Value of Consistent Condom Use: A Study of Sexually Transmitted Disease Prevention Among African American Adolescent Females

A recent report suggested that evidence was insufficient to evaluate the degree of benefit provided by condoms against most sexually transmitted diseases (STDs).1 However, in vitro evidence2–4 suggests that transmission of the largest STD pathogens (i.e., bacterial and protozoans) can be prevented by condom use. This study prospectively determined the association between African American adolescent females’ condom use and their acquisition of biologically confirmed infection with Chlamydia trachomatis, Neisseria gonorrhoeae, or Trichomonas vaginalis.

METHODS

Study Sample

From December 1996 through April 1999, sexually active African American adolescent females (aged 14–18) were recruited from multiple venues (medical clinics and high schools) in low-income neighborhoods of Birmingham, Ala. The study achieved an 85.7% baseline participation rate. Ninety percent returned for the 6-month follow-up assessment (n = 468). Of these adolescents, 390 (83%) engaging in penile-vaginal sex between the 2 periods constituted the sample.

Data were collected at the University of Alabama Family Medicine Clinic and were gathered as part of a randomized controlled trial of an HIV prevention program.

Study Design

The design was modeled from a recent publication describing necessary conditions for testing hypotheses about condom effectiveness.5 To establish an infection-free cohort at baseline, we tested adolescents with DNA amplification measures (for C trachomatis, N gonorrhoeae) and culture for T vaginalis. Those with positive test results were notified and treated. All treatments were verified.

Six months later, interviewers assessed adolescents’ frequency of engaging in penile-vaginal intercourse since baseline and the number of times condoms were used during these sexual encounters. The latter measure was divided by the former to create a distribution representing proportional condom use. This distribution was dichotomized to compare consistent (100%) users with remaining adolescents with respect to STD incidence as measured by repeated testing.

Laboratory Measures

At both assessment periods, adolescents provided 2 (self-collected) vaginal swabs. One was placed in a transport tube (Abbott LCx Probe System; Abbott Laboratories, Abbott Park, Ill) and tested for C trachomatis and N gonorrhoeae DNA by ligase chain reaction.6,7 Studies have confirmed the high sensitivity and specificity of these procedures.6–9 The other swab was used to inoculate culture medium for T vaginalis (InPouch TV test; BioMed Diagnostics, San Jose, Calif).10

RESULTS

At baseline, 28.2% had positive test results for at least 1 STD. Six months later, adolescents provided information about their sexual behaviors after the baseline assessment. The mean frequency of penile-vaginal sex was 13.5 episodes (SD = 24.2) and did not vary between those subsequently acquiring and not acquiring an STD (t = .40; P = .69). Fifty-one percent of the adolescents reported 100% condom use between the 2 assessment periods. The mean proportion of condom use among those not using condoms 100% of the time was 42%. At follow-up, valid specimens were obtained for 380 adolescents; 90 (23.7%) tested positive for at least 1 STD (chlamydia = 16%; gonorrhea = 3%; trichomoniasis = 9%).

Among adolescents reporting 100% condom use, 17.8% had positive test results. By comparison, 30.0% of those reporting less than consistent condom use had positive test results. This difference was significant (relative risk ratio = 1.69; 95% confidence interval = 1.16, 2.46; P = .005) and remained significant after adjusting for whether adolescents had positive test results at baseline and the effects of having more than 1 sexual partner during the recall period (adjusted odds ratio = 1.85; 95% confidence interval = 1.13, 3.04; P = .01).

DISCUSSION

This study used a design specifically suggested for investigating condom effectiveness,5 and the findings suggested that using condoms consistently could result in substantial, but not complete, protection against STD acquisition. Adolescents not consistently using condoms were about twice as likely to acquire an STD.

The study compared consistent (100%) users with all others. Precedent for this was established in seminal studies showing the protective value of condoms against HIV infection.11,12 Thus, as suggested also by the Centers for Disease Control and Prevention,13 our findings support the public health practice of promoting consistent condom use as a primary strategy to reduce STD incidence (efficacy of community- and clinic-based programs to achieve this goal has been well documented14).

Although the findings are encouraging, 17.8% of the adolescents acquired an STD despite 100% condom use. At least 3 forms of bias could have inflated the percentage with positive test results despite 100% use, thereby underestimating the protective effects. For example, evidence suggests that young people make multiple errors when using condoms.15–17 Thus, one likely reason STDs occurred among 100% users is that condoms were used incorrectly. Furthermore, if condoms were not used from start to finish of intercourse (possibly a common phenomenon18), then the observed protective effect, again, was underestimated. Finally, if adolescents felt a social desirability bias when responding to interview questions about condom use, this also would inflate the number misclassified as 100% users. This misclassification would underestimate protective effects.

Other causes of study bias would result in random error (random error attenuates effect but does not bias in one direction). One example is possible treatment failure. To compensate, and to control for the potential effect of STD history on subsequent infections (as suggested in a recent study19), findings included an odds ratio adjusted for STD diagnosis at baseline. Another potential cause of random error may be recall bias relative to frequency of sex and condom use. Finally, it should be noted that findings are limited by convenience sampling and a sample size that precluded separate analyses of the 3 STDs. Further research is needed with larger samples that adequately represent diverse populations.