Table 2

 Epidemiological evidence for STI prevention effectiveness of diaphragms*

AuthorsYearStudy populationNoDesignOutcomeKey finding
 >Odds ratio (95% CI)
*Includes observational studies that presented risk estimates adjusted for potential confounding factors. Five additional studies examined effects of physical barrier method use on chlamydia, gonorrhoea and/or trichomoniasis; though, because they defined a combined barrier method measure for analysis (for example, male condom, diaphragm and nonoxynol-9 containing spermicides) these investigations, as published, are not informative in assessing the effectiveness of particular methods (Park et al,74 Quinn and O’Reilly,75 McCormick et al,76 Berger et al,77 Keith et al78).
†0/77 diaphragm users infected compared to 20% of non-users of contraception.
‡Chlamydia assessed in only 35 of 227 diaphragm users.
§Odds ratio for use of diaphragm ever. Estimate for current use not significant: OR = 0.5, 95% CI: 0.2 to 1.9.
Magder et al231988STD clinic1031Cross sectionalChlamydia0†
Rosenberg et al251992STD clinic4162Cross sectionalChlamydia‡0.25 (0.05 to 1.36)
Gonorrhoea0.32 (0.16 to 0.65)
Austin et al241984STD clinic1781Case-controlGonorrhoea0.45 (0.15 to 0.3)
Bradbeer et al261987Sex workers100Cross sectionalGonorrhoea0.36 (p<0.05)
Kelaghan et al271982Hospital1481Case-controlPID0.4 (0.2 to 0.7)
Wolner-Hanssen et al281990Health clinics880Case-controlPID0.3 (p = 0.005)
Cramer et al291987Hospital4116Case-controlTubal infertility0.5 (0.3 to 0.7)
Becker et al301994Women’s health clinics538Case-controlCervical neoplasia0.3 (0.2 to 0.6)§
Hildesheim et al311990Hospital with community controls1267Case-controlCervical cancer<5 years’ use: OR = 0.9 (0.6, 1.3); 5+ years’ use: OR = 0.8 (0.4, 1.6)
Wright et al321978Family planning clinics17 032Prospective cohortCervical neoplasia0.12 p<0.01
191Case-control0.23 p<0.05