Condom Use and the Risk of Recurrent Pelvic Inflammatory Disease, Chronic Pelvic Pain, or Infertility Following an Episode of Pelvic Inflammatory Disease

Abstract

Among 684 sexually active women with pelvic inflammatory disease (PID) followed up for a mean of 35 months, we related contraceptive use to self-reported PID recurrence, chronic pelvic pain, and infertility. Persistent use of condoms during the study reduced the risk of recurrent PID, chronic pelvic pain, and infertility. Consistent condom use (about 60% of encounters) at baseline also reduced these risks, after adjustment for confounders, by 30% to 60%. Self-reported persistent and consistent condom use was associated with lower rates of PID sequelae.

Pelvic inflammatory disease (PID), the clinical condition representing inflammation of the pelvic organs, is common1 and can result in PID recurrence, chronic pelvic pain, and infertility.2,3 Prevention of the bacterial sexually transmitted diseases (STDs) that cause PID is a cornerstone of efforts to reduce morbidity from PID and its sequelae.4,5

Condom use prevents acquisition of viral STDs, including HIV. However, because no prospective data show that condoms are effective against transmission of bacterial STDs,6 controversy surrounds their use in primary prevention.7,8

Within the PID Evaluation and Clinical Health Study, a multicenter, follow-up study of women with PID,9 we assessed the relation between condom use and PID-related morbidity.

METHODS

The methods of subject recruitment, data collection, and follow-up have been reported elsewhere.9,10 In brief, women aged 14 to 37 years were recruited from 13 US sites between March 1996 and February 1999. Enrolled women met clinical criteria for suspected PID, including pelvic discomfort, pelvic organ tenderness, leukorrhea, mucopurulent cervicitis, and untreated gonococcal or chlamydial cervicitis. This analysis includes the 684 women who were sexually active at baseline and who had at least 1 follow-up visit.

In a standardized in-person interview, we asked about the use of oral contraceptives, hormonal implants or injections, intrauterine devices (used by only 15 women and thus not reported), diaphragms, spermicides, cervical caps, female condoms, and male condoms by a partner. More than 1 method could be selected. About half (53%) of the women reported baseline use of barrier methods of contraception, 92% of which was condom use. Condom use was considered to be consistent if a woman reported use with at least 6 of the last 10 sexual encounters.

Every 3 to 4 months, telephone interviews were repeated. Follow-up information was available for 85% of the cohort after a mean of 35 months. Outcomes included (1) self-reported recurrent PID (subsequent to the baseline episode), with medical record verification (in 68% of cases); (2) chronic pelvic pain, defined as consistent self-reports of at least 6 months’ duration; and (3) infertility, defined as the proportion of women without a β–human chorionic gonadotropin–confirmed pregnancy among the subgroup of women who reported no effective contraception (no contraception, natural family planning, or rhythm method) or rare use of barrier contraception for an aggregate of at least 12 months.

Baseline differences between groups were analyzed with χ² tests. Frequencies and unadjusted relative risks of recurrent PID, chronic pelvic pain, and infertility were calculated by comparing use with nonuse of condoms and consistent with nonconsistent use of condoms at each follow-up time point. Persistence (the percentage of all interviews in which condoms were used) was divided into quartiles. Analyses were repeated to compare women reporting use of condoms alone (without concurrent use of another method) with those reporting use of no effective method (including withdrawal, natural family planning, and none). Finally, we calculated the risks of outcomes among users and nonusers of other methods of contraception.

Separate logistic regression models for each outcome adjusted for age (continuous), number of live births (continuous), educational attainment (did not complete high school, high school graduate or equivalent, any education beyond high school), race (Black, White, other), nonmonogamy at baseline (yes or no), new partner in the past month at baseline (yes or no), gonococcal or chlamydial cervicitis at baseline (yes or no), number of study visits (continuous), and other methods of contraception. Adjusted odds ratios, derived from these models, estimated the adjusted relative risks.

RESULTS

Most of the women enrolled in the PID Evaluation and Clinical Health cohort were Black (74%), were aged 24 years or younger (66%), and had no more than a high school education (76%). Cervical infection with Neisseria gonorrhoeae was identified in 15% of the women, Chlamydia trachomatis was identified in 16%, and both were found in 6%.

Rates of recurrent PID, chronic pelvic pain, and infertility were highest among nonpersistent condom users (25% to less than 50% of reports) and lowest among persistent condom users (75%–100% of reports) (Figure 1 ▶).

FIGURE 1—

Adjusted relative risks for each condition, by percentage of interviews at which condom use was reported.

Note. P values are for trend.

^aReference group.

After adjustment for covariates, the relative risks for consistent condom users compared with nonusers were 0.5 (95% confidence interval [CI] = 0.3, 0.9) for recurrent PID, 0.7 (95% CI = 0.5, 1.2) for chronic pelvic pain, and 0.4 (95% CI = 0.2, 0.9) for infertility (Table 1 ▶). Users of other barrier methods were at reduced risk, albeit nonsignificant, for developing recurrent PID. Use of oral contraceptives or medroxyprogesterone was not associated with significantly elevated or reduced risks of the PID sequelae studied.

TABLE 1—

Percentages of and Relative Risks (RRs) for Recurrent Pelvic Inflammatory Disease (PID), Chronic Pelvic Pain, and Infertility After an Episode of PID, by Contraceptive Method Use in the 4 Weeks Prior to Baseline: 1996–1999

	Condition
	Recurrent PID				Chronic Pelvic Pain				Infertility
	n	%	RR	Adjusted RRa (95% CI)	n	%	RR	Adjusted RRa (95% CI)	n	%	RR	Adjusted RRa (95% CI)
Condoms
No	324	16.7		1.0	300	36.7		1.0	132	54.5		1.0
≤ 5–10 times	156	16.0	1.0	0.8 (0.5, 1.5)	142	31.0	0.9	0.8 (0.5, 1.3)	59	45.8	0.8	0.7 (0.4, 1.5)
≤ 6–10 times	204	8.8	0.5	0.5 (0.3, 0.9)	187	26.7	0.7	0.7 (0.5, 1.2)	46	34.8	0.6	0.4 (0.2, 0.9)
Other barrierb
No	650	14.8		1.0	597	32.7		1.0	226	48.7		1.0
Yes	34	2.9	0.2	0.2 (0.02, 1.1)	32	28.1	0.9	0.8 (0.3, 1.7)	11	45.5	0.9	1.2 (0.3, 4.8)
Oral contraceptives
No	606	14.4		1.0	556	33.1		1.0	217	47.0		1.0
Yes	78	12.8	0.9	0.9 (0.4, 1.8)	73	27.4	0.8	0.8 (0.4, 1.4)	20	65.0	1.4	3.2 (1.1, 9.4)
Medroxyprogesterone
No	605	14.5		1.0	556	33.3		1.0	214	50.9		1.0
Yes	79	11.4	0.8	0.6 (0.3, 1.2)	73	26.0	0.8	0.5 (0.3, 0.9)	23	26.1	0.5	0.5 (0.2, 1.4)
No effective methodc
No	453	12.1		1.0	416	26.8		1.0	134	41.8		1.0
Yes	231	18.2	1.5	1.4 (0.5, 3.6)	213	39.9	1.4	1.7 (0.8, 3.6)	103	57.3	1.4	2.3 (0.6, 8.3)

Note. CI = confidence interval.

^aAdjusted for age, number of live births, race, nonmonogamy at baseline, new partner at baseline, gonococcal or chlamydial cervicitis at baseline, education, number of study visits, and all other forms of contraception other than that under consideration.

^bDiaphragms, spermicides, cervical cap, or female condom.

^cNo contraception, natural family planning, or withdrawal.

Similar associations were found when comparing women who, at baseline, reported use of only a single method of contraception with women who reported use of no effective contraceptive method (data not shown). For example, adjusted relative risks for consistent condom use compared with use of ineffective methods were 0.6 for recurrent PID, 0.6 for chronic pelvic pain, and 0.3 for infertility. Excluding women who reported a history of PID at baseline, restricting our analysis to women with baseline evidence of endometritis or gonococcal or chlamydial upper genital tract infection, and including only recurrent PID based on verified medical record reports had little effect on these estimates.

DISCUSSION

A limited number of cross-sectional and case–control studies have examined the effectiveness of condoms in preventing the acquisition of N gonorrhoeae or C trachomatis among women, with mixed results.11–19 An additional 2 case–control studies and 1 cross-sectional study reported reductions in the occurrence of PID and tubal infertility among condom users, but this was significant in only 1 study.20–22

This analysis of the PID Evaluation and Clinical Health cohort lends strength to the literature on condom use and the prevention of PID and its sequelae. This study had several strengths: reports of condom use preceded the occurrence of outcomes, sample size was large, adjustment for confounding was made, a geographically diverse cohort was enrolled, and outcomes were validated.

The greatest weakness of this analysis was the reliance on self-report for contraceptive use, which may have resulted in an underestimation of the true association.16,23 Concurrent use of spermicides also may have reduced the observed protective effect because nonoxynol 9–containing spermicides may facilitate the risk for acquisition of STDs.24

These prospective data support the use of condoms for the prevention of PID sequelae.