Failure Rates of Male and Female Condoms Fall with Use

D. Hollander

First published online:

Male and female condoms have fairly low rates of slippage or breakage, and such failure declines as users' experience with the method grows, according to findings from a large observational study of women attending two STI clinics in the southern United States in 1995- 1998.1 For example, 3% of all female condoms slipped out of the woman's vagina; the rate was 11% at first use, but it fell steadily to less than 1% if the method had been used 15 times or more. Similarly, 3% of male condoms broke during use—7% among first-time users, compared with 2% among those who had used male condoms 15 times or more. Multivariate analyses confirmed the association suggested by these rates.

To be eligible for the study, women had to be aged 18-35 and not pregnant or planning to become pregnant within the next six months. At their first study visit, participants completed an interview and received instruction on recording their sexual activity and condom use on a diary form. They then watched a videotape providing instruction on condom use, attended a skills-oriented counseling session and were offered an opportunity to practice inserting the female condom under a nurse clinician's guidance. Women received a supply of either male condoms or female condoms with male condoms as a backup. They were asked to return for six follow-up visits at four-week intervals. At each visit, women returned their sexual diary, answered questions about their sexual and contraceptive behavior in the previous 30 days, attended an individual counseling session and received a new supply of condoms.

The analyses were based on 869 women who used at least one condom during follow-up. The majority of these women were black and low-income; their median age was 24, and their median number of years of schooling was 12. Three in five women had used male condoms in the 30 days before entering the study, but only one in four had used them every time they had intercourse. Three women had ever used a female condom, and 95% practiced inserting the device at the clinic.

During the study period, the women used 7,895 female condoms and 12,253 male condoms. In all, 0.1% of female condoms broke, and 6% slipped (in 3% of cases, the condom slipped out of the vagina; in 3%, the outer ring of the device slipped in). Three percent of male condoms broke, and 1% slipped off the penis. The researchers used logistic regression to assess factors associated with condom failure, controlling for a range of user characteristics measured at baseline and over time.

Compared with women who were 30 or older when they entered the study, younger age-groups had higher odds of having a female condom slip out (odds ratios, 2.1-2.6), but not of having the device slip in. Other baseline characteristics (e.g., education, lifetime number of partners and STI history) were not significantly associated with the likelihood of slippage.

Factors measured over time showed a number of associations with slippage of the female condom. Notably, the rates of slipping out and slipping in were substantially above average at first use (11% and 8%, respectively) but fell precipitously thereafter; for women who had used the female condom 15 or more times, both rates were less than 1%. Likewise, in the regression analyses, the odds of slippage were significantly elevated for first-time users compared with those who had used the device 15 or more times (odds ratios, 18.5 for slipping out and 19.7 for slipping in). Previous slippage of a female condom also was associated with increased odds of both kinds of slippage (2.7-2.8). In addition, having more than one sexual partner was associated only with an elevated likelihood of having a female condom slip out (1.5); having a casual partner and having had other problems with a female condom, only with an increased risk of the device's slipping in (2.0 and 2.3, respectively). Women who had had a female condom break were at increased risk of having one slip out (3.7); this factor was not examined with respect to slipping in.

Women's baseline characteristics had more associations with male condom failure than with female condom failure. Those who were younger than 20 were at risk of experiencing breakage (odds ratio, 1.9). Married women had increased odds of reporting slippage (1.9), as did those who had had a large number of partners (2.0 both for women reporting 6-7 partners and for those reporting 13 or more). Participants reporting alcohol or drug use in the past 30 days had a reduced likelihood of having a male condom slip (0.6).

Again, some of the most striking differences in failure rates were linked to experience with the method: The male condom breakage rate fell from 7% among first-time users to 2% among those who had used the method at least 15 times; the slippage rate dropped from 3% to 0.4%. In the regression analyses, the odds of breakage and slippage were significantly higher among first-time users than among those with the most experience (odds ratios, 6.0 and 7.9, respectively). Women who had had a condom break in the past were at increased risk of experiencing another breakage (3.6), and those who had had a male condom slip had elevated odds of repeating that experience (8.3). Other problems with male condoms and prior breakage of a female condom also were associated with an increased risk of slippage (2.8 and 3.1, respectively).

The researchers acknowledge three potential limitations of their study: The trial was not randomized, attrition was high and because participants received intensive instruction about the methods, they may have experienced lower failure rates than would typically be the case. However, the investigators contend that the study's strengths—its large size, short recall periods and multiple data collection methods—more than make up for these weaknesses.

In conclusion, the researchers note that either condom used correctly should provide protection against STIs. Nevertheless, they caution that for some individuals at high risk of acquiring an STI, condom failure rates may be unacceptably high, and health care providers should consider counseling such individuals to abstain from vaginal intercourse.

—D. Hollander