Study	Reason for exclusion
Al‐Karaawi 2001	Retrospective analysis of cumulative exposure to bacteraemia following various dental procedures in children with severe congenital heart disease but no cases of endocarditis
Anonymous 1992	Economic analysis of the cost‐effectiveness of using prophylactic antibiotics using same data as Bonhomme 1992
Archard 1966	2 case studies of high risk patients developing endocarditis after dental treatment with antibiotic prophylaxis
Bayliss 1983	Not all cases at risk and no controls
Bennis 1995	No control group
Bhat 1996	Retrospective analysis of 28 cases of endocarditis, no controls
Biron 1997	Case report
Bonhomme 1992	Economic analysis of the cost‐effectiveness of using prophylactic antibiotics based on published data
Caretta 1988	No control group
Clemens 1982	Assessment of the effect of mitral valve prolapse on risk of endocarditis (rather than assessment of the effect of prophylaxis), case‐control design
Conner 1967	Participants not at high risk of endocarditis
Gersony 1977	Cohort study, but it was not stated how many patients had preceding dental treatment, only two cases with preceding dental treatment and no prophylaxis
Herr 1976	Case report (German)
Hess 1983	All children with cardiac disease received antibiotic prophylaxis before dental extraction, no controls
Horstkotte 1986	Retrospective study of a group of people at high risk of endocarditis who had had appropriate prophylaxis for medical and dental interventions, and a group of people at similar risk who did not have appropriate prophylaxis for such interventions. It was not possible to ascertain how many of the cases or controls had had dental interventions, and the source of the 2 groups is unclear.
Imperiale 1990	Case‐control study: people with endocarditis (cases) who died were excluded, although the mortality rate in the cases was much higher (20%) than was likely in the control group, thus making the 2 groups incomparable.
Khairat 1966	CCT, but participants not at high risk of endocarditis and no relevant outcomes measured
Lacassin 1995	Case‐control study: people with endocarditis (cases) who died were excluded, although the mortality rate in the cases was much higher (20%) than was likely in the control group, thus making the 2 groups incomparable.
Lauridson 1984	Case reports
Lecointre 1981	Cohort study of patients having dental extractions but all patients received antibiotics
McGowan 1978	Letter on failures of prophylaxis on a case by case basis, not RCT, CCT, cohort or case‐control design
McGowan 1982	Case reports
Pogrel 1975	Retrospective study of cases of endocarditis but no controls
Rahn 1988	Serological study of bacteraemia following penicillin versus administration and tooth extraction
Rahn 1993	Not an assessment of antibiotic prophylaxis (concerned with adjunctive use of antiseptic solution)
Schirger 1964	Case series
Shanson 1980	No at‐risk patients; examined serum levels of amoxicillin in healthy volunteers
Strom 1998b	Case‐control study based in the USA of 273 hospital patients with endocarditis. Not all the cases (38%) or controls (6%) had a previously known risk of endocarditis.
Tozer 1966	No dental interventions, and participants not at high risk of endocarditis
Tzukert 1984	Same group of patients as Tzukert 1986
Van der Meer 1992b	Epidemiological study of endocarditis in the Netherlands, no controls
Woodman 1985	Basic science research paper
Yoshimura 1985	Cohort study of 17 patients undergoing dental extractions; all received antibiotics

CCT: controlled clinical trial; RCT: randomised controlled trial