Pulmonary embolism: Are we there yet?

RADIOLOGY—ORIGINAL ARTICLE jmiro_2372 270..281

Pulmonary embolism: Are we there yet?Aanchal Agarwal,1 Jennifer Persaud,1,2 Rafal Grabinski,1 Dean Rabinowitz,1 Alexandra Bremner3 andRichard Mendelson1

1Division of Imaging Services, Royal Perth Hospital, Perth, 2University of Notre Dame, Fremantle, and 3School of Population Health, University of

Western Australia, Nedlands, Western Australia, Australia

A Agarwal MBBS(Hons); J Persaud BSc(Hons),

MSc, MCSP, MAPA; R Grabinski FRANZCR,

MBBS; D Rabinowitz FRANZCR, MBBS;

A Bremner DipEd, GradDipAppStats, BSc,

PhD; R Mendelson MBChB, MRCP, FRCR,

FRANZCR.

CorrespondenceDr Aanchal Agarwal, Division of Imaging

Services, Royal Perth Hospital, Perth, WA 6001,

Australia.

Email: [email protected]

Conflict of interest: None.

Submitted 7 June 2011; accepted 26

September 2011.

doi:10.1111/j.1754-9485.2012.02372.x

Abstract

Introduction: Clinical prediction rules (such as Wells model) are a reliableassessment tool for diagnostic work-up of suspected pulmonary embolism(PE). When used as part of a clinical algorithm and in combination with aD-Dimer, the model can safely exclude PE in low-risk groups and indicatewhen further investigations are unnecessary. The purpose of this study was toinvestigate the level of adherence to local diagnostic imaging guidelines forsuspected PE and to ascertain the impact of interventions.Methods: Retrospective search of all patients referred from the EmergencyDepartment (ED) of Royal Perth Hospital for computed tomography pulmo-nary angiography (CTPA) or V/Q scan between 11 September 2005 to 10March 2006 (pre-intervention) and 1 January 2008 to 31 March 2008 (post-intervention) was conducted. The guidelines on ‘Diagnostic Imaging Path-ways’ were considered as gold standard. Interventions included orienting EDdoctors to guidelines and modified request forms for mandatory completion ofWells score. A prevalence- and bias-adjusted kappa (PABAK) score analysedthe level of agreement between documentation on notes (R-score) and stamp(S-score).Results: Thirty-five per cent (n = 187) and 22% (n = 109) deviated from thepathway pre-intervention and post-intervention, respectively (13% absolutereduction; P = 0.017). Stamp compliance was only 55% despite mandatory fill-ing requirement. PABAK for ‘PE as most likely diagnosis’ was 0.25 for V/Q groupand – 0.26 for CTPA. In addition, 44/60 (73%) had an intermediate or highS-score, yet only 11 of those 44 had a matched intermediate to high R-Score.Conclusions: Interventions reduced inappropriate practice but did not eliminateit completely. Compliance issues may be managed in the future via theintroduction of electronic request linked to decision support.

Key words: computed tomography pulmonary angiography; D-Dimer; pulmo-nary embolism; thromboembolism; ventilation perfusion scintigraphy.

Background

Clinical prediction rules provide a reliable and non-invasive tool to evaluate the need for diagnostic work-upfor patients presenting with suspected pulmonary embo-lism (PE).1–5 Based on the history and clinical findings,clinical prediction rules determine the pretest probabilityof the disease being present and indicate when furthertests are required.

The Wells model (Table 1) is an example of an inter-nationally validated set of clinical prediction rules used in

the diagnosis of PE.4–10 When used as part of a clinicalalgorithm and in combination with a D-Dimer blood test,the Wells model can safely exclude PE in low-risk groupsand indicate when further investigations for PE areunnecessary.11–13

Despite the existence of the Wells model and theprecept that imaging should be carried out only whenindicated,11 it is apparent that guidelines are increasinglybeing substituted by computed tomography pulmonaryangiography (CTPA) or Ventilation Perfusion (V/Q)scans as the primary step in diagnosing PE.12,13 This is

bs_bs_banner

Journal of Medical Imaging and Radiation Oncology 56 (2012) 270–281

© 2012 The AuthorsJournal of Medical Imaging and Radiation Oncology © 2012 The Royal Australian and New Zealand College of Radiologists270

concerning, especially given the radiation exposure andhigh costs associated with these studies and that inap-propriate imaging leads to risks without benefit andthreatens effective allocation of resources.14,15

There are a number of formats in which clinical rulesand guidelines are disseminated and promoted, yetconversion into uptake in clinical practice remains achallenge. ‘Diagnostic Imaging Pathways’ (DIP) (http://www.imagingpathways.health.wa.gov.au) is an onlineclinical decision support tool and educational resource forreferrers and providers of diagnostic imaging.16 Its aim isto provide clinicians and imaging specialists with quickand easy access to evidence-based and consensus-approved pathways in order to encourage appropriateuse of diagnostic imaging.

DIP on ‘Pulmonary Embolism in the HaemodynamicallyStable Patient’ (Fig. 1) promotes use of Wells score asa primary screening tool to stratify patients into low-,intermediate- or high-risk categories for the disease.Further investigations, if appropriate, are then mappedout accordingly.17

Despite the established presence of DIP and numerousvalidated guidelines, there is evidence confirming thatthe existence of guidelines alone does not neces-sarily result in their acceptance and uptake in clinicalpractice.18–20 Hence, enforcement of guideline adherencein practice is arguably a necessary solution to addressthe problem. Hence, this study aimed to investigate thelevel of uptake and adherence to local diagnostic imagingguidelines for suspected PE and to evaluate whether theintroduction of more mandatory processes for approvalof CTPA or V/Q requests would reduce inappropriateimaging.

Methods

Study design

A retrospective audit was performed, followed by anintervention and a re-audit to complete the audit cycle.

A retrospective computer-based search was con-ducted, using Emergency Department InformationSystem and the Radiology Information System. The

search captured consecutive patients referred by theEmergency Department (ED) of Royal Perth Hospital(RPH) for CTPA or VQ scan between 11 September2005 and 10 March 2006 (6 months). Exclusion criteriawere: (i) patient less than 18 years of age at the timeof referral; (ii) pregnancy; (iii) D-Dimer or imagingrequested for diagnosis other than PE.

The guidelines on DIP are evidence based and basedon multidisciplinary consensus.17 Therefore, deviationfrom the guidelines (Fig. 2) was considered to be aninappropriate event (Table 2). The medical notes werehand searched for proof of adherence to the PE DIP,including use of the Wells’ criteria. Where no Wells scorewas recorded, a retrospective score (R-score) was cal-culated based on the documented clinical details in themedical notes.

Intervention was in the form of a 1-day teachingsession to orient ED junior medical officers (residentsand interns) to the guidelines on DIP, along with regularreminders to all ED clinical staff at daily handovers. Aspecial stamp outlining the Wells score (Fig 3) was intro-duced on the radiology request forms for mandatoryreferrer completion for each CTPA or V/Q scan request.Incomplete or unstamped request forms were returnedto the requesting medical officer for appropriate comple-tion and resubmission. An investigator, blinded to thedetails of the study, recorded the specific item(s) of theWells’ criteria that were ticked on the stamp and calcu-lated the score (S-score).

Re-audit was conducted on the data obtained post-intervention between 01 January 2008 and 31 March2008 (3 months). A retrospective calculation of the Wellsscore (R-score) from patients’ notes was also performed.

The study was a joint collaboration between the RPHEmergency and Radiology departments and was con-ducted in the ED. As the nature of the study was clinicalaudit, local ethics committee approval was not required.

Data analysis

A chi-square analysis compared pre-intervention andpost-intervention groups for the number of patients withat least one incidence of non-compliance to the pathway.A prevalence- and bias-adjusted kappa (PABAK) test21

Table 1. Summary of the Wells model for predicting the pretest probability

of pulmonary embolism (PE)

Characteristic Score

Previous PE or deep vein thrombosis 1.5

Heart rate >100 beats/min 1.5

Recent surgery or immobilisation 1.5

Clinical signs of DVT 3.0

Alternative diagnosis less likely than PE 3.0

Haemoptysis 1.0

Cancer 1.0

PE, pulmonary embolism.

Table 2. Criteria for interpreting inappropriate incidences

• Any deviation from the PE DIP

• Chest X-ray (CXR) report: (abnormal CXR findings or patients with chronic

lung disease should have CTPA rather than VQ)

• Serum creatinine levels (VQ scans were not considered inappropriate if

there was an abnormal CXR or chronic lung disease with an accompanying

serum creatinine level of >100 mmol/L for males and 90 mmol/L for

females).

• D-Dimer result (�40 mg/mL was indicative of a positive result and

appropriate for further investigation)

Pulmonary embolism

© 2012 The AuthorsJournal of Medical Imaging and Radiation Oncology © 2012 The Royal Australian and New Zealand College of Radiologists 271

analysed the agreement between the observationsrecorded on the stamp versus those recorded on themedical notes.

Statistical significance was set to 5% and the StatisticsPackage for Social Science (SPSS) version 16 was used(SPSS Inc., Chicago, IL, USA) for data analysis.

Results

Pre-intervention phase

The pre-intervention group included 187 subjects(females = 103, males = 84, mean age = 55 � 18 years,

Fig. 1. Diagnostic Imaging Pathway for PE as displayed currently in DIP.

Agarwal et al.


Fig. 2. Diagnostic Imaging Pathway for PE adopted for the purpose of study.

Pulmonary embolism


CTPA = 89, V/Q = 98). In total, there were 73 inappro-priate incidences among 66 (35%) patients becauseof improper use of D-Dimer, V/Q scan or CTPA investi-gation, or because a V/Q scan was carried out as theinvestigation of choice despite an abnormal chest X-rayresult (Table 3). An overview of patient managementand investigation results for the pre-intervention groupmay be viewed in Figure 4a (V/Q) and 4b (CTPA).

Post-intervention phase

There were 109 subjects in the post-intervention group(females = 58, males = 51, mean age = 59 � 20 years,CTPA = 76, V/Q = 33). In total, there were 27 inappro-priate incidences among 24 (22%) patients because ofinappropriate use of D-Dimer, V/Q scan or CTPA inves-tigation, or because a V/Q scan was carried out as the

investigation of choice despite an abnormal chest X-rayresult. An overview of investigations performed in thepost-intervention group may be viewed in Figure 4c(V/Q) and 4d (CTPA).

Between-group analysis

A between-group analysis confirmed that the reductionin the proportion of patients who had at least one inci-dence of inappropriate deviation from the pathway from35% pre-intervention to 22% post-intervention wassignificant (P = 0.017).

Stamp results

Radiology request forms for 60/109 (55%) contained acompleted stamp detailing the Wells score (named theS-Score). V/Q scan request forms for 8/33 (24%) con-tained a stamp versus 52/76 (68%) for the CTPA group.

Although 44/60 (73%) patients had an intermediateor high S-Score, only 11 of those 44 patients had amatched intermediate to high R-Score.

The remaining 33/44 patients with a mismatchedintermediate or high S-Score and low R-Score all under-went imaging (30 had CTPA and 3 had V/Q scan), andtwo of these were confirmed positive for PE. The fre-quencies for each positively identified Wells’ criterion inthe medical notes and the stamped request form aredisplayed in Table 4.

The PABAK scores revealed that ‘PE as most likelydiagnosis’ had very poor levels of agreement betweenR-scores and S-scores (k = 0.25 for the V/Q group andk = -0.26 for the CTPA group). In comparison, R-scoresand S-scores for other objective criteria (such as pre-vious history of thromboembolism, heart rate >100,history of cancer, recent immobilisation, etc.) were notsignificantly different (Table 4).

Discussion

There was only a 55% compliance rate for stampcompletion despite the mandatory filling requirement.

Wells’ Criteria for Pre-test Probability of a Pulmonary Embolism

Clinical signs and symptoms of DVT 3.0

Recent immobilisation ≥ 3/7 or 1.5

surgery in past 4/52

Previous DVT or PE 1.5

Heart Rate >100 beats per minute 1.5

Haemoptysis 1.0

recnaC 1.0

PE as/more likely as alternative 3.0

diagnosis

Heart Rate >100 beats per minute 1.5

Fig. 3. The tick-box style stamp as it appeared on the imaging request form.

Table 3. Levels of inappropriate incidences for the pre-intervention and post-intervention groups

Pre-intervention

(n = 187)

Post-intervention

(n = 109)

Wells score in medical notes 21 (11%) 11 (10%)

Low Wells score 135 (72%) 80 (73%)

Intermediate to high Wells score 52 (28%) 29 (27%)

Total patients adhering to pathway 121 (65%) 85 (78%)

Total patients deviating from pathway 66 (35%) 24 (22%)

No inappropriate incidences per person

• 1 incidence per person 60 21

• 2 incidences per person 5 3

• 3 incidences per person 1 0

Total No inappropriate incidences 73 27

Agarwal et al.


This could be explained by several factors including thepossibility that the ED staff resisted following the guide-lines (perhaps due to non-belief in the validity of guide-lines, patient expectations for further investigations and

time constraints to access the guidelines on the DIPwebsite or to complete the stamp). Perceived lack ofaccessibility of the stamps is another possible factor as itwas reported that stamps were sometimes difficult

Wells score (R-score)

N=98

Chest x-ray=98; 4/98

abnormal

Low risk (66/98) Intermediate/high risk

(32/98)

D-dimer test? D-dimer test?

No (5/66)Yes (61/66) Yes (24/32) No (8/32)

D-dimer result?

-ve (6/61) +ve (55/61) -ve (9/24) +ve (15/24)

VQ=6

Normal=3

*Low prob=3

VQ=55,Normal=32

Low prob=18

*I.med prob=2High prob=3

VQ=5

Normal=3

Low prob=1

High prob=1

VQ=9

Normal=7

Low prob=2

VQ=15,

Normal=4

Low prob=3

I.med prob=5High prob=3

VQ=8

Normal=5

Low prob=1

*I med prob= 2

*CTPA=1

(-ve=1) *CTPA=2

(-ve=2)

Normal VQ: USS=1(-ve =1)

CTPA plus USS =1

(-ve =1)

Low VQ: USS=1(+ve =1)

CTPA=1(-ve =1)

High VQ: USS=1(+ve =1)

CTPA plus USS =1

(-ve =1)

*CTPA=2

(-ve=1,

+ve=1)

D-dimer result?

a

Fig. 4. (a) Pathway summarising investigation and outcomes for VQ pre-intervention group. (Low Prob, low probability of PE; I’med, intermediate probability of PE;

High Prob, high probability of PE; USS, ultrasound scan). Asterisk (*) links diagnostic outcome of the VQ scan to further investigations in the same patient. Coloured

boxes indicate inappropriate event excluding chest X-ray. (b) Pathway summarising investigation and outcomes for CTPA pre-intervention. (c) Pathway summarising

investigation and outcomes for VQ post-intervention group. (Low Prob, low probability of PE; I’med, intermediate probability of PE; High Prob, high probability of PE).

Asterisk (*) links further investigation outcomes to the outcome of the VQ scan in the same patient. (d) Pathway summarising investigation and outcomes for CTPA

post-intervention. Asterisk (*) links further investigation outcomes of the CTPA scan in the same patient. US, ultrasound scan.

Pulmonary embolism


to locate. There was also some anecdotal suggestionthat pressure was placed upon radiographers to acceptunstamped request forms.

The observed 13% post-intervention reduction in in-appropriate incidences may be attributed to increasedawareness and understanding of the existing guidelines.However, the possibility of a Hawthorne effect cannot beruled out, i.e. ED staff was aware that audit was beingcarried out.

Despite strategies to improve compliance with guide-lines, inappropriate practice could not be totally elimi-nated. This outcome reinforces conclusions drawnfrom previous studies that availability of guidelines doesnot necessarily translate into change in practice.21–23

Furthermore, in the post-intervention phase, 19 (17.5%)patients had inappropriate D-Dimer and six (5%) under-went CTPA or V/Q scan despite a negative D-Dimerand a low pretest probability. Compliance issues may be

Low Risk (69/89)

D-dimer Test?

Yes (58/69)

CTPA = 10 PE -ve = 9 PE +ve = 1

No (11/69)

CTPA = 48 PE -ve = 41 PE ive = 6 Inconclusive = 1

CTPA = 11 PE -ve = 11

Intermediate/high Risk

(20/89)

D-dimer Test?

Yes (9/20) No (11/20)

D-dimer Result?

-ve (2/9)

+ve (7/9)

CTPA = 2 PE -ve = 2

CTPA = 7 PE -ve = 3 PE +ve = 4

CTPA = 11 PE -ve = 9 PE +ve =1 Inconclusive = 1

D-dimer Result?

-ve (10/58)

+ve (48/58)

Wells Score (R-Score) N=89

b

Fig. 4. Continued.

Agarwal et al.


Wells Score (R-Score) N=33

Low Risk (27/33)

D-dimer Test?

Yes (26/27)

VQ = 3 Normal = 2 *I’med Prob =1

No (1/27)

VQ = 24 Normal = 23 High Prob = 1

VQ = 1 Normal = 1

Intermediate/high Risk

(6/33)

D-dimer Test?

Yes (3/6) No (3/6)

D-dimer Result?

-ve (1/3)

+ve (2/3)

VQ = 1 Normal = 1

VQ = 2 Normal = 1 Low Prob = 1

VQ = 3 Normal = 3

D-dimer Result?

-ve (3/26)

+ve (24/26)

Chest x-ray = 33 3/33 with abnormal chest

x-ray

USS=1 (+ve = 1) *CTPA =1 (-ve = 1)

c

Fig. 4. Continued.

Pulmonary embolism


managed in the future via the introduction of electronicrequest forms linked to decision support. This idea haspreviously been proposed14,22 based on the premise thatrequestors can receive alerts to ‘inappropriate’ requestswhich may not be overridden without qualification ofspecific clinical indications22 or following discussion withthe radiologist.

Of the 44 patients with high S-scores, only 11 hadmatching high R-scores. This might indicate that theremaining 33 (75%) patients were ‘overscored’ on thestamp. This suggests poor documentation of correct

scores in the medical notes, lack of expertise, ‘defensivemedicine’ or ‘gaming’ on the part of referring doctors tofacilitate requests for CTPA or V/Q scan. This issue isalready of concern among radiologists.23

All 33 of the ‘overscored’ patients had inappropriateCTPA or V/Q scans. Based on the assumption that theR-score was a more accurate representation of theclinical presentation in this group, they should haveundergone a D-Dimer test with subsequent imaging onlyif the D-Dimer was raised. Accordingly, by appropriatelyfollowing the guidelines, it is very unlikely that a patient

Low risk

(53/76)

Intermediate/high

risk (23/76)

D-dimer test? D-dimer test?

Yes(46/53) No(7/53) Yes (8/23) No (15/23)

D-dimer result? D-dimer result?

Well Score(R-Score)

N=76

-ve (3/46) +ve(43/46) -ve (1/8) +ve(7/8)

CTPA=3

PE ve =3

CTPA =43

PE ve =37

PE +ve = 6

CTPA =7

PE ve =5

PE +ve = 2

CTPA =1

PE ve =1

CTPA =7

*PE ve =5

PE +ve = 1

Inconclusive

=1

CTPA =15

PE ve =11

PE +ve = 4

USS=1

PE-ve=1

d

Fig. 4. Continued.

Agarwal et al.


with a PE would have been missed in this group ofpatients, but additionally, those patients with low pretestprobabilities and negative D-Dimers would have avoidedan unnecessary CTPA or V/Q scan. This would haveresulted in a reduction in health costs, time savings bothon the part of radiographers to perform the test as wellas radiologists to interpret the images and, most im-portantly, reduction in the exposure of the patients tounnecessary radiation.

The discordance noted between the R-scores andS-scores was strongly associated with the ‘PE as mostlikely diagnosis’ criterion of the Wells score. This para-meter is heavily weighted (3 points) and can elevate apatient’s risk category, thereby providing potential jus-tification for an imaging request. Also, the subjectivenature of the criterion may provide opportunity forabuse or application of inappropriate rationale in orderto trigger imaging, and the inter-observer differencecertainly supports this possibility. Universal adoption ofa less subjective tool may lead to a more satisfactorymethod of triaging patients with suspected PE. Therevised Geneva score and the two-tier Wells score havebeen suggested as appropriate alternatives10,24,25 andboth are validated tools.10,26

Our study had some limitations. The retrospectivelycalculated Wells scores relied on information containedin the medical notes. However, it is possible that therewas additional clinical information or factors known orsuspected by the ED clinicians that were not docu-mented, for example, time constraints limiting documen-tation, or misconceptions that completion of the stampwas sufficient documentation.

Several investigators were involved in the calculationof the retrospective Wells score. This could be a sourceof bias in the interpretation of the retrospective score, inparticular, the interpretation of ‘PE as most likely diag-nosis’. This may explain the disconcordance betweenR-scores and S-scores and the low agreement reflectedby the PABAK analysis. An improvement in clinical pre-diction rule uptake may have been temporary due to aHawthorne effect. Our study did not include a follow-up

audit to evaluate whether the improvement in appropri-ate imaging persisted after the intervention.

It is to be noted that DIP is continuously updated asnew literature becomes available and the ‘PulmonaryEmbolism in Haemodynamically Stable’ patient pathwayhas changed since the introduction of the 2008 Europeanguidelines on Diagnosis and Management of Acute Pul-monary Embolism.25 According to the guidelines, a highlysensitive D-Dimer assay safely excludes PE in patientswith a low or moderate clinical probability, while a mod-erately sensitive assay excludes PE only in patients witha low clinical probability. Hence, the current DIP pathwaydichotomises the patients with intermediate pretestprobability into those having a highly sensitive or amoderately sensitive D-Dimer assay (Fig 1). This is likelyto mitigate the problem of inappropriate imaging refer-rals as patients with intermediate probability will then beinvestigated with only a D-Dimer test when a high sen-sitivity assay is available which if negative will withholdany further investigations for PE. Similarly, with theavailability of a moderately sensitive assay, the inter-mediate probability patients would proceed directly toimaging.

We did not quantify the time spent by the radiogra-phers and imaging specialists performing and interpret-ing the images. This would have been useful to estimatethe time wasted on inappropriate imaging. We also didnot quantify the amount of harmful radiation exposureinappropriately experienced by patients.

While our study did not have a control group, thesample population was representative of the targetpopulation, and our study was able to show a statisticallysignificant reduction in inappropriate practice followingintervention.

Conclusions

The findings of this study indicate that failure to adhereto the DIP when requesting radiological investigation forsuspected PE is prevalent in the ED. The introduction ofa more mandatory process for the approval of VQ/CTPA

Table 4. Discordant pairs for R-scores vs. S-scores and the PABAK score

Wells’ criterion No. of positively identified

criteria (VQ group)

PABAK

(VQ Group)

No. of positively identified

criteria (CTPA group)

PABAK

(CTPA group)

(n = 8) (n = 52)

S-score R-score S-score R-score

Previous PE or DVT 0 0 1 4 5 0.89

Heart rate >100 beats per minute 1 1 1 10 14 0.62

Recent surgery/immobilisation 2 1 0.75 (P = 0.64) 7 2 0.81

Clinical signs of DVT 0 0 1 7 5 0.85

Haemoptysis 0 0 1 3 0 -0.89

Cancer 0 0 1 5 4 0.88

PE as/more likely than alternative diagnosis 5 2 0.25 (P = 0.206) 34 7 0.269 (P = 0.62)

Pulmonary embolism


requests resulted in an improvement in referrals and adecrease in unnecessary and inappropriate investi-gations. The use of educational sessions, promotionalmaterial and a mandatory stamp appears to be a simpleintervention to improve imaging request behaviour andpatient outcomes.

The adoption of a more mandatory process of imagingrequests by ED clinicians has a number of prerequisitesin order to be successful. It requires acceptance from EDsenior staff and willingness by the department to coop-erate with imaging guidelines. In the first instance, thisrequires a review by senior staff in both Radiology andEDs of the evidence behind the guidelines and recogni-tion of common goals of using the pathway (i.e. feweradverse events, better resource allocation and patientadvocacy). It requires frequent education sessions withnew staff as well as review sessions, and advocates inboth departments to educate, promote and police theuse of the tool.

An electronic request form with an inbuilt DIP may bethe best way to ensure mandatory use of the guideline,but flexible options would be necessary to allow accept-able and justifiable deviations from the pathway. Afuture study may involve the introduction of an elec-tronic request system and a review similar to our own.

References

1. Agnelli G, Becattini C. Acute pulmonary embolism.N Engl J Med 2010; 363: 266–74.

2. Miniati M, Bottai M, Monti S. Comparison of 3 clinicalmodels for predicting the probability of pulmonaryembolism. Medicine 2005; 84: 107–14.

3. Stein PD, Woodard PK, Weg JG et al. Diagnosticpathways in acute pulmonary embolism:recommendations of The PIOPED II Investigators.Am J Med 2006; 119: 1048–55.

4. Wells PS, Anderson DR, Rodger M et al. Derivation ofa simple clinical model to caegorize patientsprobability of pulmonary embolism: increasing themodels utility with the SimpliRED d-dimer. ThrombHaemost 2000; 83: 416–20.

5. Wells PS, Anderson DR, Rodger M et al. Excludingpulmonary embolism at the bedside withoutdiagnostic imaging: management of patients withsuspected pulmonary embolism presentign to theemergency department by using a simple clinicalmodel and D-Dimer. Ann Intern Med 2001; 135:98–107.

6. Iles S, Hodges AM, Darley JR et al. Clinicalexperience and pre-test probability scores in thediagnosis of pulmonary embolism. Q J Med 2003;96: 211–5.

7. Kearon C, Ginsberg JS, Douketis J et al. Anevaluation of d-dimer in the diagnosis of pulmonaryembolism. Ann Intern Med 2006; 144: 812–21.

8. Kruip MJHA, Sohne MM, Nijkeuter M et al. On behalfof the Christopher Study Investigators. A simple

diagnostic strategy in hospitalized patients withclinically suspected pulmonary embolism. J InternMed 2006; 260: 459–66.

9. Perrier A, Roy P-M, Aujesky D et al. Diagnosingpulmonary embolism in outpatients with clinicalassessment, D-dimer measurement, venousultrasound, and helical computed tomography: amulticenter management study. Am J Med 2004;116: 291–9.

10. van Belle A, Büller H, Huisman M et al. ChristopherStudy Investigators. Effectiveness of managingsuspected pulmonary embolism using an algorithmcombining clinical probability, d-dimer testing,and computed tomography. JAMA 2006; 295:172–9.

11. Stein PD, Sostman HD, Bounameaux H et al.Challenges in the diagnosis acute pulmonaryembolism. Am J Med 2008; 121: 565–71.

12. Pistolesi M. Pulmonary CT angiography in patientssuspected of having pulmonary embolism: casefinding or screening procedure? Radiology 2010;256: 334–7.

13. Remy-Jardin M, Pistolesi M, Goodman LR et al.Management of suspected acute pulmonaryembolism in the era of CT angiography: a statementfrom the fleischner society1. Radiology 2007;245: 315–29. Review Article.

14. Bairstow PJ, Mendelson R, Dhillon R, Valton F.Diagnostic imaging pathways: development,dissemination, implementation, and evaluation. Int JQual Health Care 2006; 18: 51–7.

15. Mendelson RM, Murray CPJ. Towards the appropriateuse of diagnostic imaging. Med J Aust 2007; 187(1): 5–6.

16. Mendelson R. Diagnostic Imaging Pathways. A ClinicalDecision Support Tool and Educational Resource forDiagnostic Imaging [Internet]. Perth (Australia):Royal Perth Hospital, 2011. [Cited Dec 2011.]Available from URL: http://www.imagingpathways.health.wa.gov.au/

17. Mendelson R. Diagnostic Imaging Pathways: AClinical Decision Support Tool and EducationalResource for Diagnostic Imaging [Internet]. Perth(Australia): Royal Perth Hospital, 2011. DiagnosticImaging Pathways-Pulmonary Embolism. [CitedJun 2010.] Available from URL: http://www.imagingpathways.health.wa.gov.au/includes/DIPMenu/pe/chart.html

18. Grimshaw JMP, Eccles MP, Tetroe J. Implementingclinical guidelines: current evidence and futureimplications. J Contin Educ Health Prof 2004;24: S31–7.

19. Grimshaw JM, Eccles MP. Is evidence-basedimplementation of evidence-based care possible?Med J Aust 2004; 180: S50–1.

20. Liu J, Wyatt JC, Altman DG. Decision tools in healthcare: focus on the problem, not the solution. BMCMed Inform Decis Mak 2006; 6 (4). Available fromURL: http://www.biomedcentral.com/1472-6947/6/4

Agarwal et al.


21. Byrt TBJ, Carlin JB. Bias, prevalence and kappa.J Clin Epidemiol 1993; 46: 423–9.

22. Bairstow PJ, Persaud J, Mendelson R, Nguyen L.Reducing inappropriate diagnostic practice througheducation and decision support. Int J Qual HealthCare 2010; 22: 194–200.

23. Mamlouk MD, vanSonnenberg E, Gosalia R et al.Pulmonary embolism at CT angiography: implicationsfor appropriateness, cost, and radiation exposurein 2003 patients1. Radiology 2010; 256:625–32.

24. Gal GL, Righini M, Roy P-M et al. Prediction ofpulmonary embolism in the emergency department:

the revised geneva score. Ann Intern Med 2006;144: 165–71.

25. Torbicki A, Perrier A, Konstantinides S et al.Guidelines on the diagnosis and management ofacute pulmonary embolism: the Task Force forthe Diagnosis and Management of Acute PulmonaryEmbolism of the European Society of Cardiology(ESC). Eur Heart J 2008; 29: 2276–315.

26. Sohne M, Kamphuisen P, van Mierlo P, Buller H.Diagnostic strategy using a modified clinical decisionrule and d-dimer test to rule out pulmonaryembolism in elderly in- and outpatients. ThrombHaemost 2005; 94: 206–10.

Pulmonary embolism


Pulmonary embolism: Are we there yet?

Documents

Transcript of Pulmonary embolism: Are we there yet?