WITH approximately 12 million cases of sexually transmitted diseases (STDs), including human immunodeficiency virus (HIV), occurring among Americans each year,1 the consistent and correct use of latex condoms remains an important component of public health prevention strategies.2,3 Three scientifically rigorous prospective studies of HIV transmission among serodiscordant heterosexual couples in Europe4,5 and Haiti6 provided strong evidence that condoms are highly effective when used with every act of intercourse. Nevertheless, findings from two of the serodiscordant couple studies5,6 and from other prospective studies7,8 suggested that, although rare, transmission of HIV or STD may occur despite consistent condom use. Such findings raised questions about the effectiveness of condom use against pregnancy and the transmission of infection.
Interpretation of these studies has been limited by the fact that condom effectiveness was not fully assessed. In the HIV serodiscordant couple studies, for example, only the consistency of use, not the correctness of use, was evaluated. Although HIV seroconversions among consistent condom users have been attributed to incorrect use of the product, a defect in the product itself, or misreporting of condom use, the true reason for condom failure cannot be determined from available data.
Other than studies of breakage and slippage,9 surprisingly little research exists on how condoms are used during vaginal intercourse, whether men experience problems during condom use,10 and the difficulties associated with proper condom use.3,11 Previous studies suggested that men experience problems using condoms effectively,12–14 although no studies systematically quantified problems that may occur when condoms are used during intercourse.
To assess how user practices influence the effective use of condoms, we conducted a retrospective study on condom-use behaviors among a sample of male university students. The primary objectives of the study were to evaluate the type of problems that men experience during condom use, to measure the extent to which these problems expose users to risks of pregnancy and infection despite condom use, and to compare exposure levels during condom use between men who use condoms consistently and men who use condoms inconsistently.
During 1993 and 1994 we recruited a sample of male students through advertisements posted on the campuses of two Georgia universities. Subjects participated in a confidential standardized interview to ascertain information about their use of condoms and condom-associated problems during vaginal intercourse. Subject eligibility criteria were male gender between 18 and 29 years of age, full-time students, and having ever used five or more condoms during vaginal intercourse.
Of 105 men who responded to the posted advertisements, 98 (93.3%) satisfied all eligibility criteria. The seven ineligible respondents had not used the minimum number of condoms required for study participation. All eligible men provided informed consent and agreed to participate in the study. Subjects received $5.00 in exchange for their participation in the study. To minimize subject recall bias and to reflect the men's most recent condom experiences, we restricted analyses to 47 subjects who reported using condoms in the month immediately preceding the study.
Subjects were asked to report both the number of times they had vaginal intercourse and the number of times they had used condoms in the last month and the last year. Participants were dichotomized as consistent condom users (100% use) or inconsistent/nonusers (<100% use) according to the level of condom use reported for the 11-month period in the last year preceding the most recent month.
To examine the prevalence of condom problems, subjects were asked to quantify the number of times they had experienced 10 various problems while using condoms in the last month. The proportion of condom uses in the last month during which each problem occurred was computed by dividing the total number of reported episodes for that problem by the total number of condoms used. Problems were classified according to whether they resulted in direct penile-vaginal contact and thus a potential risk of pregnancy or disease transmission.
Six problems were considered to result in no direct penile-vaginal contact: (1) “started to put a condom on inside-out, then flipped it over before putting it on and using it” (although this behavior may indirectly expose partners to infectious secretions from preejaculatory fluid, including HIV15,16); (2) “lost erection before or after a condom was placed on the penis”; (3) “experienced allergic reactions or irritations from a condom”; (4) “completely unrolled a condom before putting it on”; (5) “removed a defective condom from a package”; and (6) “tore a condom with a fingernail, jewelry, or ring (or other sharp object) prior to use.”
The remaining four problems were considered to result in direct penile-vaginal contact: (1) “started intercourse without a condom, then stopped to put it on”; (2) “broke a condom during intercourse or withdrawal”; (3) “started intercourse with a condom, then removed it and continued intercourse”; and (4) “had a condom completely fall off during intercourse or withdrawal.”
To determine the overall level of transmission risk during condom use, we assessed the number of times that each subject was potentially exposed to risks of pregnancy or infection during condom use from problems resulting in direct penile-vaginal contact. Because multiple problems may have occurred during the use of a single condom (e.g., starting intercourse without a condom before putting it on, followed by condom breakage), problems resulting in exposure could be unintentionally double counted, if these problems were incorrectly assumed by the investigators to have occurred during the use of different condoms. Because condom problems could not be linked with specific acts of intercourse, we computed the minimum number of condom uses in the last month during which each subject may have been at risk for pregnancy or disease transmission. For example, a subject who used 10 condoms during the last month and twice reported breakage, twice reported slippage, and three times reported removing condoms before ejaculation would be counted as having only three condoms resulting in potential transmission risk, although as many as seven condoms may have actually resulted in potential transmission risk. Thus, the minimum number of condoms resulting in exposure for a subject was equal to the maximum number of occurrences reported for the most common problem for that subject. Although reports of multiple problems during condom use from a single subject could underestimate subjects' actual risk, this situation was highly unusual because 45 of 47 subjects (95.7%) in the study reported either one or no problems.
Generalized estimated equations methodology was used to account for the correlation of condom use problems within individual subjects in the computation of standard errors (SDs) and 95% confidence intervals (CIs) for the prevalence of condom problems.17,18 Generalized estimated equations methodology was also used to compute standard errors and 95% CIs for the mean level of potential exposure during condom use between consistent and inconsistent users. All analyses were conducted using SAS statistical software (SAS Institute; Cary, NC).
The mean age of the 47 men who reported condom use in the prior month was 22.3 ± 3.4 years (± SD; range, 18-29). Twenty-four subjects (51.1%) were white, 20 (42.6%) were African-American, and 3 (6.4%) were Asian-American. Two men (4.3%) reported using condoms to prevent pregnancy, 2 men (4.3%) to prevent infection, and 43 men (91.5%) for both purposes. On average, men reported their first act of vaginal intercourse at 15.5 ± 2.4 years of age (range, 10-21). Twenty men (42.6%) reported using a condom at first vaginal intercourse, and the mean time between first condom use and study participation was 5.8 ± 3.1 years. Approximately two thirds of subjects (63.8%) reported using a condom at last vaginal intercourse. Among all condom users, the mean number of condoms used in the last month was 5.7 (median, 4; range, 1-25). Men who used condoms inconsistently in the last 11 months reported more acts of vaginal intercourse (8.8 versus 6.6) and fewer condom uses (5.6 versus 6.1) in the most recent month than men who used condoms consistently. Neither difference between consistent and inconsistent users was statistically significant.
Problems Experienced During Condom Use
Altogether, the 47 men used a total of 270 condoms in the month preceding the study. Table 1 presents the frequency of condom problems reported in the last month on a percondom basis (with 95% CIs) according to risk for direct penile-vaginal contact. Among the problems posing no direct risk of contact, the most common problem was using a condom that had been put on inside-out and then flipped over (13.0%), followed by loss of erection (12.2%), allergic reaction or irritation (2.2%), and complete unrolling of a condom before use (0.7%). No defective condoms were removed from a package, and none were torn before use.
Among the four problems that posed a direct risk of penile-vaginal contact, the most frequent was starting intercourse without a condom and then stopping to put one on (7.8%), followed by breakage during intercourse or withdrawal (4.1%), starting intercourse with a condom and then removing it and resuming intercourse (2.6%), and experiencing complete slippage during intercourse or withdrawal (1.5%).
Risks of Pregnancy and Infection During Condom Use
After computing the prevalence of each condom problem, we then measured the degree to which users were exposed to risks of pregnancy and infection during condom use in the last month. Sixteen of 47 men (34.0%, 95% CI, 20.9-49.3) reporting any condom use in the last month potentially exposed themselves and their partners to risks of pregnancy and infection as the result of breakage, slippage, or failing to use a condom throughout intercourse. Accounting for the possibility that multiple problems may have been experienced during the same condom encounter, a minimum of 35 of 270 total condom uses (13.0%, 95% CI, 7.4-18.5) in the prior month potentially exposed users to risks of pregnancy or infection.
Consistent users and inconsistent users had comparable levels of transmission risk while using condoms for vaginal intercourse. Of 178 condoms used in the last month by the 32 inconsistent users, 26 (14.6%, 95% CI, 8.0-21.2) resulted in exposure to risks of pregnancy or infection despite condom use. Of 92 condoms used in the last month by the 15 consistent users, 9 (9.8%, 95% CI, 1.0-18.5) resulted in exposure to risks of pregnancy or infection despite condom use. The difference in the percentage of condom uses resulting in potential exposure was not statistically different from zero.
This study provides one of the first in-depth evaluations of the prevalence of problems that men experience during condom use. Despite their exploratory nature, these data indicate a sobering level of exposure to potential risks of pregnancy and STDs, including HIV, during condom use. Altogether, at least 13% of condom uses resulted in exposure to risks of unprotected intercourse because of breakage, slippage, or failure to use condoms throughout intercourse. In the last month, 33% of consistent users were potentially exposed to risks of infection and pregnancy during condom use. Moreover, 1 of every 10 condoms used by these men resulted in a potential transmission risk.
These data hold important methodologic implications for the assessment of studies of condom-use effectiveness against HIV infection, STDs, and pregnancy. Exposures to unprotected intercourse during condom use provide another plausible, although little researched, explanation for reports of these outcomes among consistent condom users. Previous reports of HIV infection5–7 and incident STD8 among consistent users have been attributed either to method failure or to biases associated with self-reported condom use measures,19,20 with little consideration to how condoms were used during intercourse. Future studies of condom effectiveness should thus carefully distinguish between user effectiveness and product effectiveness by measuring whether condoms were used both consistently and correctly.
Condom effectiveness depends on the skill level, experience, and motivation of the user,21,22 even for men who use condoms consistently. Four of nine transmission risks among consistent users were preventable if the condom had been put on the penis before genital contact, used throughout intercourse, and not removed until after ejaculation. User techniques also may have contributed to episodes of breakage and slippage among consistent users. One subject with two breakages reported oil-based lubricant use with condoms, and another subject with two slippages reported delayed withdrawal after ejaculation. Both practices are contrary to instructions for condom use.
These findings are subject to a number of limitations. First, the small sample size and self-selected nature of our sample preclude generalization of our findings to the larger population of U.S. men and may have limited our ability to detect a statistically significant difference between consistent and inconsistent condom users. Moreover, the process for recruiting subjects for this study, through posted advertisements, may have resulted in an overrepresentation of subjects who believed they had interesting or unusual experiences to relate to investigators. Second, subject responses may also have been influenced by the social desirability effect if, contrary to instructions, subjects intentionally underreported or overreported problems experienced during condom use to please investigators. Third, despite our short recall period of 1 month, the retrospective nature of this study and its reliance on self-reported condom-use measures introduces the possibility of subject recall bias of both the number of condoms used and number of problems experienced. Fourth, our conservative approach to estimating actual transmission risk during condom use underestimates the true prevalence of exposure. However, the effect of this approach was negligible in this study because only two subjects (4.3%) experienced multiple types of condom problems resulting in direct penile-vaginal contact. If every possible episode of exposure had been counted as an independent transmission risk, the overall prevalence of exposure would have only increased from 13.0% to 15.9%.
The population in this study consisted of self-selected well-educated men with an average of more than 5 years of condom experience. Given the level of exposure to unprotected intercourse during condom use in this population, we suspect that exposure levels may be even greater in less-experienced less-educated populations, such as new condom users and adolescents. Research studies evaluating condom effectiveness in these populations would be useful both for identifying problem condom users and targeting education efforts toward men who use condoms ineffectively.
Even occasional problems experienced during condom use, such as breakage, slippage, or loss of erection, may decrease user confidence in condoms and cause users to misuse condoms intentionally or abandon condom use altogether. Men who experience problems with condoms, such as breakage, may also fail to communicate this information to their sexual partners.23 These issues warrant further study, given low condom use levels reported in national studies of adolescents and young adults.24–26 In view of these findings, education campaigns, package inserts, and other media vehicles should explicitly discuss proper condom use, acknowledging that problems with condoms may occur, so that men and women can learn to be effective condom users.
1. Eng TR, Butler WT, eds, for the Institute of Medicine. The Hidden Epidemic: Confronting Sexually Transmitted Diseases. Washington, DC: National Academy Press, 1997.
2. Cates W, Jr., Stone KM. Family planning, sexually transmitted diseases and contraceptive choice: a literature update-Part I. Fam Plann Perspect 1992; 24:75-84.
3. Centers for Disease Control and Prevention. Update: barrier protection against HIV infection and other sexually transmitted diseases. MMWR 1993; 42:589-591, 597.
4. De Vincenzi I. A longitudinal study of human immunodeficiency virus transmission by heterosexual partners. European Study Group on Heterosexual Transmission of HIV. N Engl J Med 1994; 331:341-346.
5. Saracco A, Musicco M, Nicolosi A, et al. Man-to-woman sexual transmission of HIV: longitudinal study of 343 steady partners of infected men. J Acqir Immune Defic Syndr 1993; 6:497-501.
6. Deschamps MM, Pape JW, Hafner A, Johnson WD, Jr. Heterosexual transmission of HIV in Haiti. Ann Intern Med 1996; 125:324-330.
7. Buchbinder SP, Douglas JM, McKirnan DJ, Judson FN, Katz MH, MacQueen KM. Feasibility of human immunodeficiency virus vaccine trials in homosexual men in the United States: risk behavior, seroincidence, and willingness to participate. J Infect Dis 1996; 174:954-961.
8. Zenilman JM, Weisman CS, Rompalo AM, et al. Condom use to prevent incident STDs: the validity of self-reported condom use. Sex Transm Dis 1995; 22:15-21.
9. Warner DL, Hatcher RA. Male condoms. In: Hatcher RA, Trussell J, Stewart F, et al., eds. Contraceptive Technology. 17th revised ed. New York: Ardent Media (in press).
10. Norris AE, Ford K. Associations between condom experiences and beliefs, intentions, and use in a sample of urban, low-income, African-American and Hispanic youth. AIDS Educ Prev 1994, 6:27-39.
11. Steiner M, Trussell J, Glover L, Joanis C, Spruyt A, Dorflinger L. Standardized protocols for condom breakage and slippage trials: A proposal. Am J Public Health 1994; 84:1897-900.
12. Darrow W. Condom use and use-effectiveness in high-risk populations. Sex Transm Dis 1989; 16:157-160.
13. Trussell J, Warner DL, Hatcher RA. Condom breakage and slippage rates. Fam Plann Perspect 1992; 24:20-23.
14. Spruyt A, Steiner MJ, Joanis C, et al. Identifying condom users at risk of breakage and slippage: findings from three international sites. Am J Public Health 1998; 88:239-244.
15. Ilaria G, Jacobs JL, Polsky B, et al. Detection of HIV-1 DNA sequences in pre-ejaculatory fluid. Lancet 1992; 340:1469.
16. Pudney J, Oneta M, Mayer K, Searge G III, Anderson D. Pre-ejaculatory fluid as potential vector for sexual transmission of HIV-1. Lancet 1992; 340:1470.
17. Liang K-Y, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986; 73:13-22.
18. Zeger SL, Liang K-Y. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986; 42:121-130.
19. Weir SS, Feldblum PJ. Condom use to prevent incident STDs [letter]. Sex Transm Dis 1996; 23:76-77.
20. Galavotti C, Cabral R, Beeker C. Condom use to prevent incident STDs [letter]. Sex Transm Dis 1996; 23:77-79.
21. Albert AE, Warner DL, Hatcher RA, Trussell J, Bennett C. Condom use among female commercial sex workers in Nevada's legal brothels. Am J Public Health 1995; 85:1514-1520.
22. Grady WR, Tanfer K. Condom breakage and slippage among men in the United States. Fam Plann Perspect 1993; 26:107-112.
23. Warner DL, Boles J, Goldsmith J, Hatcher RA. Disclosure of condom breakage to sexual partners. JAMA 1997; 278:291-292.
24. Tanfer K, Grady WR, Klepfinger DH, Billy JOG. Condom use among U.S. men, 1991. Fam Plann Perspect 1993; 25:61-66.
25. Pleck JH, Sonenstein FL, Ku L. Changes in adolescent males' use of and attitudes toward condom use, 1988-1991. Fam Plann Perspect 1993; 25:106-109,117.
26. Kann L, Warren CW, Harris WA, et al. Youth risk behavior surveillance-United States 1995. MMWR 1996; 45(SS-4):16-18.