Summary of findings
Description of the condition
Infective endocarditis is a rare disease caused by infected vegetations (growths), which often occur on previously damaged or congenitally malformed cardiac valves or endocardium (heart chamber lining). The infecting organisms are usually bacteria but less commonly are fungi, particularly of the Candida species, which may enter the blood via a number of portals. Bacterial endocarditis (BE) is infective endocarditis caused by bacteria which enter the blood (bacteraemia). Bacteria may enter the blood through a variety of portals but especially mucosal surfaces. The gingiva and periodontal ligament, which surrounds all teeth, are almost constantly experiencing a degree of inflammation and as such are a potential point of entry for bacteria within the blood. Indeed, it has been demonstrated that everyday activities such as toothbrushing cause bacteraemia (Lucas 2000; Roberts 1999).
There is a generally accepted incidence of BE of approximately 10 per 100,000 of the population per year. Though rare, it is a life-threatening condition with a mortality of up to 30%, even with antibiotic therapy (Delahaye 1995; Netzer 2000; Verheul 1993). In the past the majority of patients who developed endocarditis had a known pre-existing cardiac defect. More recently, however, this trend has shifted with nearly half of the cases of endocarditis having no known previous cardiac disease (Duval 2012; Hoen 2002; Hoen 2006). Common cardiac conditions that put people at risk include previous endocarditis, prosthetic heart valves, valvular stenosis, ventricular septal defect and valvular damage following rheumatic fever (Danchin 2005; Farook 2012). People with some of these conditions have a higher risk of developing endocarditis, namely those with previous endocarditis and prosthetic heart valves (Durack 1994). The above conditions either cause changes in the surface of the heart lining (endocardium) or changes in the blood flow, which damage the endocardium and enable organisms in the blood to adhere and multiply forming bacterial vegetations. This leads to a severe systemic illness as well as having direct effects on the functioning of the heart. Fragments of the vegetations may break away and become lodged elsewhere in the circulation (embolism).
The incidence of BE is low and the proportion of cases arising as a result of dental treatment is arguable, estimated to be as low as 4% (Gendron 2000; Gunneroth 1984) and as high as 64% of cases of BE (Bennis 1995). Although dental procedures are commonly implicated in the causation of endocarditis, the number of cases where the temporal relationship can be demonstrated ranges between only 4% and 7.5% of cases (Gendron 2000).
Description of the intervention
Most dental procedures cause bacteraemia and it was believed that this may lead to BE. There has previously been a well established practice of administering antibiotics, typically penicillins, to individuals who are at risk of developing BE prior to procedures during which a bacteraemia may develop. Early guidelines supported this practice (Dajani 1997; EWP 1993). The rationale for this was that a high circulating dose of antibiotic will prevent the development of an infected vegetation on damaged endocardium and thus prevent endocarditis. However, since the publication of the first version of this review in 2004, some authorities have questioned the routine use of antibiotics for endocarditis prophylaxis (Strom 1998), arguing that the adverse effects of antibiotics may outweigh the potential benefits. For example, in France, one study stated that patients receiving penicillin were five times more likely to die from an anaphylactic (allergic) reaction than from endocarditis (Bor 1984), however this study only considered people with mitral valve prolapse who are not at greatly increased risk of endocarditis. Across Europe, the US and Australia there have been radical changes to the guidelines, moving away from giving antibiotics to all at risk patients to only advising that they are given to those at high risk (Farook 2012). The definition of 'high risk' varies across American, European and Australian regulatory bodies. More recently, in England and Wales, the National Institute for Health and Care Excellence (NICE) published guidance that was more restrictive, stating that no antibiotics were required for any interventional procedure (dental or other surgical site) (NICE 2008). Despite marked reductions in antibiotic prophylaxis for BE over the past 10 years, no increase in its incidence has been shown (Duval 2012; Thornhill 2011).
Why it is important to do this review
Antibiotic prophylaxis before invasive procedures has been a key strategy for preventing BE for several decades, and remains so in many parts of the world (Thornhill 2011). Antibiotic therapy for the treatment of BE is not in question; without antibiotic therapy, infective endocarditis is fatal (Durack 1994). However, the use of antibiotic prophylaxis for the prevention of BE remains controversial.
To determine whether prophylactic antibiotic administration, compared to no such administration or placebo, before invasive dental procedures in people at risk or at high risk of bacterial endocarditis influences mortality, serious illness or the incidence of endocarditis.
To determine whether the effect of dental antibiotic prophylaxis differs in people with different cardiac conditions predisposing to raised risk of endocarditis, and in people undergoing different high risk dental procedures.
If there was no evidence from randomised controlled trials or cohort studies on whether prophylactic antibiotics affected mortality or serious illness, and there was evidence from these or case-control studies suggesting that prophylaxis with antibiotics reduced the incidence of endocarditis, then the following would have also been assessed: whether the harms of prophylaxis with single antibiotic doses such as with penicillin (amoxicillin 2 g or 3 g) before invasive dental procedures, compared with no antibiotic or placebo, in people at high risk of endocarditis did not differ from the benefits in prevention of endocarditis.
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (RCTs) and controlled clinical trials (CCTs) would be included where available. Due to the low incidence of bacterial endocarditis it was anticipated that few such trials would be found. Cohort and case-control studies were included where suitably matched control or comparison groups had been studied.
Types of participants
RCTs and CCTs
Studies must have included adults or children, or both, who had any of the following pre-existing cardiac defects (that is patients known to be at risk): congenital heart defects, a history of rheumatic fever, and those at high risk with prosthetic heart valves (tissue and mechanical), or who had had endocarditis previously. People with pacemakers (and no other risk factors) were excluded.
The dental procedures which the patients may have undergone in the studies included: supragingival and subgingival scaling of teeth, extensive restorations of teeth, endodontics, and oral surgery including dental extractions. Procedures performed under local and general anaesthetic were considered.
Types of interventions
RCTs and CCTs
The prime intervention assessed was the administration of an antibiotic, compared with no such administration or placebo, before a dental procedure. Studies in which an antibiotic was administered post-operatively were included if this was part of a protocol in which the antibiotic was administered pre-operatively. The antibiotics could be administered by oral, intravenous or intramuscular routes but not topically.
Co-interventions may have included the following procedures: the use of mouthwash pre-operatively or the mechanical cleaning of teeth.
Types of outcome measures
RCTs and CCTs
The primary outcome measures were the following.
(1) Mortality or serious adverse events (from any cause) requiring hospital admission.
(2) The development of endocarditis following any dental procedure in a defined time period.
(3) The development of endocarditis due to other non-dental causes.
The secondary outcome measures were the following.
(1) Any recorded adverse events to the antibiotics.
(2) Cost implications of antibiotic provision for prophylaxis compared with the cost of care of those extra patients who develop bacterial endocarditis (BE).
Assessment of harms would have included all studies where potentially serious (such as would be expected to result in hospitalisation) or fatal side effects of a single antibiotic dose had been reported or assessed. Studies included would have been any studies already included in the review, as well as randomised controlled trials, cohort studies, case-control studies, uncontrolled trials, case series and case reports.
Characteristics of included cohort studies and case-control studies
Cohort studies to be included would fulfil the following criteria.
Participants were people at high risk of endocarditis (as above). Their progress was followed (no minimum time period) and invasive dental procedures carried out; use (or not) of prophylactic antibiotics at these visits and occurrence or not of BE, death or serious illness were recorded (as a minimum). It would have been possible to compare incidence of BE, and death or serious illness in those who received invasive dental procedures with and without antibiotics.
Case-control studies included fulfilled the following criteria.
The groups that were compared included a group of people at high risk of endocarditis who did develop BE, and a group of people at high risk of endocarditis who did not develop BE. Information was provided on the numbers of people in each group who had undergone invasive dental procedures within a (stated) set period, and the numbers who had received antibiotic prophylaxis concomitant with that dentistry. Post hoc it was decided that studies which excluded cases when they died due to endocarditis would be excluded as up to 30% of people who contract endocarditis will die of it and these participants may be different from those who do not die.
Search methods for identification of studies
For the identification of studies to be included or considered for this review, we developed detailed search strategies for each database that was searched. These were based on the search strategy developed for MEDLINE (OVID) but were revised appropriately for each database. The search strategy used a combination of controlled vocabulary and free text terms. Details of the MEDLINE search are provided in Appendix 1. We opted not to use a study design filter as the yield from the subject search was small.
We searched the following electronic databases:
- The Cochrane Oral Health Group's Trials Register (to 21 January 2013) (Appendix 2);
- The Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2012, Issue 12) (Appendix 3);
- MEDLINE via OVID (1946 to 21 January 2013) (Appendix 1);
- EMBASE via OVID (1980 to 21 January 2013) (Appendix 4).
No restrictions on language or date of publication were applied in the searches of the electronic databases.
Searching other resources
We searched the following trials registers for ongoing studies (to January 2013):
Only handsearching done as part of the Cochrane Worldwide Handsearching Programme and uploaded to CENTRAL was included (see the Cochrane Masterlist for details of journal issues searched to date).
Data collection and analysis
Inclusion and exclusion criteria
Study titles and abstracts obtained from the search were screened for inclusion independently by two review authors. The review authors were not blinded to the authors, institution or journal. Full text papers that were retrieved were similarly screened for inclusion independently by two review authors. The inclusion criteria were as stated above. Disagreements over inclusion were resolved by discussion; if agreement could still not be reached it was intended that the paper would be taken to a third party, but this did not occur.
Data and quality information were independently extracted by two review authors onto a custom designed data collection form. Briefly, in addition to bibliographic details of the paper, the key items of data were the study design, country of origin, details of the antibiotic intervention, study population details including risk factors and type of dental procedure. The outcome data collected from RCTs and CCTs included number of deaths; number of hospital admissions; the number of serious illnesses that would be expected to result in hospital admission; the number of cases of endocarditis; any other adverse events noted; and the number of people originally randomised to each group. The outcome data collected from cohort studies would have included the same information as for RCTs plus adjusted odds ratios or risk ratios and information on which factors were adjusted for. The outcome data collected from case-control studies included the adjusted odds ratio of a person at high risk of endocarditis having had antibiotic prophylaxis before invasive dentistry before either developing endocarditis (cases) or not (controls). Authors were contacted for further details of their studies as well as to assess inclusion, where necessary.
Assessment of study quality
Included studies were to be ranked according to study design: RCT, CCT, cohort study, case-control study.
All included studies were assessed for risk of bias by two review authors using the Cochrane risk of bias assessment tool for non-randomised studies. Domains assessed for each included study are: sequence generation, allocation concealment, confounding, blinding of outcome assessment, completeness of outcome data, risk of selective outcome reporting, and risk of other potential sources of bias. For the primary outcome the following factors have been identified as important confounders; other sources of fluoride, social class; ethnicity and residential history while the use of other fluoride sources is the only confounder considered to be relevant to the secondary outcome.
A description of the risk of bias domains was tabulated for each included trial, along with a judgement of low, high or unclear risk of bias.
A summary assessment of the risk of bias for the primary outcome (across domains) across studies would have been undertaken (Higgins 2011) had data allowed. Within a study, a summary assessment of low risk of bias was given when there was a low risk of bias for all key domains, unclear risk of bias when there was an unclear risk of bias for one or more key domains, and high risk of bias when there was a high risk of bias for one or more key domains.
Where data appeared ambiguous or incomplete, the study authors would have been contacted.
It was planned that data on the number of patients with each outcome event, by allocated treatment group (RCTs) or quantile (cohort studies), would be sought. We aimed to calculate a pooled estimate of the treatment effect for each outcome (separately) across RCTs, CCTs, cohort studies and case-control studies in a random-effects meta-analysis as an odds ratio (the ratio of odds in the prophylaxis group to the odds in the no prophylaxis group), since the odds ratio is the only good measure of association that works across prospective studies and case-control studies (Fleiss 1981). Heterogeneity between trial results was to be tested for using a standard Chi
For amoxicillin, if sufficient data were available, then the interventions of 2 g and 3 g amoxicillin would be considered separately. Also, if appropriate, subgrouping would be used to explore the effects of different underlying causes of at risk and high risk status for endocarditis, and of different invasive dental techniques.
It was also planned that sensitivity analysis would be carried out on any pooled analyses, removing studies where there appeared to be significant differences in risk factors for endocarditis between the groups compared and where this had not been adequately adjusted for.
Description of studies
A total of 1390 references were identified by the above search strategy. Assessment of the titles and abstracts, where available, resulted in 122 references of potential relevance; all of which were obtained in full text (Figure 1). A further 86 were rejected since they were clearly discussion papers, editorials or guidelines. A total of 36 references were assessed for inclusion, in duplicate. Of these full papers, 35 were eventually excluded (see Characteristics of excluded studies for further details). In particular, one potentially useful cohort study was identified but excluded as it was not possible to separate out those who received prophylaxis, or not, before dental treatment (Gersony 1977). Several case-control studies were identified where only some of the cases were at high risk of endocarditis before dental treatment (Strom 1998B; Strom 2000). It was evident that there were many review articles, commentaries and guidelines but few primary studies.
|Figure 1. Study flow diagram.|
No randomised controlled trials (RCTs) were identified, nor were any other controlled trials (quasi-randomised, historically controlled) or cohort studies identified. Three case-control studies were initially included (Imperiale 1990; Lacassin 1995; Van der Meer 1992) however on discussion with editors of the Cochrane Oral Health Group it was decided that two of these, Imperiale 1990 and Lacassin 1995, might have been severely biased and so should be excluded. In these studies people with endocarditis (cases) who died, about 20% of potential cases in both studies, were not included. Excluding this group, whose characteristics may have been very different from those cases who survived endocarditis, from the cases but not from the controls (where appreciably fewer people were likely to have died) will have made the characteristics of the two groups very different. Thus only one case-control study was included in this review, Van der Meer 1992.
The Van der Meer 1992 study collected details of all of the 349 people who developed definite native-valve endocarditis in the Netherlands over a two-year period (first November 1986 to first November 1988). Cases were eligible if they had previously had congenital heart disease, coarctation of the aorta, rheumatic or other valvular dysfunction, or mitral valve prolapse with mitral regurgitation. They had to have undergone a medical or dental procedure that required prophylaxis within 180 days of the onset of symptoms of endocarditis. Proxy responders (spouses or general practitioners) were used where cases were too ill to be interviewed or had died.
Controls had not been diagnosed with endocarditis but had one of the cardiac conditions and were outpatients at cardiology departments of one of five hospitals. Controls were matched for age (within the same five year age category) and had undergone a medical or dental procedure within 180 days of their interview. A random sample of potential controls was drawn, and where there were at least four controls per case all were contacted. Where there were fewer than four controls a further random sample was drawn.
For both groups, all information about invasive procedures and use of prophylaxis was checked with medical or dental specialists and pharmacists.
Risk of bias in included studies
The included study used a case-control design and was therefore at risk of the usual sources of bias associated with this observational study design. The study was judged to be at risk of selection bias and outcome reporting bias (Characteristics of included studies; Figure 2). There were some concerns regarding attrition bias as cases who were very ill or who died were included in the analysis via the use of proxy responders, however this did not occur for the 53/889 controls who died. In addition, there were concerns that it was unclear how similar the groups were with regard to type of cardiac risk factor, gender, or type of dental intervention.
|Figure 2. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.|
Overall, the observational and retrospective nature of the design conferred a high risk of bias.
Effects of interventions
In the included case-control study (Van der Meer 1992), of the 349 people with definite native-valve endocarditis 197 had previous heart disease (proxy responders were interviewed for 10 of these). Of these, 54 had undergone a medical or dental procedure with an indication for prophylaxis within the past 180 days, of whom in six a causal relationship was ruled out as it was unlikely that the agent isolated from the blood originated in the area of the procedure. Of the 48 people with endocarditis that were left, 44 had undergone a dental procedure which the paper identified as having a definite (24) or possible (20) indication for prophylaxis (none of these cases had used a proxy responder). Indications for definite prophylaxis were dental extractions and dental root work, while indications for possible prophylaxis were defined as dental scaling.
Of 889 potential controls who were sent an introductory letter, 689 were ineligible (53 had died, 29 had a prosthetic heart valve, 62 could not be located, 102 could not be contacted by phone, and 418 had not undergone an invasive dental or medical procedure within the past 180 days). The remaining 200 were interviewed by phone two to five days later; 181 of these controls had undergone a dental procedure with definite (79) or possible (102) indications for prophylaxis.
Seven of 24 cases and 16 of 79 controls had had appropriate prophylaxis for a dental procedure requiring definite prophylaxis within 180 days.
The characteristics of the cases and controls were not well described as those who had received a dental procedure (rather than a medical one) were not separated out in the publication (the separated data were provided by Professor Van der Meer). The median time between a dental procedure requiring definite prophylaxis and onset of endocarditis was 10 days in the cases, and the median time between a dental procedure requiring definite prophylaxis and interview was 71 days in the controls (data missing for 12 controls). The procedure was apical surgery in one case (4%) and one control (1%), dental avulsion in one case (4%) and 12 controls (15%), dental extraction in nine cases (38%) and 15 controls (19%), dental abscess in one case (4%) and one control (1%), removal of subgingival calculus in three cases (13%) and eight controls (33%), removal of calculus plus polishing of teeth in six cases (25%) and 34 controls (43%), and root canal therapy in three cases (13%) and eight controls (10%).
Including the cases and controls undergoing either definite or possible indication for prophylaxis, and including the four cases and 19 controls who underwent a non-dental indication, 69% of cases were male and 55% of controls. The median age of this larger group was 41 years for cases and 40 for controls (the controls were age-matched).
Seven of 21 cases and 9 of 46 controls had had appropriate prophylaxis for a dental procedure requiring definite prophylaxis within 90 days. Seven of 44 cases and 17 of 181 controls had had appropriate prophylaxis for a dental procedure requiring definite or possible prophylaxis within 180 days. Seven of 32 cases and 9 of 100 controls had had appropriate prophylaxis for a dental procedure requiring definite or possible prophylaxis within 90 days. No information was presented on the adjunctive use of mouthwash.
In each of these ways of assessing the data the proportion of those receiving prophylaxis was greater in the cases than in the controls. Translating the data so that we assessed the odds of developing endocarditis in those receiving prophylaxis compared with those not receiving prophylaxis we find an odds ratio (OR) which was not significantly different from the OR for any of the groupings (OR 1.62, 95% confidence interval (CI) 0.57 to 4.57 for those with a definite indication for prophylaxis over 180 days).
Only four cases developed endocarditis following non-dental medical interventions (within the past 180 days and with pre-existing cardiac indications for the use of prophylaxis), so assessment of the effects of prophylaxis in these cases was not possible.
It was unclear whether antibiotic prophylaxis was effective or ineffective against bacterial endocarditis in people at risk who were about to undergo an invasive dental procedure.
No studies were located that assessed mortality, serious adverse events requiring hospital admission, other adverse events or cost implications of treatment.
Because no significant protective effect of antibiotic prophylaxis was seen against endocarditis, we did not do a wide ranging search to pool information on the potential harmful effects of antibiotic prophylaxis as pre-specified in the protocol.
Summary of main results
This review update has identified no further studies meeting the review's inclusion criteria.
The one included case-control study, which included all of the people in the Netherlands who developed endocarditis following an invasive dental procedure while at known cardiac risk over a two-year period (24 individuals who underwent a procedure definitely requiring prophylaxis, and a further 20 that may possibly have required prophylaxis), provides no conclusive evidence about whether antibiotic prophylaxis is effective or ineffective against bacterial endocarditis in high risk individuals about to undergo an invasive dental procedure.
A non-statistically significant difference for the development of endocarditis over 180 days in those receiving prophylaxis compared with those not receiving prophylaxis was shown for those individuals with a definite indication for prophylaxis (OR 1.62, 95% CI 0.57 to 4.57). If we include data from the two excluded case-control studies (Imperiale 1990; Lacassin 1995) (excluded due to inappropriate comparisons, see Characteristics of excluded studies) a total of 59 participants developed endocarditis of the 67 people receiving appropriate prophylaxis and 148 people not receiving prophylaxis across all three studies (random-effects OR 0.56, 95% CI 0.15 to 2.15).
There are currently insufficient primary data to know whether antibiotic prophylaxis before invasive dental procedures in people at high risk of endocarditis does actually prevent endocarditis, deaths or other serious illness.
For the original version of this review we canvassed opinion among some healthcare workers with an interest in evidence based care and a separate group of dentists who had attended an evidence based practice course. Those healthcare workers and dentists who were blinded to the details of the intervention and condition, but shown the level of evidence and the non-significant odds ratio, stated that in this circumstance they would attempt to seek out further evidence (randomised controlled trials (RCTs), cohort studies etc) or involve their patients in the decision about whether to use the intervention, or both. However, those dentists who were not blinded, who were told that the topic was endocarditis prophylaxis using penicillin, stated that they would use prophylaxis despite the paucity of evidence and cited medico-legal reasons for this decision. It will be interesting how the National Institute for Health and Care Excellence (NICE) guideline (NICE 2008) is adhered to not just by dentists but other healthcare professionals to which it applies. A recent study examined the legal situation of claimants who had contracted endocarditis against the dentists who had recently treated them (Martin 2007). The authors concluded that as long as dentists adhere to current published guidelines there is little course for redress and it is eminently defensible in a court of law.
Overall completeness and applicability of evidence
As the usefulness of prophylaxis could not be established, we have not examined in detail the harms of antibiotic administration; this would be a systematic review in itself. Such a review would, however, be extremely valuable and could potentially be used by a wide spectrum of research workers and other systematic reviews. In the absence of a systematic review on the harms of penicillins, the most authoritative source is Meyler's Side Effects of Drugs (Aronson 2006). The range of potential side effects from the administration of antibiotics is vast, largely with a hypersensitive aetiology, but some direct toxic effects may also occur.
The effects of the NICE guidance on the incidence of endocarditis in the UK are going to be monitored using the Hospital Episode Statistics (HES). Tracked over a number of years and compared with the pre-NICE guideline era the incidence of endocarditis might be one method to answer this conundrum, albeit in a rather crude fashion.
Another under explored area is the cost of prophylaxis, both in terms of finance and health. The financial cost to health services of providing large quantities of prophylactic antibiotics must be weighed against the cost of treating patients who develop endocarditis, which, although significant, occurs in appreciably fewer patients than those potentially at risk. The health costs need considering, particularly the potential harms of the administration of antibiotics compared to the development of endocarditis. The involvement of health economists would be beneficial. This was explored in depth in the recent NICE guidance (NICE 2008).
Despite the varying guidelines produced over the years and the recent significant change recommended by NICE, it is important for medical and dental practitioners to remember that patients remain at risk of developing endocarditis. Many patients will develop endocarditis with organisms that have originated from the oral cavity. Whilst there is no evidence that dental treatment is directly related to the development of the disease or not, nor that prophylactic antibiotics can prevent the development of the disease or not, it would appear logical to recommend that the highest level of oral health should be achieved and maintained in at risk patients.
Quality of the evidence
The overall quality of the evidence is low, coming from a single study at high risk of bias. It would be useful to have evidence about the effectiveness of antibiotic prophylaxis of endocarditis in dentistry from higher levels of evidence. As the incidence of endocarditis is so low, a randomised controlled trial run over two years would require approximately 60,000 patients with a cardiac risk factor for endocarditis to be included (a cohort study over 10 years would require approximately 18,000 patients). Such a trial would require an intense international effort.
A larger, well conducted case-control study might be more feasible but would still require a large effort and multicentre participation. If including every endocarditis case in the Netherlands for two years produces only 24 appropriate cases then the area or time span covered will be very large indeed. Selection of appropriate controls is probably the most challenging aspect; ideally, as in Van der Meer's study, they would have had dental treatment in a predefined time and be matched very closely for sex, age and type of cardiac risk factor. Additionally, neither cases nor controls should be excluded for death or serious illness (use of proxy respondents would be ideal and this would require retrospective identification of controls as well as ongoing prospective identification of cases) and dental records would be available and be explicit about the use (or not) of prophylaxis. Full details would be collected on other factors which may compound the risk such as general well being, coexisting medical problems, socio-economic status and oral health status.
Agreements and disagreements with other studies or reviews
There is general agreement that there is little scientific evidence to support the effectiveness of antibiotic prophylaxis for the prevention of bacterial endocarditis (Duval 2012; Farook 2012; Thornhill 2011). This lack of evidence has led to variations in guideline recommendations with regard to who should or should not be prescribed antibiotic prophylaxis and who is or is not considered high risk for bacterial endocarditis. However, one area where most guidelines are in agreement is with regard to the need for regular dental surveillance to promote good oral hygiene, thus reducing the need for invasive dental procedures and subsequently reducing the risk of bacterial endocarditis (Chambers 2012).
Implications for practice
There remains no evidence that antibiotic prophylaxis is either effective or ineffective against bacterial endocarditis in people considered at risk who are about to undergo an invasive dental procedure. It is not clear whether the potential harms and costs of penicillin administration outweigh any beneficial effect.
Implications for research
A randomised controlled trial (RCT) would only be feasible in extensive areas of very centralised and organised health care, due to the large numbers of participants with risk factors for endocarditis required. A well designed multicentre cohort or case-control study is possible but would still require a very large and co-ordinated effort, and a great deal of attention would need to be paid to recruiting suitable control participants. These would be most feasible to perform in an area where registers exist so that it is possible to identify all people with current risk factors for endocarditis, and would randomise them or follow their dental histories in detail and identify outcomes completely and accurately.
A systematic review of the harms and costs associated with antibiotic use is needed, and such a review may be useful to assess the harms for systematic reviews of a number of different interventions. It would be important to assess the effects of type of antibiotic, route of administration, dose, previous history of reaction and duration of use on the side effects and adverse events experienced by people on antibiotic therapy.
Our gratitude goes to Dr Van der Meer (Academic Medical Center, University of Amsterdam) for splitting his data into people who had or had not had a relevant dental intervention, and to Dr Imperiale (Yale University School of Medicine) for his helpful information on his study. Thanks also to the Cochrane Non-Randomised Methods Group, and the Harms subgroup, for support, discussion and ideas about the review. We are grateful to Luisa Fernandez Mauleffinch of the Cochrane Oral Health Group for excellent support, Anne Littlewood (Cochrane Oral Health Group), Margaret Burke (Cochrane Heart Group) and Vittoria Lutje (Cochrane Infectious Diseases Group) for kindly searching their trials registers for us, and Marco Esposito, Helen Worthington, Sylvia Bickley, Jan Clarkson, Paul Coulthard, Jayne Harrison, David Rickard, Robin Richardson, Jan van der Meer, Robin Seymour and Jeremy Bagg for constructive comments on the manuscript.
Data and analyses
This review has no analyses.
Appendix 1. MEDLINE (OVID) search strategy
1. exp Endocarditis/
3. (endocardium adj5 (inflamm$ or infect$)).tw.
4. (ABE or SABE).ti,ab.
6. exp Dental prophylaxis/
7. exp Dentistry, operative/
8. exp Endodontics/
9. exp Oral surgical procedures/
10. ((oral or tooth or teeth) adj5 (surg$ or extract$ or restor$ or invas$ or scale or scaling or polish$ or endodontic$ or "root canal" or apicectom$ or apicoectom$)).tw.
11. (dental or dentist$).tw.
13. Antibiotic prophylaxis/
14. (antibiotic$ or anti-biotic$ or antimicrobial$ or anti-microbial$).tw.
15. Anti-bacterial agents/
16. exp Penicillins/
17. (amoxicillin$ or amoxycillin$ or amoxil or actimoxi or clamoxyl or hydroxyampicillin or penamox or trimox or wymox).tw.
18. ("penicillin v$" or apocillin or beromycin or betapen or fenoxymethylpenicillin or "Pen VK" or phenoxymethylpenicillin or "V-Cillin K" or vegacillin).tw.
20. (clindamycin or chlolincocin or cleocin or "Dalacin C").tw.
22. (cephalexin or cefalexin$ or ceporexine or Palitrex or cephahexin$).tw.
24. (azithromycin or azadose or azitrocin or azythromycin or gozal or sumamed or toraseptol or ultreon or vinzam or zentavion or zithromax or zitromax).tw.
26. (clarithromycin or biaxin).tw.
28. (cefazolin or ancef or cefamedin or cefamezine or cephamezine or cephazolin or gramaxin or kefzol or totacef).tw.
30. (ceftriaxone or benaxona or cefatriaxone or cefaxona or ceftrex or ceftriaxon$ or lendacin or longacef or longaceph or rocefalin or rocefin or rocephin$ or tacex or terbac).tw.
32. 5 and 12 and 31
Appendix 2. Cochrane Oral Health Group's Trials Register search strategy
#1 (endocarditis) AND (INREGISTER)
#2 ((endocardium and (inflamm* or infect*)):ti,ab) AND (INREGISTER)
#3 ((ABE or SABE):ti,ab) AND (INREGISTER)
#4 (#1 or #2 or #3) AND (INREGISTER)
#5 ((antibiotic* or anti-biotic* or antimicrobial* or anti-microbial*):ti,ab) AND (INREGISTER)
#6 ((antibacterial* or anti-bacterial*):ti,ab) AND (INREGISTER)
#7 ((penicillin* or amoxicillin* or amoxycillin* or amoxil or actimoxi or clamoxyl or hydroxyampicillin or penamox or trimox or wymox):ti,ab) AND (INREGISTER)
#8 ((apocillin or beromycin or betapen or fenoxymethylpenicillin or "Pen VK" or phenoxymethylpenicillin or "V-Cillin K" or vegacillin):ti,ab) AND (INREGISTER)
#9 ((clindamycin or chlolincocin or cleocin or "Dalacin C"):ti,ab) AND (INREGISTER)
#10 ((cephalexin or cefalexin* or ceporexine or Palitrex or cephahexin*):ti,ab) AND (INREGISTER)
#11 ((azithromycin or azadose or azitrocin or azythromycin or gozal or sumamed or toraseptol or ultreon or vinzam or zentavion or zithromax or zitromax):ti,ab) AND (INREGISTER)
#12 ((clarithromycin or biaxin):ti,ab) AND (INREGISTER)
#13 ((cefazolin or ancef or cefamedin or cefamezine or cephamezine or cephazolin or gramaxin or kefzol or totacef):ti,ab) AND (INREGISTER)
#14 ((ceftriaxone or benaxona or cefatriaxone or cefaxona or ceftrex or ceftriaxon* or lendacin or longacef or longaceph or rocefalin or rocefin or rocephin* or tacex or terbac):ti,ab) AND (INREGISTER)
#15 (#5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13 or #14) AND (INREGISTER)
#16 (#4 and #15) AND (INREGISTER)
Appendix 3. Cochrane Central Register of Controlled Trials search strategy
#1 [mh Endocarditis]
#3 (endocardium near/5 (inflamm* or infect*))
#4 (ABE or SABE):ti,ab
#5 #1 or #2 or #3 or #4
#6 [mh "Dental prophylaxis"]
#7 [mh ^"Dentistry, operative"]
#8 [mh Endodontics]
#9 [mh "Oral surgical procedures"]
#10 ((oral or tooth or teeth) near/5 (surg* or extract* or restor* or invas* or scale or scaling or polish* or endodontic* or "root canal" or apicectom* or apicoectom*))
#11 (dental or dentist*)
#12 #6 or #7 or #8 or #9 or #10 or #11
#13 [mh ^"Antibiotic prophylaxis"]
#14 (antibiotic* or anti-biotic* or antimicrobial* or anti-microbial*)
#15 [mh ^"Anti-bacterial agents"]
#16 [mh Penicillins]
#17 (amoxicillin* or amoxycillin* or amoxil or actimoxi or clamoxyl or hydroxyampicillin or penamox or trimox or wymox)
#18 ("penicillin v*" or apocillin or beromycin or betapen or fenoxymethylpenicillin or "Pen VK" or phenoxymethylpenicillin or "V-Cillin K" or vegacillin)
#19 (clindamycin or chlolincocin or cleocin or "Dalacin C")
#20 (azithromycin or azadose or azitrocin or azythromycin or gozal or sumamed or toraseptol or ultreon or vinzam or zentavion or zithromax or zitromax)
#21 (clarithromycin or biaxin)
#22 (cefazolin or ancef or cefamedin or cefamezine or cephamezine or cephazolin or gramaxin or kefzol or totacef)
#23 (ceftriaxone or benaxona or cefatriaxone or cefaxona or ceftrex or ceftriaxon* or lendacin or longacef or longaceph or rocefalin or rocefin or rocephin$ or tacex or terbac)
#24 (cephalexin or cefalexin* or ceporexine or Palitrex or cephahexin*)
#25 #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21 or #22 or #23 or #24
#26 #5 and #12 and #25
Appendix 4. EMBASE (OVID) search strategy
1. exp Endocarditis/
3. (endocardium adj5 (inflamm$ or infect$)).tw.
4. (ABE or SABE).ti,ab.
6. exp Dental surgery/
7. exp Endodontics/
8. exp Oral surgery/
9. ((oral or tooth or teeth) adj5 (surg$ or extract$ or restor$ or invas$ or scale or scaling or polish$ or endodontic$ or "root canal" or apicectom$ or apicoectom$)).tw.
10. (dental or dentist$).tw.
12. Antibiotic prophylaxis/
13. (antibiotic$ or anti-biotic$ or antimicrobial$ or anti-microbial$).tw.
14. Antiinfective agent/
15. exp Penicillin derivative/
16. (amoxicillin$ or amoxycillin$ or amoxil or actimoxi or clamoxyl or hydroxyampicillin or penamox or trimox or wymox).tw.
17. ("penicillin v$" or apocillin or beromycin or betapen or fenoxymethylpenicillin or "Pen VK" or phenoxymethylpenicillin or "V-Cillin K" or vegacillin).tw.
19. (clindamycin or chlolincocin or cleocin or "Dalacin C").tw.
21. (cephalexin or cefalexin$ or ceporexine or Palitrex or cephahexin$).tw.
23. (azithromycin or azadose or azitrocin or azythromycin or gozal or sumamed or toraseptol or ultreon or vinzam or zentavion or zithromax or zitromax).tw.
25. (clarithromycin or biaxin).tw.
27. (cefazolin or ancef or cefamedin or cefamezine or cephamezine or cephazolin or gramaxin or kefzol or totacef).tw.
29. (ceftriaxone or benaxona or cefatriaxone or cefaxona or ceftrex or ceftriaxon$ or lendacin or longacef or longaceph or rocefalin or rocefin or rocephin$ or tacex or terbac).tw.
31. 5 and 11 and 30
Appendix 5. US National Institutes of Health Trials Register search strategy
endocarditis AND dental AND prophylaxis
endocarditis AND dental AND prophylaxis
Appendix 6. metaRegister of Controlled Trials search strategy
endocarditis AND dental AND prophylaxis
endocarditis AND dental AND prophylaxis
Last assessed as up-to-date: 21 January 2013.
Protocol first published: Issue 3, 2002
Review first published: Issue 2, 2004
Contributions of authors
Anne-Marie Glenny: led the review update, screened update search results, amended background and discussion to reflect recent literature, added risk of bias tables and summary of findings table.
Richard Oliver: initiated the review, involved with the design and writing of the protocol, searching, duplication of assessment of titles and abstracts, inclusion and exclusion of full text papers, data extraction and quality assessment of included studies, data analysis and interpretation.
Lee Hooper: Lee was involved in the design of the first published version of the review, writing of the protocol, searching, duplication of assessment of titles and abstracts, inclusion and exclusion of full text papers, data extraction and quality assessment of included studies, data analysis and interpretation.
Graham J Roberts: provided background information for the protocol and review.
Helen Worthington: provided methodological and statistical support; screened update search results.
Declarations of interest
Anne-Marie Glenny: no interests to declare.
Richard Oliver: no interests to declare.
Graham J Roberts: no interests to declare.
Lee Hooper: no interests to declare.
Helen V Worthington: no interests to declare.
Sources of support
- Central Manchester and Manchester Children's University Hospitals NHS Trust, UK.
- The University of Manchester, UK.
- Eastman Dental Institute, UK.
- Manchester Academic Health Sciences Centre (MAHSC), UK.The Cochrane Oral Health Group is supported by MAHSC and the NIHR Manchester Biomedical Research Centre
- Department of Health Cochrane Review Incentive Scheme 2007, UK.
- Cochrane Oral Health Group Global Alliance, UK.All reviews in the Cochrane Oral Health Group are supported by Global Alliance member organisations (British Orthodontic Society, UK; British Society of Paediatric Dentistry, UK; British Society of Periodontology, UK; Canadian Dental Hygienists Association, Canada; National Center for Dental Hygiene Research & Practice, USA; New York University College of Dentistry, USA; and Royal College of Surgeons of Edinburgh, UK) providing funding for the editorial process (http://ohg.cochrane.org/)
- National Institute for Health Research (NIHR), UK.CRG funding acknowledgement:
The NIHR is the largest single funder of the Cochrane Oral Health GroupDisclaimer:
The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the NIHR, NHS or the Department of Health
Medical Subject Headings (MeSH)
MeSH check words
* Indicates the major publication for the study