifi c characteristics was coded as 0 (absent) or 1 (present). To simplify interpretation, we rounded the regression coeffi cients to develop an overall score value. The scores for individual prognostic variables were added to form the prognostic score for clinical failure.

65

Chapter 4

Characteristic Odds ratio 95% CI P-value

Gram negative pneumonia 3.15 1.39-7.13 0.01

Altered mental status 3.79 2.03-7.08 <0.00

Arterial pH < 7,35 4.46 1.57-12.70 0.01

Heart failure 0.32 0.10-1.03 0.06

Arterial pO2< 60 1.77 0.97-3.21 0.06

Table 5

Multivariate analysis

Altered mental status +2

Gram negative infection +2

PH < 7.35 +3

PO2 < 60 +2

Heart-failure -1

Minimum score: -2 points.

Maximum score: +8 points

Table 6

Prediction rule for early clinical failure in hospitalized CAP

Non-failure (%) Failure (%) Total (patients)

-2 – 1 points 82 18 159

2-4 points 53 47 91

5-8 points 20 80 10

Table 7

cut-off points of prediction rule

Results

From July 2000 to June 2003, 273 patients with severe CAP were included. Thirteen patients were subsequently excluded from analysis: 6 patients withdrew informed consent, 6 patients had co-infections outside the respiratory tract needing antibiotic treatment, and in 1 patient data on early failure were incomplete, leaving 260 patients for analysis.

Of these 260 patients, 180 (69%) were responders and 80 (31%) had early clinical failure. Reasons for failure were death in 5 of these 80 (6%) patients and need of mechanical ventilation before day in another 5 patients (6%). Other reasons for failure at day 3 were

66

Chapter 1Chapter 4

respiratory rate > 25/minute in 34 (42%), oxygen-saturation < 90% in 28 (35%), fever in 19 (24%) altered mental state in 20 (25%) and hemodynamically instability in 4 (5 %) patients. One feature was present in 6 (57%) patients, multiple features were present 34 (43%) patients.

In univariate analysis total PSI score, APACHE II score, Glasgow Coma Score, presence of mental state change, arterial pH < 7.35 and arterial pO2 < 60 mm Hg were signifi cantly more often present in patients with early failure. A history of chronic heart failure was more often found in early responders (table 1). Causative micro-organisms were identifi ed in 134 (52 %) patients, in 90 (50 %) responders and in 45 (56%) patients with early failure (table 2). Most frequently isolated micro-organism was S. pneumoniae in both groups. Gram-negative micro-organisms were more often identifi ed in patients with early failure. (OR= 2.21, 95% CI 1,04 – 4.69) Most frequently instituted initial therapy was beta-lactam monotherapy (table 3). Initial therapy with macrolide / beta-lactam combinations was associated with early clinical failure in univariate analysis Twenty-eight patients received discordant therapy, but this was not associated with early clinical failure (OR=1.14, 95% CI 0.48-2.73). Reasons for discordant treatment were serological evidence of infection with M. pneumoniae, C. pneumoniae or L. pneumophila in patients treated with macrolides or quinolones.In multivariate analysis the most simple predictive model with highest predictive value included altered mental state, Gram negative infection, arterial pH < 7.35, pO2 < 60 and absence of history of heart failure as independent predictors of early clinical failure (area under the receiver operating characteristics curve: 0.73, 95% CI: 0.66 –0.80).

A prediction rule was derived from the multivariate model in which a score was assigned to the presence of each variable. The predicted

probability of outcome was determined as 1/(1 + e-LP), where the linear predictor (LP) = -1.64+ (1.15 × Gram negative infection)+ (1.33 × altered mental status) + (1.50×pH<7.35) + (-1.13 ×heart failure) + (0.57 x Arterial PO2<60 mm Hg). A prognostic score for each patient (minimum -2 to maximum 8 points), refl ecting the

67

Chapter 4

probability of early failure, was calculated by adding the scores of relevant characteristics. Patients with a cut-off point of < 2 points had a chance for early clinical failure of 18 %, whereas patients with a cut-off point of > 5 points had an 80 % chance for failure. (table 6 and 7)

In the follow-up period, patients with early clinical failure had a signifi cant higher chance of all-cause death and had longer lengths of hospital stay: 9 patients with early clinical failure died (12%, the 5 patients who died before day 3 not included), whereas 8 (4,4%) early clinical responders died (Odds ratio: 2.93, 95% CI: 1.09-7.92). Length of hospital stay of patients with early clinical failure was 13.4 ± 5.3 days as compared to 9.6 ± 4.7 days in early responders. (mean difference: 3.8 days, 95% CI: 2.6 – 5.4 days)

Discussion

Clinical, biochemical and microbiological information that are usually routinely assessed in patients with CAP can be used to determine early clinical failure in patients hospitalized with severe CAP. Early clinical failure was associated with increased mortality and longer stay in hospital. Close monitoring of patients at high risk for treatment failure may prevent unnecessary deaths, ICU admissions and associated costs.The strengths of the current study are the prospectively collected data, strict criteria for inclusion and the focus on patients with a high risk of early clinical failure. In addition, early failure was evaluated with previous defi ned and generally accepted criteria. Until now, no studies have been performed that analyzed risk factors for early clinical failure in patients with severe CAP.

Gram negative infection, arterial pH < 7.35 mm Hg, arterial p02 < 60 mm Hg, altered mental state on admission and absence of a history of chronic heart failure were independent predictors of early clinical failure in patients treated for severe CAP. Gram negative infection was also found by other investigators as an important cause for either severe CAP or early clinical failure. Patients with aspiration, previous

68

Chapter 1Chapter 4

hospitalization or pulmonary co-morbidity are especially at risk for Gram negative infections (26-28). Hypoxemia, acidosis and confusion may all indicate tissue hypoperfusion due to oxygen defi ciency and is associated with high mortality in patients with CAP (29). As hypoxemia and acidosis can only be determined with proper arterial blood-gas analysis, the importance of performing this diagnostic procedure at admission should be stressed out.Remarkably, a history of chronic heart failure was inversely related with early clinical failure. Although it is possible that some patients with a history of heart failure were admitted because of congestive heart failure, which is sometimes diffi cult to differentiate from pneumonia on a chest X-ray, all patients had evidence of new or progressive infi ltrates on chest radiograph, fulfi lled the clinical criteria for pneumonia and had a PSI score of at least 90, or they fulfi lled the ATS criteria for severe CAP.Although heart failure has been associated with adverse clinical outcomes, in patients with CAP it has also been associated with a high risk for viral pneumonia (28,30). It is, therefore, possible that viral respiratory infections with a less severe course were more prevalent in patients with heart failure. Indeed, in 68% of patients with heart failure no microbial cause of infection was detected, as compared to 45% of patients without heart failure (p=0.02).

In univariate analysis the combination of beta-lactams and macrolides were associated with early failure. Although in the Netherlands most patients hospitalized with severe CAP on regular wards are treated with β-lactam antibiotics alone, combination therapy is sometimes used for more severe cases of SCAP, as judged by the treating physician. It is therefore possible that the patients with high risk of clincal failure recieved combination therapy.There was no statistically signifi cant difference in the prevalence of discordant treatment between failing and non-failing patients. In the majority of patients with discordant treatment an atypical pathogen, mostly Chlamydia pneumoniae or Mycoplasma pneumoniae, was detected, which was not treated with a macrolide or fl uoroquinolone. Most of these pathogens were C. pneumoniae or M. pneumoniae, which usually cause infections with a mild course. Moreover, diagnosing

69

Chapter 4

atypical infections by means of serological investigations is diffi cult, especially in case of C. pneumoniae (31). The possible infl uence of penicillin resistance of S. pneumoniae on early clinical failure could not be investigated in the present study, as true pneumococcal resistance to penicillins in the Netherlands is < 1%.

Our study also has several limitations. While explicit defi nitions of clinical response are lacking, criteria for clinical stability usually include normalization of heart rate, systolic blood pressure, respiratory rate, temperature, oxygenation status and mental status (12,32). We, therefore, used these criteria to defi ne clinical failure. Evidently, the use of other criteria may lead to other outcomes. This is illustrated by the lower rates of failures in previous studies on this topic, in which other criteria for early failure were used, such as lack of response or worsening of clinical and/or radiologic status at 48 to 72 hours, requiring changes in therapy and/or performance of invasive procedures for diagnostic and therapeutic purposes (26,33). Furthermore, the higher failure rate in our study can also be explained by limiting the inclusion to patients with severe CAP.

In patients who fail to respond to initial therapy, incorrect diagnosis, inadequate host-related responses, drug-related problems or pathogen-related problems as antibiotic resistant pathogens, should be considered. Which feature is most important for early clinical failure has not yet been determined. In our study, strict diagnostic criteria for CAP were used for inclusion, discordant therapy appeared not to be associated with early clinical failure and all isolates of S. pneumoniae were susceptible for β-lactam antibiotics. Nevertheless, still 31% of the patients fulfi lled the criteria for early clinical failure. Therefore, inadequate host response is probably the most important factor for early treatment failure in our study cohort, possibly due to genetic predisposition as recently suggested. (34-36).

In summary, clinical, biochemical and microbiological information that are usually routinely assessed in patients with CAP can be used to determine early clinical failure in patients hospitalized with severe CAP. Close monitoring of patients at high risk for treatment failure may prevent unnecessary deaths, ICU admissions and associated

70

Chapter 1Chapter 4

costs. Early identifi cation of patients at low risk for early failure may assist physicians in scheduling treatment strategies, for example to switch to oral antibiotics early in the treatment process. The prognostic variables identifi ed in our study are easy to assess and might be helpful in these treatment decisions. The prediction rule derived from our data needs validation in other cohorts of severe CAP.

71

Chapter 4

Reference List

(1) Marrie TJ. Epidemiology of community-acquired pneumonia in the elderly. Semin Respir Inf 1990; 5:260-8.

(2) Almirall J, Morato I, Riera F et al. Incidence of community-acquired pneumonia and Chlamydia pneumonia infection: a prospective multicenter study. Eur Respir J 1993;6:14-8.

(3) Lave JR, Fine MJ, Sankey SS et al. Hospitalized pneumonia. Outcomes, treatment patterns, and costs in urban and rural areas. J Gen Intern Med 1996;11:415-21.

(4) Woodhead MA, Macfarlane JT, McCracken JS et al. Prospective study of the aetiology and outcome of pneumonia in the community. Lancet 1987;i:671-4.

(5) Guest JF, Morris A. Community-acquired pneumonia: the annual cost to the National Health Service in the United Kingdom. Eur Respir J 1997;10:1530-4.

(6) British Thoracic Society Research Commity and Public Health Laboratory Service: The aetiology, management and outcome of severe community-acquired pneumonia on the intensive care unit. Respir Med 1992;86:7-13.

(7) Torres A, Serra-Batlles J et al. Severe community-acquired pneumonia: epidemiology and prognostic factors. Am Rev Respir Dis 1991;144:312-8.

(8) Fine MJ, Smith MA, Carson CA, Mutha SS, Sankey SS, Weissfeld LA et al. Prognosis and outcomes of patients with community-acquired pneumonia. A meta-analysis. JAMA 1996; 275(2):134-141.

(9) Ramirez JA, Vargas S, Ritter GW, Brier ME. Early switch from intravenous to oral antibiotics and early hospital discharge: a prospective observational study of 200 consecutive patients with community-acquired pneumonia. Arch Int Med 1999;159:2449-2454.

(10) Ramirez JA, Bordon J. Early switch from intravenous to oral antibiotics in hospitalized patients with bacteremic Streptococcus pneumoniae community-acquired pneumonia. Arch Int Med 2001;161:848-850.

(11) Ramirez JA. Switch therapy in adult patients with pneumonia. Clin Pulm Med 1995;2:327-333.

(12) Halm EA, Fine MJ, Marrie TJ, Coley CM, Kapoor WN, Obrosky DS et al. Time to clinical stability in patients hospitalized with community-

72

Chapter 1Chapter 4

acquired pneumonia: implications for practice guidelines. JAMA 1998; 279(18):1452-1457.


(14) Lim WS, van der Eerden MM, Laing R, Boersma WG, Karalus N, Town GI et al. Defi ning community acquired pneumonia severity on presentation to hospital: an international derivation and validation study. Thorax 2003; 58(5):377-382.

(15) Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classifi cation system. Crit Care Med 1985; 13(10):818-829.

(16) Angus DC, Marrie TJ, Obrosky DS, Clermont G, Dremsizov TT, Coley C et al. Severe Community-acquired Pneumonia: Use of Intensive Care Services and Evaluation of American and British Thoracic Society Diagnostic Criteria. Am J Respir Crit Care Med 2002; 166(5):717-723.

(17) Oosterheert JJ, Bonten MJ, Hak E, Schneider MM, Hoepelman AI. Severe community-acquired pneumonia: what’s in a name? Curr Opin Infect Dis 2003; 16(2):153-159.


(19) Offi cial Statement of the American Thoracic Society. Guidelines for the management of adults with community-acquired pneumonia. Am J Resp Crit Care Med 2001; 163:1730-1754.

(20) Weingarten SR, Riedinger MS, Varis G et al. Identifi cation of low-risk hospitalized patients with pneumonia. Implications for early conversion to oral antimicrobial therapy. Chest 1994;105(4):1109-15.


(22) Stout JE, Yu VL. Legionellosis. N Engl J Med 1997; 337(10):682-687.

(23) Hosmer DW, Lemeshow S. Applied logistic regression. New York: John Wiley & Sons, 1998:135-75.

(24) Hanley JA, Mc Neill BJ. The meaning and use of the area under

73

Chapter 4

the receiver operating characteristic curve (ROC. Radiology 1982; 143:29-36.

(25) Hak E, Wei F, Nordin J, Mullooly J. Development and validation of a clinical prediction rule for hospitalisation due to pneumonia or infl uenza or death during infl uenza epidemics among community-dwelling elderly persons. JID 2004; 189:450-8.

(26) Roson B, Carratala J, Fernandez-Sabe N, Tubau F, Manresa F, Gudiol F. Causes and factors associated with early failure in hospitalized patients with community-acquired pneumonia. Arch Intern Med 2004; 164(5):502-508.

(27) Arancibia F, Bauer TT, Ewig S, Mensa J, Gonzalez J, Niederman MS et al. Community-acquired pneumonia due to gram-negative bacteria and pseudomonas aeruginosa: incidence, risk, and prognosis. Arch Intern Med 2002; 162(16):1849-1858.

(28) Ruiz M, Ewig S, Torres A, Arancibia F, Marco F, Mensa J et al. Severe community-acquired pneumonia. Risk factors and follow-up epidemiology. Am J Respir Crit Care Med 1999; 160(3):923-929.

(29) Mortensen EM, Coley CM, Singer DE, Marrie TJ, Obrosky DS, Kapoor WN et al. Causes of death for patients with community-acquired pneumonia: results from the Pneumonia Patient Outcomes Research Team cohort study. Arch Intern Med 2002; 162(9):1059-1064.

(30) de Roux A, Marcos MA, Garcia E, Mensa J, Ewig S, Lode H et al. Viral community-acquired pneumonia in nonimmunocompromised adults. Chest 2004; 125(4):1343-1351.

(31) Tuuminen T, Palomaki P, Paavonen J. The use of serologic tests for the diagnosis of chlamydial infections. J Microbial Methods 2000; 45: 265-279.


(33) Ortqvist A, Kalin M, Lejdeborn L, Lundberg B. Diagnostic fi beroptic bronchoscopy and protected brush culture in patients with community-acquired pneumonia. Chest 1990; 97(3):576-582.

(34) Roy S, Knox K, Segal S et al. MBL genotype and risk of invasive pneumococcal disease: a case-control study. Lancet 2002;359:1569-1573.

(35) Waterer GW, Quasney MW, Cantor RM, Wunderink RG. Septic shock and respiratory failure in community-acquired pneumonia have different TNF polymorphism associations. Am J Respir Care Med 2001; 163(7):1599-1604.

74

Chapter 1Chapter 4

(36) Waterer GW, ElBahlawan L, Quasney MW, Zhang Q, Kessler LA, Wunderink RG. Heat shock protein 70-2+1267 AA homozygotes have an increased risk of septic shock in adults with community-acquired pneumonia. Crit Care Med 2003; 31(5):1367-1372.

5How good is the evidence for

the recommended empiric

antimicrobial treatment of

patients hospitalised because

of community-acquired

pneumonia?

J Antimicrob Chemother. 2003; 52 (4): 555-63

JJ Oosterheert, MJM Bonten, E Hak,MME Schneider, IM Hoepelman

76

Chapter 1Chapter 5

Abstract

Background: For years, monotherapy with a β-lactam antibiotic (Penicillin, Amoxicillin or 2nd generation cephalosporin) was recommended as empirical therapy for patients with community-acquired pneumonia (CAP). A combination of a β-lactam and a macrolide antibiotic was only recommended for patients with severe CAP needing intensive care treatment or when atypical pathogens i.e. Legionella pneumophila, Mycoplasma pneumoniae and Chlamydia pneumoniae, were strongly suspected. New guidelines however, recommend a combination of a β-lactam antibiotic plus a macrolide or monotherapy with a fl uoroquinolone for all patients hospitalised with CAP. We evaluated if treatment with a β-lactam plus macrolide or quinolone monotherapy is truly superior to β-lactam treatment alone. Methods: We systematically reviewed available studies, retrieved from Medline and by hand-searching reference lists from recent reviews and guidelines on the effectiveness of recommended empiric antimicrobial treatment of patients hospitalised because of CAP. Results: Eight relevant studies were selected. In 6 studies signifi cant reductions in mortality were found, in one study a reduction in hospital length of stay was found and in one study no benefi cial effects could be demonstrated for treatment regimens with fl uoroquinolone monotherapy or combinations of β-lactams and macrolides. The benefi cial value of macrolides or fl uoroquinolones might be caused by a large and mainly unrecognised role of atypical pathogens in the aetiology of CAP, anti-infl ammatory effects of macrolides or resistance to β-lactams of the most important pathogens. However, the studies supporting the recommended treatment regimen were designed as non-experimental cohort studies. As a consequence, the results may have been infl uenced by confounding by indication. In addition, the outcomes showed several inconsistencies. Conclusion: A randomised controlled trial is warranted to circumvent the methodological fl aws in the designs of the currently available studies. Since the addition of macrolides or treatment with fl uoroquinolones may lead to enhanced antibiotic resistance, increased side effects and health care-related costs, such a fundamental change in the treatment of CAP should be based on valid data.

77

Chapter 5

IntroductionInitial therapy for patients with community-acquired pneumonia (CAP) is mostly empiric. Previous British and North-American guidelines recommended initial treatment with benzyl penicillin, amoxicillin or another β-lactam antibiotic for uncomplicated pneumonia. The addition of a macrolide to the initial management was not recommended unless there was either a strong suspicion of atypical pneumonia caused by Legionella pneumophila, Mycoplasma pneumoniae, or Chlamydia pneumoniae or severe pneumonia requiring admission to an intensive care unit (ICU). (1-3)Recent North-American publications, however, have suggested that combination therapy consisting of a β-lactam antibiotic plus a macrolide or monotherapy with one of the newer fl uoroquinolones in the initial management of all patients hospitalised with CAP who do not require admission to an ICU reduces both mortality rates and length of hospitalisation. Based on these fi ndings, the British and American Thoracic Societies and the Infectious Diseases Society of America (IDSA) have revised their guidelines for the treatment for CAP and now recommend either treatment with a β-lactam antibiotic plus a macrolide or treatment with a fl uoroquinolone for all patients hospitalised because of CAP. (4;5) However, such a strategy could lead to an increase in costs, an increase in adverse events related to antibiotic use and an increase in the induction of antibiotic resistance. Here we will discuss whether there is suffi cient scientifi c evidence to justify such a fundamental change in the empirical treatment of patients admitted with CAP. We review recent peer reviewed reports to determine whether, compared to monotherapy with β-lactams, initial therapy with a β-lactam plus a macrolide or quinolone monotherapy reduces mortality or length of stay among adult patients hospitalised with CAP. In addition, we explore possible etiological reasons that might explain the results.

Material and methods

We performed a systematic Medline search in which the following MESH terms were used: ‘community-acquired pneumonia’ or ‘pneumococcal pneumonia’ in combination with either ‘antibiotics, empiric’, ‘antibiotics, combinations’, ‘antibiotics macrolides’, or

78

Chapter 1Chapter 5

‘antibiotics, quinolones’ combined with ‘mortality’, ‘length of stay’ or ‘treatment outcome’. Additionally, we hand-searched references of relevant articles and guidelines.We only included original peer reviewed articles published in the period between January 1997 and April 2003 regarding adult patients hospitalised because of community-acquired pneumonia and dealing with the question whether treatment with β-lactam plus macrolide or quinolone monotherapy is associated with a reduction in mortality or length of stay compared to β-lactam treatment alone. We are unaware of any relevant studies before January 1997. We excluded articles in which children were the subject of study, articles in which no clear comparisons were made between fl uoroquinolones or β-lactams plus macrolide and β-lactam treatment alone, articles of CAP in HIV-infected patients, nosocomial pneumonia or CAP managed in an out-patient setting. In addition, we explored possible etiological reasons that may explain the results and assigned quality-levels as defi ned by Sackett to the reviewed studies.(6)

Results

The initial literature-search yielded 135 articles. One-hundred and sixteen studies were excluded based on their titles and abstracts and another 12 were excluded after the original articles had been reviewed. The reasons for exclusion are given in fi gure 1. Eventually, we found eight studies dealing with the question whether atypical coverage by means of combinations of macrolides and β-lactams or fl uoroquinolones in the initial treatment of patients requiring hospitalisation because of CAP is associated with better outcomes. (7-14) These studies are summarized in table 1. Six studies found a signifi cant reduction in all-cause mortality for patients treated with combinations of β-lactams plus macrolides or with monotherapy of a fl uoroquinolone (7;14) In one study no mortality-reduction could be demonstrated, (8) and in another study, including 76 patients, a signifi cant reduction in hospital length of stay (LOS) was found for the 12 patients treated with a macrolide. (13) In addition, Dudas et al. found a reduction in LOS for patients treated with combination therapy in multivariate analysis. (9) Reductions in LOS could not be

79

Chapter 5

demonstrated in two other studies, (10;13) and were not analysed in the remaining studies. (11;12;14;15) The role of atypical pathogens in the aetiology of CAP was not addressed in any study and the susceptibility of causative micro-organisms to antimicrobial therapy was analysed in only two studies. (7;14) Possible disadvantages resulting from increased macrolide or fl uoroquinolones use, such as an increase in costs, an increase in side effects as a result of antibiotic usage, and an increase in the development of resistance were not investigated in any study.

Initial result 135

Exclusion based on abstract and title■ Study did not specifi cally compare fl uoroquinolone

monotherapy or combinations of macrolides and beta-lactams to beta lactam monotherapy 34

■ Study did not include antibiotics 22■ Study included children 1■ Study included HIV-infected patients 1■ Study was review, letter or case-report 40■ Study only addressed severe CAP 3■ Study only included out-patients 2■ Study investigated other diagnosis than CAP 13

Original articles reviewed 19

Exclusion after original article had been reviewed■ Study did not specifi cally compare fl uoroquinolone

monotherapy or combinations of macrolides and beta-lactams to beta lactam monotherapy 6

■ Study also included nosocomial pneumonia 1■ Study only performed cost-analysis 1■ Study also included out-patients 1■ Study only included out-patients 2

Articles reviewed in this paper 8

Figure 1

Search results and reasons for exclusion

80

Chapter 1Chapter 5

Design and patient selection

Seven studies (7;8;10-13) were designed as retrospective cohort studies. Only Dudas et al. studied patients prospectively. In this study, the diagnosis of CAP was based on a chest X-ray, made within 72 hours of hospitalisation, but an infi ltrate was not required for a defi nite diagnosis of pneumonia.(9) Furthermore, accurate patient inclusion can also be questioned in the studies by Gleason et al. and Houck et al. (10;11) since administrative databases of health-care insurance companies were used to identify patients with possible pneumonia. Three other studies used culture results to select patients. (12;14;15)

Validation of available studies

We assigned evidence levels to the reviewed studies, based on the defi nitions of Sackett. (6) (table 2). Based on these evidence levels, a graded recommendation can be formulated for the treatment regimen under study. Considering the non-randomised design, the results of the reviewed studies should be categorized as level III evidence. Since newly formulated recommendations for the initial therapy of CAP are supported only by level III studies, they should be graded as level D recommendations.

Discussion

Patient selection bias

A potential drawback of retrospective studies is that the reported benefi cial effects of empirical coverage of atypical pathogens could be the result of confounding by indication that may be present when patients are selected for certain prescribed antibiotic regimens based on the severity of their clinical condition. (16;17) Except for Legionella pneumophila, atypical pneumonia tends to follow a less severe course and is more often found in younger persons than bacterial pneumonia. Therefore, it is possible that physicians considered atypical microorganisms as causative agents and added macrolides or fl uoroquinolones to the empirical treatment, in case

81

Chapter 5

of milder presentations or in relatively younger persons. Clues for this form of confounding bias are found in the study of Gleason et al. (10) Antibiotic regimens containing macrolides were more often prescribed in patients with a low risk of mortality. Regimens containing a β-lactam/β-lactamase inhibitor without a macrolide were more often prescribed in high-risk patients. In contrast, in Burgess’ study, combination therapy was more often prescribed in high-risk patients. (8) In Gleason’s study, (10) combinations of β-lactam / β-lactamase inhibitors plus a macrolide were associated with high risk of mortality whereas combinations of cephalosporins and macrolides were associated with a better survival. Although statistical corrections were made, these cannot correct entirely for this form of confounding bias. (16) The outcomes suggest that cephalosporins are more effective than penicillins, but another observational study of 460 patients with pneumococcal pneumonia suggested the contrary.(18) Furthermore, in Martinez’ study, a benefi cial effect of macrolides was not demonstrated in univariate analysis but was revealed only after using a stepwise logistic regression method including all putative prognostic factors. Patients receiving macrolide treatment were more often in shock and more often admitted to an ICU. However, other adverse prognostic features, such as ultimately or rapidly underlying fatal disease, having received cancer chemotherapy, comorbidity, and nosocomial infection were more prevalent in patients receiving treatment without a macrolide. (7)

Other reports did not confi rm the fi ndings of these retrospective analyses: in a prospective randomised study a newer fl uoroquinolone (moxifl oxacin) was more effective than amoxicillin / clavulanic acid with or without a macrolide. However, a signifi cant survival benefi t in these 628 patients was not found. There were no data on how many and which patients received macrolides.(19) Besides, in a recent Scandinavian study, initial therapy with narrow spectrum antibiotics was not associated with worse outcomes.(20) The question whether treatment with monotherapy with a β-lactam agent is inferior to the combination with a macrolide or monotherapy with a quinolone remains, therefore, unanswered.

82

Chapter 1Chapter 5Tab

le 1

Rec

ent

liter

ature

on a

dditio

nal

val

ue

of m

acro

lides

or

fl uoro

quin

olo

nes

in t

he

initia

l m

anag

emen

t of CAP

Au

tho

rN

um

ber

of

pati

en

tsP

ati

en

t ca

teg

ory

Stu

dy d

esi

gn

En

dp

oin

tsR

esu

lts

Co

mm

en

t

Mufs

on

(1

999)

(12)

423

Bac

tera

emia

S.

pneu

monia

eRet

rosp

ective

non-

exper

imen

tal co

hort

st

udy

30-d

ay

mort

ality

Mort

ality

*Reg

imen

s w

ith m

acro

lide

vs.

regim

ens

without

mac

rolid

e 1983-1

987 O

R:

0,5

6 (

95%

CI:

0,0

6 -

5,5

7)

1988-1

992 O

R:

0,2

7 (

95%

CI:

0,0

3 -

2,3

6)

1993-1

997 O

R:

0,1

6 (

95%

CI:

0,0

3 -

0,8

6)

• Aet

iolo

gy

No d

ata

on a

typic

al p

athogen

s•

Res

ista

nce

No d

ata

on s

usc

eptibili

ty o

f S.p

neu

monia

e•

Sev

erity

of CAP

rela

ted t

o c

hoic

e of

antibio

tics

No d

ata

• Le

ngth

of st

ayN

o d

ata

• Le

vel of ev

iden

ce:

III

Gle

ason

(1999)

(10)

12.9

45

65 y

ears

of ag

e or

old

er,

pat

ients

sel

ecte

d

from

dat

abas

es o

f hea

lthca

re fi n

ance

org

aniz

atio

n

Ret

rosp

ective

non-

exper

imen

tal co

hort

st

udy

30-d

ay

mort

ality

Mort

ality

com

par

ed t

o 3rd

gen

erat

ion

cephal

osp

orin

2nd g

ener

atio

n c

ephal

osp

orin +

mac

rolid

e O

R:

0,7

1 (

95%

CI:

0,5

2-0

,96)

3rd

gen

erat

ion c

ephal

osp

orin +

mac

rolid

eO

R:

0,7

4 (

95%

CI:

0,6

0-0

,92)

Quin

olo

ne

O

R:

0,6

5 (

95%

CI:

0,4

3-0

,94)

β-la

ctam

/ β

-lac

tam

ase

inhib

itor

+

mac

rolid

eO

R:

1,7

7 (

95%

CI:

1,2

8-2

,46)

• Aet

iolo

gy

No d

ata

on a

typic

al p

athogen

s•

Res

ista

nce

No d

ata

• Sev

erity

of CAP

rela

ted t

o c

hoic

e of

antibio

tics

Cep

hal

osp

orin p

lus

mac

rolid

e m

ore

oft

en p

resc

ribed

in low

ris

k pneu

monia

, β-

lact

am a

ntibio

tics

more

oft

en

pre

scribed

in h

igh r

isk

CAP

• Le

ngth

of st

ayN

o d

iffe

rence

s in

len

gth

of st

ay•

Leve

l of ev

iden

ce:

III

Sta

hl

(1999)

(13)

76

Patien

ts s

elec

ted fro

m

char

t re

view

.Ret

rosp

ective

non-

exper

imen

tal co

hort

st

udy

Length

of

stay

Mea

n len

gth

of st

ayReg

imen

s w

ith m

acro

lide

vs.

regim

ens

without

mac

rolid

e:

2,7

5 d

vs.

5,3

d (

p=

0,1

)

• Aet

iolo

gy

Lim

ited

dat

a on a

typic

al p

athogen

s•

Res

ista

nce

No d

ata

• Sev

erity

of CAP

rela

ted t

o c

hoic

e of

antibio

tics

No d

ata

• Le

vel of ev

iden

ce:

III

83

Chapter 5

Dudas

(2000)

(9

)2963

Phys

icia

n-p

resu

med

CAP,

but

for

incl

usi

on

no infi ltra

te o

n c

hes

t X-r

ay w

as n

eces

sary

Prosp

ective

, non-

exper

imen

tal co

hort

st

udy

Mort

ality

Mort

ality

β-la

ctam

/ β

-lac

tam

ase

inhib

itor

plu

s

mac

rolid

e vs

. β-

lact

am:

OR:

0,4

(9

5%

CI:

0,2

-0,8

)

• Aet

iolo

gy

No d

ata

• Res

ista

nce

No d

ata

• Sev

erity

of CAP

rela

ted t

o c

hoic

e of

antibio

tics

Low

er a

ge,

lac

k of IC

U a

dm

itta

nce

an

d s

hort

del

ay t

o a

dm

inis

trat

ion o

f an

tibio

tics

wer

e al

so a

ssoci

ated

with

low

er m

ort

ality

• Le

ngth

of st

ayβ-

lact

am /

β-l

acta

mas

e in

hib

itor

plu

s m

acro

lide

indep

enden

tly

asso

ciat

ed

with d

ecre

ased

LO

S•

Leve

l of ev

iden

ce:

III

Burg

ess

(2000)

(8)

213

Patien

ts s

elec

ted

from

dat

abas

es o

f hea

lthca

re fi n

ance

org

anis

atio

n

Ret

rosp

ective

non

exper

imen

tal

cohort

stu

dy

Mort

ality

Mort

ality*

3rd

gen

. Cep

hal

osp

orin w

ith m

acro

lide

vs.

3rd

gen

. Cep

hal

osp

orin w

ithout

mac

rolid

e

OR:

0,2

7 (

95%

CI:

0,0

3 –

2,6

7)

• Aet

iolo

gy

No e

tiolo

gic

inve

stig

atio

ns

per

form

ed•

Res

ista

nce

No d

ata

• Sev

erity

of CAP

rela

ted t

o c

hoic

e of

antibio

tics

Patien

ts t

reat

ed w

ith m

acro

lide

had

a

signifi

cant

low

er a

ge

• Le

ngth

of st

ay

3rd

gen

erat

ion c

ephal

osp

orin p

lus

mac

rolid

e vs

. 3er

d g

ener

atio

n

cephal

osp

orin:

5,2

+/-

2,8

d v

s. 5

,2 +

/- 3

,4 d

(not

signifi

cant)

• Le

vel of ev

iden

ce:

III

Houck

(2001)

(11)

10.0

69

65 y

ears

of ag

e or

old

er,

pat

ients

sel

ecte

d

from

dat

abas

es o

f hea

lthca

re fi n

ance

org

aniz

atio

n

Ret

rosp

ective

non-

exper

imen

tal co

hort

st

udy

Mort

ality

Mort

ality

(1993)

β-la

ctam

plu

s m

acro

lide

vs.

stan

dar

d

ther

apy

OR:

0,4

2 (

95%

CI:

0,2

5-0

,69)

No fav

oura

ble

effec

t in

1995 a

nd 1

997

• Aet

iolo

gy

No e

tiolo

gic

al inve

stig

atio

ns

per

form

ed•

Res

ista

nce

No d

ata

• Sev

erity

of CAP

rela

ted t

o c

hoic

e of

antibio

tics

No d

ata

• Le

ngth

of st

ay

No d

ata

• Le

vel of ev

iden

ce:

III

84

Chapter 1Chapter 5

Wat

erer

(2001)

(14)

225

Posi

tive

blo

od c

ulture

fo

r S.

pneu

monia

eRet

rosp

ective

non-

exper

imen

tal co

hort

st

udy

Mort

ality

Mort

ality

Mono t

her

apy

vs.

Dual

ther

apy

OR:

3,0

(95%

CI:

1,2

-7,6

)

• Aet

iolo

gy

All

case

s S.

pneu

monia

e. N

o d

ata

on

atyp

ical

pat

hogen

s•

Res

ista

nce

All

stra

ins

wer

e su

scep

tible

for

initia

ted

ther

apy.

• Sev

erity

of CAP

rela

ted t

o c

hoic

e of

antibio

tics

No d

ata

• Le

ngth

of st

ayN

o d

ata

• Le

vel of ev

iden

ce:

III

Mar

tinez

(2

003)

(7)

409

Posi

tive

blo

od c

ulture

fo

r S.

pneu

monia

e

Ret

rosp

ective

non-

exper

imen

tal co

hort

st

udy

Mort

ality

Mort

ality

Rec

eipt

of em

piric

al m

acro

lide

ther

apy

vs.

no r

ecei

pt

of em

piric

al m

acro

lide

ther

apy:

OR:

0.4

(95%

CI:

0.1

7-0

.92)

• Aet

iolo

gy

All

case

s S.

pneu

monia

e. N

o d

ata

on

atyp

ical

pat

hogen

s•

Res

ista

nce

17%

ery

thro

myc

in r

esis

tance

18%

pen

icill

in r

esis

tance

8%

both

ery

thro

myc

in a

nd p

enic

illin

re

sist

ance

• Sev

erity

of CAP

rela

ted t

o c

hoic

e of

antibio

tics

Patien

ts r

ecei

ving m

acro

lide

more

oft

en in s

hock

and a

dm

itte

d t

o I

CU

, oth

er a

dve

rse

pro

gnost

ic fea

ture

s m

ore

pre

vale

nt

in p

atie

nts

not

rece

ivin

g

mac

rolid

e •

Length

of st

ayN

o d

ata

• Le

vel of ev

iden

ce:

III

OR=

odds-

ratio,

CI=

confi den

ce inte

rval

* O

dds-

ratio’s

wer

e re

calc

ula

ted b

ased

on d

ata

giv

en in t

he

public

atio

ns

85

Chapter 5

Possible explanations for the reported outcomes

Atypical pathogens

The causative role of atypical pathogens in the aetiology of CAP is relatively unknown, but could be substantial. A recent report found that CAP was associated with the presence of atypical pathogens in 20% of patients.(21) Many other prospective studies, however, failed to identify an important role of atypical pathogens, despite extensive serological testing. (5) Moreover, annual variation of for example Legionella pneumophila and Mycoplasma pneumoniae, possible as a result of epidemics, could have infl uenced the results of Houck’s study, (11) in which the additional value of macrolides was not consistent over the years. Several arguments can be formulated against the hypothesis that undiagnosed (co-) infections with atypical microorganisms are responsible for the benefi ts of this broad-spectrum empirical coverage. First, the increased mortality associated with treatment with a macrolide plus a β-lactam / β-lactamase inhibitor, as found in one study, (10) is contradictory to this hypothesis. Second, the absence of such a benefi t for patients treated with a quinolone, as found in another study ,(14) also fails to confi rm such a hypothesis. Third, a number of prospective randomised trials reported no benefi t for treatment of macrolides or quinolones as compared to β-lactam antibiotics. (22-26)Whether treatment of atypical pathogens is the explanatory mechanism for the favourable results remains therefore undetermined.

Resistance to antibiotics

The survival benefi t and reduced length of stay associated with regimens with combination therapy could also be explained by resistance of S. pneumoniae against β-lactam antibiotics. Unfortunately, in only 2 of the 7 studies susceptibility of microorganisms for initiated therapy was determined. In one of these, all streptococci were susceptible to prescribed monotherapy. In the USA, 30% of pneumococci show

86

Chapter 1Chapter 5

reduced susceptibility for penicillin in some areas, (27) but this does not seem to infl uence survival in patients with pneumonia. In previous studies of bacteriemic pneumococcal pneumonia mortality risks for patients treated with β-lactams were comparable for those infected with susceptible and non-susceptible pneumococci. (28-33) Furthermore, for β-lactams, it has been shown that maintaining serum levels above MIC for approximately 40% of the dosing interval achieves a good clinical cure for S. pneumoniae. As penicillin-resistance is no absolute resistance, this can be achieved with higher dosages of β-lactams antibiotics.(34;35) It is therefore unlikely that β-lactam resistance can explain the favourable results. In areas with limited resistance, like in the Netherlands and the UK, (36)decreased susceptibility or resistance against S. pneumoniae will hardly be relevant. In fact, resistance to macrolides probably is a greater and more rapidly increasing problem. (37;38) Importantly, clinical failure due to resistance, arisen during treatment, has already been reported for the newer fl uoroquinolones, (39) and may jeopardize the use of this class of agents in the future.

Synergy

The favourable outcomes could also be explained by synergy of macrolides and β-lactam antibiotics. Such a synergistic effect, however, has never been demonstrated and in an animal model this combination even showed antagonism. (40) In theory, antagonism could have lead to the high mortality rate for the combination of β-lactam / β-lactamase inhibitors and macrolides in Gleason’s study, (10) but combinations of other β-lactam antibiotics and macrolides were associated with a reduction of mortality in this study. Synergy would therefore not be a likely explanation for the reduction in mortality as seen in the studies.

Anti-infl ammatory effect of macrolides

Macrolides show anti-infl ammatory effects, which property is used, for example, in the treatment of diffuse panbronchiolitis. The mechanism that causes the modulation of infl ammation in the acute phase is not yet entirely clear. In vitro, macrolides can decrease the production

87

Chapter 5

of pro-infl ammatory cytokines and expression of endothelin-1, inhibit the production of super-oxide and diminish the adherence of pneumococci to respiratory epithelium.(41) If the favourable outcome of short-term macrolide treatment in CAP results from anti-infl ammatory effects, however, remains unknown. Moreover, unless quinolones have similar anti-infl ammatory properties, the survival benefi t associated with quinolone treatment would then remain unexplained.

Evidence level Defi nition

I

II

III

IV

V

Large, randomised trials with clear cut results; low risk of false-positive error or

false-negative error

Small, randomised trials with uncertain results; moderate to high risk of false-

positive and/or false negative error

Nonrandomised, contemporaneous controls

Nonrandomised, historical controls and expert opinion

Case series, uncontrolled studies and expert opinion

Grading of guideline statement Defi nition

ABCDE

Supported by at least 2 level I investigationsSupported by only 1 level I investigationSupported by level II investigations onlySupported by at least 1 level III investigationSupported by level IV or V evidence

Table 2

Description of levels of evidence and grading of guideline statements (6)

88

Chapter 1Chapter 5

Conclusion

In the studies discussed above, treatment with a macrolide plus a β-lactam antibiotic or monotherapy with a quinolone showed reductions in mortality or length of stay. However, the non-experimental design of the reviewed studies may have resulted in confounding by indication and this may have infl uenced the results signifi cantly. The current advise to use empirical treatment with either a β-lactam / macrolide combination or monotherapy with a new quinolone for patients hospitalised with CAP is based on studies of level III evidence. Moreover, the results are inconsistent and do not reveal a mechanism that explains the favourable results. Therefore, given the current evidence it cannot be concluded that the addition of a macrolide or monotherapy with one of the newer fl uoroquinolones should become the standard of care for patients admitted to the hospital because of CAP. In addition, possible disadvantages, such as an increase in costs, an increase in side effects as a result of antibiotic usage, and an increase in the development of resistance were not investigated. Widespread implementation of treatment regimes including unnecessary use of macrolides and fl uoroquinolones could lead to increase in antibiotic pressure which may enhance antibiotic resistance dramatically. (42;43) To determine the costs and benefi ts of adding a macrolide or a fl uoroquinolones to the initial treatment of patients with CAP, large, prospective, randomised studies are necessary.

89

Chapter 5

Reference List

(1) Guidelines for the management of community-acquired pneumonia in adults admitted to hospital. The British Thoracic Society. Br J Hosp Med 1993; 49(5):346-350.

(2) Niederman MS, Bass JB, Jr., Campbell GD, Fein AM, Grossman RF, Mandell LA et al. Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. American Thoracic Society. Medical Section of the American Lung Association. Am Rev Respir Dis 1993; 148(5):1418-1426.

(3) Bartlett JG, Breiman RF, Mandell LA, File Jr TM. Community-acquired pneumonia in adults: guidelines for management. Clinical Infectious Diseases 1998; 26:811-838.



(6) Sackett DL. Rules of evidence and clinical recommendations on the use of antithrombotic agents. Chest 1989; 95(2 Suppl):2S-4S.

(7) Martinez JA, Horcajada JP, Almela M, Marco F, Soriano A, Garcia E et al. Addition of a Macrolide to a beta-Lactam-Based Empirical Antibiotic Regimen Is Associated with Lower In-Hospital Mortality for Patients with Bacteremic Pneumococcal Pneumonia. Clin Infect Dis 2003; 36(4):389-395.

(8) Burgess DS, Lewis JS. Effect of macrolides as part of initial empiric therapy on medical outcomes for hospitalized patients with community-acquired pneumonia. Clin Ther 2000; 22(7):872-878.



90

Chapter 1Chapter 5





(15) Simpson JC, Macfarlane JT, Watson J, Woodhead MA. A national confi dential enquiry into community acquired pneumonia deaths in young adults in England and Wales. British Thoracic Society Research Committee and Public Health Laboratory Service. Thorax 2000; 55(12):1040-1045.

(16) Grobbee DE, Hoes AW. Confounding and indication for treatment in evaluation of drug treatment for hypertension. BMJ 1997; 315(7116):1151-1154.

(17) Hak E, Verheij TJ, Grobbee DE, Nichol KL, Hoes AW. Confounding by indication in non-expirimental evaluation of vaccine effectiveness: the example of prevention of infl uenza complications. J Epidemiol Community Health 2002; 56(12):951-955.

(18) Kalin M, Ortqvist A, Almela M, Aufwerber E, Dwyer R, Henriques B et al. Prospective study of prognostic factors in community-acquired bacteremic pneumococcal disease in 5 countries. J Infect Dis 2000; 182(3):840-847.

(19) Finch R, Schurmann D, Collins O, Kubin R, McGivern J, Bobbaers H et al. Randomized controlled trial of sequential intravenous (i.v.) and oral moxifl oxacin compared with sequential i.v. and oral co-amoxiclav with or without clarithromycin in patients with community-acquired pneumonia requiring initial parenteral treatment. Antimicrob Agents Chemother 2002; 46(6):1746-1754.

(20) Hedlund J, Ortqvist A, Ahlqvist T, Augustinsson A, Beckman H, Blanck C et al. Management of patients with community-acquired pneumonia treated in hospital in Sweden. Scand J Infect Dis 2002; 34(12):887-892.

(21) Lim WS, Macfarlane JT, Boswell TCJ, Harrison TG, Rose D, Leinonen M et al. Study of community acquired pneumonia aetiology (SCAPA) in

91

Chapter 5

adults admitted to hospital: implications for management guidelines. Thorax 2001; 56:296-301.

(22) Aubier M, Verster R, Regamey C, Geslin P, Vercken JB, Sparfl oxacin European Study Group. Once-daily sparfl oxacin versus high-dosage amoxicillin in the treatment of community-acquired, suspected pneumococcal pneumonia in adults. Clin Infect Dis 1998; 26:1312-1320.

(23) File Jr TM, Segreti J, Dunbar L, Player R, Kohler R, Williams RR et al. A multicenter, randomized study comparing the effi cacy and safety of intravenous and/or oral levofl oxacin versus ceftriaxone and/or cefuroxime axetil in treatment of adults with community-acquired pneumonia. Antimicrobial Agents and Chemotherapy 1997; 41(9):1965-1972.

(24) Genne D, Siegrist HH, Humair L, Janin-Jaquat B, de Torrente A. Clarithromycin versus amoxicillin-clavulanic acid in the treatment of community-acquired pneumonia. Eur J Clin Microbiol Infect Dis 1997; 16(11):783-788.

(25) Norrby SR, Petermann W, Willcox PA, Vetter N, Salewski E. A comparative study of levofl oxacin and ceftriaxone in the treatment of hospitalized patients with pneumonia. Scand J Infect Dis 1998; 30:397-404.

(26) Tremolieres F, Kock F de, Pluck N, Daniel R. Trovafl oxacin versus high-dose amoxicillin (1g three times daily) in the treatment of community-acquired bacterial pneumonia. Eur J Clin Microbiol Infect Dis 1998; 17:447-453.

(27) Centers for Disease Control and Prevention. Geographic variation in penicillin resistance in Streptococcus pneumoniae - selected sites, United States, 1997. MMWR Morb Mort Wkly Rep 1999; 48:656-661.

(28) Pallares R, Linares J, Vadillo M, Cabellos C, Manresa F, Viladrich PF et al. Resistance to penicillin and cephalosporin and mortality from severe pneumococcal pneumonia in Barcelona, Spain. N Engl J Med 1995; 333(8):474-480.

(29) Deeks SL, Palacio R, Ruvinsky R, Kertesz DA, Hortal M, Rossi A et al. Risk factors and course of illness among children with invasive penicillin-resistant Streptococcus pneumoniae.The Streptococcus pneumoniae Working Group. Pediatrics 1999; 103(2):409-413.

(30) Friedland IR. Comparison of the response to antimicrobial therapy of penicillin- resistant and penicillin-susceptible pneumococcal disease. Pediatr Infect Dis J 1995; 14(10):885-890.

(31) Sanchez C, Armengol R, Lite J, Mir I, Garau J. Penicillin-resistant

92

Chapter 1Chapter 5

pneumococci and community-acquired pneumonia. Lancet 1992; 339(8799):988.

(32) Klugman KP. Clinical relevance of in-vitro resistace to penicillin, ampicillin, amoxycillin and alternative agents, for the treatment of community-acquired pneumonia caused by Streptococcus pneumoniae, Haemophilus infl uenzae and Moraxella catarrhalis. J Antimicrob Chemother 1996; 38 (suppl A):133-140.

(33) Moroney JF, Fiore AE, Harrison LH, Patterson JE, Farley MM, Jorgensen JH et al. Clinical outcomes of bacteremic pneumococcal pneumonia in the era of antibiotic resistance. Clin Infect Dis 2001; 33:797-805.

(34) Craig WA. The hidden impact of antibacterial resistance in respiratory tract infection. Re-evaluating current antibiotic therapy. Respir Med 2001; 95 Suppl A:S12-S19.

(35) File TM, Jr., Jacobs MR, Poole MD, Wynne B. Outcome of treatment of respiratory tract infections due to Streptococcus pneumoniae, including drug-resistant strains, with pharmacokinetically enhanced amoxycillin/clavulanate. Int J Antimicrob Agents 2002; 20(4):235-247.

(36) Vardhan MS, Allen KD. Epidemiology of Penicillin-resistant Pneumococci in a Merseyside Health District over a 14-year Period. J Infect 2003; 46(1):23-29.

(37) de Neeling AJ, Overbeek BP, Horrevorts AM, Ligtvoet EE, Goettsch WG. Antibiotic use and resistance of Streptococcus pneumoniae in The Netherlands during the period 1994-1999. J Antimicrob Chemother 2001; 48(3):441-444.

(38) File TM, Jr., Tan JS. International guidelines for the treatment of community-acquired pneumonia in adults: the role of macrolides. Drugs 2003; 63(2):181-205.


(40) Johansen HK, Jensen TG, Dessau RB, Lundgren B, Frimodt-Moller N. Antagonism between penicillin and erythromycin against Streptococcus pneumoniae in vitro and in vivo. J Antimicrob Chemother 2000; 46(6):973-980.

(41) Wales D, Woodhead M. The anti-infl ammatory effects of macrolides. Thorax 1999; 54(Suppl 2):S58-S62.

(42) Neuhauser MM, Weinstein RA, Rydman R, Danziger LH, Karam G, Quinn JP. Antibiotic resistance among gram-negative bacilli in US intensive care units: implications for fl uoroquinolone use. JAMA

93

Chapter 5

2003; 289(7):885-888.

(43) McCormick AW, Whitney CG, Farley MM, Lynfi eld R, Harrison LH, Bennett NM et al. Geographic diversity and temporal trends of antimicrobial resistance in Streptococcus pneumoniae in the United States. Nat Med 2003; 9(4):424-430.

94

Chapter 1Chapter 5

6Introducing British or

North American Guidelines

for Community-Acquired

Pneumonia in The Netherlands

Predicted effects on antibiotic use and

adequacy of treatment

Submitted

JJ Oosterheert, MJM Bonten, MME Schneider, IM Hoepelman

96

Chapter 1Chapter 6

Abstract

Introduction: Recent treatment guidelines for community-acquired pneumonia (CAP) recommend combinations of beta-lactams and macrolides or monotherapy with fl uoroquinolones for patients admitted to hospital. We evaluated changes in antibiotic use and adequacy of treatment when British or North-American guidelines would be implemented in the Netherlands.Patients and Methods: Patients admitted for mild, moderate and severe CAP were prospectively evaluated. Adequacy and volume of antibiotic use based upon BTS, ATS or IDSA guidelines were predicted and compared to current practice. Results: In 248 patients, current antibiotic use was 3087 Defi ned Daily Doses (DDD). Antibiotic use would be increase with 38% when based on ATS guidelines, with 23% when based on IDSA guidelines and with 21% when based on BTS guidelines. Antibiotic use would increase most for cases of moderate CAP. Unnecessary atypical coverage would increase from 32% to 56%, whereas uncovered atypical infections would decrease from 57% to 33%. This would lead to incremental antibiotic costs of € 1.750.000,- to € 3.500.000 in the Netherlands. Conclusions: Implementation of new British or North-American guidelines for the empirical treatment of CAP in the Netherlands would lead to a considerable increase in antibiotic use and costs, especially in patients with moderate CAP.

97

Chapter 6

Introduction

Community-acquired pneumonia (CAP) is a common disease which is associated with considerable mortality and morbidity and accounts for high antibiotic consumption. (1;2) Initial antibiotic therapy for CAP is preferably targeted to the causative micro-organism, but as this cannot be predicted upon clinical, laboratory and radiographic fi ndings, initial therapy is mostly empiric covering several possible causative micro-organisms. (3-5) S. pneumoniae is by far the most common isolated pathogen. (6-8) Formerly, recommended empirical therapy for uncomplicated pneumonia was benzyl penicillin, amoxicillin or another β-lactam antibiotic. The addition of a macrolide or a quinolone to the initial treatment was not recommended unless there was a strong suspicion of atypical pneumonia based on history or laboratory fi ndings or in case of severe pneumonia requiring mechanical ventilation. (9-11) Recently, outcomes of non-randomized studies have suggested that initial treatment with combinations of beta-lactam and macrolide antibiotics or initial treatment with a fl uoroquinolone improves patient outcome when compared to treatment with beta-lactam antibiotics alone. Possible explanations for these fi ndings are increasing resistance of respiratory pathogens to penicillins, a large but undetected role of C. pneumoniae, M. pneumoniae and L. pneumophila in the etiology of CAP, or anti-infl ammatory effects of macrolides. (12-18) Although controversies remain regarding the design of these studies, British and North-American infectious diseases and thoracic societies have changed their guidelines for the empirical treatment of CAP and now recommend initial therapy with combinations of beta lactams and macrolides or monotherapy with quinolones for patients hospitalized because of CAP. (19-23) However, monotherapy with penicillin remained adequate as initial therapy for CAP in European studies even at the end of the last century (24-26) Antibiotic resistance of S. pneumoniae isolates are still low in most European countries, including the Netherlands (www.earss.rivm.nl), and strains with reduced susceptibility can still be adequately treated with higher doses of penicillins. (27-30) Therefore, guidelines from the European continent including Dutch guidelines, though published in 1998, still recommend betalactam monotherapy as initial management for CAP treated in general wards. (31-33)

98

Chapter 1Chapter 6

Currently, new treatment guidelines for CAP in the Netherlands are under consideration. Should a strategy for treating CAP-patients with combinations of beta-lactams and macrolides or monotherapy with fl uoroquinolones for this group of patients be introduced in The Netherlands, this will evidently affect antibiotic use. Adverse consequences of an increase in antibiotic use include an increase in drug-costs, a possible increase in side effects related to antibiotic treatment and an increase in selection pressure that enhances the development of resistance. (34) Favorable effects may be an increase of adequate treatment or improvement of clinical outcomes. Therefore, in the current study, we estimated the change in antibiotic use, antibiotic-related costs and adequacy of initial therapy, as compared to the present standard of care, for the scenario that initial treatment of CAP is based on British Thoracic Society (BTS), Infectious Diseases Society of America (IDSA), or American Thoracic Society (ATS) guidelines in the Netherlands.

Material and methods

Patients

Consecutive episodes of CAP in patients admitted to the University Medical Center Utrecht from September 2000 till January 2003 were studied. CAP was defi ned as a new of progressive infi ltrate on chest X-ray and two or more of the following criteria: cough, production of purulent sputum, rectal temperature above 38oC or below 36oC, auscultatory fi ndings consistent with pneumonia, leucocytosis (>10,000/mm3) or CRP > 3x upper limit of normal. (35) Patients with cystic fi brosis, neutropenic patients (<0,5 x 109 neutrophils / l) and patients in which another infection needed treatment were excluded from the analysis. Patients were categorized into three categories. Mild CAP was defi ned as Fine class I - III, moderate CAP was defi ned as Fine class IV or V or fulfi lling the ATS-criteria for severe CAP, but not needing intensive care treatment and severe CAP was defi ned as treated in the intensive care unit. Demographic and clinical data were collected. In addition, points were assigned according to BTS-guidelines for the presence of confusion, urea > 7 mmol/l, respiratory rate > 30/ min and systolic blood pressure <90 mm Hg or diastolic

99

Chapter 6

blood pressure <60 mm Hg. For the presence of each feature, one point was assigned. (“CURB”-classifi cation). (23) Microbiology

Microbiological evaluations were performed following standard procedures and on request of the treating physician. Sputum samples and blood samples were collected, cultured and evaluated following standard procedures. Micro-organisms cultured in blood or sputum were recorded. In addition, Binax NOW-tests were used to detect urinary antigen for Legionella pneumophila and S. pneumoniae. Collected acute and convalescent serology samples were evaluated for Mycoplasma pneumoniae, L. pneumophila and Chlamydia pneumoniae. Any non-contaminating micro-organism cultured from a blood or sputum sample or detected by urinary antigen testing was considered a cause for the episode of pneumonia. For Mycoplasma pneumoniae, a fourfold or greater increase in titer in paired sera or a single titer of greater than or equal to 1:40 was considered indicative of infection. (36) (Immune fl uorescence agglutination, Serodia-MycoII ®, Fujirebio, inc.) For Legionella pneumophila, a fourfold increase in the antibody titer to 1:128 or greater, or single titers of 1:256 or more were considered evidence of Legionella pneumonia.(37) For Chlamydia pneumoniae, detection of IgM above established values, seroconversion of IgG between acute and convalescence samples, high amounts of IgG in single titers or a combination of these factors were considered serological evidence of infection, according to the manufacturers instructions. (ELISA, Savyon Diagnostics Ltd)

Antibiotic use

Empirical antibiotic treatment was guided by Dutch guidelines. (31) Comparisons were made between antibiotic use in the present standard of care and predicted treatment based on “CURB”-classifi cations as recommended in the BTS guidelines for the initial treatment of CAP, predicted treatment based on IDSA-guidelines and predicted treatment based on ATS- guidelines. (19;22;23) In the calculations based on “CURB”-classifi cation, we assigned patients

100

Chapter 1Chapter 6

with 0 or 1 CURB points, without additional features (age > 50y, presence of comorbid illness, saturation < 92%, or bilateral infi ltrates) monotherapy with a beta-lactam agent. Patients with 1 CURB point with additional features, and patients with 2 or more CURB-points were assigned the combination of ceftriaxone with erythromycin. For the calculations based IDSA guidelines, we assigned beta-lactam monotherapy to patients in Fine class II and III (as outpatient therapy) and the combination of ceftriaxone and erythromycin to patients in Fine class IV and V. For the calculations based on ATS guidelines, we assigned macrolide monotherapy to patients without ATS-defi ned modifi ers (age > 65 years , beta-lactam therapy within the past 3 months, alcoholism, immune-suppressive illness, multiple medical comorbidities, exposure to a child in day care center, residence in a nursing home, underlying cardiopulmonary disease, multiple medical comorbidities, recent antibiotic therapy, structural lung disease, cortocosteroid therapy, broad-spectrum antibiotic therapy for > 7 d in the past month, malnutrition) and the combination of ceftriaxone and erythromycin to patients with ATS-defi ned modifi ers. (22)Calculations were based on a standard length of treatment (3 days for azithromycin and 10 days for all other antibiotics) and were expressed in defi ned daily doses (DDD, WHO defi nition, www.whocc.no/atcddd/). Drug costs were determined using 2003 Dutch cost-price levels as indicated in the Dutch Farmacotherapeutic Guide (Netherlands National Health Insurance Council, edition 2004).

Results

Patients

Two-hundred and forty-eight patients admitted because of CAP were included (Table 1). Fifty (20%) patients were admitted to the intensive care unit, 107 (43%) were diagnosed with moderate CAP and 91 (37%) were diagnosed with mild CAP. In the total population, 82 (33%) patients fulfi lled 0 “CURB”- criteria, 35 (14%) 1 “ CURB” criterion without additional features, 61 (25%) 1 “CURB” criterion with additional features and 70 (29%) had 2 or more “CURB” criteria. One-hundred and twenty one (49%) patients were in Fine class IV and V and 155 (63%) had ATS-defi ned modifi ers.

101

Chapter 6

Totaln (% ) or mean ± sd

Severe Moderate Mild

Total 248 50 (20) 107 (43) 91 (37)

Men 145 (59) 33 (66) 74 (69) 38 (42)

Age (y) 59.5 ±18 58.8 ± 17 66.5 ± 15 51.4 ± 17

Fine 92.6 ± 36 109.4 ± 34 111.1 ± 29 61.5 ± 20

Leucocytes (109 / l) 15.0 ± 7 14.3 ± 7 15.6 ± 8 14.7 ± 6

Respiratory rate (/min) 27.2 ± 9 31.9 ± 10 28 ± 9 24.2 ± 7

In hospital mortality 32 (13) 21 (42) 8 (8) 3 (3)

LOS (days) 12.4 ±11 16.4 ± 18 10.8 ± 6 8.2 ± 4

ICU-LOS (days) - 9.7 ± 14 - -

CURB-class*

0 82 (33) 9 (18) 15 (14) 58 (64)

1 35 (14) 11 (22) 6 (5) 18 (20)

1+ 61 (25) 12 (24) 41 (38) 8 (9)

2 54 (22) 15 (30) 32 (29) 7 (8)

3 14 (6) 3 (6) 11 (10)

4 2 (2) 0 (0) 2 (2)

Fine-class

2 72 (29) 9 (18) 9 (8) 54 (59)

3 55 (22) 6 (12) 13 (12) 37 (41)

4 81 (32) 18 (36) 62 (58)

5 40 (16) 17 (34) 23 (22)

Initial therapy, patients

Bl 153 (62) 8 (16) 83 (78) 62 (68)

Ch 5 (2) 0 (0) 0 (0) 5 (5)

M 7 (3) 1 (2) 3 (3) 3 (3)

Bl+M 58 (23) 21 (42) 18 (17) 19 (21)

Bl+Ag 13 (5) 11 (22) 1 (1) 1 (1)

Bl+M+Ag 3 (1) 3 (6) 0 (0) 0 (0)

Other 9 (4) 6 (12) 2 (2) 1 (1)

Initial therapy, DDDBetalactamsChinolonesMacrolidesOthertotal

234070602753087

83050164161060

10501680111229

4602702048798

Table 1 Patient characteristics* CURB: 1 point each for confusion, urea > 7 mmol/l, respiratory rate > 30, systolic blood pressure < 90 or diastolic blood pressure < 60 mm Hg. “+” constitutes the presence of additional features age > 50y, presence of comorbid illness, saturation < 92%, or bilateral infi ltrates** Bl = beta-lactam antibiotic, M=macrolide antibiotic, Ch = fl uoroquinolone, Ag = aminoglycoside

102

Chapter 1Chapter 6

Total (n=248)n (%)

ICU (n=50)

Moderate (n=107)

Mild(n=91)

S. pneumoniae 50 (20) 14 (28) 31 (29) 5 (6)

E. coli 9 (4) 3 (6) 4 (4) 2 (2)

S. aureus 11 (4) 4 (8) 6 (6) 1 (1)

H. infl uenzae 6 (2) 2 (4) 2 (2) 2 (2)

S. oralis 2 (1) 2 (4) 0 0

P. aeruginosa 5 (2) 3 (6) 0 2 (2)

M. catharralis 6 (2) 1 (2) 3 (3) 2 (2)

M. morganii 4 (2) 3 (6) 0 1 (1)

L. pneumophila 5 (2) 2 (4) 2 (2) 1 (1)

C. pneumoniae 14 (6) 0 11 (10) 3 (3)

M. pneumoniae 2 (1) 0 1 (1) 1 (1)

K. pneumoniae 4 (2) 2 (4) 0 2 (2)

Infl uenza virus 5 (2) 2 (4) 0 3 (3)

Other 23 (9) 5 (10) 12 (11) 6 (7)

Mixed 18 (7) 10 (20) 5 (5) 3 (3)

Unknown 115 (46) 16 (32) 45 (42) 54 (59)

Tabel 2

Etiology

Clinical outcome

Thirty-two (13 %) patients died: twenty-one of 50 (42%) patients initially admitted to ICU, 8 (8%) of 107 patients with moderate to severe CAP and 3 (3%) of 91 patients with mild CAP. Two (2%) patients, initially admitted to a general ward, clinically deteriorated and needed mechanical ventilation in ICU, one of them died. Twelve (38%) deaths were directly attributed to pneumonia, sepsis or respiratory insuffi ciency and 20 (62%) patients had other causes of death. Initial treatment in these deceased patients appeared to be adequate with current treatment: only one patient (3%) that died had evidence of atypical infection (Legionella pneumophila infection was diagnosed based on a positive urinary antigen test) and this patient was initially treated with amoxicillin + clavulanic acid, erythromycin and ciproxin. The other patients that died had infections with “typical” bacteria (n=15, 47%), viral pathogens (n=1, 3%) or had unknown causes

103

Chapter 6

(n=15, 47%). They were treated with beta-lactam monotherapy in 12 (38%) patients, the combination of a betalactam plus a macrolide in 9 (28%) patients and other combinations in 11 (34%) patients.

Microbiological outcome

The most frequently isolated pathogen was S. pneumoniae in 50 (20%) patients. As penicillin resistance of S. pneumoniae is < 1% in the Netherlands, all pneumococcal isolates were considered penicillin susceptible. Other bacteria were isolated in 73 (29%) patients. Twenty-one (9%) patients had evidence of ‘atypical’ infections: 2 (4%) patients admitted to ICU had CAP due to L. pneumophila, 14 (13%) patients with moderate CAP had atypical infection (M. pneumoniae 1 (1%), C. pneumoniae 11 (10%) , L. pneumophila 2 (1%)) and 5 (6%) patients with mild CAP had atypical infection. (M. pneumoniae 1 (1%), C. pneumoniae 3 (4%), L. pneumophila 1 (1%)). No etiological cause could be identifi ed in 115 (46%) episodes of CAP. (Table 2)

Antibiotic use and costs

Current antibiotic use consisted of monotherapy with a beta-lactam antibiotic in 153 (62%) patients, a macrolide in 7 (3%) and a fl uoroquinolone in 5 (2%) patients. Eighty-fi ve (23%) patients received combinations of beta-lactams and macrolides or other combinations (18 (7%)) Combination therapy was more frequently prescribed for patients initially admitted to ICU. (21 of 50 patients (42%) in ICU), versus 21% and 17% in patients with mild or moderate CAP, respectively (p<0.01). Based on 10 day antibiotic courses (Azithromycin once daily 500 mg, for 3 days), total antibiotic consumption was 3087 DDD Predicted antibiotic use, should therapy be based on BTS-guidelines, would increase to 3790 DDD (95% CI: 3645 - 3943) , an increase of 23% (95% CI: 17 - 28%). Treatment based on IDSA-guidelines would lead to an antibiotic use of 3740 DDD (95% CI: 3594 - 3894) (+21%, 95% CI: 16-26) and ATS-guidelines would increase antibiotic use to 4270 DDD (95% CI: 4132 - 4399) (+ 38%, 95% CI: 33 - 43). The increase in antibiotic consumption would be most apparent among patients with moderate CAP (increase of antibiotic use of 1229 DDD to 1930 DDD (+57%, BTS), to 1920 DDD

104

Chapter 1Chapter 6

(+57%, IDSA) or to 2050 (+66%, ATS)) (fi gure 1)Calculated average drug-costs per patient treated according to current practice was € 252 and would be € 382 with BTS guidelines, € 362 with IDSA guidelines and € 475 with ATS guidelines. In our cohort, applying new British or North-American guidelines would lead to an estimated increase in drug costs between € 27.000,- and € 55.000,-. With annual hospitalisations because of CAP in the Netherlands of 16.000, (38) the estimated annual increase in drug costs would be between € 1.750.000,- to € 3.500.000,-.

3087

3740

1060 1060920

11201229

1930 19202050

798 800 8501000

3790

4270

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

Total,

curre

nt p

racti

ce

Total,

BTS

Total,

IDSA

Total,

ATS

Mild

, cur

rent

pra

ctice

Mild

, BTS

Mild

, IDSA

Mild

, ATS

Mod

erat

e, cu

rrent

pra

ctice

Mod

erat

e, B

TS

Mod

erat

e, ID

SA

Mod

erat

e, A

TS

Sever

e, cu

rrent

pra

ctice

Sever

e, B

TS

Sever

e, ID

SA

Sever

e, A

TS

Figure 1

Estimates of antibiotic use in current practice, and based on BTS-, IDSA- or ATS-guidelines

Adequacy of atypical treatment when based on new British or North-American guidelines

With current practice, eighty patients received initial antibiotic therapy that covered atypical pathogens. However, 71 (89%) of these patients had no evidence of infection with M. pneumoniae, C. pneumoniae or L. pneumophila (unnecessary atypical coverage: 29% of the total population). On the other hand, of the 21 patients with atypical infection, 12 (57%) did not receive initial antibiotic treatment to cover these pathogens. (uncovered atypical infection). Unnecessary atypical coverage would increase to 49%, 63% and 53% with IDSA,

105

Chapter 6

ATS and BTS guidelines respectively, whereas uncovered atypical infection would decrease to 38% , 28% and 33% with IDSA, ATS and BTS guidelines, respectively. (table 3) With the incidence of atypical infections as found in this study, unnecessary atypical coverage of approximately 60 patients with beta-lactam / macrolide combinations would be needed to prevent one untreated atypical infection.

n=248

Current practiceN/total(%)

IDSAN/total(%)

p(compared to current practice)

ATSN/total(%)


BTSN/total(%)


Atypical coverage

80/248(32)

121/248(49) <0.01 155/248

(63) <0.01 131/248(53) <0.01

Unnecessary atypical coverage*

71/227(31)

108/227(48) <0.01 140/227

(62) <0.01 122/227(54) <0.01

Uncovered atypical infection*

12/21(57)

8/21(38) 0.08 6/21

(28) 0.01 7/21(33) 0.03

Table 3

Treatment of atypical pathogens in current practice, when guided by BTS-, ATS- or IDSA-

guidelines

* based on 21 (8%) cases of demonstrated atypical infections in the study cohort.

Unnecessary atypical coverage: patients without infection with atypical pathogens, but treated

with antibiotics that cover these infections.

Uncovered atypical coverage: patients with infections with atypical pathogens, but not treated

with antibiotics that cover these infections.

Discussion

Our analyses demonstrate that implementation of British or North-American guidelines in the Netherlands would result in a considerable increase in antibiotic use, which is most obvious in patients with moderate CAP, needing treatment in general hospital wards. This also applies to countries with similar empirical treatment strategies, such as the Nordic countries. (24;25) The adverse consequences are obvious: an increase in drug-costs of over € 1.750.000 in the Netherlands, a possible increase in side effects and an increase in selective pressure that enhances development of antibiotic resistance.

106

Chapter 1Chapter 6

Macrolide resistance of pneumococci has already risen to 7% in the Netherlands, (http://www.earss.rivm.nl) and any further rise may limit its use for empirical treatment in the near future. Bacterial resistance to fl uoroquinolones is still low, but can develop easily even during treatment. (34;39;40) To justify an increase in antibiotic use, the evidence must, therefore, be clear and convincing. Consequently, new treatment guidelines should either lead to increased patient survival, or have better cost-effectiveness. Increased patient survival has not been unequivocally demonstrated: the new recommendations for beta-lactam plus macrolide combinations or treatment with fl uoroquinolones are based on retrospective, non-randomized studies which only qualify for “level III”-evidence. (41),(42) The fi rst of these studies was published in 1999 (13). Publication bias may have limited reporting of studies failing to demonstrate outcome differences between beta-lactam monotherapy and combinations of beta-lactams and macrolides or monotherapy with fl uoroquinolones. Up till now, only 2 studies with “negative” outcomes have been published. (43;44) In our study, infections with C. pneumoniae, M. pneumoniae or L. pneumophila were infrequent and clinical outcome for patients with evidence of atypical infection, but not treated with macrolides or fl uoroquinolones was favorable: 9 of 107 patients with moderate CAP and atypical infection admitted to a general ward, were initially not treated for these micro-organisms and all were discharged in good health after a mean of 10,7 days. The estimated cost increase may be even higher in reality. In the cost-estimations based on the treatment of BTS guidelines, we assigned mono therapy to patients in CURB class 0 and 1. However, BTS guidelines only recommend mono therapy with a betalactam if the patient has not received therapy prior to admission or is being admitted for social reasons. Therefore, we may have underestimated the total number of DDD’s given when using the BTS guidelines. The contribution of atypical pathogens may vary in different settings, but can be substantial. (7;14;23) In addition, diagnosing atypical infections by means of serological investigations is diffi cult, especially in case of C. pneumoniae. (45) If serological evidence of C. pneumoniae represents self-limiting or previous infection, the number of untreated atypical infections is lower. On the other hand, because of the diffi culties in diagnosing atypical infections it is possible that some patients with

107

Chapter 6

unknown etiology have had undetected atypical infections. Treatment of these infections may increase cost-effectiveness of combinations of beta-lactams and macrolides. Overestimation of the cost increase may have occurred because as well, as we were unable to take the subjective clinical judgment into account. Some patients with 1 CURB point plus additional features may be classifi ed as having “non-severe” in stead of “severe” pneumonia, therefore receiving beta-lactam plus a macrolide instead of ceftriaxone plus a macrolide. Current guidelines are mainly based on severity of presentation of CAP-patients and not on the presumed causative pathogen. More adequate initial treatment may be realized by using rapid diagnostic testing. For L. pneumophila and S. pneumoniae rapid urinary antigen assays are available. The essay for L. pneumophila is highly specifi c and its use is now recommended for all patients with severe CAP. (23;46) Sensitivity of the S. pneumoniae assay, however, is only 66-80% which is probably not good enough to guide clinical care. (47-50) In the near future, rapid diagnostic tests by means of real-time PCR will become available, but their clinical value has yet to be determined. (51;52)Although guidelines and critical pathways can help in reducing the use of institutional resources without causing adverse effects on the well-being of patients, (53) in practice, non-adherence to national guidelines occurs frequently. Some factors that are associated with non-adherence to CAP guidelines are the presence of active co-morbidities, the primary care physician’s wish for hospitalization, clinical interpretation, patient preference or inadequate home support. (54) Consequently, clinical judgment and social factors remain important in making site-of care and treatment decisions. (55) Furthermore, the effect on clinical outcomes of adhering to guidelines is not clearly demonstrated. In a recent study of 295 CAP-patients, adherence to ATS-guidelines was associated with a lower risk of mortality only for severe CAP. For the other risk classes, no signifi cant differences in mortality or length of hospital stay were found. (56) Another study of 8975 CAP patients showed no effect on length of hospital stay or mortality for patients initially treated for atypical pathogens, as recommended in North-American and British treatment guidelines as compared to monotherapy with beta-lactams.

108

Chapter 1Chapter 6

(44)In conclusion, implementing new British or North-American guidelines for the treatment of CAP will be associated with an increase in antibiotic use. The clinical benefi t of these new strategies, however, still remains to be determined in randomized trials.

109

Chapter 6

Reference List









(9) Guidelines for the management of community-acquired pneumonia in adults admitted to hospital. The British Thoracic Society. Br J Hosp Med 1993; 49(5):346-350.

(10) Niederman MS, Bass JB, Jr., Campbell GD, Fein AM, Grossman RF, Mandell LA et al. Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. American Thoracic Society. Medical Section of the American Lung Association. Am Rev Respir Dis 1993; 148(5):1418-1426.

110

Chapter 1Chapter 6

(11) Bartlett JG, Breiman RF, Mandell LA, File Jr TM. Community-acquired pneumonia in adults: guidelines for management. Clinical Infectious Diseases 1998; 26:811-838.




(15) Martinez JA, Horcajada JP, Almela M, Marco F, Soriano A, Garcia E et al. Addition of a Macrolide to a beta-Lactam-Based Empirical Antibiotic Regimen Is Associated with Lower In-Hospital Mortality for Patients with Bacteremic Pneumococcal Pneumonia. Clin Infect Dis 2003; 36(4):389-395.






(21) Mandell LA, Bartlett JG, Dowell SF, File TM, Jr., Musher DM, Whitney C. Update of practice guidelines for the management of community-

111

Chapter 6

acquired pneumonia in immunocompetent adults. Clin Infect Dis 2003; 37(11):1405-1433.



(24) Kirk O, Glenthoj J, Dragsted UB, Helweg-Larsen J, Aru TM, Benfi eld TL et al. Penicillin as empirical therapy for patients hospitalised with community acquired pneumonia at a Danish hospital. Dan Med Bull 2001; 48(2):84-88.

(25) Hedlund J, Ortqvist A, Ahlqvist T, Augustinsson A, Beckman H, Blanck C et al. Management of patients with community-acquired pneumonia treated in hospital in Sweden. Scand J Infect Dis 2002; 34(12):887-892.

(26) Bohte R, Wout JW van ‘t, Lobatto S, Blussé van Oud Alblas A, Boekhout M, Nauta EH et al. Effi cacy and safety of azithromycin versus benzylpenicillin or erythromycin in community-acquired pneumonia. European Journal of Clinical Microbiology and Infectious Diseases 1995; 14:182-187.

(27) Friedland IR. Comparison of the response to antimicrobial therapy of penicillin- resistant and penicillin-susceptible pneumococcal disease. Pediatr Infect Dis J 1995; 14(10):885-890.

(28) Klugman KP. Clinical relevance of in-vitro resistace to penicillin, ampicillin, amoxycillin and alternative agents, for the treatment of community-acquired pneumonia caused by Streptococcus pneumoniae, Haemophilus infl uenzae and Moraxella catarrhalis. J Antimicrob Chemother 1996; 38 (suppl A):133-140.

(29) File TM, Jr., Jacobs MR, Poole MD, Wynne B. Outcome of treatment of respiratory tract infections due to Streptococcus pneumoniae, including drug-resistant strains, with pharmacokinetically enhanced amoxycillin/clavulanate. Int J Antimicrob Agents 2002; 20(4):235-247.



112

Chapter 1Chapter 6

(32) ERS Task Force Report. Guidelines for management of adult community- acquired lower respiratory tract infections. European Respiratory Society. Eur Respir J 1998; 11(4):986-991.

(33) van Kasteren ME, Wijnands WJ, Stobberingh EE, Janknegt R, van der Meer JW. [Optimizing of antibiotics policy in the Netherlands. III. SWAB guidelines for antimicrobial therapy in adults hospitalized with bronchitis. Foundation Antibiotics Policy Work Group]. Ned Tijdschr Geneeskd 1998; 142(46):2512-2515.

(34) McCormick AW, Whitney CG, Farley MM, Lynfi eld R, Harrison LH, Bennett NM et al. Geographic diversity and temporal trends of antimicrobial resistance in Streptococcus pneumoniae in the United States. Nat Med 2003; 9(4):424-430.


(36) Jacobs E. Serological diagnosis of Mycoplasma pneumoniae infections: a critical review of current procedures. Clin Infect Dis 1993; 17 Suppl 1:S79-82.:S79-S82.


(38) Furth R van. Pneumokokkeninfecties en pneumokokkenvaccinaties in de 21e eeuw. Ned Tijdsch Med Microbiol 2000; 8:4-8.


(40) Neuhauser MM, Weinstein RA, Rydman R, Danziger LH, Karam G, Quinn JP. Antibiotic resistance among gram-negative bacilli in US intensive care units: implications for fl uoroquinolone use. JAMA 2003; 289(7):885-888.

(41) Oosterheert JJ, Bonten MJ, Hak E, Schneider MM, Hoepelman IM. How good is the evidence for the recommended empirical antimicrobial treatment of patients hospitalized because of community-acquired pneumonia? A systematic review. J Antimicrob Chemother 2003; 52(4):555-563.

(42) Sackett DL. Rules of evidence and clinical recommendations on the use of antithrombotic agents. Chest 1989; 95(2 Suppl):2S-4S.

(43) Burgess DS, Lewis JS. Effect of macrolides as part of initial empiric therapy on medical outcomes for hospitalized patients with community-acquired pneumonia. Clin Ther 2000; 22(7):872-878.

113

Chapter 6

(44) Frei CR, Koeller JM, Burgess DS, Talbert RL, Johnsrud MT. Impact of atypical coverage for patients with community-acquired pneumonia managed on the medical ward: results from the United States Community-Acquired Pneumonia Project. Pharmacotherapy 2003; 23(9):1167-1174.

(45) Tuuminen T, Palomaki P, Paavonen J. The use of serologic tests for the diagnosis of chlamydial infections. J Microbiol Methods 2000; 42(3):265-279.

(46) Birtles RJ, Harrison TG, Samuel D, Taylor AG. Evaluation of urinary antigen ELISA for diagnosing Legionella pneumophila serogroup 1 infection. J Clin Pathol 1990; 43(8):685-690.

(47) Dowell SF, Garman RL, Liu G, Levine OS, Yang YH. Evaluation of Binax NOW, an assay for the detection of pneumococcal antigen in urine samples, performed among pediatric patients. Clin Infect Dis 2001; 32(5):824-825.

(48) Gutierrez F, Masia M, Rodriguez JC, Ayelo A, Soldan B, Cebrian L et al. Evaluation of the Immunochromatographic Binax NOW Assay for Detection of Streptococcus pneumoniae Urinary Antigen in a Prospective Study of Community-Acquired Pneumonia in Spain. Clin Infect Dis 2003; 36(3):286-292.


(50) Roson B, Fernandez-Sabe N, Carratala J, Verdaguer R, Dorca J, Manresa F et al. Contribution of a urinary antigen assay (Binax NOW) to the early diagnosis of pneumococcal pneumonia. Clin Infect Dis 2004; 38(2):222-226.



(53) Marrie TJ, Lau CY, Wheeler SL, Wong CJ, Vandervoort MK, Feagan BG. A controlled trial of a critical pathway for treatment of community- acquired pneumonia. CAPITAL Study Investigators. Community-Acquired Pneumonia Intervention Trial Assessing Levofl oxacin. JAMA 2000; 283(6):749-755.

114

Chapter 1Chapter 6

(54) Halm EA, Atlas SJ, Borowsky LH, Benzer TI, Metlay JP, Chang YC et al. Understanding physician adherence with a pneumonia practice guideline: effects of patient, system, and physician factors. Arch Intern Med 2000; 160(1):98-104.

(55) Lim WS, Macfarlane JT. Importance of severity of illness assessment in management of lower respiratory infections. Curr Opin Infect Dis 2004; 17(2):121-125.

(56) Menendez R, Ferrando D, Valles JM, Vallterra J. Infl uence of deviation from guidelines on the outcome of community-acquired pneumonia. Chest 2002; 122(2):612-617.

7An algorithm to determine

cost-savings of targeting

antimicrobial therapy based

on the results of rapid

diagnostic testing

J Clin Microbiol. 2003: 41 (10): 4708-13

JJ Oosterheert, MJM Bonten, E Buskens, MME Schneider, IM Hoepelman

116

Chapter 1Chapter 7

Abstract

A rapid diagnosis of pneumococcal pneumonia may allow earlier use of narrow-spectrum antimicrobial therapy. It is however unknown whether rapid diagnostic testing in patients hospitalised with community-acquired pneumonia (CAP) is cost-saving. Therefore, an algorithm to calculate costs associated with diagnosis and treatment of CAP was formulated.Subsequently, the algorithm was applied to clinical data of one hundred and twenty two consecutively admitted patients with CAP, from whom sputum samples were Gram stained and urine was tested for Streptococcus pneumoniae antigen. Costs of initial antimicrobial therapy, personnel and materials were measured. When compared to the most expensive empiric regimen, rapid diagnostic testing would result in a cost saving per patient (PP) of € 3.51 for GS and of € 8.11 for urinary pneumococcal antigen testing. When compared to the cheapest regimen Gram staining would increase cost by € 2.25 PP and urinary antigen testing by € 24.26 PP. In our setting the use of rapid diagnostic testing would not be cost saving. Cost savings depend however on price differences of different antibiotic choices and the proportion of evaluable and positive samples.

117

Chapter 7

Introduction

In addition to therapeutic effi cacy, rising costs have become a major concern in treating patients with serious infections. An important factor associated with high costs in treating these patients is the use of unnecessary broad spectrum antibiotics. Therefore, apart from microbiological and therapeutic considerations, from an economical perspective, strategies to decrease the unnecessary use of these agents are wanted. In an attempt to cover all suspected pathogens, initial antibiotic therapy is often broad spectrum. Results of microbiological investigations can help in targeting antimicrobial therapy to isolated pathogens, an approach known as ‘streamlining’ (1). However, results from diagnostic procedures such as microbiological cultures or serological tests, have a delay of days to weeks and are, therefore, not suitable to guide therapy in an early stage of the disease. Other diagnostic procedures, however, yield almost instantaneous results and could be potentially useful in guiding initial antimicrobial therapy.

In this analysis, we evaluate potential cost-savings associated with the use of rapid diagnostic tests to guide initial antimicrobial therapy in patients hospitalised with community-acquired pneumonia (CAP). As the causative micro-organism cannot be predicted from clinical, laboratory or radiological fi ndings (2;3;4), initial antimicrobial therapy is mostly empiric covering different potential pathogens. Among different pathogens, S. pneumoniae is the most prevalent causative micro-organism, and found in up to 40% of episodes (5-7). Especially in areas with low resistance rates, S. pneumoniae can be adequately treated with narrow spectrum antibiotics, such as penicillin G or amoxicillin instead of more broad spectrum agents like cefriaxone both with or without a macrolide, or levofl oxacin (8). Diagnostic procedures which can be used in the diagnostic workup of CAP that provide results within minutes are sputum Gram staining and antigen testing for pneumococci in urine. Advantages of sputum Gram staining include its wide availability and low costs. However, adequate sputum samples cannot always be obtained, either because there is no sputum production or because samples are not adequate for evaluation. Furthermore, sensitivity and specifi city are unknown,

118

Chapter 1Chapter 7

some bacteria cannot be identifi ed and a uniform defi nition of a positive stain does not exist. (9;10) For these reasons, the use of sputum Gram stain is controversial: its use is recommended by the Infectious Diseases Society of America (IDSA), but not by the American Thoracic Society (ATS) (11;12) Another way to rapidly diagnose pneumococcal pneumonia is urinary antigen testing. An immunochromatographic test, the NOW S. pneumoniae urinary antigen test (Binax, Inc. Portland, Maine) detects the C polysaccharide wall antigen common to all S. pneumoniae strains (13), with results being available within 15 minutes. Preliminary results suggest that this method has an 90 - 100% specifi city and 74% sensitivity (14). Another report showed a lower specifi city: test results could also be positive in patients who are nasopharyngeal carriers of pneumococci. (15)A rapidly established diagnosis of pneumococcal pneumonia may result in cheaper empiric antibiotics. However, it is unknown whether the potential cost savings outweigh the costs for personnel and materials. Therefore, we developed a simple algorithm to assess the potential costs and savings associated with rapid diagnostic testing for pneumococcal pneumonia using sputum Gram stain or a urinary pneumococcal antigen test and evaluated cost-savings in 122 consecutively admitted patients with CAP.

Patients and methods

Patients and setting

The study was approved by the local ethics committee and all patients provided informed consent to participate. The University Medical Centre Utrecht is a 1042 bed tertiary care hospital. The department of medicine consists of 2 general internal medicine wards and 1 ward for acute medicine and infectious diseases. Together they accounted for 3036 admissions in 2001. All patients hospitalised with CAP between November 2000 and November 2002 on internal medicine wards in the University Medical Centre Utrecht with severe CAP (Fine class IV, V (16) or fulfi lling the criteria for ‘severe community-acquired pneumonia’ as defi ned by the American Thoracic Society (11) ) were included. Patients which needed mechanical ventilation in an intensive care unit were not included. Initial therapy, age and

119

Chapter 7

severity of pneumonia as defi ned by Fine (16) were documented. CAP was defi ned as a new or progressive infi ltrate on chest X-ray and two or more of the following criteria: cough, production of purulent sputum, rectal temperature above 38°C or below 36 °C, auscultatory fi ndings consistent with pneumonia, leucocytosis (>10,000/mm) or CRP > 3x upper limit of normal. Patients with cystic fi brosis, neutropenic patients (<0,5 x 109 neutrophils / l) and patients in which another infection needed treatment, were excluded. All patients were encouraged to provide a sputum and urine sample, however, we only used samples provided within 24h of hospitalisation in the analysis.

Microbiological assessment

Sputum samples were evaluated in the microbiological laboratory and Gram stained according to standard techniques. Sputum samples were considered evaluable if no more than 10 squamous epithelial cells and more than 20 neutrophils per low power fi eld were visible, and were considered positive for pneumococci when >10 Gram positive cocci per LPF were present as the predominant organism. To identify pneumococcal urinary antigen we used the NOW® Streptococcus pneumoniae urinary antigen test, provided by Binax Inc., Portland. Serology samples for Chlamydia pneumoniae, C. psittaci, Legionella pneumophila and Mycoplasma pneumoniae, blood cultures and sputum cultures were obtained and evaluated according to standard procedures. In addition, to identify L. pneumophila, we used a urinary antigen test (Binax NOW®).

Cost assessment

Antimicrobial costs were based on the actual cost prices of the antibiotics paid by the department of Clinical Pharmacy of the University Medical Centre in Utrecht. We calculated the potential cost-reduction when therapy would be streamlined based on the results of rapid diagnostic tests before culture results become available. Therefore, only antimicrobial costs for the fi rst three days of therapy were calculated. Duration of preparation and handling of medication were measured twice for all relevant antibiotics. Average costs per antibiotic and per combination of antibiotics for three days

120

Chapter 1Chapter 7

were calculated, which included costs for personnel (nurses wages for the time used for preparation and administration) and used materials (needles, syringes, antibiotics, intravenous solutions etc.), as described previously (17). Diagnostic costs (i.e. costs for preparation and examining Gram stains and performing urinary antigen test) were based on the hospital’s tarif system, which includes wages for personnel and costs for material. To evaluate the potential cost-reduction, we formulated an algorithm.

The algorithm to analyse cost reduction

In the algorithm for cost reduction it was assumed that antimicrobial therapy would be streamlined based on a positive urinary pneumococcal antigen test or a positive Gram stain. Only samples obtained on the fi rst day of hospitalisation and of suffi cient quality for microbiological analysis were used. Patients without a positive test, either because sampling was not performed at all or not performed within 24h hospitalisation, or because test results were negative, received broad-spectrum therapy. The difference in total costs between targeted narrow-spectrum therapy based on positive diagnostic tests and empiric and broad-spectrum therapy was defi ned as cost reduction, which could be expressed as:Cost Reduction = Costs for empirical therapy – (Costs for targeted therapy in patients with positive test results + Costs for empirical therapy in patients with negative test results + Costs of diagnostic procedures)In formula:Cost Reduction = Np * CEmp – { PEv * PPos * CTarg + [Np – (PEv * PPos)] * CEmp + CDx * NTests}With Np = number of patients; CEmp = costs of empirical therapy; PEv = % adequate samples, PPos = % positive adequate samples; CTarg = costs of targeted therapy; ΔP = Price difference between empiric therapy and targeted therapy; CDx = diagnostic costs of the test; Ntests = number of tests that can be performed.Resolution of this equation results in:Cost Reduction = (Np * PEv * PPos * ΔP) - (Ntests * CDx) {1}

121

Chapter 7

Sensitivity analysis

A sensitivity analysis was performed by calculating the cost outcomes when varying the amount of evaluable samples (PEv), the amount of samples positive for pneumococci (PPos) and the price difference between broad spectrum and targeted antibiotics (ΔP).

Results

Patients One hundred and twenty two consecutive patients (84 male) admitted with CAP were evaluated (Np=122). The mean age of the population was 67.20 (Standard deviation (SD) 14.50 years, range 28-96) with a mean Fine score of 110.79 (SD 28.17, range 45-195). 67 (54.9%) fell in PSI risk class IV, 27 patients in risk class V (22.1) and 26 patients (21.7%) fulfi lled the ATS-criteria for ‘severe community-acquired pneumonia’ (11). Initial therapy consisted of amoxicillin / clavulanic acid in 68 (56%) patients, in twelve patients in combination with a macrolide, of ceftriaxone in 48 (39%) patients, in fi ve patients in combination with a macrolide. One patient was switched from amoxicillin / clavulanic acid to erythromycin plus rifampicine as soon as a urinary antigen test indicated Legionella pneumophila infecion, 1 patient received trimethoprim / sulfamethoxazol and 1 patient was treated with ciprofl oxacin.Twenty-eight (23%) patients had received prior antibiotic treatment before hospital admission. Ultimately, a causative agent for CAP was identifi ed in 54 of 122 (44%) patients (Table 1). In another 12 (10%) patients a positive urinary antigen test was the only indicator of S. pneumoniae infection. Sputum samples of 52 patients (43%) (Ntests = 52) were Gram stained during the fi rst day of hospitalisation. Of these 52 Gram stains, 23 (19% of all patients) were evaluable (PEv = 0.19), and in ten samples Gram positive cocci could be identifi ed. However, Gram positive cocci were considered the predominant micro-organism in only seven of these seven samples (7/52, PPos=0.13). In one sample another predominant micro-organism was identifi ed and in two samples multiple pathogens were present. Eighty-fi ve patients (70.0%) (Ntests = 85, PEv = 0.70) provided urine samples, 23 of which

122

Chapter 1Chapter 7

(23/85, PPos = 0.27) were positive for pneumococcal antigen.

Microbiological cause Total no. of patiens (%) Detection with

S. pneumoniae

27 (22.1%)(12 (9.8%) of wich had only positive urinary antigen test)

Sputum cultureBlood cultureUrinary antigen test

9 (7.4%)8 (6.6%)23 (18.9%)

H. infl uenzae 3 (2.5%) Sputum culture

S. aureus 8 (6.6%) Sputum cultureBlood culture

8 (6.6%)2 (1.6%)

K. pneumoniae 3 (2.5%) Sputum culture

M. catharralis 2 (1.6%) Sputum culture 2 (1.6%)

C. pneumoniae 12 (9.8%)

L. pneumophila 3 (2.5%) SerologyUrinary antigen test

3 (2.5%)2 (1.6%)

M. pneumoniae 2 (1.6%) Serology 2 (1.6%)

E. coli 5 (4.1%) Sputum cultureBlood culture

4 (3.3%)1 (0.8%)

Citrobacter freudii, corynebacterium, Streptcoccus oralis, Enterobacter cloacae, P. aeruginosa

1 (0.8%)

No microbiological cause 68 (55.4%)

Multiple pathogens 13 (10.7%)

Table 1

Ultimate microbiological outcome.

Cost calculations

Costs per dosage of antibiotic in our hospital ranged from € 0.80 for penicillin G to € 35.00 for ceftriaxone. Average material costs (needles, syringes, infusion fl uids etc.) per dosage were € 7.51. Average time for preparing and dispensing antibiotics were 4 minutes 25 seconds (ranging from 4 minutes to 4 minutes 50 seconds) per dosage, which would mean an average nurses wage of € 0.89 per dosage prepared. When including the number of dosages per day and preparation

123

Chapter 7

and handling costs amoxicillin was cheapest (€ 34.53 per day) and penicillin G, given 6 times daily, was most expensive (€ 55.23 per day). Calculated costs of combinations of therapy per day were € 91.15 for amoxicillin / clavulanic acid combined with erythromycin and € 95.82 for ceftriaxone combined with erythromycin. (Table 2)The costs for performing a sputum Gram stain were € 2.42 (€ 0.78 material costs and € 1.64 personnel costs). The costs for performing a urinary antigen test for pneumococcal pneumonia were € 21.39 (€ 15.80 material costs and € 5.59 wages for personnel).

Antibiotic Dosis / frequency

Total preparation costs for 3 days*

Drug-cost for 3 days

Total cost for 3 days

Costs per patient per day

Penicillin G 1 milj U q 4h 151.20 14.50 165.70 55.23

Amoxicillin 1000 mg q 8h 75.60 27.98 103.58 34.53

Amoxicillin / clavulanic acid 1200 mg q 8h 75.60 40.60 116.20 38.73

Ceftriaxone 2000 mg q 24h 25.20 105.00 130.20 43.40

Augmentin / erythromycin

1200 mg q 8h / 1000 mg q 8h

151.20 122.26 273.46 91.15

Ceftriaxone / erythromycin

2000 mg q 24h / 1000 mg q 8h

100.80 186.66 287.46 95.82

Augmentin / azithromycin

Ceftriaxone / azithromycin

1200 mg q 8h/ 500 mg q 24h orally

2000 mg q 24h / 500 mg q 24h orally

75.60

25.20

57.56

121.96

133.16

147.16

44.39

49.05

Average costs of antibiotics instituted

137.17 45.72

Table 2

Antibiotic preparation and total costs for different antibiotic regimens in euro.

* preparation costs were measured: average preparation plus administration time was 4’25” per

dosage, wich means nurses wages of € 0.89 per dosage.

average material costs (needles, syringes, intravenous solutions) were € 7.51 per dosage

prepared . Total preparation costs per dosage were on average.

124

Chapter 1Chapter 7

Sputum Gram stain

Using the data of our patient population and formula {1} results in the following algorithm for cost-reduction:(122 * 0.19 * 0.13 * ΔP) – (52 * 2.42) = When targeted therapy consists of amoxicillin, given 3 times daily (€ 103.58) and recommended initial therapy would consist of the most expensive empiric regimen (ceftriaxone and erythromycin, ΔP = € 183.88) cost reduction would be € 428.26 for the total population (€ 3.51 per patient). However, when targeted therapy is compared to the actual average costs for initial therapy as prescribed in our population (ΔP = € 33.55) Gram staining would cost € 24.74 (€ 0.20 per patient)When targeted therapy with Penicillin G, given 6 times daily (€ 165.70), is compared to initial therapy with the most expensive empiric regimen (ceftriaxone and erythromycin, ΔP = € 121.76) Gram staining would result in a total cost saving of € 241.07 (€ 1.98 per patient), but when compared to the cheapest regimen (amoxicillin / clavulanic acid , ΔP = € -49.50), Gram staining would cost € 275.00 (€ 2.25 per patient). When compared to the average costs for therapy as prescribed in our patient population (ΔP = € -28.75) Gram staining would cost € 211.93 (€ 1.74 per patient).

Urinary pneumococcal antigen testing

For urinary testing, formula {1} results in the following algorithm for cost reduction(122 * 0.70 * 0.27 * ΔP) – (85 * 21.39) =When targeted therapy consists of amoxicillin (€ 103.58), given 3 times daily and recommended initial therapy would consist of most expensive empiric regimen (ceftriaxone and erythromycin, ΔP = € 183.88 ) cost reduction is € 2421.76 ( € 19.85 per patient). However, when targeted therapy is compared to the actual average costs for initial therapy as prescribed in our population, (ΔP = € 33.55) urine antigen testing would cost € 1044.55 (€ 8.56 per patient)When therapy with Penicillin G, given 6 times daily (€ 165.70), is compared to the most expensive empiric regimen (ceftriaxone and erythromycin, ΔP = € 121.76) urinary antigen testing would result

125

Chapter 7

in a cost saving of € 989.39 (€ 8.11 per patient). In contrast, when compared to the cheapest regimen (amoxicillin / clavulanic acid, ΔP =-49.50), urinary antigen testing would cost € 2959.52 (€ 24.26 per patient). When compared to the average costs for initial therapy ( ΔP= € 28.57), urinary antigen testing would cost € 2476.91 (€ 20.30 per patient)The cost calculations are displayed in table 3.

Variable Gram stain

Urinary antigen test for pneumococci

Number of patients

Np 122 122

Proportion Evaluable samples PEv 0.19 0.71

Proportion Positive samples PPos 0.13 0.26

Price difference*

When Pencillin G as targeted therapy compared to • Ceftriaxone / erythromycin• Amoxicillin / clavulanic acid• Average antibiotic costs

When Amoxicillin as targeted therapy compared to • Ceftriaxone / erythromycin• Average antibiotic costs

∆P

121.76-49.50-28.75

183.8833.55

121.76-49.50-28.75

183.8833.55

Number of tests performed Ntests 52 85

Costs for diagnostic procedure CDx 2.42 21.39

Cost reduction*

When Penicillin G as targeted therapy compared to

• Ceftriaxone / erythromycin• Amoxicillin / clavulanic acid• Average antibiotic costs

When Amoxicillin as targeted therapy compared to

• Ceftriaxone / erythromycin• Average antibiotic costs

1.98-2.251.75

9.680.85

8.11-24.26-20.30

19.85-8.56

Table 3

Cost calculations

Formula used: Cost Reduction = (Np * PEv * PPos * ∆P) - (Ntests * CDx)

* in € per patient

126

Chapter 1Chapter 7

Sensitivity analysis

Evidently, when the prevalence of samples positive for pneumocci is high (Ppos), when more sputum samples are evaluable (PEv) or when price differences between initial and targeted therapy are greater (ΔP), cost outcome of performing rapid diagnostic tests will be infl uenced. As is clear from the algorithm, the infl uence of Ppos, PEv and ΔP are of equal importance on cost-outcome. A sensitivity analysis was performed by calculating the cost outcomes when varying these parameters. The associations between the price differences in targeted and non-targeted therapy, the proportion of evaluable and positive sputum samples and the resulting cost per patient per day are depicted in Figure 1. For example, when targeted therapy with amoxicillin is compared with the most expensive empiric regimen in our setting, 8.1% of the patients needs to have a positive urinary antigen test to reduce costs.

Figure 1

Sensitivity analysis for sputum Gram stain

Legend:

■ X-axis shows the variation in price difference between broad-spectrum therapy (BST)

and targeted therapy (NST). Negative fi gures mean that NST is more costly than BST.

■ Y-axis shows the cost reduction per patient per day

■ Z-axis shows the proportion of patients with positive test results (Pev*Ppos).

The numbers vary from 2.5% - 17.5%.

127

Chapter 7

Discussion

We have formulated an algorithm to calculate potential cost-savings when using rapid diagnostic testing to target empirical antimicrobial therapy for CAP. The use of Gram staining and urine antigen tests appeared not to reduce health-care associated costs in our situation. The cost reduction is infl uenced by price differences between targeted therapy and non-targeted therapy and the proportion of positive test results. The algorithm provides a means to determine potential cost-savings in any given setting and can also be applied when new rapid diagnostic tests are evaluated.The lack of cost reduction in our setting is explained by the small amount of patients (19%) which were able to provide a useful sputum sample, the small amount of samples (Gram stain 13%; urinary antigen 27%) being positive for pneumococci and the small price-difference of narrow spectrum and broad spectrum therapy. The cost reduction would increase when more samples were evaluable and positive. Reported percentages of adequate and positive samples have ranged from 24-39%, depending on the time interval between admittance and processing of samples and supervision during collection. (18)In addition, when the cost difference of broad spectrum and targeted therapy increases, cost reduction also increases. In settings where empirical therapy is more expensive, streamlining may have a higher impact on cost reduction. In our hospital, recommended empirical treatment for patients hospitalised with CAP consists of monotherapy with a β-lactam agent. Addition of a macrolide is not recommended, unless pneumonia is severe, needing admission to an intensive care unit, or when a strong suspicion of atypical pneumonia exists. (8) Cost reduction will also increase if broad-spectrum therapy would be associated with extra costs, for example due to adverse events.Several scenarios will result in a lower cost reduction than estimated in our study. The possibility of false positive results ((14), for example when staphylococci, although a rare case for community-acquired pneumonia, are falsely identifi ed as streptococci(19) and the inability of Gram stains and urinary antigen tests to identify atypical (co-)infections could result in inappropiate antimicrobial therapy and lower clinical cure rates. Furthermore, when test results have no impact on treatment decisions, cost savings will never be achieved.

128

Chapter 1Chapter 7

The low impact of microbiological investigations on treatment decisions is noted by several authors. (18;20) And evidently, cost reduction will decrease when cost difference of broad spectrum and targeted therapy decreases, for example as a result of once daily dosing regimens of broad-spectrum therapy, instead of regimens containing penicillin G six times daily.This study was designed to investigate the cost-benefi t of streamlining initial antibiotic therapy when using rapid diagnostic tests for pneumococcal pneumonia. Because of this perspective, we did not take in account other possible disadvantages of using unnecessary broad spectrum therapy, such as antimicrobial effectiveness or long term effects of antibiotics on antimicrobial resistance. Whether targeted therapy is more effective than broad-spectrum therapy is yet unclear. Recent analyses have suggested that initial therapy with a β-lactam and a macrolide antibiotic increases survival in CAP, even when pneumococci are causative micro-organisms (21;22;23;24;25;26). From this point of view, early recognition of a causative micro-organism would not be benefi cial. However, these studies are retrospective and possible subject to prescription bias, showed inconsistencies in reported outcomes and provided no data whether targeted therapy based on the results of microbiological investigations infl uenced patient outcomes. (27;28) In addition, unnecessary use of broad spectrum antibacterial agents enhances induction of antimicrobial resistance. In theory, fi nancial investment in methods allowing rapid streamlining of antibiotic therapy may outweigh future costs associated with treatment of less suspectible micro-organisms. In conclusion, we showed that using sputum Gram stain or urinary antigen test to streamline initial therapy in patients hospitalised with CAP, would not be associated with cost savings in our setting. However, clinical effi cacy of different antibiotics and long-term effects on antimicrobial susceptibility were not included. Moreover, differences in costs of empirical treatment, and the proportion of evaluable and positive tests may lead to different amounts of cost-reduction. Our algorithm is an easy tool to calculate such cost-reduction.

129

Chapter 7

Reference List

(1) Kuijper EJ, Van Der MJ, De Jong MD, Speelman P, Dankert J. Usefulness of gram stain for diagnosis of lower respiratory tract infection or urinary tract infection and as an aid in guiding treatment. Eur J Clin Microbiol Infect Dis 2003; 22(4):228-234.








(9) Reed WW, Byrd GS, Gates RH Jr, et al. Sputum Gram’s stain in community acquired pneumonia: A meta analysis. West J Med 1996; 165:197.

(10) Smith PR. What diagnostic tests are needed for community-acquired pneumonia. Medical Clinics of North America 2001; 85(6):1381-1396.

130

Chapter 1Chapter 7



(13) Skov Sorensen UB, Henrichsen J. Cross-reactions between pneumococci and other streptococci due to C polysaccaride and F antigen. J Clin Microbiol 1987; 25:1854-1859.


(15) Dowell SF, Garman RL, Liu G, Levine OS, Yang YH. Evaluation of Binax NOW, an assay for the detection of pneumococcal antigen in urine samples, performed among pediatric patients. Clin Infect Dis 2001; 32(5):824-825.


(17) Hoepelman IM, Rozenberg-Arska M, Verhoef J. Comparison of once daily ceftriaxone with gentamicin plus cefuroxime for treatment of serious bacterial infections. Lancet 1988; 1(8598):1305-1309.

(18) Ewig S, Schlochtermeier M, Goke N, Niederman MS. Applying sputum as a diagnostic tool in pneumonia: limited yield, minimal impact on treatment decisions. Chest 2002; 121(5):1486-1492.

(19) Dominguez J, Gali N, Blanco S, Pedroso P, Prat C, Matas L et al. Detection of streptococcus pneumoniae antigen by a rapid immunochromatographic assay in urine samples. Chest 2001; 119(1):243-249.

(20) Waterer GW, Jennings SG, Wunderink RG. The impact of blood cultures on antibiotic therapy in pneumococcal pneumonia. Chest 1999; 116:1278-1281.


(22) Gleason PP, Meehan TP, Fine JM, Galusha DH, Fine MJ. Associations between initial antimicrobial therapy and medical outcomes for

131

Chapter 7

hospitalized elderly patients with pneumonia. Arch Intern Med 1999; 159:2562-2572.





(27) Macfarlane J. Severe pneumonia and a second antibiotic. Lancet 2002; 359(9313):1170-1172.

(28) Dowell SF. The best treatment for pneumonia. New clues but no defi nitive answers. Arch Intern Med 1999; 159:2511-2512.

132

Chapter 1Chapter 7

8Impact of rapid viral diagnosis

by real-time PCR in patients

with lower respiratory tract

infectionsSubmitted

JJ Oosterheert, AM van Loon, R Schuurman, IM Hoepelman,E Hak, S Thijsen, G Nossent, MME Schneider, WMN Hustinx,MJM Bonten

134

Chapter 1Chapter 8

Abstract

Rationale: Rapid diagnostic tests with high sensitivity for lower respiratory tract infection (LRTI) could lead to improved patient care and reduce unnecessary antibiotic use and associated costs. Objectives: Evaluate the feasibility and clinical and economic impact of the use of rapid Taqman-PCR for diagnosis of respiratory viruses and atypical pathogens in patients hospitalized with LRTIMethods: and measurements In a multi-center randomized trial, real-time PCRs for respiratory viruses and atypical pathogens were performed immediately on admission in addition to conventional diagnostic procedures. Real-time PCR results of patients in the intervention group were reported to the treating physician, results of patients in the control group were not available. Diagnostic yields, feasibility and costs of real-time PCR in routine diagnostic work-up of LRTI were determined. Main results: 107 consecutive patients (mean age 63.6 ± 16.3 y) admitted to hospital for antibiotic treatment of LRTI were included. Most frequently detected pathogens were Infl uenzavirus (n=14), S. pneumoniae (n=8), Coronavirus (n=6), S. aureus (n=5) and Rhinoviruses (n=5). Real time PCR increased the diagnostic yield from 23 (21%) to 47 (43%) compared to conventional diagnostic tests. The presence of viral pathogens detected by PCR was associated with winter season, less infi ltrates on chest X-ray, lower CRP levels and shorter duration of complaints. Reporting of real-time PCR results resulted in partial or total cessation of antibiotic treatment in 6 patients (11%, 95% CI: 2-19), but overall antibiotic use was comparable in the intervention and the control group (12.3 vs. 10.3 d). Application of real-time PCR increased treatment and diagnostic costs with € 318.17 per patient. Conclusions: Implementation of real-time PCR for the etiological diagnosis of LRTI increased the diagnostic yield considerably, but did not reduce antibiotic use or costs.

135

Chapter 8

Introduction

The health and economic burden from community-acquired lower respiratory tract infections (LRTI) such as community-acquired pneumonia (CAP) and exacerbations of chronic obstructive pulmonary disease is large and continues to increase. (1-4) Empirical antimicrobial therapy of LRTI is based on the expected etiological role of bacteria such as S. pneumoniae, H. Infl uenzae, S. aureus, L. pneumophila, C. pneumonia and M. pneumoniae. (5-7) Due to improved sensitivity of laboratory techniques, respiratory viruses recently have been identifi ed as more frequent and important causes of severe LRTI than previously assumed. (8-12) Rapid identifi cation of a viral etiology of LRTI may improve effective patient management by withholding antibiotic treatment, institution of antiviral therapy or implementation of infection control measures to prevent transmission. (13) Rapid assessment of viral and bacterial etiology is now possible with novel sensitive and highly specifi c Taqman-based real-time PCRs. (10;14) However, these diagnostic tests are costly and the diagnostic yield and feasibility of implementation in routine diagnostic work-up has not been evaluated suffi ciently. Therefore, a randomized controlled trial to evaluate feasibility and clinical and economic impact of the use of rapid Taqman-PCR for diagnosis of respiratory viruses and atypical pathogens in patients hospitalized with LRTI was conducted.

Material and Methods

Setting and study population

A multicenter, randomized clinical trial was conducted in a 1042-bed university hospital and a 627-bed teaching hospital. The trial included patients with LRTI who were referred to one of these hospitals by their general practitioner (GPs).Between November 2002 and March 2004, all consecutive patients aged 18 or older who were admitted to one of the participating hospitals and who needed immediate antimicrobial treatment for LRTI were eligible for inclusion. LRTI was defi ned as two of the following: onset or increase of cough, sputum production, shortness of breath,

136

Chapter 1Chapter 8

wheezing, chest pain, focal or diffuse signs on chest examination, and one constitutional symptom including fever, confusion, sweating, headaches, leucocytosis. (15) Pneumonia was defi ned as a LRTI with a new or progressive infi ltrate on chest x-ray. Patients with severe immunosuppression (neutropenia < 0.5 x 109 neutrophils / liter; CD4-count < 200/mm3), concurrent non-respiratory infection needing antibiotic treatment, severe LRTI needing treatment in an intensive care unit and who were expected not to comply with the study procedures were excluded. The study was approved by the medical ethical committees of both hospitals and all patients provided written informed consent.

Study procedures

Patients were included within 24 hours after admission to the hospital. Demographic data, information on previous treatments, duration of symptoms and presence of high-risk co-morbidity were collected, laboratory tests (white blood cell counts (WBC) and C-reactive proteine (CRP)), and a chest X-ray were performed.

Pathogen detection

Sputum, when available, and blood samples were cultured and processed following standard microbiological procedures. An urinary antigen assay was used for detection of Legionella pneumophila. (Binax Now®, Portland, Maine). Within 24h of admission, a nose and throat swab was collected, and transported immediately in viral transport medium to the laboratory. After vortexing for 10 s and centrifugation at 2,000xg for 15 min., the supernatant was used for virus isolation and real-time PCR. For virus isolation, conventional as well as shell-vial cultures of tertiary monkey kidney cells and of human diploid fi broblast cells were inoculated with 0.1 ml of clinical specimen and incubated for a maximum of 14 days. Conventional cultures were examined twice weekly for the development of a cytopathological effect (CPE). In positive cultures, virus was identifi ed by immunofl uorescence with monoclonal antibodies to Adenoviruses, Infl uenzaviruses A and B, Respiratory syncytial virus and Parainfl uenza viruses 1-3 (DAKO

137

Chapter 8

Imagen). Rhinoviruses were identifi ed by acid-lability testing. In shell-vial cultures an immunofl uorescence test using the above-mentioned monoclonal antibodies was performed after two days of culture, irrespective of the development of a CPE.

Identifi cation of respiratory pathogens by real-time PCR analysis

Detection of viral and atypical pathogens was performed in parallel, using real-time PCR assays specifi c for: Infl uenzavirus A and B, Respiratory Syncitial Virus A and B, Coronavirus OC43 and 229E, Para-infl uenzavirus 1 to 4, Rhinoviruses, pan-Adenoviruses, Mycoplasma pneumoniae, Chlamydia pneumoniae and Legionella pneumophila. Real-time PCR procedures were performed as described previously (10;14;16) Briefl y, after addition of generic internal virus controls to monitor for inhibition of extraction and amplifi cation, 200 μl of nose and throat swab supernatant was used for total nucleic acid (RNA plus DNA) extraction with HighPure Nucleic Acid extraction columns (Roche Diagnostics, Branchburg NJ USA). Subsequently, the purifi ed RNA was used for cDNA synthesis using random hexamers (Applied Biosystems, Foster City, CA, USA). Thereafter, purifi ed DNA and cDNA were used as input in individual real-time PCR reactions and amplifi ed using the Sequence Detection System 7700 (Applied Biosystems, Foster City, CA, USA). Detection of amplifi ed products was performed using pathogen specifi c Fam-labelled TaqMan probes. All pathogens were detected in parallel amplifi cation reactions in one assay run. In addition, positive- and negative controls were included in each run for each individual pathogen. An assay result was validated based on preset Ct-criteria for both the internal control and the positive controls.

Randomization and diagnostic intervention

All patients were randomly allocated to the intervention or the control group by a computer generated table. In the intervention group, results of the real-time Taqman PCR had to be reported as soon as possible, but at least within 48 hours, to the responsible clinicians. To mimic real-life situations, decisions regarding treatment changes

138

Chapter 1Chapter 8

following results of PCR analysis were left at the discretion of the physician. Results of PCR analysis of control patients were not made available for treating physicians. Physicians complied with the hospital guidelines described in the hospital antibiotic formulary. Changes in antibiotic treatment and reasons for changes were recorded.

Intervention group (n=55)

Control group (n=52)

Age (year) 65.3 ± 14.6 61.7 ± 17.9

Men 30 (55%) 27 (52%)

Clinical features

Diagnosis Pneumonia 28 (51%) 27 (52%)

Diagnosis exacerbation COPD 12 (22%) 10 (19%)

Diagnosis other LRTI 15 (27%) 15 (29%)

Temperature (oC) 38.6 ± 0.99 38.5 ± 0.99

Respiratory rate (/min) 23.6 ± 7.3 23.6 ± 6.6

Leucocytes (109/L) 15.9 ± 11.7 14.0 ± 8.3

Lymphocytes (109/L) 9.0 ± 8.2 8.9 ± 8.6

C-reactive protein (mg/L) 145 ± 164.8 133.4 ± 124.8

Cough present 44 (80%) 47 (90%)

Sputum production present 34 (62%) 38 (73%)

Previous antibiotic treatment 23 (42%) 12 (23%)

Fine score 89.8 ± 27.3 86.1 ± 32.5

Duration of LRTI symptoms 4.7 ± 3.8 5.4 ± 6.4

Initial antibiotic treatment

Beta lactam 33 (60%) 38 (73%)

Beta lactam + macrolide 9 (16%) 6 (12%)

Beta lactam + aminoglycoside 3 (5%) 0 (0%)

Fluoroquinolone 4 (7%) 3 (6%)

Macrolide 3 (5%) 2 (4%)

Other 3 (5%) 3 (6%)

Co-morbidities

COPD 18 (33%) 23 (44%)

Diabetes Mellitus 4 (7%) 6 (12%)

Malignancy 11 (20%) 11 (21%)

Table 1

Patient characteristics, expressed in mean ± standard deviation or numbers (%)

Abbreviations:

COPD: chronic obstructive pulmonary disease

LRTI: lower respiratory tract infection

139

Chapter 8

Intervention Group n=55

Control Group n=52

Patients with etiology identifi ed by conventional

diagnostic techniques10 (18) 13 (25)

S. pneumoniae 3 (5) 5 (10)

S. aureus 2 (4) 1 (2)

H. infl uenzae 1 (2) 2 (4)

E. coli 1 (2)

P. aeruginosa 1 (2)

K. pneumoniae 2 (4) 1 (2)

M. catharralis 2 (4)

L. pneumophila 1 (2) 1 (2)

Other 1 (2) 3 (6)

Mixed bacterial etiology* 3 (5) 1 (2)

Patients with etiology identifi ed by of Taq-man based

real-time PCR11 (20) 13 (25)

Infl uenzavirus 7 (13) 7 (13)

Coronavirus 2 (4) 3 (6)

Rhinovirus 2 (4) 1 (2)

Para-infl uenzavirus 1 (2)

RS-virus 1 (2)

Adenovirus 1 (2)

Mixed viral etiology* 1 (2)

Patients with etiology identifi ed by virus culture 10 (18) 6 (12)

Infl uenzavirus 3 (5) 5 (10)

Herpes simplex virus 3 (5)

Enterovirus 4 (7) 1 (2)

Patients with etiology identifi ed by both conventional

methods and real-time PCR3 (5) 3 (6)

Parainfl uenzavirus + H. parainfl uenzae 1 (2)

Rhinovirus + E. coli 1 (2)

Rhinovirus + S. aureus 1 (2)

Infl uenzavirus + coronavirus + H. Infl uenzae 1 (2)

Infl uenzavirus + S. aureus + M. catharralis 1 (2)

Coronavirus + Legionella pneumophila 1 (2)

No Cause 31 (56) 23 (44)

Table 2

Results of etiologic investigations (n (%) )

* All mixed infections are also counted as individual pathogens.

140

Chapter 1Chapter 8

Follow up and outcome measurements

All relevant micro-organisms cultured from blood or sputum samples, detected by urinary antigen testing, virus cultures or real-time PCR were considered a cause of LRTI. Patients were followed for a maximum of 28 days. In-hospital clinical data such as diagnostic procedures performed and antibiotics used were recorded. If patients were discharged within 28 days after admission, clinical outcome and health-care related costs made after discharge were recorded at the out-patient clinic. A change in antibiotic treatment based upon the results of PCR was defi ned as the primary outcome measure.

Cost calculations

The health care perspective was used to calculate costs associated with the diagnostic intervention. Costs per patient were calculated by multiplying the resource use by the unit costs. Costs of medical drugs were determined using Dutch 2003 cost-prices (Netherlands National Health Insurance Council (CVZ, edition 2004). Other costs included hospital stay which was estimated at € 512,- per day and included standard diagnostic procedures. (Manual for Cost Research, CVZ, the Netherlands, 2000) Costs for real-time PCRs including labor costs and costs of reagents, depreciation equipment and overhead costs (20%) were € 330.78 per sample. Other diagnostic costs were based on prices as indicated by the “The Netherlands College for Tariffs in Healthcare (CTG)” in 2004.

Sample size calculation and statistical analysis

To demonstrate a reduction in antibiotic treatment from 100% to 80% (α=0.05; 1-β = 0.80) 100 persons would be needed. Statistical analysis was performed according to the intention-to-treat principle. Differences between the comparison groups for continuous variables were evaluated by means of Student’s T-tests for normally distributed variables and by Mann-Withney-U test for skewed distributed continuous variables. Chi-square tests were used to test for differences in proportions between the two groups in categorical variables.

141

Chapter 8

Results

One hundred and seven patients (mean age 63.6 ± 16.3 years, mean Fine score 88.0 ± 29.9) were included of whom 55 were randomized to the intervention group and 52 to the control group. Baseline characteristics were comparable in both study groups, but there was a tendency that more patients had received previous antibiotic treatment in the intervention group. (table 1)

PCR positiveN=29

PCR negativeN= 78

OR (95% CI) p

Infi ltrate present on chest x-ray 10 (34%) 45 (58%) 0.39 (0.16-0.94) 0.05*

Admitted in winter season 19 (66%) 25 (32%) 4.03 (1.64-9.92) 0.00*

COPD present 14 (48%) 27 (35%) 1.76 (0.74-4.19) 0.25

Sputum production present 21 (72%) 51 (65%) 1.39 (0.54-3.55) 0.64

Cough present 27 (93%) 64 (82%) 2.95 (0.62-13.89) 0.25

Has had previous antibiotic treatment 11 (38%) 24 (31%) 1.43 (0.58-3.51) 0.49

Difference (95% CI)

Duration of complaints 3.79 5.52 1.73 (0.74-3.39) 0.04*

CRP level 89.8 158.7 68.9 (6.78-131.05) 0.03*

Leucocytes 12.4 15.9 3.50 (-0.96-7.97) 0.12

Age 62.5 63.9 1.39 (-5.65-8.44) 0.70

Temperature 38.5 38.6 0.11 (-0.31-0.54) 0.60

Table 3

Patient characteristics associated with positive PCR results: results of multivariate logistic

regression analysis (cont.)

(other tested characteristics included Fine score, respiratory rate, lymphocytes, blood urea level,

serum natrium level, serum glucose level, gender, presence of coronary artery disease, presence

of renal disease, presence of cerebrovascular disease, presence of diabetes mellitus, presence of

malignancy and sore throat present. None were predictive of PCR positivity)

* included in multivariate analysis

Etiology

In the total patient population, an etiological diagnosis could be made in 53 patients (50%). Conventional diagnostic tests (blood and sputum cultures and urinary antigen testing) yielded a potential

142

Chapter 1Chapter 8

pathogen for LRTI in 23 (21%) patients. Real-time PCR increased the diagnostic yield with another 24 (22%) to 47 (43%) patients. In 6 patients (6%), both standard cultures and real-time PCR yielded pathogens. (see table 2). Most frequently detected pathogens were Infl uenzavirus in 16 patients (15%), S. pneumoniae in 8 patients (7%), Coronavirus in 6 patients (6%) and S. aureus in 5 patients (5%). No infections with C. pneumoniae, M. pneumoniae or L. pneumophila were detected. There were no signifi cant differences in etiology in both study groups. Patients with a virus identifi ed by PCR had signifi cantly shorter duration of complaints, lower CRP levels, less frequently infi ltrates on chest X-ray and were admitted more frequently in the winter season than PCR negative patients. (table 3) In multivariate analysis, the best predictive model for prediction of PCR-positivity included all of these features (Area under the curve of receiver operating characteristic curve: 0.74, 95% Confi dence Interval 0.63-0.86).

Feasibility of PCR analysis and impact on treatment decisions

In the intervention group, results of real-time PCR were reported after a mean of 30 (± 13) hours after sampling, and yielded positive results in 14 patients (25%). Based on PCR results antibiotic treatment was modifi ed in 6 patients of the intervention group (11%). In 4 of these six patients, treatment for a possible M. pneumoniae, C. pneumoniae and L. pneumophila infection was discontinued because of negative PCR results. In each of these 4 patients beta-lactam therapy was continued and no defi nitive etiological diagnosis was made. In the 2 other patients, antibiotic treatment was discontinued when PCR was positive for coronavirus and infl uenzavirus. As no treatment adaptations were made in the control group, the relative reduction in completed antibiotic courses was 4% (95% CI: -1% - 9%), and if treatment adaptations with continued beta-lactam treatment are included this is 11% (95% CI: 2% - 19%). All six patients in whom treatment was adapted were clinically cured by day 28. Four patients (7%) were put in barrier isolation when PCR results yielded infl uenzavirus, and two of them (4%) received treatment with a neuramidase inhibitor.

143

Chapter 8

Average resource use

Unit cost in €

Average costs

Intervention group

Control group

Intervention group

Control group

Hospitalization days 9.0 8.9 512 4608 4557

Real time PCR 1 0 331 331 0

Additional diagnostic procedures

CT-thorax / Pulmonalis angiogram 0.07 0.08 164 11.48 13.12

Additional blood-gases 0.11 0.14 4.05 0.44 0.57

Additional blood cultures 0.04 0.12 23.15 0.93 2.78

Additional sputum cultures 0.06 0.08 8.68 0.52 0.69

Spirometry 0.07 0.04 15 1.05 0.60

Bronchoscopy 0.13 0.04 301 39.13 12.04

Total diagnostic procedures 384.55 29.80

Antibiotic treatment days 12.3 10.3 15 184.50 154.50

Total hospitalisation, diagnostic and treatment costs 5177.05 4741.30

Tabel 4

economical outcome

Clinical and economic evaluation

Three patients died in each study group (5.5% in the intervention group, 5.8% in the control group). The mean of duration of antimicrobial treatment was comparable in both study groups: 12.3 ± 8.6 and 10.3 ± 5.7 days in the intervention and control group respectively. (p=0.93) Average costs for antibiotics were €169.80 ± 203.30 per patient. Total antibiotic costs were comparable for both study groups. (see table 4) Treatment with neuramidase inhibitors added € 18.10 per treated patient to the costs in the intervention group. Importantly, the use of real-time PCR had no effect on the length of hospital stay. (9.0 ± 4.8 days in the intervention group, 8.9

144

Chapter 1Chapter 8

± 4.6 days in the control group) and did not reduce extra diagnostic procedures to exclude or confi rm other related diagnoses with comparable numbers of additional cultures, blood-gas analyses, CT-(angio)-scans or bronchoscopies performed in both study groups. Total costs per patient (hospitalization, diagnostic procedures and treatment) were € 5117 in the intervention group and € 4741 in the control group. (table 4)In the intervention group, actual antibiotic cost savings attributed to the reporting of PCR results were € 699.30 compared to continued treatment in these patients. Opposed to these savings, total costs for PCR testing were € 18.193. Thus, if applied to all patients, real time PCR would increase costs on average with € 318.17 per patient, which is 6,3% of treatment and diagnostic costs of patients in the intervention group.

Discussion

Implementation of real-time PCR in the diagnostic work-up of patients hospitalized with community-acquired LRTI increased the etiologic diagnosis from 21% to 43%. However, clinical management of patients hardly changed. Antibiotics were partially or totally discontinued in only 6 patients (11%) and rapid diagnosis of viral LRTI did not reduce hospital length of stay, the number and costs of other diagnostic procedures, antibiotic use or antibiotic costs. Despite advances in healthcare, LRTI is still one of the leading causes of hospital admissions and mortality and is associated with considerable antibiotic use and health-care related costs. For example, CAP is responsible for approximately 500 000 hospitalisations in the United States each year and the annual costs of treating these patients is approximately $9.7 billion. (1;2;17;18) In addition, there is considerable overuse of antibiotics in LRTI especially for viral infections. (19) Unnecessary antibiotic use is regarded as a driving force in the global rise of antibiotic resistance. The clinical value of conventional diagnostic methods as microbiological culturing or Gram staining in guiding treatment of LRTI is limited because of low sensitivity and considerable delay. (20-24) Due to small fractions of positive samples, Gram staining of sputum probably cannot lead to great decreases in antibiotic use or costs. (25) Therefore, improvement of diagnostic

145

Chapter 8

methods and policies to decrease antibiotic use in patients with LRTI are necessary. Recently, a promising method has been evaluated: the use of procalcitonin serum levels in patients with clinical symptoms of LRTI resulted in a signifi cant decrease in antibiotic use without adverse effect on patient outcomes, especially among patients with acute bronchitis. (26) In theory, rapid detection of respiratory viruses might also result in clinical and economical benefi ts. In a before-after study, rapid diagnosis of viral LRTI resulted in a 52% reduction in antibiotic use in children as compared to the preceding year, but no statistical signifi cant fi nancial benefi t was achieved for adult patients in another study with a similar design. (13;27) Drawbacks of these studies were the use of immune fl uorescence techniques which are less sensitive for virus detection in adult patients (28;29) and the use of historical controls for evaluation of costs and outcomes. In addition, comparisons were made only for those patients with viral infections identifi ed. In the present study, the use of sensitive real-time PCRs in all patients admitted with LRTI did not gain economical benefi ts as compared to a contemporary control group. Cost-effectiveness of real-time PCR could be improved by increasing the diagnostic yield or decreasing its costs. Increasing diagnostic yield could be accomplished by adding more pathogens to the test panel, such as the recently discovered human metapneumovirus and coronaviruses (12;30) or by performing real-time PCR on sputum samples rather than on nose-throat swabs. However, only 49% of patients in our study-cohort produced adequate sputum samples. As the duration of symptoms at the time of PCR was inversely related to the likelihood of test positivity, performing real-time PCR earlier in the course of LRTI might also increase the diagnostic yield. In the Netherlands, most patients are fi rst seen by their general practitioner and only those patients with a more severe clinical presentation or not responding to empirical treatment are referred to the hospital. Whether real-time PCR will be cost-benefi cial in outpatient clinics or in general practice populations remains to be determined. Alternatively, costs might be lowered by reducing the numbers of selected pathogens in the test panel or by selecting the patients to be examined. For example, patients admitted in the winter season, with a recent onset of symptoms, low CRP levels and absence of infi ltrates

146

Chapter 1Chapter 8

on chest X-ray were most likely to have viral pathogens detected. Limiting real-time PCR tests to these high-risk patients may increase cost-effectiveness. Finally, physicians were hesitant to discontinue antibiotic treatment even if PCR yielded a virus, possibly because bacterial culture results were available only after PCR results. It is therefore doubtful that more rapid reporting of results, which may be possible in the near future with more expensive automated systems, will lead to a better cost-effectiveness. Real-time PCR might have been more cost-effective if clinicians would have been less reluctant to change clinical management based on test results. Studies with protocolized and more rigorous patient management are needed to address this issue.In conclusion, although rapid detection of respiratory viruses by means of real-time PCR increased the etiologic yield considerably, test results hardly infl uenced clinical management of patients and did not reduce additional diagnostic procedures, antibiotic use, antibiotic costs or length of hospital stay.

Acknowledgements

Financial support: Effi ciency Research Program of the Association of Academic Hospitals and the Dutch Health Insurance Council. Grant no. 01233

147

Chapter 8

Reference List




(4) Pneumonia and infl uenza death rates--United States, 1979-1994. MMWR Morb Mortal Wkly Rep 1995; 44(28):535-537.




(8) Ewig S, Torres A. Severe community-acquired pneumonia. Clin Chest Med 1999; 20(3):575-587.

(9) Vabret A, Mourez T, Gouarin S, Petitjean J, Freymuth F. An outbreak of coronavirus OC43 respiratory infection in Normandy, France. Clin Infect Dis 2003; 36(8):985-989.


(11) de Roux A, Marcos MA, Garcia E, Mensa J, Ewig S, Lode H et al. Viral community-acquired pneumonia in nonimmunocompromised adults.

148

Chapter 1Chapter 8

Chest 2004; 125(4):1343-1351.

(12) Hoek vdL, Pyrc K, Jebbink MF, Vermeulen-Oost W, Berkhout RJ, Wolthers KC et al. Identifi cation of a new human coronavirus. Nat Med 2004; 10(4):368-373.

(13) Barenfanger J, Drake C, Leon N, Mueller T, Trout T. Clinical and fi nancial benefi ts of rapid detection of respiratory viruses: an outcomes study. J Clin Microbiol 2000; 38(8):2824-2828.


(15) Macfarlane J, Prewett J, Rose D, Gard P, Cunningham R, Saikku P et al. Prospective case-control study of role of infection in patients who reconsult after initial antibiotic treatment for lower respiratory tract infection in primary care. BMJ 1997; 315(7117):1206-1210.

(16) van Elden LJ, van Loon AM, van der BA, Hendriksen KA, Hoepelman AI, van Kraaij MG et al. Applicability of a real-time quantitative PCR assay for diagnosis of respiratory syncytial virus infection in immunocompromised adults. J Clin Microbiol 2003; 41(9):4378-4381.


(18) Lave JR, Lin CJ, Fine MJ, Hughes-Cromwick P. The cost of treating patients with community-acquired pneumonia. Seminars in respiratory and critical care medicine 1999; 20(3):189-197.

(19) Ball P, Baquero F, Cars O, File T, Garau J, Klugman K et al. Antibiotic therapy of community respiratory tract infections: strategies for optimal outcomes and minimized resistance emergence. J Antimicrob Chemother 2002; 49(1):31-40.

(20) Waterer GW, Jennings SG, Wunderink RG. The impact of blood cultures on antibiotic therapy in pneumococcal pneumonia. Chest 1999; 116:1278-1281.

(21) Ewig S, Schlochtermeier M, Goke N, Niederman MS. Applying sputum as a diagnostic tool in pneumonia(*) : limited yield, minimal impact on treatment decisions. Chest 2002; 121(5):1486-1492.

(22) Reed WW, Byrd GS, Gates RH Jr, et al. Sputum Gram’s stain in community acquired pneumonia: A meta analysis. West J Med 1996; 165:197.

(23) Smith PR. What diagnostic tests are needed for community-acquired

149

Chapter 8

pneumonia. Medical Clinics of North America 2001; 85(6):1381-1396.

(24) Corbo J, Friedman B, Bijur P, Gallagher EJ. Limited usefulness of initial blood cultures in community acquired pneumonia. Emerg Med J 2004; 21(4):446-448.

(25) Oosterheert JJ, Bonten MJ, Buskens E, Schneider MM, Hoepelman IM. Algorithm to determine cost savings of targeting antimicrobial therapy based on results of rapid diagnostic testing. J Clin Microbiol 2003; 41(10):4708-4713.

(26) Christ-Crain M, Jaccard-Stolz D, Bingisser R, Gencay MM, Huber PR, Tamm M et al. Effect of procalcitonin-guided treatment on antibiotic use and outcome in lower respiratory tract infections: cluster-randomised, single-blinded intervention trial. Lancet 2004; 363(9409):600-607.

(27) Woo PC, Chiu SS, Seto WH, Peiris M. Cost-effectiveness of rapid diagnosis of viral respiratory tract infections in pediatric patients. J Clin Microbiol 1997; 35(6):1579-1581.

(28) Matthey S, Nicholson D, Ruhs S, Alden B, Knock M, Schultz K et al. Rapid detection of respiratory viruses by shell vial culture and direct staining by using pooled and individual monoclonal antibodies. J Clin Microbiol 1992; 30(3):540-544.

(29) Weinberg GA, Erdman DD, Edwards KM, Hall CB, Walker FJ, Griffi n MR et al. Superiority of reverse-transcription polymerase chain reaction to conventional viral culture in the diagnosis of acute respiratory tract infections in children. J Infect Dis 2004; 189(4):706-710.

(30) van den Hoogen BG, de Jong JC, Groen J, Kuiken T, de Groot R, Fouchier RA et al. A newly discovered human pneumovirus isolated from young children with respiratory tract disease. Nat Med 2001; 7(6):719-724.

150

Chapter 1Chapter 8

9Costs and effects of

early switch from IV to

oral treatment in severe

community-acquired

pneumonia: a multi center

randomized trialSubmitted

JJ Oosterheert, MJM Bonten, MME Schneider, E Buskens, JWJ Lammers, WMN Hustinx, MHH Kramer, J Prins, PTJ Slee,K Kaasjager, IM Hoepelman

152

Chapter 1Chapter 9

Abstract

Background: Early conversion of intravenous (IV) to oral antibiotic therapy in the treatment of mild and moderate community-acquired pneumonia (CAP) is assumed to reduce costs of antibiotics and length of hospital stay (LOS). However, effi cacy and economic data from randomized trials are lacking and this strategy has not been evaluated for patients hospitalized with severe CAP (SCAP) Methods: Patients hospitalized in non-ICU wards with SCAP (Fine IV, V, or ATS-criteria for severe CAP) in 5 hospitals in the Netherlands were randomized to either 3 days IV therapy followed by 7 days oral antibiotics (intervention group) when clinically stable at day 3 (respiratory rate <25 /min, hemodynamically stable, resolution of fever, no mental confusion and able to take oral medication) or to 7 days IV antibiotics (control group). Choices of antibiotics refl ected local treatment policies. Clinical cure and treatment costs both inside and outside the hospital were evaluated. Analyses were performed according to the intention-to-treat principle. Results: 302 patients were randomized (mean age 69.5 ± 14.0, mean Fine score 112.7 ± 26.0), 150 in the intervention group and 152 in the control group. In the intervention group, 106 (71%) were eligible for the IV to oral antibiotic switch at day 3. Mortality at day 28 was 4 % in the intervention group and 6% in the control group (mean difference: -2%, 95% CI -8 – -3%). Clinical cure at day 28 was 83% in the intervention group and 85% in the control group (mean difference -1.7, 95% CI –7 – 10%) . Duration of IV therapy and length of hospital stay were reduced in the intervention group, with mean differences of 3.4 days (95% CI 2.8 – 3.9 days) and 1.9 days (95% CI 0.6 – 3.2 days), respectively. Calculated cost-reductions with early switch therapy were € 680 (95% CI: € 1180 – €190) per patient (including € 118 for antibiotics and € 542 for hospital stay)Conclusions: Early conversion from IV to PO antibiotic in patients hospitalized with severe CAP is safe, decreased length of hospital stay by 2 days and reduced treatment costs by € 680 (17%).

153

Chapter 9

Introduction

Community-acquired pneumonia (CAP) is a common and potentially fatal infection with high healthcare costs. (1-3) Initial antibiotic treatment for patients needing hospitalization usually is administered intravenously to warrant optimal tissue levels. The duration of intravenous therapy is an important determinant of length of hospital stay (LOS) for these patients. (4) In conventional treatment approaches, intravenous therapy is continued until defi nite clinical cure has been achieved. A switch to treatment with oral antibiotics may allow early discharge, reduce drug costs and decrease risk of complications associated with the use of intravenous catheters. Possible disadvantages of this strategy are treatment failures due to low compliance or inadequate absorption of oral medication, resulting in relapses or reinfections, readmissions or deaths. Furthermore, early discharge may lead to increased burden for family-members or health-care professionals outside the hospital. The concept of early transition from intravenous to oral antibiotics in the treatment of CAP has been evaluated before, though only in patients with mild to moderately severe CAP, rarely in a randomised design and always without cost-effectiveness analyses. (4-14) For patients with severe CAP, effects of early switch approach on outcome, LOS, treatment costs and healthcare resource use have not been determined in randomized trials. In recent European and North-American treatment guidelines there are no rigid recommendations of the timing of switch to oral treatment in severe CAP, mainly because the only available evidence is coming from non-randomized studies among patients with mild episodes of CAP and low risks of complications. (15-19) Therefore, a multi-center randomized trial was conducted to evaluate the cost-effectiveness of an early intravenous to oral switch strategy compared to a 7 day intravenous treatment regimen in patients with severe CAP.

Patients and Methods

Study design

A multi-center, randomized open label clinical trial was performed in 2

154

Chapter 1Chapter 9

university medical centers and 5 teaching hospitals in the Netherlands. The study was approved by the medical ethics committees of all participating hospitals and all patients provided written informed consent prior to enrollment.After inclusion in the trial, treatment allocation was established through an independent central randomization center. Stratifi ed randomization per center was based on computer generated tables. Patients were either randomized to the intervention group, where clinically stable patients were switched from IV to oral antibiotics on the 3rd day of hospitalization to complete a total of 10 days antibiotic treatment, or to the control group, who received a standard regimen of 7 days of intravenous therapy. Clinical stability was defi ned as respiratory rate < 25/min, O2 saturation > 90% or arterial pO2>55mmHg, hemodynamically stable, > 1oC decrease in temperature in case of fever, absent mental confusion and the ability to take medication orally (10). Antibiotic choices were left to the discretion of the attending consultant and were based on Dutch treatment guidelines (20).

Patients

Adult patients (age 18 or above) with severe CAP admitted to general hospital wards were eligible for inclusion in the study. Pneumonia was defi ned as a new or progressive infi ltrate on a chest X-ray plus at least two of the following criteria: cough, sputum production, rectal temperature > 38oC or < 36.1oC, auscultatory fi ndings consistent with pneumonia, leucocytosis (>10.000/mm3, or >15% bands), C-reactive protein > 3 times the upper limit of normal, positive blood culture or positive culture of pleural fl uid. (21) Severe pneumonia was defi ned as Fine class IV or V or fulfi lling the ATS-criteria for severe community-acquired pneumonia. (16;22) Patients with cystic fi brosis, a history of colonization with Gram negative bacteria due to structural damage to the respiratory tract, malfunction of the digestive tract, life expectancy of less than 1 month due to underlying diseases, infections other than pneumonia needing antibiotic treatment, severe immunosuppression (neutropenia (<0,5 x 109 / l) or a CD4 count < 200 / mm3) and needing mechanical ventilation in an intensive care unit were excluded.

155

Chapter 9

Severe CAPN=302

133standard care(7 days IV therapy)

Excluded from analysis: 37

� 9 other diagnosis (m. wegener, urosepsis, tuberculosis,cholangitis, pulmonary embolism, sinusitis, nefritis, urinary tract infection,pneumocystis carinii pneumonia)

� 4 protocol violation� 7 deaths before d3� 6 admitted to intensive care before day 3� 11 withdrawn consent

265 randomized

132switch therapy protocol(3 days IV therapy)

Figure 1

Flow of patients throughout the trial

Baseline, follow up and outcome measurements

Patients were followed for a maximum of 28 days. On admission, a complete physical examination, chest radiography and blood sampling for arterial blood gas analysis and for hematologic and biochemical analysis were performed. Demographic data, initial intravenous therapy and clinical data were recorded. During the follow-up period, in-hospital clinical data were recorded. Clinical stability was evaluated after 3 days of intravenous therapy in both study groups and discharge criteria (temperature < 37.8 oC, saturation >92%, normal blood pressure, heart rate < 100/min, respiratory rate < 25/min, absence of mental confusion and ability to take medication orally) were evaluated daily thereafter. If patients were discharged within 28 days after admission they were asked to return to the out-patient clinic 28 days after inclusion, where history, physical examination, blood chemistry analysis and a chest X-ray were taken and health-care consumption after discharge was recorded. Questionnaires were used to evaluate the effects on health care resource use outside the hospital (recorded daily after discharge). Quality of life was determined with the short form health survey questionnaire (www.sf-36.org) recorded at days 3, 10 and 28 of the

156

Chapter 1Chapter 9

study. Additional questionnaires were used to measure the effects of an early discharge on the workload for family members (questionnaire at day 28 of study) and effects of administration route on freedom of movement, adverse events and compliance (questionnaire at day 7 of study). Treatment failures were defi ned as death, clinical deterioration (clinical worsening such that the patient needed mechanical ventilation, re-administration of intravenous antibiotics after a switch to oral therapy, readmission for pulmonary reinfection after discharge, increase in temperature after initial improvement) or hospitalization at day 28 of the study. Clinical cure was defi ned as discharged in good health without signs and symptoms of pneumonia and no treatment failures during the follow-up period. (21)

Microbiological analyses

Sputum samples and blood samples were collected, cultured and evaluated following standard procedures. Micro-organisms cultured in blood or sputum were recorded. In addition, Binax NOW-tests were used to detect urinary antigen for Legionella pneumophila and S. pneumoniae. Acute and convalescent serology samples were collected and evaluated for Mycoplasma pneumoniae, L. pneumophila and Chlamydia pneumonia. Any non-contaminating micro-organism cultured from a blood or sputum sample or detected by urinary antigen testing was considered a cause for the episode of pneumonia. For Mycoplasma pneumoniae, a fourfold or greater increase in titer in paired sera or a single titer of greater than or equal to 1:40 was considered indicative of infection. (23) (Immune fl uorescence agglutination, Serodia-MycoII ®, Fujirebio, inc.) For Legionella pneumophila, a fourfold increase in the antibody titer to 1:128 or greater, or single titers of 1:256 or more were considered suggestive of Legionella pneumonia.(24) For Chlamydia pneumoniae, detection of IgM above established values, seroconversion of IgG between acute and convalescence samples, high amounts of IgG in single titers or a combination of these factors were considered serological evidence of infection, according to the manufacturers instructions. (ELISA, Savyon Diagnostics Ltd)

157

Chapter 9

Intervention Group n=150

Control Groupn=152

Patient characteristicsMen 102 (68%) 97 (64%)

Age (y) 69.9 ± 13.8 69.0 ± 14.2

Nursing home patients 7 (5%) 5 (3%)

Fine score 111.6 ± 26.3 113.7 ± 25.8

Fine classIIIIIIVV

11 (7%)14 (9%) 93 (62%)32 (21%)

7 (5%)7 (11%)111 (73%)23 (15%)

Leukocytes (109 / l ) 17.2 ± 9.9 15.7 ± 8.5

C-Reactive protein (mg/l) 209.1 ± 151.6 199.2 ± 151.4

Heart Rate (/min) 103.3 ± 22.0 108.2 ± 23.0

Respiratory Rate (/min) 26.5 ± 8.7 26.6 ± 8.5

Temperature (oC) 38.5 ± 1.2 38.6 ±1.2

Oxygen saturation (%) 93.3 ± 5.3 92.1 ± 8.3

Po2 (mm Hg) 68.4 ± 20.9 67.3 ± 23.1

Presenting symptoms

Myalgia 43 (30%) 44 (29%)

Nausea 34 (23% 40 (27%)

Diarhoea 14 (10%) 25 (17%)

Headache 30 (21%) 39 (26%)

Dyspnea 129 (87%) 131 (87%)

Chest pain 54 (36%) 47 (31%)

Sore throat 12 (8%) 16 (11%)

Productive Cough 94 (63%) 90 (60%)

Hemoptoe 16 (11%) 18 (12%)

Confusion 34 (23%) 43 (29%)

Fever 100 (67%) 93 (61%)

ComorbiditiesNeoplasmLiver diseaseHeart failureCerebrovascular diseaseRenal disease

32 (21%)0 (0%)20 (13%)11 (7%)16 (10%)

35 (23%)3 (2%)17 (11%)16 (11%)46 (30%)

Initial therapyAmoxicillin ± clavulanic acidCephalosporin (2nd and 3rd gen)FluoroquinoloneAmoxi ± ca + macrolideCeph + macrolideOther

90 (60%)28 (20 %)0 (0%)15 (10%)4 (3%)13 (9%)

84 (55%)31 (20%)1 (1%)11 (7%)8 (5%)17 (11%)

Table 1

Patient baseline characteristics

158

Chapter 1Chapter 9

Economic evaluation

The societal perspective was used to calculate direct medical costs associated with the treatment in both study groups. Costs were assessed in 2002 euros. Costs per patient were calculated by multiplying resource use by the unit costs. Resource use during hospital stay was measured for diagnostic and therapeutic interventions, consultations of medical or paramedic specialists and antibiotic use. Resource use outside the hospital was evaluated using questionnaires which recorded contacts with general practitioners, specialists, extra diagnostic procedures, use of medication, readmissions, home care and other disease-related costs. Additional costs associated with specifi c diagnostic tests were based on tariffs. Costs of hospital stay, diagnostic procedures and specialist consultations, were calculated using unit costs as determined within the realm of the Dutch guidelines for pharmaco-economic analyses (25). Costs for readmissions and reinfections were assessed specifi cally for the study cohort and included costs for extra diagnostic procedures, treatment and hospitalisations. Costs of antibiotics prescribed were estimated using Dutch 2002 formulary cost-prices (26)

Intervention groupN=150

Control groupN=152

S. pneumoniae 29 (19%) 47 (31%)

S. aureus 7 (5%) 5 (3%)

H. infl uenzae 3 (2%) 3 (2%)

M. catharalis 5 (3%) 0 (0%)

C. pneumoniae 8 (5%) 7 (5%)

M. pneumoniae 2 (1%) 6 (4%)

L. pneumophila 3 (2%) 6 (4%)

Other 17 (11%) 24 (16%)

Unknown cause 84 (56%) 71 (47%)

Multiple pathogens 10 (7%) 13 (9%)

Table 2

Identifi ed micro-organisms

159

Chapter 9

Sample size and statistical analysis

To demonstrate equivalence in effi cacy of the two treatment groups, initially, the sample size was set at 250 patients in each study group based on an expected cure rate of 85% in the IV group and acceptance of a 75% cure rate in the switch group. (α=0.05, 2 sided; 1-β=0.80). The absolute difference in cure rate including 95% CI was calculated. Equivalence was rejected if the lower limit of the CI exceeded -10%. Eventually, with less than anticipated enrolment, a 15% lower effectiveness in the intervention group could be excluded with α=0.05 and 1-β=0.80. Analyses were performed on an intention-to-treat basis. Differences in continuous variables were presented as absolute differences with their corresponding 95% CI´s. Dichotomous data were compared with chi-square statistics. Alike cure rate differences in proportions were presented including 95% CI´s. For cost-calculations, arithmetic means of both study groups were compared. The uncertainty surrounding the cost-calculations was evaluated by means of standard bootstrap techniques. (27) Ultimately, the balance between costs and effects was compared for both strategies and incremental costs per therapy failure averted at 28 days were calculated. The uncertainty surrounding the incremental cost-effectiveness ratio was again evaluated by means of bootstrapping.

Results

Characteristics of the patients and treatment assignment

Between July 2000 and March 2004, 302 consecutive patients were randomized: 150 were assigned to receive a standard course of 7 days intravenous treatment and 152 were randomized to the early switch group. (Figure 1) Slower than expected enrollment lead to the inclusion of less than the precalculated 500 patients. Characteristics of clinical symptoms, co-morbidities and severity of infection were comparable in both study groups. More than 80% of patients in the study population were in Fine class IV or V. Most patients received empirical monotherapy with either amoxicillin or amoxicillin with

160

Chapter 1Chapter 9

clavulanic acid (57.6 %) or a cephalosporin (20%). (table 1) Most frequently identifi ed micro-organism was S. pneumoniae (25%, table 2). Atypical pathogens were detected in 11 % of patients. (table 2) Before day 3, 37 (12%) patients were excluded from analysis, leaving 132 patients for analysis in the intervention group and 133 in the control group. Reasons for exclusion before day 3 were rejection of the initial diagnosis of CAP with another diagnosis established (n=9), withdrawn consent (n=11), protocol violation (n=4), need of ICU admission for mechanical ventilation (n=6) or death (n=7). After 3 days of intravenous therapy, 108/132 (81%) patients in the intervention group were switched to oral therapy of whom 102 (94.4%) received amxocillin + clavulanic acid 625 mg q8h. In the control group, 5 patients did not receive intravenous antibiotics for the full length of 7 days due to phlebitis associated with intravenous treatment but these patients were included in the intention to treat analysis.

Clincal outome InterventionN=150

ControlN=152 Difference 95% CI

Mortality after day 3 5 (4%) 8 (6%) 2% -3% - 8%

Clinical cure 110 (83.3%) 113 (85.0) -1.7% -10% - 7%

Clinical failure 22 (16.7%) 20 (15.0%) -1.7% -10% - 7%

Clinical cure but still hospitalized 7 (5%) 6 (5%) 0% -4% - 6%

Clinical deterioration 9 (6%) 8 (6%) 0,1% -5% - 6%

Death 5 (3%) 8 (5%) 2% -3% - 8%

Clinical deterioration+death 14 (9%) 16 (11%) 1% -1% - 6%

LOS 9.6 ± 4.9 11.5 ± 5.0 1.9 days 0.6 – 3.2

Duration of IV therapy 3.6 ± 1.5 7.0 ± 2.0 3.4 days 2.8 – 3.9

Table 3

Clinical outcome

Clinical outcome

Twenty two (16.7%) and 20 (15.0%) patients had treatment failures in the control group and the intervention group respectively (mean

161

Chapter 9

difference -1.7%, 95% CI -10% to 7%) (Table 3). In the intervention group, 7 (5%) patients were still hospitalized at day 28, 9 (7%) had clinical deterioration and 5 (4%) died. In the control group, 6 (5%) patients were still hospitalized, 8 (6%) had clinical deterioration and 8 (6%) had died. Based on the number of patients included, a 15% lower effectiveness of early switch, compared to standard intravenous therapy could be excluded (α=0.05 and 1-β=90).

-2.000-1.500-1.000

-5000

5001.0001.5002.000

-0,25 -0,15 -0,05 0,05 0,15 0,25

Therapy failure at day 28 IV vs SAT

Incr

emen

tal c

ost

sIV

vs

SA

T (€

)

90%

68%

Figure 2

Incremental costs vs. effects of sequential antibiotic therapy vs. intravenous therapy

Note:

90% of data points below x-axis, i.e. 90% certainty with regard to cost savings

68% of data points to the right of the y-axis, i.e. 68% certainty with regard to marginal

difference in effi cacy

Duration of intravenous therapy was signifi cantly shorter in the intervention group (3.6 ± 1.5 vs. d 7.0 ± 2.0 d, mean difference 3.4 days 95% CI 2.8 - 3.9 days) (Figure 3). Average time to meet the discharge criteria was 5.2 ± 2.9 days in the early switch group vs. 5.7 ± 3.1 days in the intravenous group. (mean difference: 0.47 days (95% CI: - 0.3 – 1.2) (table 3) Total lengths of hospital stay (LOS) were 9.6 ± 4.9 and 11.5 ± 5.0 days days for patients randomized to the early switch group and standard group, respectively (mean difference 1.9 days (95% CI: 0.6 – 3.2 days). (table 3) In the subset

162

Chapter 1Chapter 9

of patients without co-morbidities those randomized to the early switch group had an average reduction of total LOS of 3.3 days (95% CI: 5.8 – 1.5), while no reduction in total length of stay was observed among patients with co-morbidities. Patients meeting discharge criteria were not always discharged immediately for several reasons. Incomplete resolution of all clinical criteria of pneumonia or co-morbid illness, the lack of continued care after discharge and the physician’s considerations were the main reasons for continued hospitalization. Overall, patients treated with oral or intravenous antibiotics experienced the same problems with regard to problems in mobility and other side effects. Quality of life of patients treated in both study groups showed no differences in the domains covered by SF-36. (Table 5)

Day of switch to oral therapy

14121086420

Pat

ient

s on

IV th

erap

y

1.2

1.0

.8

.6

.4

.2

0.0

-.2

Treatment group

Iv therapy

Switch group

Figure 3

Fraction of patients on IV therapy in both treatment groups

Economic evaluation

Clinical effi cacy in both treatment groups was considered equal.

163

Chapter 9

Accordingly, initially a cost-minimization analysis was performed.

VolumeUnit costs(€)

Resource use IV-group

Costs (use * unit cost)

Resource use Switch group

Costs(use * unit cost)

Hospital days 285 11.5 3278 9.6 2736

Diagnostic proceduresPleural punctionBronchoscopiesCT-thoraxBroncho Alveolar LavageChest X-ray in follow up

4130116430135

0.090.080.050.020.17

3.6924.088.206.025.95

0.070.040.0500.28

2.8712.048.2009.80

MedicationIV antibioticsPO antibioticsExtra antibiotics

451.401.40

7.00.60.06

3150.840.08

3.64.00.13

1625.600.18

ConsultsFysiotherapistsPulmonary physicianOther specialistGeneral Practicioner

4011011017

1.170.320.360.75

46.8035.2039.6012.75

0.980.290.420.80

39.2031.9046.2013.60

Other care givers Home nursingFysiotherapist at home

2540

00.08

03.20

0.130.10

3.254

Homecare 25 0.48 12 1.25 31.25

Family care 9 0.20 1.80 0.35 3.15

Treatmentfailure associated costsReinfectionReadmissionSwitch back to IVTemperature raiseMechanical ventilation

45515326066028153

0.020.01

0.010.01

9.1015.32

6.0281.53

0.020.020.020.010.01

9.1030.6412.126.0281.53

Total direct medical costs

3918.02 ± 1991

3236.53 ± 2070

Table 4

Resource use and costs

Average total costs were € 3236.53 ± 2070 in the intervention group and € 3918.02 ± 1991 in the control group. (mean difference: € 681 (95% CI € 1176 – € 187, 17% reduction). Hospitalization costs accounted for 85% of total costs and were € 2736 vs. € 3278 in

164

Chapter 1Chapter 9

the intervention and control group, respectively (mean difference: € 542, 17% reduction). Costs for antibiotic treatment were reduced from € 315.92 in the control group to € 167.78 in the intervention group. (mean difference: € 148.14, 47% reduction) However, out-of-hospital costs for treatment and patient care were higher in the intervention group. (€ 35.78 vs. € 61.98, mean difference: € 26.20, 42%) (table 4).

Intervention group

Control group

Avg SD SD difference 95% CI

Day 3

physical functioning 22,7 24,5 19,4 23,1 3,4 -3,4 10,2

social functioning 43,3 30,0 43,8 29,0 -0,5 -9,1 8,2

role emotional 39,3 44,1 41,6 47,2 -2,3 -15,6 11,0

mental health 63,8 24,5 62,6 19,7 1,2 -5,1 7,5

vitality 33,9 19,0 32,7 20,6 1,2 -4,4 6,9

pain 48,0 36,7 51,2 33,0 -3,2 -13,0 6,7

general health 43,9 24,6 40,8 22,8 3,1 -3,7 9,8

health change 68,0 25,4 63,8 24,4 4,2 -2,8 11,2

Day 10


social functioning 57,9 27,3 50,8 29,9 7,0 -2,1 16,2

role emotional 48,5 46,0 48,7 47,0 -0,1 -15,1 14,8

mental health 70,6 20,6 69,0 19,8 1,6 -4,6 7,7

vitality 41,2 22,0 40,1 21,9 1,1 -5,6 7,8

pain 72,9 28,6 65,8 30,9 7,1 -2,0 16,3

general health 44,0 23,4 42,8 22,8 1,2 -5,9 8,2

health change 70,7 22,2 67,5 23,3 3,3 -3,6 10,1

Day 28


social functioning 61,1 27,5 58,2 29,1 3,0 -6,4 12,4

role emotional 53,5 46,7 56,0 47,6 -2,5 -17,6 12,5

mental health 71,9 21,3 70,8 21,0 1,1 -5,6 7,9

vitality 48,3 22,3 49,2 21,4 -1,0 -8,0 6,0

pain 78,4 26,5 70,1 26,9 8,4 0,0 16,7

general health 44,2 25,8 40,9 22,2 3,4 -4,4 11,1

health change 65,4 19,1 61,8 23,6 3,5 -3,2 10,2

Table 5

Quality of life scores (SF 36)

165

Chapter 9

As can be observed in fi gure 2, bootstrap analysis showed that sequential antibiotic therapy reduced costs with a certainty level of 90%, i.e. 90% of the bootstrap replicates indicate cost savings for sequential antibiotic therapy. Similarly, fi gure 2 reveals that early switch might be associated with a marginal difference in effectiveness. (68% certain) The estimated incremental cost-effectiveness ratio of sequential antibiotic therapy as compared to standard care was € 15.500 per prevented therapy failure. This ratio indicates that € 15.500 is saved at the expense of one additional therapy failure. One could also think of this ratio as investments needed to prevent one therapy failure in case standard therapy would be maintained.

Discussion

Patients hospitalized with severe CAP can be managed safely and more effi ciently by an early switch from IV to oral medication. Based on predefi ned criteria, 81% of patients could be switched to oral antibiotics on day 3. Early conversion to oral therapy reduced overall treatment costs by 17% (€ 681), antibiotic costs by 47% (€ 148.14) and length of hospital stay by 1.91 days. Bootstrap analysis showed that € 15.500 needs to be invested to prevent one therapy failure. Patients without co-morbidity benefi ted most from an early conversion to oral antibiotics.This is the fi rst randomized study evaluating costs and effects, both inside and outside the hospital of early switch to oral therapy in patients admitted because of severe community-acquired pneumonia. Our fi ndings appear, at least to some extent, to be generalizable, as the average length of stay in the control group, etiology and mortality rate of 7% are very similar to other cohorts of patients with severe CAP. (8;14) With an estimated annual number of hospitalisations for CAP of 16 000 in the Netherlands with about 40% of them being severe, nationwide implementation of an early IV-oral switch strategy would lead to a cost-reduction of € 4.358.400,- in our country.The results from this study are in line with other reports that suggested the cost-benefi t of early IV-oral switch programs in patients with non-severe CAP. In comparison to these studies, however, the current study does provide sound evidence. Most of the previous studies had a non-randomized design (4;5;9), were restricted to specifi c patient

166

Chapter 1Chapter 9

populations of veterans or moderately ill patients (6;28), had small sample sizes, (8) or patients were switched late in the course of the disease, e.g. after 2-3 consecutive days without fever. (8;14) Moreover, effects on costs made outside the hospital have not been evaluated in any of the previous studies. Our study also has limitations. The number of patients included was lower than calculated prior to study start. Nevertheless, we feel that because the results show only minimal and non-signifi cant differences in therapy failure rates, a more than 10% lower effectiveness of early switch is highly unlikely. Moreover, mortality-rates were even lower in the intervention group. Thus, unfeasibly high patient numbers would be needed to demonstrate a possible mortality disadvantage for patients treated with early switch. Furthermore, the delicate balance between mimicking real-life situations and strict protocol adherence can easily be disturbed in randomized trials in which treatment processes are evaluated. The effects of switch therapy may, therefore, have been overestimated for two reasons. First, in case of clinical stability at day 3, protocol dictated a switch to oral antibiotic therapy. It remains uncertain how many patients would have been switched if the decision had been left to the discretion of the treating physician. With growing confi dence in the safety of an early switch strategy this effect might well decrease, but initially, some reluctance of physicians may be anticipated. Secondly, the minimal length of IV treatment for the control group of 7 days was also dictated by protocol. Obviously, shorter lengths of intravenous treatment in the control group would have decreased the benefi ts of early switch, provided that failure rates would have remained the same. However, controlled trials that specifi cally have addressed the question how long pulmonary infections should be treated are lacking and the optimal duration of treatment of SCAP remains unknown. On the other hand, the effects of switch therapy could also have been underestimated for two reasons. First, discharge was not protocolized and the physician’s considerations for continued hospitalization were important reasons for delayed discharge of clinically stable patients. Again, with growing confi dence in the concept of early switch therapy, this phenomenon may decrease over time, leading to further costs-savings. Although clinical instability on discharge is associated

167

Chapter 9

with adverse clinical outcomes, clinical deterioration after reaching clinical stability is rare. (29) In our study, only 3 (2%) patients were restarted on IV therapy after being switched to oral therapy. Yet, optimal standards for discharge remain to be established. Second, by protocol, a switch to oral therapy was only allowed when patients were clinically stable on day 3. Possibly, patients that are clinically stable before day 3 can be switched earlier, which could further enhance the benefi ts of this strategy. Based on the results of this trial and the evidence mounting from previous non-randomized studies, early conversion to oral antibiotics should be implemented in clinical practice not only for patients with mild or moderately severe episodes of CAP, but also for those with severe CAP that do not need treatment in the intensive care unit.

168

Chapter 1Chapter 9

Reference List




(4) Ramirez JA, Vargas S, Ritter GW, Brier ME, Wright A, Smith S et al. Early switch from intravenous to oral antibiotics and early hospital discharge: a prospective observational study of 200 consecutive patients with community-acquired pneumonia. Arch Intern Med 1999; 159(20):2449-2454.

(5) Rhew DC, Weingarten SR. Achieving a safe and early discharge for patients with community- acquired pneumonia. Med Clin North Am 2001; 85(6):1427-1440.

(6) Siegel RE, Halpern NA, Almenoff PL, Lee A, Cashin R, Greene JG. A prospective randomized study of inpatient iv. antibiotics for community-acquired pneumonia. The optimal duration of therapy. Chest 1996; 110(4):965-971.

(7) Omidvari K, de Boisblanc BP, Karam G, Nelson S, Haponik E, Summer W. Early transition to oral antibiotic therapy for community-acquired pneumonia: duration of therapy, clinical outcomes, and cost analysis. Respir Med 1998; 92(8):1032-1039.

(8) Castro-Guardiola A, Viejo-Rodriguez AL, Soler-Simon S, Armengou-Arxe A, Bisbe-Company V, Penarroja-Matutano G et al. Effi cacy and safety of oral and early-switch therapy for community- acquired pneumonia: a randomized controlled trial. Am J Med 2001; 111(5):367-374.

(9) Ramirez JA, Srinath L, Ahkee S, Huang A, Raff MJ. Early switch from intravenous to oral cephalosporins in the treatment of hospitalized patients with community-acquired pneumonia. Arch Intern Med 1995; 155(12):1273-1276.

(10) Weingarten SR, Riedinger MS, Hobson P, Noah MS, Johnson B. Evaluation of a pneumonia practice guideline in an interventional

169

Chapter 9

trial. Am J Respir Crit Care Med 1996; 153(3):1110-1115.

(11) Rhew DC, Riedinger MS, Sandhu M, Bowers C, Greengold N, Weingarten SR. A prospective, multicenter study of a pneumonia practice guideline. Chest 1998; 114(1):115-119.

(12) Marrie TJ, Lau CY, Wheeler SL, Wong CJ, Vandervoort MK, Feagan BG. A controlled trial of a critical pathway for treatment of community- acquired pneumonia. CAPITAL Study Investigators. Community-Acquired Pneumonia Intervention Trial Assessing Levofl oxacin. JAMA 2000; 283(6):749-755.

(13) Ramirez JA, Bordon J. Early switch from intravenous to oral antibiotics in hospitalized patients with bacteremic community-acquired Streptococcus pneumoniae pneumonia. Arch Intern Med 2001; 161(6):848-850.

(14) van der Eerden MM, de Graaff CS, Vlaspolder F, Bronsveld W, Jansen HM, Boersma WG. Evaluation of an algorithm for switching from IV to PO therapy in clinical practice in patients with community-acquired pneumonia. Clin Ther 2004; 26(2):294-303.







170

Chapter 1Chapter 9



(23) Jacobs E. Serological diagnosis of Mycoplasma pneumoniae infections: a critical review of current procedures. Clin Infect Dis 1993; 17 Suppl 1:S79-82.:S79-S82.


(25) Oostenbrink. Manual for cost research. Amstelveen: College voor Zorgverzekeringen, 2005.

(26) Farmacotherapeutic Guide. 2003 ed. Amstelveen: College voor Zorgverzekeringen, 2003.

(27) Efron B, Thibshirani RJ. An introduction to the bootstrap. New York: Chapman & Hall, 1993.

(28) Siegel RE, Halpern NA, Almenoff PL, Lee A, Cashin R, Greene JG. A prospective randomized study of inpatient iv. antibiotics for community-acquired pneumonia. The optimal duration of therapy. Chest 1996; 110(4):965-971.

(29) Halm EA, Fine MJ, Kapoor WN, Singer DE, Marrie TJ, Siu AL. Instability on hospital discharge and the risk of adverse outcomes in patients with pneumonia. Arch Intern Med 2002; 162(11):1278-1284.

10Summary and general

discussion

172

Chapter 1Chapter 10

General discussion

In addition to therapeutic effi cacy, controlling costs and avoiding unnecessary antibiotic use have become important issues in the treatment of serious infections. Despite advances in health-care and improvements in antibiotic therapies, lower respiratory tract infections have remained an important cause for morbidity and mortality. This is demonstrated in an analysis of pneumonia hospitalizations and deaths in the Netherlands described in chapter 2: due to increases in the number of elderly and presence of co-morbidities, morbidity and pneumonia-related death has increased in the past 10 years with over 50%. Treating episodes of lower respiratory tract infections are associated with considerable costs and high antibiotic consumption, especially in patients needing hospitalization. The aim of this thesis, therefore, was to evaluate diagnostic and treatment strategies to decrease costs and control antibiotic use in patients hospitalized with lower respiratory tract infections.

Early identifi cation of patients with high risk for death or a complicated course of lower respiratory tract infections can guide decisions of how and where to treat patients, thus preventing costs for unnecessary hospitalisations or ICU admissions. However, adequate risk stratifi cation for CAP is diffi cult. In chapter 3 this complexity was demonstrated by discussing frequently used criteria for “severe” CAP. We evaluated whether differentiation based on these defi nitions refl ects variation in etiology or risk-factors and could, therefore, aid in diagnostic approaches and treatment strategies. We concluded that using current defi nitions of SCAP, distribution of pathogens causing non-severe and severe CAP appear to be comparable and extra diagnostic efforts to identify pathogens in SCAP have not shown to improve adequacy of therapy. Therefore, current risk classifi cations are not adequate enough to guide pneumonia management. However, initially, an approach of broader spectrum of antibiotic coverage is frequently used in these patients. This can lead to unnecessary antibiotic use, induction of resistance, and an increase of costs and adverse events. For that reason, in chapter 4, a prediction rule based on routine clinical, biochemical and microbiological information was developed

173

Chapter 10

that can help physicians to determine which patients hospitalized with severe CAP are at risk for early clinical failure. The presence of Gram negative bacteria in sputum culture, mental state change, arterial pH <7.35 and pO2<60 mmHg on admission were independent risk factors for clinical failure to initiated therapy in an analysis of hospitalized CAP patients. Close monitoring of these patients may eventually prevent unnecessary ICU admissions, complications or deaths. Further investigations should address the question whether the use of newer risk classifi cations can indeed improve the quality of pneumonia management as compared to current standards.

The use of antibiotics is the cornerstone of treatment for lower respiratory tract infections, but the optimal initial antibiotic treatment strategy for patients hospitalized because of CAP is not yet clarifi ed. Formerly, monotherapy with a β-lactam antibiotic was considered an adequate empirical therapy for patients hospitalized with CAP. However, recently, clinical guidelines from North-America and Great Britain have recommended a combination of a β-lactam antibiotic plus a macrolide or monotherapy with a fl uoroquinolone as empirical therapy for all hospitalized patients. Putative benefi cial effects are an increased patient survival but possible disadvantages include an increase in antibiotic use and treatment costs. In a systematic review presented in chapter 5, the evidence for this recommendation was evaluated. It turned out that unequivocal evidence of increased patient survival is not available and mainly based on studies in which patient prescription bias may have played a substantial role. In chapter 6 the increase in antibiotic use and treatment costs when introducing this treatment strategy recommended by North American or British guidelines in the Netherlands was quantifi ed. Antibiotic use and treatment costs will increase with over 20% in the total CAP population, increasing up to over 50% in patients with moderate CAP, whereas the adequacy of therapy hardly changes. Therefore, it was concluded that current evidence for the use of combination therapy or treatment with fl uoroquinolones in CAP is not strong enough to justify a place in routine clinical practice and should not be introduced in the Netherlands unless, well-designed, randomized controlled trials provide evidence for increased patient survival or better cost-effectiveness. However, such trials would encounter

174

Chapter 1Chapter 10

diffi culties regarding sample size, selection of endpoints and blinding. Future studies should consider these points carefully. Based on average lengths of stay in our studies, a feasible study design would be a double blind, randomized trail including 125 patients in both treatment groups. Such study could detect a one day difference in length of hospital stay.

The results of rapid microbiological tests can help in targeting the antimicrobial therapy to detected pathogens early in the course of treatment, thereby allowing the discontinuation or streamlining of antibiotics. For example, pneumococcal pneumonia, which can be treated with inexpensive small spectrum penicillins, can be diagnosed upon the results of Gram staining of sputum or urinary antigen testing. In chapter 7 an algorithm was developed to determine cost savings of targeting antimicrobial therapy on the results of rapid diagnostic testing. Variables constituting the algorithm are the prevalence of pneumococcal infections, drug costs and the costs for diagnostic procedures. The algorithm provides a means to determine potential cost-savings in any given setting and can also be applied when new rapid diagnostic tests are evaluated. Application of this algorithm to a population of patients hospitalised with CAP showed that cost-savings will be minimal with the use of Gram staining of sputum or urinary antigen testing to diagnose pneumococcal pneumonia.

In addition to targeting antimicrobials to detected pathogens, cessation of empirically initiated antibiotic treatment may be justifi ed if viruses are identifi ed as causative micro-organisms of respiratory infections. Early identifi cation of viral infection is possible with novel Taq-man real-time PCR techniques. The advantages of these tests are the rapid availability of results, as well as a possible higher sensitivity in detecting organisms causing lower respiratory tract infections as compared to viral culture or serological examinations. Potential disadvantages include increased costs. The impacts on patient care however, had never been determined. In chapter 8 we evaluated the clinical and economical impact of rapid viral diagnosis by Taq-man real-time PCR in patients hospitalized with lower respiratory tract infections in a randomized trial. In this study, the use of real-time PCR indeed increased the diagnostic yield as compared to

175

Chapter 10

routine diagnostic procedures and viral infections were detected in over 20% of LRTI. However, rapid identifi cation of these viruses was not associated with reductions in antibiotic use or treatment costs if applied to all patients admitted with LRTI. Hence, real-time PCR should not become part of the routine work-up for every patient admitted to hospital because of LRTI at this time, unless the cost-effectiveness could be improved by increasing the diagnostic yield, the impact on treatment decisions or decreasing its costs.

Choosing which diagnostic tool can be used in routine diagnostic work-up is, among other arguments, based on cost-effectiveness considerations. As demonstrated, not every promising diagnostic strategy is cost-effective when used in routine clinical practice. The clinical use of rapid diagnostic test was not cost-saving in our studies and hardly led to reductions in antibiotic use. Nevertheless, further development and studies of these rapid diagnostic strategies are justifi ed for several reasons. First, in theory, as unnecessary use of (broad-spectrum) antibiotics enhances the induction of antimicrobial resistance, fi nancial investment in methods that allow rapid streamlining or cessation of antibiotics early in the course of the disease may outweigh the costs associated with the future treatment of infections caused by less susceptible micro-organisms. Second, rapid diagnostic methods may become more cost-effective if clinicians are less reluctant to change clinical management based on results of new rapid diagnostic tests. Therefore, studies with protocolized and more rigorous patient management based on rapid diagnostic tests are needed to evaluate the impact on development of antimicrobial resistance and the costs and effects on patient management.

A frequently suggested strategy to decrease costs of treatment for patients hospitalized because of CAP is an early switch from intravenous to oral therapy. In theory, an early switch to oral antibiotics reduces intravenous drug costs and allows an early discharge from hospital. Potential disadvantages are treatment failures and an increase in treatment costs outside the hospital. Large, prospective randomized trials evaluating the effects of this strategy on outcome of severe CAP and costs both inside and outside the hospital had not been performed. In chapter 9 the fi rst randomized study evaluating costs

176

Chapter 1Chapter 10

and effects, both inside and outside the hospital, of early switch to oral therapy in patients admitted because of severe community-acquired pneumonia was described. As compared to a standard treatment of 7 days intravenous antibiotic therapy, a strategy of early conversion from intravenous to oral treatment in patients hospitalized with severe CAP was associated with comparable clinical outcomes and signifi cantly reduced length of hospital stay and overall costs for treatment. Reductions of lengths of stay were most obvious in patients admitted without co-morbidities. Therefore, the treatment strategy of an early switch to oral antibiotics in patients needing hospitalization because of CAP is proven to be safe and leads to a signifi cantly reduces treatment costs. Future steps towards more effi cient management of CAP may include even shorter courses of antimicrobial treatment, especially in patients without co-morbidities and implementation of early switch strategies in clinical practice.

11Samenvatting in

het Nederlands

178

Chapter 1Samenvatting in het Nederlands

Naast therapeutische effectiviteit zijn kostenbeheersing en het verminderen van onnodig antibioticumgebruik belangrijke aspecten geworden bij de behandeling van ernstige infecties. Ondanks vooruitgang in de gezondheidszorg en verbeteringen in antibiotische behandelingen, zijn lagere luchtweginfecties nog altijd een belangrijke oorzaak voor morbiditeit en mortaliteit. In hoofdstuk 2 is dit aangetoond door middel van een analyse van ziekenhuisopnames en sterfte vanwege pneumonie in Nederland: door toename van de populatie ouderen en de aanwezigheid van co-morbiditeit is de pneumonie-gerelateerde sterfte in de afgelopen 10 jaar met 50% gestegen. Behandeling van deze lagere luchtweginfecties gaat gepaard met hoge kosten en veelvuldig gebruik van antibiotica, vooral bij oudere patiënten die opgenomen worden in het ziekenhuis. Het doel van dit proefschrift was daarom om diagnostische en therapeutische strategieën voor kostenbeheersing en vermindering van antibioticumgebruik te evalueren bij patiënten opgenomen in het ziekenhuis vanwege lagere luchtweginfecties.

Vroege herkenning van patiënten met een hoog risico op een gecompliceerd beloop van lagere luchtweginfecties kan helpen bij beslissingen hoe en waar patiënten te behandelen, en kan zo kosten voor onnodige ziekenhuisopnames of intensive care opnames voorkomen. Adequate risicoclassifi catie voor buiten het ziekenhuis opgelopen pneumonie (community-acquired pneumonia, CAP) is echter moeilijk. In hoofdstuk 3 werd deze complexiteit aangetoond door de meest gebruikte defi nities voor ernstige CAP te bediscussiëren. We evalueerden of onderscheid gebaseerd op deze defi nities ook een verschil in etiologie of risicofactoren betekent en daarom zou kunnen helpen bij de diagnostiek en behandeling van CAP. We concludeerden dat met de huidige defi nities voor ernstige CAP, de verdeling van pathogenen bij ernstige en minder ernstige CAP vergelijkbaar is en dat extra diagnostiek bij ernstige CAP om pathogenen te identifi ceren niet heeft geleid tot betere behandeling. Daarom zijn de huidige risicoclassifi caties niet goed genoeg om de behandeling van pneumonie te sturen. Echter, bij aanvang van de behandeling wordt bij deze patiënten wel vaak een behandeling ingesteld met een breder spectrum aan antibiotica. Dit kan leiden tot onnodig antibioticumgebruik, gepaard met een toename in kosten en

179

Samenvatting in het Nederlands

bijwerkingen en inductie van resistentie. Om die reden is in hoofdstuk 4 een predictieregel ontwikkeld die gebaseerd is op routine klinische, biochemische en microbiologische informatie en artsen kan helpen om te bepalen welke patiënten opgenomen met een ernstige pneumonie risico lopen op een geprotraheerd beloop. De aanwezigheid van Gram negatieve bacteriën in sputum, verwardheid, arteriële pH < 7.35 en een pO2 < 60 mm Hg bij opname waren onafhankelijke risicofactoren voor ongunstig beloop in een analyse van patiënten met CAP opgenomen in het ziekenhuis. Nauwlettend monitoren van deze patiënten kan uiteindelijk leiden tot minder onnodige opnames op de intensive care, minder complicaties en minder sterfgevallen. Vervolgstudies moeten aantonen of deze nieuwere risicoclassifi caties inderdaad de kwaliteit van behandeling van CAP kunnen verbeteren.

Het gebruik van antibiotica is de hoeksteen van de behandeling van lagere luchtweginfecties, maar de optimale initiële antibiotische behandeling voor patiënten die moeten worden opgenomen in het ziekenhuis vanwege CAP is nog niet opgehelderd. Tot voor kort werd monotherapie met een β-lactam antibioticum een adequate initiële behandeling geacht. Recent echter hebben klinische richtlijnen vanuit Noord-Amerika en Groot-Brittannië de combinatie van een β-lactam antibioticum met een macrolide of monotherapie met een fl uoroquinolone geadviseerd als empirische therapie voor alle patiënten opgenomen vanwege CAP. Een mogelijke gunstig effect is een verbeterde overleving voor patiënten, een mogelijk ongunstig effect is een toename in gebruik van antibiotica en behandelingskosten. In een systematische review die gepresenteerd is in hoofdstuk 5 is het bewijs voor deze behandelingsstrategie geëvalueerd. Het bleek dat eenduidig bewijs voor een verbeterde overleving niet aanwezig is en hoofdzakelijk gebaseerd is op studies waarbij “patient prescription bias” een belangrijke rol heeft gespeeld. In hoofdstuk 6 is de toename in antibioticumgebruik en behandelingskosten gekwantifi ceerd als deze door de Noord-Amerikaanse en Britse richtlijnen geadviseerde behandelingsstrategie wordt geïmplementeerd in Nederland. Het antibioticumgebruik en de behandelingskosten zullen met meer dan 20% stijgen in de gehele CAP-populatie, met een toename van meer dan 50% bij patiënten met een matig-ernstige CAP. Daarom

180


is geconcludeerd dat het huidige bewijs voor het gebruik van fl uoroquinolonen of de combinatie van een β-lactam antibioticum met een macrolide nog niet voldoende is om een plaats in de routinematige behandeling van CAP in Nederland te rechtvaardigen en niet moet worden ingevoerd, tenzij goed opgezette gerandomiseerde trials bewijs hebben geleverd voor een verbeterde overleving of een betere kosteneffectiviteit. Echter, de opzet van zulke trials gaat gepaard met moeilijkheden betreffende het aantal patiënten dat geïncludeerd moet worden, het bepalen van de eindpunten en blindering. Toekomstige studies moeten zorgvuldig deze punten overwegen. Gebaseerd op de gemiddelde ligduur van patiënten in onze studies zou een haalbare studie een dubbel blind, gerandomiseerde studie zijn met 125 patiënten in beide studie-armen. Zo’n studie zou één dag verschil in ligduur met voldoende zekerheid kunnen aantonen.

De resultaten van microbiologische sneltesten kunnen ertoe leiden dat de antibiotische therapie in een vroeg stadium van de behandeling wordt toegespitst op aangetoonde pathogenen. Bijvoorbeeld, een pneumokokkenpneumonie, die behandeld kan worden met goedkope smal spectrum penicillinen, kan gediagnosticeerd worden met behulp van Gram kleuring van het sputum of antigeentesten in de urine. In hoofdstuk 7 werd een algoritme ontwikkeld om kostenbesparingen te bepalen van het toespitsen van antimicrobiële therapie gebaseerd op de resultaten van microbiologische sneltesten. Variabelen waaruit het algoritme bestaat zijn de prevalentie van pneumokokkeninfecties, medicijnkosten en kosten voor de diagnostiek. Het algoritme is een methode om potentiële kostenbesparingen te bepalen in verschillende settings en kan ook gebruikt worden als nieuwe sneltesten geëvalueerd moeten worden. Toepassing van dit algoritme op een populatie van patiënten opgenomen met CAP liet zien dat kostenbesparingen minimaal waren bij gebruik van Gram preparaten van het sputum of urine antigeentesten om pneumokokkenpneumonie te diagnosticeren.

Naast het toespitsen van antibiotica op aangetoonde pathogenen, kan staken van empirisch gestarte antibiotica gerechtvaardigd zijn als virussen worden geïdentifi ceerd als de oorzaak van luchtweginfecties. Vroege identifi catie van virusinfecties is mogelijk met nieuwe Taq-

181

Samenvatting in het Nederlands

man real-time PCR technieken. De voordelen van deze tests zijn de snelle beschikbaarheid van de resultaten en een vermoedelijk hogere sensitiviteit in het detecteren van organismen bij lagere luchtweginfecties, in vergelijking met viruskweken of serologische bepalingen. Mogelijke nadeel is een toename in kosten. De effecten op patiëntenzorg waren nog niet eerder bepaald. In hoofdstuk 8 hebben we in een gerandomiseerde studie de klinische en economische effecten bepaald van virale sneldiagnostiek door middel van Taq-man real-time PCR bij patiënten opgenomen vanwege lagere luchtweginfecties. In deze studie verhoogde het gebruik van Taq-man real-time PCR inderdaad de diagnostische opbrengst in vergelijking met routine diagnostische procedures en werden virussen bij meer dan 20% van de lagere luchtweginfecties gevonden. Echter, het snel aantonen van deze virussen leidde niet tot een afname in antibioticumgebruik of behandelingskosten als de sneldiagnostiek wordt toegepast op elke patiënt opgenomen met een lagere luchtweginfectie. Real-time PCR kan daarom nog niet een deel uit gaan maken van de routine work-up van patiënten opgenomen in het ziekenhuis vanwege een lagere luchtweginfectie, tenzij de kosteneffectiviteit verbeterd kan worden door hogere diagnostische opbrengst, grotere effecten op behandeling of lagere kosten.

De keuze welke diagnostische methode routinematig gebruikt moet worden is naast andere argumenten gebaseerd op kosteneffectiviteitoverwegingen. Zoals aangetoond, is niet elke veelbelovende diagnostische strategie kosteneffectief bij routinematig gebruik. Sneldiagnostiek in een klinische setting was niet kostenbesparend in onze studies en leidde nauwelijks tot vermindering van antibioticumgebruik. Desalniettemin, verdere ontwikkeling en onderzoek van sneldiagnostiek is verantwoord om verschillende redenen. Ten eerste, omdat onnodig gebruik van (breed-spectrum) antibiotica resistentie van micro-organismen induceert, kan, in theorie, investering in methoden die snelle behandeling met smaller spectrum antibiotica of staken van antibiotica vroeg in het beloop van de behandeling kunnen bewerkstelligen, opwegen tegen de kosten voor de kosten geassocieerd met de toekomstige behandeling van infecties die veroorzaakt worden door minder gevoelige micro-organismen. Ten tweede kan sneldiagnostiek kosteneffectiever

182


worden als clinici bereidwilliger zijn de behandeling aan te passen op basis van de resultaten van de nieuwe sneltests. Daarom zijn studies met een meer geprotocolleerde behandeling nodig om de effecten op de ontwikkeling van resistentie en kosten en effecten van de behandeling te evalueren.

Een veelvuldig gesuggereerde methode om kosten te verminderen voor de behandeling van patiënten opgenomen vanwege CAP is een vroege switch van intraveneuze naar orale therapie. In theorie vermindert een vroege switch naar orale therapie kosten voor intraveneuze antibiotica en maakt het een vroeg ontslag uit het ziekenhuis mogelijk. Mogelijke nadelen zijn falen van de behandeling en een toename in behandelingskosten buiten het ziekenhuis. Grote, prospectief gerandomiseerde studies die de effecten van deze strategie op uitkomsten van ernstige CAP en op kosten zowel binnen als buiten het ziekenhuis evalueren zijn niet uitgevoerd. In hoofdstuk 9 wordt de eerste gerandomiseerde studie beschreven die de kosten en effecten, zowel binnen als buiten het ziekenhuis evalueert bij patiënten in het ziekenhuis opgenomen vanwege een ernstige pneumonie. Vergeleken met een standaard behandeling van 7 dagen intraveneuze therapie, was een behandeling met een vroege switch van intraveneuze naar orale therapie geassocieerd met vergelijkbare klinische uitkomsten en een signifi cante afname van opnameduur en totale kosten voor behandeling. De reductie in opnameduur was het meest uitgesproken bij patiënten zonder co-morbiditeit. Een behandelingsstrategie met een vroege switch van intraveneuze naar orale therapie bij patiënten opgenomen vanwege CAP is daarom veilig gebleken en leidt tot een signifi cante vermindering in behandelingskosten. Toekomstige stappen naar een effi ciëntere behandeling van CAP kunnen bestaan uit de implementatie van vroege switch strategieën in de klinische praktijk of zelfs kortere antibiotische behandeling, voornamelijk bij patiënten zonder co-morbiditeit.

12Dankwoord

184

Chapter 1Dankwoord

Nu dit boekje nagenoeg voltooid is, is het wel zo eerlijk om stil te staan bij die mensen die direct of indirect een bijdrage hebben geleverd aan de voltooiing ervan.

Allereerst wil ik de patiënten die meegedaan hebben aan de klinische onderzoeken danken voor hun inzet, vertrouwen en tijd. Zonder jullie medewerking was een onderzoek als deze natuurlijk nooit mogelijk geweest.

Als tweede gaat een woord van dank uit aan de promotoren, prof dr. I.M. Hoepelman en prof. dr. M.J.M Bonten. Andy, gedurende de periode dat ik onderzoek gedaan heb, heb ik het gevoel gehad dat je me ruimte en vertrouwen gaf, waarbij jij de grote lijn steeds goed in de gaten hield. Wat ik het meest waardeer is dat je de beloftes die je me hebt gedaan altijd bent nagekomen, waardoor ik een groot respect voor je heb ontwikkeld.Marc, ook voor jou heb ik grote bewondering. Niet alleen vanwege je vermogen om me te motiveren voor en te stimuleren in het onderzoek maar ook vanwege je snelheid van denken, je altijd kritische noot en je laagdrempeligheid voor overleg.

Uiteraard ben ik ook mijn co-promotor, dr. M.M.E. Schneider, dank verschuldigd.Margriet, je inzet, overzicht en heldere kijk zijn in deze onderzoeksperiode een belangrijke stimulans geweest waar ik ook nog eens veel van geleerd heb.

Ook ben ik zeer dankbaar voor de nooit afl atende inzet en betrokkenheid van de onderzoeksverpleegkundigen, met name Marianne Versloot en Helga Kragten.Marianne, je nauwkeurigheid en trouw heb ik als een grote zegen ervaren in de afgelopen periode. Als meest directe collega ben je gedurende het hele onderzoek een grote steun geweest. Helga, hoewel je voortijdig de onderzoeksgroep hebt verlaten, ben ook jij door je opbouwende kritiek en steun onmisbaar geweest. Ook Marieke Voortman en Karen van Schuppen als tijdelijke onderzoeks-verpleegkundigen verdienen een woord van dank.

185

Dankwoord

Voor de epidemiologische, economische en virologische onderdelen van het onderzoek zijn Eelko Hak, Erik Buskens en Ton van Loon belangrijke schakels geweest.Eelko, altijd was je bereid tot overleg en je enthousiasme voor de epidemiologie heeft aanstekelijk op me gewerkt.Erik, ondanks de opstartproblemen bij de “SAT-P” studie denk ik dat we, mede door jouw hulp, tevreden kunnen zijn met het eindresultaat. Ton, door je nauwkeurigheid en betrokkenheid bij met name de “VAZ”-studie heb ik het als prettig ervaren om met je samen te werken.

Dank gaat ook uit naar de “vrouwenvleugel” waardoor het leven als onderzoeker niet zo eenzaam en saai is geweest als het wel eens voorgespiegeld wordt. Fieke, Ilja, Irene, Mirelle, Saskia, Stefan, Ruby: door jullie voelde ik me moreel gesteund, dankzij jullie heb ik dagelijks gezellig geluncht en uiteraard ben ik door jullie een stuk wijzer geworden over het moeilijke leven van de vrouw in het algemeen en van vrouwen die met me samenwerken in het bijzonder.De kamergenoten uit het “Kippenhok”, ben ik dankbaar omdat ondanks onze krappe behuizing er toch ruimte was om serieus en in stilte hard te werken. Martine Hoogewerf heeft met het uitwerken van een van de hoofdstukken van dit proefschrift een belangrijke bijdrage geleverd, waarvoor dankzegging uiteraard op zijn plaats is. De samenwerking liep mijns inziens gesmeerd.

De input van de bezoekers van de onderzoeksbijeenkomsten zijn een waardevolle aanvulling geweest. Met name wil ik noemen Martin Bootsma en Karin Schurink.

Verder dank ik alle betrokkenen bij het onderzoek: stafl eden infectieziekten, de longartsen en met name prof. dr. J.W.J. Lammers, G. Nossent, F. Teding-van Berkhout, E. van de Graaf, de arts-assistenten longziekten, internisten en arts-assistenten interne geneeskunde van het UMC Utrecht en andere deelnemende ziekenhuizen, met name de lokale coördinatoren van de SAT-P studie Willem Hustinx, Aline Stades, Paul Hamberg, Peter Slee, Karin de Boer, Barteld de Jong, Karin Kok, Karin Kaasjager, Mark Kramer, Dylan de Lange, Marleen Willems, Peter

186

Chapter 1Dankwoord

Schneeberger, Jan Prins, Rachida El Moussaoui, Peter Speelman, en de verpleegkundigen van de betrokken afdelingen. Rob Schuurman, microbiologen, laboranten van de medische microbiologie, met name Alie McArthur, virologen en laboranten van de afdeling virologie van het UMC Utrecht dank ik voor hun geduld bij de analyses. Marianne van den Haak en Hanneke den Breijen van het Juliuscentrum dank ik voor hun voortreffelijke hulp bij het datamanagement.

Erg prettig was ook het meeleven van vrienden waarvan ik met name mijn paranymfen Remco Groenewold en Koen Both wil bedankten voor hun oprechte belangstelling en hulp bij de organisatorisch aspecten van deze promotie. Ook aan mijn vriendin heb ik grote steun ervaren. Annemarie, ik hoop dat ik ook na deze onderzoeksperiode nog steeds van je belangeloze inzet en de rust die je me geeft kan genieten.

Uiteraard dank ik ook mijn broertje Anne Jan, met name voor de IT ondersteuning en de relativerende opmerkingen.

Last but zeker not least wil ik bedanken mijn ouders, Klaas en Gretha Oosterheert, voor hun steun op de achtergrond. Jullie hebben me altijd het gevoel gegeven ergens een veilig thuis te hebben waar ik zondermeer en belangeloos op terug kan vallen. Zonder de kansen die jullie me geboden hebben was voltooiing van dit onderzoek nooit mogelijk geweest.

13Curriculum Vitae

188

Chapter 1Curriculum Vitae

189

Jan Jelrik Oosterheert is geboren op 25 juni 1976 in Zwolle. In 1994 behaalde hij het gymnasium diploma aan het Carolus Clusius College in Zwolle. Aanvankelijk uitgeloot voor de studie geneeskunde, startte hij vervolgens met de studie klassieke talen en culturen aan de Rijksuniversiteit Groningen. Eind 1994 kon door naplaatsing toch met de studie geneeskunde aan de Universiteit Utrecht worden begonnen. Het doctoraal werd in 1998, het artsexamen in augustus 2000 behaald. Na het artsexamen werkte hij gedurende ruim een jaar als assistent geneeskundige niet in opleiding op de afdeling interne geneeskunde van het Eemland ziekenhuis (nu Meander Medisch Centrum) in Amersfoort. In januari 2002 werd gestart met het onderzoek beschreven in dit proefschrift onder begeleiding van prof. dr. I.M. Hoepelman en prof. dr. M.J.M. Bonten. In september 2004 is gestart met de opleiding interne geneeskunde in het Meander Medisch Centrum (opleider dr. A. van de Wiel)

Curriculum Vitae

Chapter 1

Diagnosis and treatment of community-acquired lower ... · improvements in therapeutic efﬁ cacy,...

Documents

Transcript of Diagnosis and treatment of community-acquired lower ... · improvements in therapeutic efﬁ cacy,...