Risk of unintended pregnancy among women with systemic lupus erythematosus

Authors


Abstract

Objective

It is recommended that women with systemic lupus erythematosus carefully time their pregnancies, but little is known about use of contraception and risk of unintended pregnancy in this population. The goal of this study was to estimate the proportion of women with lupus at risk for unintended pregnancy.

Methods

We surveyed 309 women with lupus, ages 18–50 years, seen at the University of Pittsburgh lupus center between January and May 2007.

Results

Of the women surveyed, 212 (69%) completed the survey. In the past 3 months, 97 (46% of 212) had faced some risk of unintended pregnancy. Of these, 53 (55% of 97) reported ≥1 occasion on which they had unprotected sex and 22 (23% of 97) reported that in the last 3 months they had unprotected sex “most of the time.” No women reported having used emergency contraception after unprotected sex. A desire to “discuss birth control with a health care provider at the lupus center” was reported by 22 (10% of 212) respondents and 16 (17% of 94) women ≤35 years old.

Conclusion

Many women cared for by our lupus center are at risk of unintended pregnancy and are interested in discussing birth control with a health care provider.

INTRODUCTION

It is recommended that women with systemic lupus erythematosus (SLE) carefully time their pregnancies because they have a higher risk of adverse outcomes of pregnancy than women without SLE (1), and pregnancies among women with active lupus are less likely to result in a live birth (2). In addition, many medications used to treat lupus may increase the risk of birth defects (3).

However, choosing a contraceptive method can be complicated for a woman with SLE. While 2 recent studies found that estrogen-containing contraceptives (4, 5) are safe for women who do not have antiphospholipid antibodies (aPL) or a recent flare in disease activity, many women with SLE carry aPL (6), and pregnancy is perhaps most dangerous for women with active SLE. In addition, many women with SLE develop hypertension and vascular disease (7), which make the use of estrogen-containing contraceptives unadvisable (8). Women with SLE are also known to be at risk of osteoporosis (9), which may lead to concern about use of progestin-only contraceptives such as medroxyprogesterone acetate, which can decrease bone mineral density (10). Intrauterine devices (IUDs) are highly effective, but misperceptions persist among some clinicians (11) about the relationship between IUDs and infection (5, 11). Barrier methods provide protection from sexually transmitted infections, but with typical use, unintended pregnancy results for 15 of every 100 women within their first year of use of condoms (12).

As pregnancy prevention is an issue that women with SLE may need to address for 20 to 30 years, a contraceptive method's cumulative risk of failure with years of use is an important consideration. Among all US women, it is estimated that half of all pregnancies are unintended (13), and that during a lifetime of use of reversible contraceptive methods the typical woman will experience 1.8 contraceptive failures (14). Unintended pregnancy has been found to be more common among women with some chronic conditions than among women without chronic conditions (15, 16), perhaps because women with chronic conditions receive less contraceptive counseling (17, 18). Because little is known about the risk of unintended pregnancy among women with SLE, this study was designed to explore plans for pregnancy and use of contraception among women with SLE.

PATIENTS AND METHODS

We surveyed all women presenting to a multidisciplinary lupus center in Pittsburgh, Pennsylvania between January and May of 2007. We distributed a 13-item, anonymous survey to 309 women, ages 18–50 years. The survey instrument was pilot tested for comprehension in our target population. Confidentiality of responses was insured by instructing women to seal their responses in an envelope. Women who were not interested in completing the survey were instructed to return the blank survey in a sealed envelope.

Responses were tabulated and descriptive statistics were generated using STATA version 8.0 (StataCorp, College Station, TX). The study was approved by the Institutional Review Board of the University of Pittsburgh.

RESULTS

A total of 212 women completed surveys, a response rate of 69%. Participants ranged in age from 18–50 years, with an average age of 34 years. Current pregnancy was reported by 6 (3%) women, while 7 (3%) reported currently trying to become pregnant. Of the remaining 199 women who were not pregnant or trying to become pregnant, 54 (27%) had been surgically sterilized, and 13 (7%) additional women reported that their partner had undergone vasectomy. Eleven (6%) women reported that they had never had sex with a man, and an additional 24 (12%) reported they had not had sex with a man for ≥3 months. Thus, in the past 3 months, 97 (46%) women had faced some risk of unintended pregnancy.

Among the 97 women at risk of unintended pregnancy, 53 (55%) reported ≥1 occasion on which they had “vaginal sex with a man without a condom or any other form of birth control,” and 22 (23%) reported that in the last 3 months they had unprotected sex “most of the time.” No women reported having used emergency contraception after unprotected sex. The contraceptive method used most frequently in the past 3 months was condoms, which were used on ≥1 occasion by 45 (46%) women (Table 1). Estrogen-containing contraceptives were used by 23 (24%) women. Use of other contraceptive methods is described in the table. No form of hormonal or barrier contraception was used at anytime by 31 (32%) of the women who reported having had sex with a man in the last 3 months.

Table 1. Use of contraception by women seen at a university lupus center, by efficacy of method*
Contraceptive methods used in the last 3 monthsNot currently pregnant or trying to conceive (n = 199)Not at risk of unintended pregnancy in last 3 months (n = 102)At risk of unintended pregnancy in last 3 months (n = 97)
  • *

    Values are the number (percentage). n/a = not applicable; IUD = intrauterine device.

  • Women could report use of more than one method in the past 3 months; for instance, 2 women used both pills and a ring, 1 woman used both a patch and a ring, 11 women used both pills and condoms, 5 women used both rhythm and withdrawal, 5 women used pills but reported no sex with a man in the last 6 months.

  • Women were considered to have been at risk of unintended pregnancy in the last 3 months if they had sex with a man, had not been sterilized, had a partner that had not had a vasectomy, did not report trying to become pregnant, and were not pregnant when surveyed.

Permanent sterilization67 (34)67 (66)n/a
 Tubal ligation, hysterectomy, or other54 (27)54 (53)n/a
 Partner with vasectomy13 (7)13 (13)n/a
Intrauterine contraception6 (3)n/a4 (4)
 Levonorgestrel IUD6 (3)n/a4 (4)
 Copper T IUD0n/a0
Progestin-only contraception11 (6)n/a7 (7)
 The shot6 (3)n/a4 (4)
 The minipill5 (3)n/a3 (3)
Estrogen-containing contraception31 (16)n/a23 (24)
 The pill28 (14)n/a22 (23)
 The patch1 (1)n/a0
 The ring5 (3)n/a2 (2)
Barrier method52 (26)n/a46 (47)
 Condoms50 (25)n/a45 (46)
 Diaphragm, cervical cap, sponge2 (1)n/a1 (1)
No barrier or hormonal method56 (28)n/a31 (32)
 Abstinent for last 3 months35 (18)35 (34)n/a
 Withdrawal27 (14)n/a22 (23)
 Rhythm8 (4)n/a6 (6)

A desire to “discuss birth control with a healthcare provider at the lupus center” was reported by 10% (22 of 212) of women, including 2 women who reported currently being pregnant. Women were more likely to want to discuss birth control if they were ≤35 years of age (16 [17%] of 94), had not been surgically sterilized or did not have a partner who had undergone vasectomy (21 [14%] of 145), or reported use of any of the following methods of contraception in the last 3 months: medroxyprogesterone acetate (2 [29%] of 7), withdrawal (7 [26%] of 27), condoms (11 [22%] of 50), or birth control pills (4 [14%] of 28).

DISCUSSION

We found that approximately half (46%) of all women seen at our lupus center had faced some risk of unintended pregnancy in the past 3 months, and that approximately a quarter of these women (23%) faced a high risk of unintended pregnancy because they had unprotected sex most of the time. We also found that 10% of all women, and 17% of women <35 years old, wanted to “discuss birth control with a health care provider at the lupus center.”

These findings suggest that comprehensive care of lupus patients should include a discussion of the safety and efficacy of available contraceptive options. Barrier contraceptives are safe for all women with SLE (19), but have 1-year failure rates with typical use that range from 15–32% (12). Estrogen-containing contraceptives, which have 1-year failure rates of 8% with typical use (12), are safe for women who do not have aPL or who have stable lupus disease activity (4), but increase health risks for women with vascular disease. Progestin-only contraceptives, such as medroxyprogesterone acetate and the minipill, are also safe for women with SLE (19), with typical-use failure rates of 3% and 8%, respectively (12), but use of these methods may decrease bone mineral density (10). Use of the levonorgestrel-containing IUD, a reversible method more effective than tubal sterilization (12), should therefore be strongly considered by women with SLE. The 1-year failure rate with typical use of a levonorgestrel-containing IUD is 0.1% (12). IUDs do not increase risk of clotting (20) or osteoporosis (21), and the levonorgestrel-containing IUD reduces menstrual blood loss and risk of anemia (22). As a result of high levels of user satisfaction (23) and notable cost-effectiveness (24) outside of the US, the IUD is currently the most widely used reversible method of contraception (25).

Unfortunately, we found that the levonorgestrel-containing IUD was used by only 3% of women surveyed. As the levonorgestrel-containing IUD is arguably a very appropriate contraceptive choice for women with SLE, it is important for rheumatologists to know that infection following IUD insertion is rare (26) and that IUDs can be safely used even by women who are immunocompromised (20). IUDs can be used by women who have previously had sexually transmitted infections or multiple sexual partners, provided they have no evidence of infection at the time of insertion (20). Insertion of an IUD is a simple office procedure and uterine perforation with IUD insertion is rare (27). IUDs can be successfully used by both nulliparous and parous women (28), with rapid return of fertility upon removal of the IUD. The levonorgestrel-containing IUD is effective for 5–7 years, after which time a new IUD can be safely inserted (29). Although some have worried that IUDs cause abortion by destroying embryos that arrive in a woman's uterus, detailed study of the mechanism of action of IUDs does not support these concerns (30). The primary disadvantages of IUDs are the need to have a trained provider insert or remove the device, and a one-time out-of-pocket cost of $400 to $800 if a woman's insurance does not cover IUDs. In addition, some women find the irregular vaginal spotting that occurs with levonorgestrel-containing IUDs (as with all progestin-only contraceptive methods) unacceptable.

Another concern raised by this study was the finding that none of the women who reported unprotected intercourse had used emergency contraception. This implies that in addition to discussing the safety and efficacy of routine methods of contraception, it is important to advise women with SLE that emergency contraception is safe to use and may decrease risk of unintended pregnancy if unprotected intercourse does occur (19). Emergency contraceptive pills containing levonorgestrel are now available in the US to women ≥18 years of age without a physician's prescription (31). While emergency contraception has been shown to have some efficacy up to 5 days after unprotected intercourse (32), it is more effective the sooner it is used. A number of professional organizations have therefore recommended that women be provided emergency contraception in advance of need (33–35). Studies have shown that women provided a supply of emergency contraception in advance of need are more likely to use it should unprotected intercourse occur, and that easy access to emergency contraception does not increase sexual risk-taking behavior (36–38). Detailed study of the mechanism of action of emergency contraception indicates that it does not disrupt implantation or induce abortion (31).

While our study of risk of unintended pregnancy among women with SLE provides insight into an area that has received little prior attention from rheumatologists, it has some limitations that deserve mention. First, the generalizability of our findings is limited by the number of women who declined to participate and by the lack of information on how responders may or may not differ from nonresponders. While we suspect that we inadvertently distributed some surveys to women who were >50 years of age, information on nonresponders is not available. Second, all data about sexual activity and use of contraception were self-reported and may be subject to social acceptability bias. However, this would imply that even more women may be at risk of unintended pregnancy than our results suggest. Third, in an effort to maximize subjects' confidence that their responses would be confidential, we did not collect sociodemographic data. As prior work has shown that contraceptive use varies across socioeconomic subgroups (39), future work to evaluate risk of unintended pregnancy among women with SLE should include more detailed assessments of socioeconomic variables. Finally, the generalizability of our findings is limited by the fact that the study sample was drawn from a single center located in Pittsburgh, Pennsylvania, where the lupus population is primarily white.

In conclusion, we found that a significant number of women cared for by our lupus center are at risk of unintended pregnancy and are interested in discussing birth control with a health care provider. These findings suggest that systems which ensure that women with SLE receive accurate information about the safety and efficacy of available contraception options (whether from their rheumatologist, generalist, or gynecologist) may prevent a significant number of unintended pregnancies that may be medically complicated.

AUTHOR CONTRIBUTIONS

Dr. Schwarz had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Schwarz, Manzi.

Acquisition of data. Schwarz, Manzi.

Analysis and interpretation of data. Schwarz, Manzi.

Manuscript preparation. Schwarz, Manzi.

Statistical analysis. Schwarz, Manzi.

Ancillary