Meta Analisis Cigarro Implante

8/11/2019 Meta Analisis Cigarro Implante

1/22

Smoking interferes with theprognosis of dental implanttreatment: a systematic reviewand meta-analysis

Strietzel FP, Reichart PA, Kale A, Kulkarni M, Wegner B, Kuchler I. Smoking interferes

with the prognosis of dental implant treatment: a systematic review and meta-analysis.J Clin Periodontol 2007; 34: 523544. doi: 10.1111/j.1600-051X.2007.01083.x.

Abstract

Aim: This systematic literature review was performed to investigate if smokinginterferes with the prognosis of implants with and without accompanyingaugmentation procedures compared with non-smokers.

Methods: A systematic electronic and handsearch (articles published between 1989and 2005; English and German language; search terms dental or oral implantsand smoking; dental or oral implants and tobacco) was performed to identifypublications providing numbers of failed implants, related to the numbers of smokersand non-smokers for meta-analysis. Publications providing statistically examined dataof implant failures or biologic complications among smokers compared with non-smokers were included for systematic review.

Results: Of 139 publications identified, 29 were considered for meta-analysis and 35for systematic review. Meta-analysis revealed a significantly enhanced risk for implantfailure among smokers [implant-related odds ratio (OR) 2.25, confidence interval(CI95%) 1.962.59; patient-related OR 2.64; CI95% 1.704.09] compared with non-smokers, and for smokers receiving implants with accompanying augmentationprocedures (OR 3.61; CI95% 2.265.77, implant related). The systematic reviewindicated significantly enhanced risks of biologic complications among smokers. Fivestudies revealed no significant impact of smoking on prognosis of implants withparticle-blasted, acid-etched or anodic oxidized surfaces.

Conclusion: Smoking is a significant risk factor for dental implant therapy andaugmentation procedures accompanying implantations.

Key words: dental implants; meta-analysis;odds ratio; smoking; success rate; survival rate

Accepted for publication 18 February 2007

Conflict of interest and source offunding statement

The authors declare that they have noconflict of interest.None of the authors received any benefit ofany kind from commercial or official par-ties related directly or indirectly to thesubject matter of this article. No grant orfunding from any party was received toprepare or perform this work.

Frank Peter Strietzel1

, Peter A.Reichart1, Abhijit Kale2, Milind

Kulkarni2, Brigitte Wegner3 and

Ingeborg Kuchler3

1Department of Oral Surgery and Dental

Radiology, Campus Virchow Clinic, Centre

for Dental Medicine, Charite Medical

University Berlin, Berlin, Germany ; 2Clinic

of General and Restorative Dentistry and

Dental Implantology, Maharashtra Medical

Foundations Joshi Hospital, Pune, India;3Campus ChariteMitte, Institute for Biometry

and Clinical Epidemiology, Charite Medical

University Berlin, Berlin, Germany

Smoking was shown to be a primary risk

factor for general health, responsible formany serious diseases, as for 90% of alllung cancers, 70% of chronic lung dis-

eases, 80% of myocardial infarctions

before the age of 50, and 30% of chronicischaemic heart diseases and strokes(Fielding 1985, La Veccia et al. 1991,Peto et al. 1996). Currently, there are anestimated 1.3 billion smokers world-wide, and 4.9 million people die fromtobacco smoking-related diseases peryear (WHO 2005).

Besides the prominent role of healthprofessionals to play in tobacco controland smoking cessation generally, certainaspects concerning current dental ther-apy should be considered in smokers forthorough patient information before oral

surgical procedures and implant therapy

planning. The increased risk of woundhealing complications (Meechan et al.1988, Miller 1988, Jones & Triplett1992, Sands et al. 1993) as well as therisk of peri-implant bone loss andincreased implant failure rates (Haaset al. 1996, Lindquist et al. 1996, Lem-ons et al. 1997) have to be emphasized.Impaired wound healing has to beexpected due to less collagen production(Jorgensen et al. 1998), reduced periph-eral blood circulation (Lehr 2000) andcompromised function of polymorpho-nuclear leucocytes and macrophages

J Clin Periodontol 2007; 34: 523544 doi: 10.1111/j.1600-051X.2007.01083.x

523r 2007 The Authors. Journal compilation r 2007 Blackwell Munksgaard


2/22

(Kenney et al. 1977, MacFarlane et al.1992). Moreover, smoking was indi-cated a significant subject-based riskfactor for periodontitis by recently pub-lished literature reviews (Palmer et al.2005, Ramseier 2005). Although notcompletely investigated, the known

mechanisms of action of smoking onperiodontitis were found to be the long-term chronic effect due to impairing thevasculature of periodontal tissues andaffection of multiple functions of neu-trophils and inflammatory responseas well as impairment of fibroblasts(Palmer et al. 2005).

Therefore, identification of the smok-ing patient is required. Patient question-naires considering general health issues,but also asking for smoking habits andalcohol consumption, seem to be neces-sary, though not generally used as a

matter of routine yet (Reichart et al.2000).Thorough patient information about

the planned treatment course and theexpected outcomes, but also about risksand risk-associated factors is necessaryto support the patients decisionmaking before implant therapy (Striet-zel 2003). Therefore, a systematic ana-lysis of the literature focusing onexpected interactions between implantprognosis and smoking seemed to benecessary to provide quantitative factsincluding the likelihood of smoking-

associated risks for implant prognosisand outcomes after augmentation pro-cedures accompanying implant surgery,to support a risk analysis before therapy,and to substantiate the patient informa-tion.

This meta-analysis and systematicreview focused on the question, if thereis a significantly enhanced risk ofimplant failures in smokers comparedwith non-smokers. The influence ofsmoking on implants inserted withaccompanying augmentation procedureswas additionally investigated.

Material and Methods

Search strategy

A systematic literature search in electro-nic databases was conducted, using thefollowing search term combinations:dental implants AND smoking,dental implants AND tobacco,oral implants AND smoking andoral implants AND tobacco.

Furthermore, a manual search wasapplied to three German-language

peer-reviewed dental journals, focusingon articles related to the effect ofsmoking on dental implant treatmentoutcomes.

Inclusion criteria

Literature search was performed toidentify meta-analyses and systematicreviews as well as randomized-con-trolled clinical trials, prospective orretrospective clinical studies, cohort stu-dies or casecontrol studies.

Publications were included for meta-analysis or systematic review, if theywere published between January 1989and December 2005 in English languageand listed in electronic databases Med-line/Pubmed or Embase, or were pub-lished in German language in Deutsche

Zahnarztliche Zeitschrift (January

1968December 2005), Zeitschrift furZahnarztliche Implantologie (January1998December 2005) and Implantolo-gie (January 1996December 2005).

Publications were included for meta-analysis, if implant survival rates or thenumber of failed implants were reportedon an implant- or patient-related basisand the number of smoking as well asnon-smoking patients were publishedand could be related to the number offailed and remaining implants, respec-tively.

Moreover, publications were

included for systematic review althoughnot meeting the inclusion criteria formeta-analysis, if odds ratios (OR), riskratios (RR) or hazard ratios (HR) forimplant failures among smokers werereported, or reports on biologic compli-cations and findings known to influencethe success of implant-prosthetic ther-apy in smokers compared with non-smokers were given, and if the resultswere statistically analysed consideringthe effects of smoking on treatmentcourse or outcome.

As definition of smokers were differ-

ent in several studies regarding quanti-ties of smoked cigarettes per day andtherefore were accordingly categorizeddifferently, any patient who smoked wasconsidered a smoker, following the defi-nition given by Wallace (2000).

Selection of studies and data extraction

Titles and abstracts of the publicationsidentified by electronic databases andhandsearch using the search terms men-tioned above were screened initially bytwo independent reviewers (F. P. S. and

A. K. or M. K.). Publications wereincluded for full text evaluation if thestudy design and content of the abstractsmet the inclusion criteria and matchedthe focused question. Agreementbetween the reviewers was determinedperforming k-statistic. Disagreements

were resolved by evaluation of the fulltexts and discussion. Final authority forselection disagreements rested withF. P. S.

Full text assessment and data extrac-tion were performed by the reviewerswithout any disagreements. The processof identification of the included studiesfrom the initial yield is described inFig. 1.

A categorization considering theduration of the observation periodsreported in the studies identified formeta-analysis was performed addition-

ally (group 1: observation period 41year; group 2:41 and45 years; group3: 45 years).

Furthermore, studies involvingimplant-prosthetic rehabilitation afteraugmentation procedures to enhancebone quantity at the implant site [sinusfloor elevation and augmentation(SFEA) and/or lateral alveolar ridgeaugmentation by guided bone regenera-tion (GBR)] providing survival data ofimplants were considered for meta-analysis to evaluate the risk of smokingconcerning augmentation procedures.

The frequency and percentage distri-bution of smokers and non-smokersregarding the frequency of implant suc-cess or failure were extracted from thestudies included for meta-analysis.Furthermore, the distributions of maleand female patients, and of implants inthe maxilla and mandible with specialrespect to the groups regarding differentobservation periods were considered.

To rule out the cumulation of indivi-dual risks, studies publishing patient-related data were considered separatelyfrom those reporting implant-related

data for meta-analysis.Considering studies publishing

implant-related data, implants were con-sidered failures if they failed to osseoin-tegrate (Balshi & Wolfinger 1999,Grunder et al. 1999, Keller et al. 1999,Kronstrom et al. 2001, Mayfield et al.2001, Kumar et al. 2002, van Steen-berghe et al. 2002), were lost orremoved for any reason (Bain & Moy1993, De Bruyn & Collaert 1994, Bain1996, Minsk & Polson 1998, De Bruynet al. 1999, Jones et al. 1999, Berge &Grnningsaeter 2000, Lambert et al.

524 Strietzel et al.

r 2007 The Authors. Journal compilation r 2007 Blackwell Munksgaard


3/22

2000, Olson et al. 2000a,b, Schwartz-Arad et al. 2000, Geurs et al. 2001,Widmark et al. 2001, Ortorp & Jemt2002, Penarrocha et al. 2002, Schwartz-Arad et al. 2002, Beschnidt et al. 2003,van Steenberghe et al. 2004, Moheng &Feryn 2005) or exceeded more than 50%bone loss (Bain & Moy 1993, Bain1996) or more than 4 mm of verticalbone defect (Berge & Grnningsaeter2000), if they failed one or more of the

criteria proposed by Smith & Zarb(1989) (Kan et al. 2002), revealed mobi-lity (without having removed implantprostheses for individual implant mobi-lity test), persistent pain, peri-implantradiolucency and/or infections attributa-ble to the implant (Gorman et al. 1994,Wallace 2000), or if they showed biolo-gic complications (peri-implantitis)(Karoussis et al. 2003).

Considering publications providingwith patient-related data, the treatmentof one patient was accounted as failureif one or more implants were considered

a failure due to the above-mentionedcriteria.

Excluded studies

Publications were excluded if they didnot meet the inclusion criteria (i.e. ifthey were considered case reports, ani-mal or in vitro experiments, educationalstatements, expert opinions), if they didnot provide implant- or patient-related

data on failures or complications inimplant treatment related to the smokinghabit, or if they did not contain materialmatching the focused question.Excluded studies and the reasons forexclusion are listed in the reference list.

Statistical analysis

A funnel plot of the log odds ratiosversus their SE was calculated to pre-vent a selected subset of studies, whichis widely accepted to detect potentialpublication bias when the actual treat-

ment effects of an intervention or theoutcomes after exposure to certain para-meters are homogeneous (Light &Pillemer 1984). Assuming that theunderlying true exposure effect in eachstudy is the same, a general non-para-metric fixed-effects selection model was

applied to the entire collection of theselected studies to estimate the extentand size of undetected reports (Hedges1992, Blettner & Schlattmann 2005).

Furthermore, a mixed random-effectsmodel was applied since deviations ofthe estimated effect sizes may also beexplained by a random error. Therefore,a mixed model for calculation of hetero-geneity was used to detect heterogeneitybetween the studies selected (Blettner &Schlattmann 2005).

For dichotomous parameters (smo-kers and non-smokers) and outcomes

(lost implants and successful implants),the estimate of the effect of the para-meter smoking was expressed as oddsratio (OR) together with the 95% con-fidence interval (CI95%) after perform-ing univariate analysis, utilizing the dataobtained from the studies providing withinformation about failed and successfulimplants related to smokers and non-smokers. According to the patient- orimplant-related data basis, the synthe-sized ORs were calculated separately.

As calculation of an OR is undefinedif one value of the cells of the cross table

is equal to zero, for the studies con-cerned, 0.5 was added to the values ofall cells as suggested by Gart & Zweifel(1967) and Fleiss (1981).

As sample sizes of the included stu-dies were different, the weighed meanvalues of the distributions of frequenciesand SD regarding smokers and non-smokers, male and female patients andimplants in the maxilla and mandiblewere calculated and compared.

Statistical calculations were per-formed using SPSS software version12.0 (SPSS Inc., Chicago, IL, USA)

and SAS software version 8.2 (SASInstitute Inc., Cary, NC, USA).

Results

Of the entire yield of 139 publicationsidentified with electronic and hand-search, 75 were finally excluded frommeta-analysis and systematic reviewafter full text assessment.

Information on failures as well assuccess proportions of implants amongsmokers and non-smokers on patient-related and/or implant-related basis

Handsearch

Result: 1 abstract

Final decision:

29 publications included for meta-analysis35 publications included for systematic review

Independent selection of abstracts:

Reviewer 1: 59 abstracts included Reviewer 2: 58 abstracts included80 abstracts excluded 81 abstracts excluded

agreement with 47 included and 70 excluded abstracts

score = 0.712

Full text assessment and discussion:

Reviewer 1: 60 publications included Reviewer 2: 59 publications included79 publications excluded 80 publications excluded

agreement with 57 included and 78 excluded publications

score = 0.985

Search with

electronic databases

Result: 138 abstracts

Fig.1 . Search strategy and results of identification and inclusion of publications considered

for meta-analysis and systematic review. The k-score expresses the agreement between thereviewers.

Smoking interferes with the prognosis of dental implant treatment 525



4/22

were retrieved for meta-analysis from29 included studies. Of these, 15 wereprospective studies (three reportingpatient-related data, nine reportingimplant-related data and three reportingboth patient- and implant-related data)and 14 were retrospective studies (threereporting patient-related data, ninereporting implant-related data and tworeporting both patient- and implant-related data).

Furthermore, 35 studies wereincluded for systematic review, report-ing on implant failures or biologiccomplications or findings known tonegatively influence the success ofimplant-prosthetic therapy and compli-cations in connection with augmentationprocedures, related to smoking habit. Ofthese 35 studies, one was a randomizedcontrolled trial, 11 were prospective and21 retrospective studies. Additionally,one meta-analysis and one systematicreview were included.

The risk of implant failures in smokers

compared with non-smokers: Results

obtained from meta-analysis

First, the existence of publication biashad to be investigated. Figure 2a and bshow the funnel plots of the ORs by theinverse of the variances of each studyincluded for meta-analysis, calculatedon an implant-related as well aspatient-related basis. Assuming thateffects of smoking are homogeneous,no asymmetry indicating missing trialswas detected. Considering the implant-

related data obtained from the studiesincluded for meta-analysis, the slope ofthe regression line was 0.014 (CI95% 0.0480.020, p50.412), and forpatient-related data, the slope of theregression line was 0.084 (CI95% 0.5290.361, p50.682), revealingno significant differences of the slopeof regression lines from zero. Therefore,no publication bias was detected, andthe sample of the selected studies was

considered adequate for meta-analysis.The overview of the characteristics ofstudies included for meta-analysis isgiven in Table 1. The weighed meanvalues of frequencies distributions ofsmokers and non-smokers, gender andimplant sites among maxilla and mand-ible of the studies considered for meta-analysis are shown in Tables 24.

Analysis of studies providing implant-

related data

Implant-related data were obtained from

18 studies. Calculations of pooled riskby applying the fixed-effects modelwithout covariates on studies providingimplant-related data showed an OR of2.25 (CI95% 1.962.59), revealing evi-dence for a significant associationbetween smoking and implant failure.The test of heterogeneity of includedstudies yielded p50.013. Because ofthe heterogeneity further calculationsusing the mixed model taking intoaccount the random effects of the stu-dies on implant-related studies revealedan OR of 2.38 (CI95% 1.932.93).

Considering the groups of differentmean observation periods as covariates,the OR for implant failures in smokerscalculated from studies with a meanobservation period of up to 1 year(n58) was 2.83 (CI95% 2.083.85).An OR of 2.25 (CI95% 1.902.67) wascalculated from studies reporting dataregarding mean observation periodsbetween more than one up to 5 years(n513). The implant-related ORs after

observation periods after up to 1 year aswell as between 1 and 5 years revealed asignificantly enhanced risk of implantfailures in smokers and did not differsignificantly (p50.1892). The OR ofimplant failures in smokers of threestudies with a mean observation periodof more than 5 years was 1.33 (CI95%0.842.11), which showed no signifi-cantly enhanced risk for implant failuresin smokers. Considering both groups ofobservation periods of more than 1 yearcombined, the ORs calculated with themixed fixed-effects model with the cov-

ariate observation period were 2.83(CI95% 2.083.85) for studies with amean observation period up to 1 yearand 2.12 (CI95% 1.802.48) for a meanobservation period of more than 1 year,which were not significantly different(p50.0939).

Two studies reporting on implantswith surfaces microstructured with acidetching and/or particle blasting (Grun-der et al. 1999, Kumar et al. 2002), asummarized OR for implant failure insmokers of 1.49 (CI95% 0.643.46) wascalculated, considering implant-related

60.0050.0040.0030.0020.0010.000.00

weighting factor (1/variance) weighting factor (1/variance)

5.00

4.00

4.00

3.00

3.00

2.00

2.00

1.00

1.00

0.00

0.00

1.00

1.00

2.00

logo

dds

ra

tio

logo

dds

ra

tio

8.006.004.002.000.00

Fig.2. (a) Funnel plot of the log OR calculated for studies providing implant-related data (

n524). (b) Funnel plot of the log OR calculatedfor studies providing patient-related data (n512).




5/22

Table1

.Character

isticsofstud

ies

include

dformeta-analys

is

Stu

dy

Implantcharacterist

ics

Stu

dydesign

Smokers

Non-s

mokers

Samplesizes

Impl.successful

Fai

lure

Impl

.successful

F

ailure

Group1:meanobservationperiodperiod

41year

Bain

(1996)

Mac

hine

dtitanium

,threadedn

Prospect.

38

9

166

1

0

223implants

78patients

Balsh

i&Wol

finger

(1999)

Mac

hine

dtitanium

,threadedn

Retrospect.

1P

1P

27P

5

P

227implants

34patients

w

De

Bruyn

&Col

laert

(1994)

Mac

hine

dtitanium

,threadedn

Retrospect.

max

illa

man

dible

7111P

3610P

7 5P

0 0P

163

43P

171

45P

3 2

P

1 1

P

244implants

61patients

208implants

56patients

Gormanetal

.(1994)

Notreporte

d

Retrospect.

604

64P

4218P

1373

208P

4

7

2

0P

2066implants

310patients

Kronstrometal

.(2001)

Mac

hine

dtitanium

,threadedn

Retrospect.

3P

9P

37P

3

1P

80patients

Kumaretal

.(2002)

Large-gr

itblaste

dan

d

acid

-etc

hed,threaded

Prospect.

261

8

899

1

5

1183implants

461patients

Schwartz-A

radetal

.(2000)

Mac

hine

dtitanium

,threadedan

d

HA-coated

,threadedan

dcy

lindric

Retrospect.

5

1

45

5

56implants

z

43patients

van

Steen

berg

heetal

.(2002)

Mac

hine

dtitan

ium

,threadedn

Prospect.

148

8

1088

1

9

1263implants

399patients

van

Steen

berg

heetal

.(2004)

Mac

hine

dtitanium

,threadedn

Prospect.

immed

iate

load

ing

(Brane-m

arknovum

),man

dible

11P

2P

28P

4

P

150implants

45patients

Widmar

ketal

.(2001)

Mac

hine

dtitanium

,threadedn

Prospect.max

illa

,SFEAgrafte

d

sitesnon-gra

fted

416 35

26179

117

7047

1

4

8 6

198implants

43patients

101implants

97implants

Group2:meanobservation

period41and45years

Bain

&Moy

(1993)

Mac

hine

dtitanium

,threadedn

Prospect.

346

44

1718

8

6

2194implants

540patients

Berge

&Grnn

ingsaeter

(2000)

Ceram

icimplants

,threaded

Prospect.

40

21

48

7

116implants

30patients

Beschni

dtetal

.(2003)

Mac

hine

dtitanium

,threadedn

Prospect.

47

4

152

1

1

214implants

76patients

Geursetal

.(2001)

Notreporte

d

Retrospect.max

illa

,SFEA

55

7

266

1

3

341implants

Grunderetal

.(1999)

Aci

d-etc

hedtitanium

,threaded

Prospect.

55

0

161

3

219implants

74patients

Jonesetal

.(1999)

Cyl

indrica

lHA

-or

TPS

-coated

implants

Prospect.

115

12P

117P

212

42P

5 2

P

343implants

63patients

Kanetal

.(2002)

HA

-coatedthreadedan

dnon-threaded

,titanium

Retrospect.max

illa

,SFEA

58

12

147

1

1

228implants

60patients

Kel

leretal

.(1999)

Mac

hine

dtitanium

,threadedn

Retrospect.max

illa

,SFEA

13

4

106

1

6




6/22

Table1

.(Contd

.)

Stu

dy

Implantcharacterist

ics

Stu

dydesign

Smokers

Non-s

mokers

Samplesizes

Impl.successful

Fai

lure

Impl

.successful

F

ailure

139implants

37patients

Lam

bertetal

.(2000)

Notreported

Retrospect.

874

85

1813

1

15

2887implants

May

fiel

detal

.(2001)

TPS-coate

dan

dmac

hinedtitanium

,

threadedan

dcy

lindric

Prospect.

SFEA,

GBR

6 1P

7 2P

2612P

0 0

P

39implants

15patients

Moheng

&Feryn

(2005)

Cyl

indrica

lTPScoated

,an

dpart

icle

-blasted

threaded

implants

Prospect.

11P

4P

75P

3

P

93patients

Olsonetal

.(2000a)

HA

-coatedan

dnon-c

oate

dtitanium

,

threadedan

dcy

lindric

Retrospect.max

illa

,SFEA

48

3

65

0

116implants

36patients

Ortorp

&Jemt

(2002)

Mac

hine

dtitanium

,threadedn

Prospect.

33P

10P

78P

5

P

729implants

126patients

Penarrochaetal

.(2002)

TPS-coate

d,threaded

Retrospect.

27P

7P

70P

1

0P

441implants

114patients

Schwartz-A

radetal

.(2002)

Notreported

Retrospect.

175

205

402

1

77

959implants

261patients

Wal

lace

(2000)

Mac

hine

dtitanium

,threadedn

Retrospect.

60

12

107

8

187implants

56patients

Group

3:meanobservat

ion

period45years

De

Bruynetal

.(1999)

Microtexturean

dHA

-coate

d,threaded

Prospect.max

illa

248P

6 2P

237P

9 6

P

62implants

23patients

Karouss

isetal

.(2003)

TPS-coate

dhollowscrews

Prospect.

26

2

81

3

112implants

53patients

Minsk

&Polson

(1998)

Notreported

Retrospect.

HRTz

non-

HRTz

269

1698

266 6

908

47253

6

0

2 2

2

1263imp

lants

116patients

z

71HRT

379non-

HRT

nBranemarkfixturesorrelatedimplanttype

s.

wAllpatientssufferedfromdiabetesmellitus.

zImmediateimplants.

Biologiccomplicationsconsideredfailure(spontaneousimplantexposure,

requiringorn

otrequiringsurgicalinterventionorcomplete

implantfailure).

zFemalepatientsabovetheageof50years,

dividedintotwogroups:patientsreceivingp

ostmenopausalhormonereplacementtherapy

(HRT)andpatientsnon-receivingHRT(non-H

RT).

P,

patient-relateddata;HA-coated

,hydroxy

apatite-coated;TPS-coated,titaniumplasmas

pray-coated;Prospect.,

prospectivestudies;R

etrospect.,

retrospectivestudies.

SFEAimplantsinsertedinconjunctionwith

sinusfloorelevationandaugmentation.

GBRimplantsinsertedinconjunctionwith

bonedefectaugmentationbyguidedbonereg

eneration.




7/22

data and observation periods of 6 and 34months, respectively. After second stagesurgery and prosthetic loading, noimplant failure was noted (Grunderet al. 1999).

Table 5 shows the ORs and CI95% ofthe studies reporting implant-related data.

Data of studies reporting on implantsinserted with accompanying augmenta-tion procedures are not included here.

Analysis of studies providing patient-

related data

Patient-related data were obtained fromten studies. Calculations of pooled riskby applying the fixed-effects modelwithout covariates on studies providingpatient-related data showed an OR of2.64 (CI95% 1.704.09). Test of hetero-geneity of included studies yielded

p5

0.07, revealing no heterogeneity ofincluded studies.Table 6 shows the ORs and CI95% of

the studies reporting patient-related data.Data of studies reporting on implantsinserted with accompanying augmenta-tion procedures are not included here.

Analysis of studies considering

augmentation procedures

Considering studies providing implant-related data for implant treatment withaccompanying augmentation procedures

as covariate, the calculation using themixed effects model procedure revealedan OR of 2.15 (CI95% 1.862.49) forimplant failures in smokers withoutaugmentation procedures (n518 stu-dies), and an OR of 3.61 (CI95% 2.265.77) for implant failures in smokersundergoing augmentation procedures(n56 studies), which was significantlydifferent (p50.039).

Although in three studies at most atendency (Geurs et al. 2001) or nosignificant associations were foundbetween smoking and enhanced fre-

quency of implant failures in augmentedsites (Keller et al. 1999, Olson et al.2000a,b), three studies showed a sig-nificantly enhanced risk of implant lossin augmented sites in smokers, whichclearly exceeded the ORs for implantfailures calculated for smokers withoutaccompanying augmentation procedures(Mayfield et al. 2001, Widmark et al.2001, Kan et al. 2002).

Only one study provided patient-related data allowing for calculation ofthe OR for implant failures in smokersundergoing augmentation procedures.T

able2

.Weighedmeanva

lues

(WMV)an

dstan

dard

dev

iations

(SD)ofpercentage

distri

butionsofsmokers

(S)an

dnon-smokers

(NS)cons

ider

ing

differentobservat

ionper

iods(group

1:observat

ion

period41year;group

2:4

1an

d45years;group

3:45years

)an

dstud

iescons

idered

Observationperiod

Implant-re

late

d

Pat

ient-r

elated

WMV(%)

SD(%)

WMV(%)

SD(%)

Group

1

S24

.0

S20

.1

1.22

NS76

.0

0.58

NS79

.9

Bain

(1996),

De

Bruyn

&Col

laert

(1994)

,

Gormaneta

l.(1994),Kumaretal

.(2002),

Schwartz-Ar

adetal

.(2000),van

Steen

berg

heetal

.(2002)

,

Widmar

ketal

.(2001)

B

alsh

i&Wol

finger

(1999)

,

D

eBruyn

&Col

laert

(1994)

,

G

ormanetal

.(1994),Kronstrometal

.(2001),

K

umaretal

.(2002)

,van

Steen

bergheetal

.(2004),

W

idmar

ketal

.(2001)

Group

2

S28

.3

0.51

S30

.4

1.60

NS71

.4

NS69

.3

Bain

&Moy

(1993)

,Beschni

dtetal

.(2003),Geursetal

.(2001)

,

Grun

dereta

l.(1999),Jonesetal

.(1999),Kanetal

.(200

2),

Kel

leretal.

(1999)

,Lam

bertetal

.(2000),May

fiel

deta

l.(2001)

,

Olsonetal.

(2000a),Schwartz-A

radetal

.(2002)

,Wallac

e(2000)

B

erge

&Grnn

ingsaeter

(2000),

G

runderetal

.(1999)

,Jonesetal

.(1999),

M

ayfiel

detal

.(2001)

,Moheng

&Feryn

(200

5),

O

rtorp

&Jemt

(2002),Penarrochaetal

.(2002)

,

S

chwartz-A

radetal

.(2002)

,Wal

lace

(2000)

Group

3

S24

.6

1.14

S28

.9

5.20

NS75

.4

NS71

.1

De

Bruynet

al.

(1999),Karouss

isetal

.(2003),Minsk

&

Polson

(1998)

D

eBruynetal

.(1999)

,Karouss

isetal

.(2003

)




8/22

Table 7 shows ORs and CI95% calcu-lated from studies reporting implant- orpatient-related data, including augmen-tation procedures.

Results obtained from the systematic

literature review

Implant failure

Five studies reporting on RR or HR forimplant failures in smokers, amongthem one study reporting on HR forimplant failures among smokers afteraugmentation procedures are listed inTable 8. Apart from one study consider-ing a 10-months observation period(Eckert et al. 2001), for smokers, theRR or HR for implant failures with or

without augmentation procedures werereported significantly enhanced.

Furthermore, three studies reportedon significantly enhanced implant fail-ure frequencies in smokers, comparedwith non-smokers. In contrast, one study(Lemmerman & Lemmerman 2005) andone meta-analysis (Bain et al. 2002)including implants with recently intro-duced implant surfaces microstructuredby acid-etching, revealed no significantinfluence of smoking on implant failurefrequency (Table 9).

The risk of peri-implant inflammation

Soft tissue and inflammatory peri-implant complications associated with

current tobacco smoking were assessedby five retrospective and two prospec-tive studies, revealing an enhanced riskfor smokers to develop peri-implant softtissue complications. One prospectivestudy investigating the effect of smok-ing on implants in a small cohort of

periodontally compromised patientsshowed an enhanced risk of implantfailures in smokers (Leonhardt et al.2003) (Table 10).

The risk of peri-implant bone loss

Thirteen studies focused on radio-graphic assessment of peri-implantbone loss. In 11 studies, a significantlyenhanced marginal bone loss was foundin smokers, compared with non-smo-kers, including two studies investigatingthe effect of smoking on marginal bone

loss around implants in patients under-going a periodontitis or peri-implantitistreatment.

Two studies considered implants withmicrostructured surfaces conditioned byparticle blasting and/or acid-etching oranodic oxidization. One of these studiesrevealed no significant differences ofmarginal bone loss around implants insmokers compared with non-smokers(Aalam & Nowzari 2005). In anotherstudy significantly more bone loss wasfound at implants in smokers with peri-apical radiographs, which was not repro-

ducible in panoramic radiographassessment, however (Penarrocha et al.2004) (Table 11).

Smoking as a risk factor for

augmentation procedures

Four retrospective studies consideredthe impact of smoking on the courseand outcome of horizontal and/or verti-cal augmentation procedures, SFEA andGBR for the treatment of alveolar ridgedefects before or simultaneously withimplantation. Although the augmenta-

tion methods and materials as well asbarrier materials used were different,three studies reported significantlymore failures and complications in smo-kers, compared with non-smokers. Thedefect filling in smokers was found lesscompared with non-smokers but showedno significant difference, however(Table 12).

In a systematic review smoking wasconfirmed as a significant risk factor forimplant failure in connection with SFEAprocedures (0.03opo0.05) (Strietzel2004).

Table 3. Weighed mean values (WMV) and standard deviations (SD) of percentage distributionsof female (F) and male patients (M) considering different observation periods (group 1:observation period 41 year; group 2:41 and 45 years) and studies considered

Observation period WMV (%) SD (%)

Group 1 F 57.0 1.80M 43.0

Balshi & Wolfinger (1999), De Bruyn &Collaert (1994), Schwartz-Arad et al. (2000),Kronstrom et al. (2001), van Steenberghe et al.(2002), van Steenberghe et al. (2004), Widmarket al. (2001)

Group 2 F 54.6 1.44M 45.3

Bain & Moy (1993), Berge & Grnningsaeter(2000), Beschnidt et al. (2003), Grunder et al.(1999), Jones et al. (1999), Kan et al. (2002),Keller et al. (1999), Mayfield et al. (2001),Moheng & Feryn (2005), Olson et al. (2000a),Penarrocha et al. (2002), [28-1]Wallace (2000)

Table 4. Weighed mean values (WMV) and standard deviations (SD) of percentage distributionsof implant sites (maxilla Mx; mandible Md) considering implant-related data and differentobservation periods (group 1: observation period41 year; group 2: 41 and45 years; group 3:45 years) and studies considered

Observation period WMV (%) SD (%)

Group 1 Mx 62.8 1.10Md 37.2

Balshi & Wolfinger (1999), De Bruyn &Collaert (1994), Kumar et al. (2002), Schwartz-Arad et al. (2000)

Group 2 Mx 51.0 0.62Md 49.0

Bain & Moy (1993), Beschnidt et al. (2003),Grunder et al. (1999), Jones et al. (1999),Lambert et al. (2000), Penarrocha et al. (2002),Wallace (2000)

Group 3 Mx 72.9Md 27.1

De Bruyn et al. (1999)




9/22

Discussion

Methods of evidence-based dentistryhave been introduced for optimizationof decision making processes in dentaldiagnostics and treatment, and for acomprehensive patient information in

preparation of diagnostic and therapeu-tic interventions, particularly beforeelective treatment. Therefore, patientsdecision making before implantpros-thetic rehabilitation should be supportedby information about risks and risk-associated factors.

This meta-analysis and systematicreview were performed to provide asummary of cumulative informationon smoking-associated risks for implantprognosis and outcomes after augmenta-tion procedures. Moreover, smokingreduction advice, given in conjunction

Table 5. Odds ratios (OR) and confidence intervals (CI95%) calculated from studies reporting implant-related data without consideration ofaugmentation procedures

nCalculated following the suggestion by Gart & Zweifel (1967) and Fleiss (1981).wFemale patients above the age of 50 years, divided into two groups: patients receiving postmenopausal hormone replacement therapy (HRT) and

patients non-receiving HRT (non-HRT).

Table 6. Odds ratios (OR) and confidence intervals (CI95%) calculated from studies reporting patient-related data without augmentation procedures

nCalculated following the suggestion by Gart & Zweifel (1967) and Fleiss (1981).




10/22

with dental health information mayhave a marked effect on smokersattitude to their habit and providesufficient incentive to cut down oreven stop smoking as well (Macgregor1996).

Although tobacco smoking is widelyaccepted as a risk factor for oral healthgenerally (Sham et al. 2003), smokingwas considered a risk factor for implanttreatment since first publication on thisissue by Bain & Moy (1993). Never-theless, the impact of consideration ofthe patients status as a smoker or non-

smoker in implant treatment planningseems not to be controversial but indis-tinct. A national questionnaire to NHS-consultants to evaluate their attitudesconcerning relevant medical and oralfactors considered in patient selectionfor dental implant treatment in 1999revealed, that among others smokingwas one of the most important medicalfactors contra-indicating implant treat-ment (Butterworth et al. 2001).

A survey among Finnish dentistsevaluating the association of variouspatient characteristics or possible con-

tra-indications for dental implant treat-ment revealed, that significantly moredentists working in the public or privatesector recommended implant therapycompared with staff of dental schoolsin case of smoking patients (p50.002).Older dentists (4049 years) were foundmore in favour for implant therapy insmokers than were younger dentists(3039 years) (Heinikainen et al.2002). Estimating smoking as a riskfactor for treatment decisions thereforemight differ among dentists. Thisimpression seems to be confirmed by

different attempts made to quantify theamount of smoked cigarettes per day:some studies included for meta-analysisand systematic review considered apatient a smoker regardless quantity ofsmoked cigarettes per day or quality oftobacco smoking (Bain & Moy 1993, DeBruyn & Collaert 1994, De Bruyn et al.1999, Jones et al. 1999, Geurs et al.2001). In other studies light and heavysmokers were distincted by quantifica-tion: patients smoking up to 10 cigar-ettes per day (Schwartz-Arad et al.2002) or less than 20 cigarettes per

day (Gruica et al. 2004) were consideredlight smokers, whereas patients smoking10 (Schwartz-Arad et al. 2002) or 20cigarettes per day or more (Gruica et al.2004) were considered heavy smokers.As quantification of smoking was dif-ferent considering the included studies,in this meta-analysis and systematicreview, any patient who smoked wasconsidered a smoker according to Wal-lace (2000). In all included studies,patient questionnaires were used todetect the smoking status. The relativelyhigh reliability of patient self-reports of

smoking habits were shown in severalstudies (Fox et al. 1989, OLoughlinet al. 2002).

Treatment outcome evaluation criter-ia were different in several includedstudies. Besides calculation of differ-ences between cumulative success ratesor implant failure frequency betweensmokers and non-smokers, frequenciesof implant failures or biologic compli-cations were recorded. The risk ofoccurrence of implant failures for smok-ing patients was expressed as OR, cal-culated from studies included into meta-

analysis by the fixed-effects model aswell as the random-effects model, andcalculated for single studies on basis ofunivariate analysis.

Our meta-analysis revealed a signifi-cantly enhanced risk for implant failureamong smokers compared with non-smokers, expressed by synthesizedORs of 2.4 considering all includedstudies providing implant-related data,or 2.6 considering all included studiesproviding patient-related data. Compar-ing the implant-related ORs for implantfailure in smokers considering different

observation periods, the risk of implantfailure for smokers ranged from 2.8 afterup to 1 year decreasing to about 2.3 upto 5 years, indicating a higher risk ofearly implant failure. Nevertheless, therisk of implant failures among smokerswas found significantly enhanced evenafter 5 years considering the findings ofthe systematic review. Here, the effectsof smoking on the enhanced frequencyof peri-implant soft tissue complica-tions, limited peri-implantitis treatmentoutcomes and enhanced peri-implantbone loss, known as factors limiting

the long-term implant prognosis wereemphasized.

In an earlier literature review onstudies reporting on threaded implantswith a machined surface (Branemarktype fixtures) predominantly, consensushad been reached that smokinghas a negative influence on implantsurvival (Esposito et al. 1998). Findingsof our meta-analysis and systematicreview were obtained from studiesinvolving threaded titanium implantswith machined, TPS- or HA-coatedsurfaces predominantly, whereas

Table 7. Odds ratios (OR) and confidence intervals (CI95%) calculated from studies reporting implant- or patient-related data, includingaugmentation procedures

nSFEA.wLateral alveolar ridge augmentation/GBR).zCalculated following the suggestion by Gart & Zweifel (1967) and Fleiss (1981).




11/22

studies including implants withmicrostructured surfaces recently intro-duced (acid etched and/or particleblasted) meeting the inclusion criteriafor this meta-analysis were rarelypublished.

One study on wide diameter-threaded

machined titanium implants revealed anon-significant HR of 2.4 for implantfailure in smokers after a 10-monthsobservation period (Eckert et al. 2001).Apart from three more studies includedinto systematic review indicating a sig-nificantly enhanced risk of implant fail-ure for smokers (Wilson & Nunn 1999,Kourtis et al. 2004, Ortorp & Jemt2004), four studies (Grunder et al.1999, Kumar et al. 2002, Aalam &Nowzari 2005, Lemmerman & Lem-merman 2005) and one systematicreview (Bain et al. 2002) on implants

with acid-etched, particle-blasted oranodic oxidized surfaces, revealed noassociations between implant failuresand smoking nor significant differencesof implant failure frequencies amongsmokers and non-smokers, respectively.Furthermore, a comparison betweenthreaded implants with machined andanodic-oxidized surfaces showed no sig-nificant influence of smoking on implantfailures for implants with an anodic-oxidized surface (Rocci et al. 2003).Although smoking was shown to be asignificant risk factor within the limits

of this meta-analysis and systematicreview referring to studies reporting onimplants with machined or TPS-or HA-coated surfaces predominantly,future studies focusing on implant sur-face condition using acid etchingand/or particle blasting should be ana-lysed regarding influence of tobaccosmoking on early as well as long-term outcomes to allow for a moresubstantiated statement concerning thisissue.

Our findings considering the implant-related OR for implant failures in smo-

kers were similar to those published byHinode et al. (2006), who performed ameta-analysis on the influence of smok-ing on osseointegrated implants, basedon implant-related data. As earlierinvestigations revealed a within-patientdependence of implant success rates(Herrmann et al. 1999, 2003), patient-related ORs were calculated separatelyto exclude cumulative effects of indivi-dual risk factors. Compared with 18studies providing implant-related data,only five prospective and five retrospec-tive studies reported patient-related dataT

able8

.Character

isticsofstud

iesreportin

gontheriskof

implant

failures

,includedintothesystemat

icreview

Stu

dy

Implantsystem

Stu

dydesign

Meanobservat

ionperiod

HRORRR

CI95%

p

Samplesizes

Bae

lum

&Ellegaard

(2004)Threa

ded,TPS

-coate

dorpart

icle

-blasted

Retrospect.

4

6years

HR2.6n

0.9

7.6

o

0.05

258impla

nts:

man

dible:

57,

max

illa:

201

140patien

tsw

Chuangetal

.(2002b)

HA

-or

TPS-c

oate

dor

uncoated

/pass

ivated

screw-l

ikes

hapew

ithfins

Retrospect.

48years

HR2.9z

1.6

5.3

o

0.01

2349imp

lants:

man

dible:

934,max

illa:

1415

677patien

ts

Eckertetal

.(2001)

Mac

hine

dtitaniumthreaded

1,z

Retrospect.

10months

HR2.4n

Notreported

5

0.16

85implan

ts:

man

dible:

57,

max

illa:

28

55patients

Moyetal

.(2005)

Different;m

ost

implants

mac

hine

dtitanium

,threaded

Retrospect.

Notreported

,

implante

dbetween

1982an

d2003

RR1.56k

1.0

2.4

o

0.05

4680imp

lants:

man

dible:

2427

,max

illa:

2253

1140patie

nts

Augmentationofalveolarcrestwidthand/orheightorinternal

orexternalsinusfloorelevationandaugm

entation

Wooetal

.2004)

HA

-or

TPS-c

oate

doruncoated

/passivate

d

screw-l

ikes

hapew

ithfins

Retrospect.

48years

HR4.4nn

2.0

9.8

o

0.001

677impla

nts:

man

dible:

252,max

illa:

425

677patien

ts

(1implantcons

ideredperpat

ient

)

nCoxsproportionalhazardmodel.

nnMultivariateCoxmodel.

wPeriodontallycompromisedpatients.

zClusteredCoxsregressionanalysis

.

Branemarkfixturesorrelatedimplanttype

s.

zBranemarkMark-I

Iwideplatformandwid

ediameterimplantswereusedexclusively.

k

Univariateanalysis

.

HR

,hazardratio;RR

,riskratio;expressing

theriskofimplantfailuresamongsmokers,comparedwithnon-smokers.




12/22

Table9

.Character

isticsofstud

iesreportingonthe

frequencyof

implant

failures

insm

okerscompare

dw

ithnon-smokers,

include

dintothesystemat

icreview

Stu

dy

Implantcharacterist

ics

Stu

dy

design

Meanobservat

ion

perio

d

Samplesizes

Eva

luat

ioncr

iter

ia,

stat

istics

,outcomes

p

Bainetal

.(2002)

Threaded

,mach

ined

titaniumn

or

acid

-etc

hed

w

imp

lants

Meta-analys

is60

.1mont

hsn

43.7

mont

hsw

3prospect

ivemu

lticenterstud

ies

(2609implantsn

in1013patients;

6prospect

ive

stud

ies

(2274imp

lants

w

in778patients

)

Imp

lantsurv

ival

Kap

lan-

Meiersurv

ivalanalys

is

3-year

CSR

Innon-smokers:

92.8

%nimp

lants

98.4

%wimplants;

Insmo

kers:

93.5

%nimp

lants

98.7

%wimplants

Nosign

ificant

differenceo

fimplant

surv

ivalwasobserved

betweensmok

ing

andnon-smok

ingpatients

Kourt

isetal

.(2004)

Cyl

indric

,TPS-c

oate

dor

threaded

,particle-

blaste

d

implants

Retrospective

4.6years

1692implants

in405pat

ients

(853implants

insmo

kers;

839implants

innon-s

mokers)

Fai

lure:

implantremovalor

loss

w2test

Signi

ficant

lyhigher

failurerate

insmo

kers

compare

dw

ithnon-s

mokers

o0.001

Lemmerman

&

Lemmerman

(2005)

Cyl

indric

,TPS-c

oate

dz;

threaded

,mac

hin

ed

titaniumn

andthreaded

,

acid

-etc

hed

w

imp

lants

Retrospective

5.3years

1003implants

in376pat

ients

146implants

z

348implantsn

655implants

w

Fai

lure:

implantremovalor

loss

ANOVA

Nocorrelat

ions

between

smok

ingan

dimplant

failures

5

0.945

Ortorp

&Jemt

(2004)

Threaded

,mach

ined

titanium

implants

RCT

5years

792implants

in126patients

Fai

lure:

imp

lantremovalor

loss

Life-ta

bleanalys

is

Signi

ficant

lyhigher

failurerate

insmo

kers

compare

dw

ithnon-s

mokers

pat

ient-r

elatedo0.01;

implant-re

late

do0.05

Wilson

&Nunn

(1999)Threaded

,TPS-c

oate

dor

mac

hine

dimplan

ts

Retrospective

Fai

ledimplants:

8months,

success-

fulimplan

ts:

3.2years

33failed

implants

in

27patients

68successfu

limp

lants

in38pat

ients

Fai

lure:

Imp

lant

lossor

50%bone

loss

RR2.5

(CI95%

1.13

5.5

5)

Log-r

anktest

Signi

ficant

lyhigher

failurerate

insmo

kers

compare

dw

ithnon-s

mokers

5

0.024

nBranemarkfixturesorrelatedthreadedtita

niumimplantswithamachinedsurface.

wThreadedtitaniumimplantswithanacid-e

tchedsurface.

zTPS-coatedcylindrictitaniumimplants.

RCT

,randomizedclinicaltrial;RR

,riskratio;expressingtheriskofimplantfailuresamongsmokers,comparedwithnon-smokers.CSR

,cumulativesurvivalrate.




13/22


14/22


15/22

without consideration of accompanyingaugmentation procedures and wereincluded into meta-analysis. Thepatient-related OR for implant failuresamong smokers was calculated 2.6 andwas found significantly enhanced,although three studies (De Bruyn et al.

1999, Penarrocha et al. 2002, van Steen-berghe et al. 2004) revealed no signifi-cantly enhanced ORs, and one study(Kronstrom et al. 2001) revealed a ten-dency of an enhanced OR for implantfailure. Nevertheless, this implicates,that even considering patient-relateddata, the risk of implant failures forsmokers should be considered criticallyin patient information.

Besides univariate analysis and con-sideration of implant failures in smokersin the meta-analysis, peri-implant para-meters as there are peri-implant mucosal

status or peri-implant bone level knownto be associated with the implant prog-nosis (Misch 1993, Buser et al. 1997,Roos et al. 1997, Sennerby & Roos1998, Schubert et al. 2001, Cochran etal. 2002) were included into the sys-tematic review concerning associationswith patients tobacco smoking status.The systematic review identified studiesinvestigating on peri-implant soft tissuesand the quality and components of theperi-implant crevicular fluid, revealing asignificantly enhanced risk for peri-implant inflammatory complications in

in periodontally compromised smokers(Table 10). Although these studiesfocused on single aspects and thereforethe supposed complex mechanisms ofsmoking side effects on the peri-implantsoft tissues might not be completelydiscovered, an enhanced risk for inflam-matory peri-implant complicationsmight be expected among others due to tobacco smoking-associatedvasoconstrictive effects at the end-arter-ial gingival vessels (Sham et al. 2003)and the significant decrease of neutro-phil elastase activity in smokers, which

might result in a reduced inflammatoryreaction in smokers (Ataoglu et al.2002). These effects seem to blur theinflammatory peri-implant reaction,whereas certain biomarkers for earlyperi-implant bone loss, as thereare pyridinoline (Oates et al. 2004) orb-glucuronidase levels (Schubert et al.2001) were found significantlyenhanced in peri-implantitis. Therefore,a regular and strict recall of smokersundergoing implant treatment is neces-sary for early detection of implant com-plications.T

able11

.(Contd

.)

Stu

dy

Implant

characterist

ics

Stu

dy

design

Meanobservat

ionperio

d

Samp

lesizes

Statist

ics,outcomes

p

Nitzanetal

.(2005)

Notreporte

d

Retrospective

42.9

months

forsmokers;

48.4months

fornon-smokers

646implants

in161patients

271implants

in59smokers;

375implants

in

102non-smokers

t-test

ANOVA

Signi

ficantly

highermean

bone

loss

aroun

dimplants

insmokers

bot

hin

the

max

illaan

dman

dible,compare

dwith

non-smokers

5

0.0

01

Penarrochaetal

.(2004)

San

d-blaste

d,large-gr

it,

acid

-etc

hed,

threaded

implants

Retrospective

1year

108implants

in42patients

47implants

in16smokers,

61implants

in26non-smo

kers

ANOVA

Assessmentofpanoram

icra

diograp

hs

showednosign

ificantassociat

ion

betweenmargina

lbone

lossan

dsm

oking

Per

iapica

lra

diographsrevealedapos

itive

linearassociat

ion

betweensmok

ing

andperi-i

mplant

bone

loss

o0

.01

Schwartz-A

radetal

.(2005a)

Notreporte

d

Retrospective

37.9

months

277implants

in61patients

50implants

ineigt

hsmokers;

227implants

in53non-smo

kers

w2test

Signi

ficantlymorecerv

ical

boneloss

at27of

50imp

lants

insmokersan

d

at54of

227implants

innon-smoke

rs

o0

.0001

Investigationsontheeffectofsmokingon

peri-implantbonelevel

inpatientsundergoingtreatmentofperiod

ontitisorperi-implantitis

Wennstrometal

.(2004a)

Particle-

blas

ted,threaded

titanium

imp

lants

Prospective

At

least1

year,

upto

5years

148implants

in51patients

137implants

in47patients

t-test

,multipleregressionmodels

Signi

ficantlyen

hanced

bone-l

evel

changearound

dental

implantsafter

a5

-yearsobservat

ionper

iodin

smo

kerscompare

dw

ithnon-smokers

5

0.0

22

Wennstrometal

.(2004b)

Particle-

blas

ted,

threadedtita

nium

implants

Prospective

5years

47patients

(15smokers,

32non-s

mokers)

Stepw

ise

backwar

dregressionanalys

is

Signi

ficantgreater

bone

loss

in

smo

kingpatientscompare

d

withnon-s

mokers

o0

.05




16/22

Smoking was also found to have asignificant negative impact on the mar-ginal bone level at implants withmachined or TPS- or HA-coated sur-faces, which again was not confirmed bytwo studies on implants with particle-blasted and acid-etched (Penarrocha

et al. 2004) or dual acid-etched oranodic-oxidized surfaces (Aalam &Nowzari 2005), showing no associationbetween smoking and marginal boneloss. As shown for implant failure,microstructured implant surfaces seemto have positive influence on marginalbone level around implants in smokerswithin the limits of this literaturereview.

The impact of smoking on the out-come of SFEA and GBR seems rarelyreported considering the studies identi-fied within this literature search. This

might be due to the fact, that smoking isoften considered an exclusion criterionfor patient recruitment for prospectivestudies on SFEA or GBR procedures.Nevertheless, our meta-analysisrevealed a synthesized OR of about 3.6and, therefore a significantly enhancedrisk of implant failures for smokersundergoing augmentation procedures.These findings based on studies report-ing outcomes after SFEA and confirmedearlier findings from a systematicreview (Strietzel 2004). Two studies onGBR and ridge augmentation proce-

dures showed a significantly enhancedrisk of graft or barrier membrane expo-sures requiring partial or complete graftor membrane removal (Strietzel 2001,Schwartz-Arad et al. 2005b). A study onoutcomes after alveolar ridge defectaugmentation using deproteinizedbovine bone mineral for defect fillingand a porcine collagen membrane forGBR revealed a reduced defect filling insmokers, which was not significantlydifferent from non-smokers (Zitzmannet al. 1999). Whether the use of collagenmembranes significantly minimizes the

risk of negative treatment outcomes forGBR in smokers, was not confirmed dueto insufficient number of available stu-dies and data, respectively.

Reviews identified within this litera-ture search confirmed, that smoking isone of the factors related to implantfailure by reporting conclusions of sev-eral studies showing that smoking isassociated with higher failure rates,complications and altered peri-implanttissue conditions (Esposito et al. 1998,Sennerby & Roos 1998, Sham et al.2003, Wood & Vermilyea 2004). AnT

able12

.Character

isticsofstud

iesreport

ingonoutcomesofaugmentationprocedures

andGBRbeforeorsimultaneouslyw

ithimp

lantat

ion

insmokerscompare

dw

ithnon-sm

okers,

included

intothe

systemat

icreview

Stu

dy

Augmentatio

nproce

dure

Observat

ionpara

meters

Samplesizes

Statist

ics,outcomes

p

Lev

inetal

.(2004)

Autogenouson

layb

onegrafts

(OBG)n

,sinus

floor

elevat

ion

andaugmentation

(SFEA)nn

Fol

low-u

pat

least6

months

postsurgery

Compl

icat

ions:

graftexposureormo

bilityn;

swel

ling

,sinus

infectionnn

64OBGin56patients;

79SFEAin72patients

ANOVA;

Pearson

scorrelat

ioncoef

ficient

Compl

icat

ions

follow

ing

OBG

occurred

in1/3ofsmokersan

din7.7%ofnon-s

mokers.

Compl

icat

ions

follow

ing

SFEAoccurred

in

23.2

%

ofsmo

kersan

d6%o

fnon-s

mokers

5

0.04

Schwartz-A

rad

etal

.(2005b)

Autogenous

bone

blockgrafts

for

horizonta

lan

d/orve

rtical

ridgeaugmentation

before

implantation

nobarr

iermem

bran

es

Compl

icat

ions:

inflammatory

symptoms,

hematoma,swel

ling,

temporaryparesthesi

a

Fai

lures:

graftexposure

/remov

al

56patients

(11smokers)

,

64autogenous

bonegra

fts

w2test

Pearson

scorrelat

ioncoef

ficient

compl

icat

ions

in50%o

fsmokersan

din23%of

non-smokers

failures

in33%of

smokersan

din7.7%ofnon-smokers

5

0.04

Str

ietzel

(2001)

Bonesu

bstitutionm

ater

ialsw

ithor

withoutautogenous

bone

GBRw

ithnon-resorbab

lebarr

ier

mem

branes

(expand

ed

polytetrafluoro-e

thylene)

for

horizonta

lalveolarri

dge

defectaugmentation

Prematuremem

braneexposure

,

requ

iringmem

brane

removal

72patients

(19smokers)

,

72augmentationsites

w2test

Therisk

forprematuremem

brane

exposure

in

smokers

issign

ificantlyen

hanced

compare

dw

ithnon-smokers

5

0.04

Zitzmann

etal

.(1999)

Augmentationo

fho

rizontal

defects

(1-,

2-,

3-wal

ldefec

ts)around

implants

,us

ing

deproteinize

dbovin

ebone

mineralan

daporcineco

llagen

barr

iermem

branefor

GBR

Defectre

duct

ionafte

rre-e

ntry

75patients

(24smokers)

,

112implants

,

112augmentationsites

Wilcoxonstest

,w

2test

Themean

defectre

ductionat

implants

inserted

in

smokerswas

foun

dless

(82%)compare

dwith

non-smokers

(88%),but

didnot

differsigni

ficant

ly.

Smok

ingwasnot

foun

dto

besigni

ficant

lyassociated

withtreatmentsuccess

(OR0.51

,CI95%

0.211.28)

5

0.41

5

0.16




17/22

earlier review revealed a significantassociation between smoking and mar-ginal bone loss, biologic complications,reduced survival rate of implants(0.001opo0.05) and the outcome ofonlay bone grafts (po0.05) as well, butno significant correlation was found

between complications after SFEA andsmoking (Levin & Schwartz-Arad2005).

Besides the general suggestion to stopsmoking regarding a protocol suggested(Bain 1996) or to quit smoking conse-quently, the peri-operative use of anti-biotics (Gorman et al. 1994) as well asadditional local risk factors as for exam-ple the use of flat instead of high coverscrews should be considered in smokersto prevent postoperative and soft tissuecomplications (Schwartz-Arad et al.2002).

Moreover, individual peculiarities ofthe patients medical history might besupposed to confound or even enhancethe risk of implant failures attributableto smoking. Three studies were includedinto meta-analysis, although confound-ing parameters might be expected due toselected patient cohorts. Although Bal-shi & Wolfinger (1999) did not find asignificant influence of smoking onosseointegration in a population of dia-betic patients undergoing metaboliccontrol, a rate of 5.7% of non-osseointe-grated implants at time of second stage

surgery might indicate diabetes mellitusas a risk factor for osseointegration. Inpostmenopausal women undergoinghormone replacement therapy (HRT) toprevent osteoporosis (Minsk & Polson1998), a significant influence of smok-ing on the frequency of implant losswas found in contrast to patients notreceiving HRT as well as in theentire cohort. Thus, a detrimental inter-action between HRT and smoking wasconcluded. Besides smoking, hypothyr-eoidism was supposed to interfere withbone physiology even if correctly medi-

cally managed. Attard & Zarb (2002)found no significant differences inosseointegration between hypothyroidpatients and a matched healthy controlgroup, but hypothyroid patients showedsignificantly more soft tissue complica-tions during the postoperativecourse and significantly more marginalbone loss compared with the controlgroup.

Therefore, medical risk factorsshould be considered critically in con-text with a smoking history in decisionmaking before implant therapy.

In conclusion, smoking was identifieda significant risk factor for dentalimplant therapy. This should beaddressed thoroughly in patient infor-mation before implant treatment.

A strict recall regime throughout thewhole treatment course to early detect

negative changes of peri-implant tissuesor implant failure is necessary. Thisrequires identification of smokers at all.

As augmentation procedures com-prise enhanced potential risks of com-plications generally, smoking should beconsidered an additional risk and there-fore indication should be consideredcautiously and critically, evaluated incontext with additional risk factorsrevealed by medical history.

A limited number of studies availableon implants with sand-blasted, largegrit, acid-etched and/or acid-etched or

anodic-oxidized surfaces did not showsignificant associations between smok-ing and implant failure and marginalbone loss. Whether these implant sur-faces indeed significantly improve out-comes in smokers, has to be confirmedincluding more studies providing dataon implant failure in relation to smokersand non-smokers and by larger samplesizes as well.

References

Blettner, M. & Schlattmann, P. (2005) Chapter

II.2: meta-analysis in epidemiology. In:

Ahrens, W. & Pigeot, I. (eds). Handbook of

Epidemiology, pp. 829859. Heidelberg:

Springer.

Buser, D., Mericske-Stern, R., Bernard, J. P.,

Behneke, A., Behneke, N., Hirt, H. P., Belser,

U. C. & Lang, N. P. (1997) Long-term

evaluation of non-submerged ITI implants.

Part I: 8-year life table analysis of a prospec-

tive multi-center study with 2359 implants.

Clinical Oral Implants Research 8, 161172.

Butterworth, C. J., Baxter, A. M., Shaw, M. J. &

Bradnock, G. (2001) The provision of dental

implants in the National Health Service hos-

pital dental services a national question-

naire.British Dental Journal 190 , 9396.

Cochran, D. L., Buser, D., ten Bruggenkate, C.

M., Weingart, D., Taylor, T. M., Bernard, J.-

P., Peters, F. & Simpson, J. P. (2002) The use

of reduced healing times on ITI implants with

a sandblasted and acid-etched (SLA) surface:

early results from clinical trials on ITI SLA

implants. Clinical Oral Implants Research

13, 144153.

Esposito, M., Hirsch, J. M., Lekholm, U. &

Thomsen, P. (1998) Biological factors con-

tributing to failures of osseointegrated oral

implants. (II). Etiopathogenesis. European

Journal of Oral Sciences 106 , 721764.

Fielding, J. A. (1985) Smoking: health effects

and control. New England Journal of Medi-

cine 313 , 491498.

Fleiss, J. L. (1981)Statistical Methods for Rates

and Proportions, 2nd edition, pp. 6167.

New York: John Wiley.

Fox, N. L., Sexton, M., Hebel, J. R. & Thomp-

son, B. (1989) The reliability of self-reports

of smoking and alcohol consumption bypregnant women. Addictive Behaviors 14,

187195.

Gart, J. J. & Zweifel, J. R. (1967) On the bias of

various estimators of the logit and its var-

iance with application to quantal bioassay.

Biometrika 54 , 181187.

Hedges, L. V. (1992) Modeling publication

selection effects in random effects models

in meta-analysis. Statistical Science 7, 246

255.

Heinikainen, M., Vehkalahti, M. & Murtomaa,

H. (2002) Influence of patient characteristics

on Finnish dentists decision-making in

implant therapy. Implant Dentistry 11, 301

307.Herrmann, I., Lekholm, U. & Holm, S. (2003)

Statistical outcome of random versus selected

withdrawal of dental implants. International

Journal of Prosthodontics 16 , 2530.

Herrmann, I., Lekholm, U., Holm, S. & Karls-

son, S. (1999) Impact of implant interdepen-

dency when evaluating success rates: a

statistical analysis of multicenter results.

International Journal of Prosthodontics 12,

160166.

Hinode, D., Tanabe, S., Yokoyama, M., Fujisa-

wa, K., Yamauchi, E. & Miyamoto, Y. (2006)

Influence of smoking on osseointegrated

implant failure: a meta-analysis. Clinical

Oral Implants Research 17 , 473478.

Jones, J. K. & Triplett, R. G. (1992) The

relationship of cigarette smoking to impaired

wound healing: a review of evidence and

implications for patient care. Journal of

Oral and Maxillofacial Surgery 50, 237239.

Jorgensen, L. N., Kallehave, F., Christensen, E.,

Siana, J. E. & Gottrup, F. (1998) Less

collagen production in smokers. Surgery

123, 450455.

Kenney, E. B., Kraal, J. H., Saxe, S. R. & Jones,

J. (1977) The effect of cigarette smoking on

human PMN leukocytes. Journal of Perio-

dontal Research 12 , 227234.

La Veccia, C., Boyle, P. & Franceschi, S.

(1991) Smoking and cancer with emphasis

on Europe. European Journal of Cancer 27,94104.

Lehr, H. A. (2000) Microcirculatory dysfunc-

tion induced by cigarette smoking. Microcir-

culation 7 , 367384.

Lemons, J. E., Laskin, D. M., Roberts, W. E.,

Tarnow, D. P., Shipman, C. Jr., Paczkowski,

C., Lorey, R. E. & English, C. (1997)

Changes in patient screening for a clinical

study of dental implants after increased

awareness of tobacco use as a risk factor.

Journal of Oral and Maxillofacial Surgery 55

(Suppl. 5), 7275.

Levin, L. & Schwartz-Arad, D. (2005) The

effect of cigarette smoking on dental implants




18/22

and related surgery. Implant Dentistry 14,

357361.

Light, R. J. & Pillemer, D. B. (1984) Summing-

Up: The Science of Reviewing Research.

Cambridge: Harvard University Press.

MacFarlane, G. D., Hetzberg, M. C., Wolff, L.

& Hardie, N. A. (1992) Refractory perio-

dontitis associated with abnormal PMN leu-

cocyte phagocytosis and cigarette smoking.Journal of Periodontology 63 , 908913.

Macgregor, I. D. M. (1996) Efficacy of dental

health advice as an aid to reducing cigarette

smoking. British Dental Journal 180, 292

296.

Meechan, J. G., Macgregor, I. D., Rogers, S. N.,

Hobson, R. S., Bate, J. P. & Dennison, M.

(1988) The effect of smoking on immediate

post-extraction socket filling with blood and

on the incidence of painful socket. British

Journal of Oral and Maxillofacial Surgery

26, 402409.

Miller, P. D. (1988) Regenerative and recon-

structive periodontal plastic surgery. Dental

Clinics of North America 32 , 287312.

Misch, C. (1993) Implant success or failure:

clinical assessment in implant dentistry. In:

Misch, C. (ed). Contemporary Implant Den-

tistry, pp. 2942. St. Louis: Mosby.

OLoughlin, J., Tarasuk, J., Difranza, J. &

Paradis, G. (2002) Reliability of selected

measures of nicotine dependence in adoles-

cents.Annals of Epidemiology 12 , 353362.

Palmer, R. M., Wilson, R. F., Hasan, A. S. &

Scott, D. A. (2005) Mechanisms of action of

environmental factors tobacco smoking.

Journal of Clinical Periodontology 32

(Suppl. 6), 180195.

Peto, R., Lopez, A. D., Boreham, J., Thun, M.,

Health, C. & Doll, R. (1996) Mortality from

smoking worldwide.British Medical Bulletin52, 1221.

Ramseier, C. A. (2005) Potential impact of

subject-based risk factor control on perio-

dontitis. Journal of Clinical Periodontology

32 (Suppl. 6), 283290.

Reichart, P. A., Kirchheim, A. & Lochte, K.-H.

(2000) Tobacco and oral health. Question-

naire about knowledge, practices and opi-

nions among dentists in Berlin. Mund-,

Kiefer- und Gesichtschirurgie 4, 4549, (arti-

cle in German).

Roos, J., Sennerby, L., lekholm, U., Jemt, T.,

Grondahl, K. & Albrektsson, T. (1997) A

qualitative and quantitative method for eval-

uating implant success: a 5-year retrospective

analysis of the Branemark implant. Interna-

tional Journal of Oral and Maxillofacial

Implants 12 , 504514.

Sands, T., Pynn, B. R. & Nenniger, S. (1993)

Third molar surgery: current concepts and

controversies. Part 2.Oral Health 83, 1921,

2730.

Schubert, U., Kleber, B.-M., Strietzel, F. P. &

Dorfling, P. (2001) CrossLaps and b-glucur-

onidase in peri-implant and gingival crevicu-

lar fluid. International Journal of Oral and

Maxillofacial Implants 16 , 252258.

Sennerby, L. & Roos, J. (1998) Surgical deter-

minants of clinical success of osseointegrated

oral implants: a review of the literature.

International Journal of Prosthodontics 11,

408420.

Sham, A. S., Cheung, L. K., Jin, L. J. & Corbet,

E. F. (2003) The effects of tobacco use on

oral health. Hong Kong Medical Journal 9,

271277.

Smith, D. E. & Zarb, G. A. (1989) Criteria for

success of osseointegrated endosseous

implants. Journal of Prosthetic Dentistry62, 567572.

Strietzel, F. P. (2003) Patients informed con-

sent prior to implant-prosthetic treatment: a

retrospective analysis of expert opinions.

International Journal of Oral and Maxillofa-

cial Implants 18 , 433439.

WHO (2005) Tobacco and Health. The role of

health professionals in tobacco control. WHO

Library Catalogue-in-Publication, pp. 78.

www.who.int/tobacco/resources/publications

Wood, M. R. & Vermilyea, S. G. (2004) A

review of selected dental literature on evi-

dence-based treatment planning for dental

implants: report of the committee on

Research in Fixed Prosthodontics of the

Academy of Fixed Prosthodontics. Journal

of Prosthetic Dentistry 92 , 447462.

References included into meta-

analysis

Bain, C. A. (1996) Smoking and implant failure

benefits from a smoking cessation protocol.



Bain, C. A. & Moy, P. E. (1993) The association

between the failure of dental implants and

cigarette smoking. International Journal of

Oral and Maxillofacial Implants 8, 609615.Balshi, T. J. & Wolfinger, G. J. (1999) Dental

implants in the diabetic patient: a retrospec-

tive study. Implant Dentistry 8 , 355359.

Berge, T. I. & Grnningsaeter, A. G. (2000)

Survival of single crystal sapphire implants

supporting mandibular overdentures.Clinical

Oral Implants Research 11, 154162.

Beschnidt, S. M., Muche, R., Krausse, A. &

Strub, J. R. (2003) Implant survival and

success rates in partially edentulous patients.

Part 1. Schweizerische Monatsschrift fur

Zahnmedizin 113, 396403.

De Bruyn, H. & Collaert, B. (1994) The effect

of smoking on early implant failure. Clinical


De Bruyn, H., Collaert, B., Linden, U., Johans-

son, C. & Albrektsson, T. (1999) Clinical

outcome of screw vent implants. A 7-year

prospective follow-up study. Clinical Oral

Implants Research 10 , 139148.

Geurs, N. C., Wang, I.-C., Shulman, L. B. &

Jeffcoat, M. K. (2001) Retrospective radio-

graphic analysis of sinus graft and implant

placement procedures from the Academy of

Osseointegration consensus conference on

sinus grafts. International Journal of Perio-

dontics and Restorative Dentistry 21, 517

523.

Gorman, L. M., Lambert, P. M., Morris, H. F.,

Ochi, S. & Winkler, S. (1994) The effect of

smoking on implant survival at second stage

surgery: DICRG interim report No. 5.

Implant Dentistry 3 , 165168.

Grunder, U., Gaberthuel, T., Boitel, N., Imo-

berdorf, M., Meyenberg, K., Andreoni, C. &

Meier, T. (1999) Evaluating the clinical

performance of the Osseotite implant: defin-

ing the prosthetic predictability.Compendium

of Continuing Education in Dentistry 20,628640.

Jones, J. D., Lupori, J., Van Sickels, J. E. &

Gardner, W. (1999) A 5-year comparison of

hydroxyapatite-coated titanium plasma-

sprayed and titanium plasma-sprayed cylin-

der dental implants.Oral Surgery, Oral Med-

icine, Oral Pathology, Oral Radiology and

Endodontology 87, 649652.

Kan, J. Y. K., Rungcharassaeng, K., Kim, J.,

Lozada, J. L. & Goodachre, C. J. (2002)

Factors affecting the survival of implants

placed in grafted maxillary sinuses: a clinical

report. Journal of Prosthetic Dentistry 87,

485489.

Karoussis, I. K., Salvi, G. E., Heitz-Mayfield, L.

A. J., Bragger, U., Hammerle, C. H. F. &

Lang, N. P. (2003) Long-term implant prog-

nosis in patients with and without a history of

chronic periodontitis: a 10-year prospective

cohort study of the ITI dental implant system.

Clinical Oral Implants Research 14, 329

339.

Keller, E. E., Tolman, D. E. & Eckert, S. E.

(1999) Maxillary antral-nasal inlay autoge-

nous bone graft reconstruction of compro-

mised maxilla: a 12-year retrospective study.



Kronstrom, M., Svenson, B., Hellman, M. &

Persson, G. R. (2001) Early implant failures

in patients treated with Branemark systemtitanium dental implants: a retrospective

study. International Journal of Oral and

Maxillofacial Implants 16 , 201207.

Kumar, A., Jaffin, R. A. & Berman, C. (2002)

The effect of smoking on achieving osseoin-

tegration of surface-modified implants: a

clinical report. International Journal of Oral

and Maxillofacial Implants 17 , 816819.

Lambert, P. M., Morris, H. F. & Ochi, S. (2000)

The influence of smoking on 3-year clinical

success of osseointegrated dental implants.

Annals of Periodontology The American

Academy of Periodontology 5 , 7989.

Mayfield, L., Skoglund, A., Hising, P., Lang, N.

P. & Attstrom, R. (2001) Evaluation follow-

ing functional loading of titanium fixturesplaced in ridges augmented by deproteinized

bone mineral. Clinical Oral Implants

Research 12 , 508514.

Minsk, L. & Polson, A. M. (1998) Dental

implant outcomes in postmenopasusal

women undergoing hormone replacement.

Compendium of Continuing Education in

Dentistry 19 , 859866.

Moheng, P. & Feryn, J. M. (2005) Clinical and

biological factors related to oral implant fail-

ure: a 2-year follow-up study. Implant Den-

tistry 14 , 281288.

Olson, J. W., Dent, C. D., Morris, H. F. & Ochi,

S. (2000a) Long-term assessment (5 to 71




19/22

months) of endosseous dental implants

placed in the augmented maxillary sinus.Annals of Periodontology The American

Academy of Periodontology 5 , 152156.

Ortorp, A. & Jemt, T. (2002) Clinical experi-

ence of CNC-milled titanium frameworks

supported by implants in the edentulous

jaw: a 3-year interim report.Clinical Implant

Dentistry and Related Research 4 , 104109.Penarrocha, M., Guarinos, J., Sanchis, J. M. &

Balaguer, J. (2002) A retrospective study

(19941999) of 441 ITI(r) implants in 114

patients followed-up during an average of 23

years. Medicina Oral 7, 144155.

Schwartz-Arad, D., Grossmann, Y. & Chaushu,

G. (2000) The clinical effectiveness of

implants placed immediately into fresh

extraction sites of molar teeth. Journal of

Periodontology 71, 839844.

Schwartz-Arad, D., Samet, N., Samet, N. &

Mamlider, A. (2002) Smoking and complica-

tions of endosseous dental implants. Journal

of Periodontology 73 , 153157.

van Steenberghe, D., Jacobs, R., Desnyder, M.,

Maffei, G. & Quirynen, M. (2002) Therelative impact of local and endogenous

patient-related factors on implant failure up

to the abutment stage. Clinical Oral Implants

Research 13 , 617622.

van Steenberghe, D., Molly, L., Jacobs, R.,

Vandekerckhove, B., Quirynen, M. & Naert,

I. (2004) The immediate rehabilitation by

means of a ready-made final fixed prosthesis

in the edentulous mandible: a 1-year follow-

up study on 50 consecutive patients. Clinical


Wallace, R. H. (2000) The relationship between

cigarette smoking and dental implant failure.

European Journal of Prosthodontics and

Restorative Dentistry 8 , 103106.Widmark, G., Andersson, B., Carlsson, G. E.,

Lindvall, A.-M. & Ivanoff, C.-J. (2001)

Rehabilitation of patients with severely

resorbed maxillae by means of implants

with or without bone grafts: a 3- to 5-year

follow-up clinical report. International Jour-

nal of Oral and Maxillofacial Implants 16,

7379.

References included for systematic

review

Aalam, A.-A. & Nowzari, H. (2005) Clinical

evaluation of dental implants with surfaces

roughened by anodic oxidation, dual acid-

etched implants, and machined implants.



Ataoglu, H., Alptekin, N. O., Haliloglu, S.,

Gursel, M., Ataoglu, T., Serpek, B. & Dur-

mus, E. (2002) Interleukin-1beta, tumor

necrosis factor-alpha levels and neutrophil

elastase activity in peri-implant crevicular

fluid. Clinical Oral Implants Research 13,

470476.

Attard, N. J. & Zarb, G. A. (2002) A study of

dental implants in medically treated

hypothyroid patients. Clinical Implant Den-

tistry and Related Research 4, 220231.

Baelum, V. & Ellegaard, B. (2004) Implant

survival in periodontally compromised

patients. Journal of Periodontology 75,

14041412.

Bain, C. A., Weng, D., Meltzer, A., Kohles, S.

S. & Stach, R. M. (2002) A meta-analysis

evaluating the risk for implant failure inpatients who smoke. Compendium of Con-

tinuing Education in Dentistry 23, 695699.

Carlsson, G. E., Lindquist, L. W. & Jemt, T.

Meta Analisis Cigarro Implante

Documents

Transcript of Meta Analisis Cigarro Implante