Since 1990, sterilization has been the most commonly used method of contraception for women in the United States.1 Performing sterilization during the postpartum hospital stay offers many advantages. The U.S. Collaborative Review of Sterilization (CREST) demonstrated that tubal occlusion by postpartum partial salpingectomy had the overall lowest failure rate of all methods of sterilization (7.5/1,000).2 Goodman and Dumas3 found that, overall, 90% of existing epidural anesthesia can be used for the procedure, thereby adding little additional anesthesia-related risk if the procedure is performed within 24 hours of the delivery. Despite its convenience, safety, and efficacy, not all women who desire postpartum sterilization actually have the surgery.
There are some international studies addressing the issue of missed postpartum sterilization. Verkuyl4 studied 284 women in Zimbabwe who did not undergo sterilization despite their desire to do so. He found that the most common explanation given by the subjects was that the attending health personnel were poorly motivated to perform the procedure. Potter and colleagues5 prospectively surveyed 1,136 Brazilian women desiring postpartum sterilization. Most women expressed a preference for postpartum sterilization over interval sterilization. Yet, patients with private insurance were much more likely to undergo postpartum sterilization than patients with public insurance.5 Therefore, in both of these studies, factors related to the health care system denied women a desired surgery.
This issue appears to be poorly studied in the United States. It would be important to know whether women who express desire for sterilization do not undergo the procedure because these women still have postpartum contraceptive needs that may not have been anticipated during antepartum care. The purpose of this study is to estimate the number of women who do not get a potentially desired surgery. The secondary aim is to identify predictors of not undergoing postpartum sterilization. If predictors were known, then we could determine whether certain populations are more vulnerable to not obtaining this surgery. Finally, we aim to generate hypotheses about why women do not undergo sterilization despite initially expressing a desire.
MATERIALS AND METHODS
We identified women who desired postpartum sterilization during their antenatal care between March 2002 and November 2003 at The University of Illinois at Chicago Medical Center. A single Oracle database serves as the repository for all prenatal and delivery records at University of Illinois at Chicago. Accuracy of data entry by health care providers is provided through an array of quality assurance procedures guided by the Joint Commission on Accreditation of Health Organizations and hospital standards. We used Structured Query Language queries to extract data which was then merged into a spreadsheet for further data analysis. We electronically searched all patient records using a wide variety of terms including “PPTL,” “TL,” “T.L.,” “postpartum tubal ligation,” “sterilization,” “permanent sterilization,” “undesired fertility,” “MFDPS” (multiparous female desires sterilization), “future fertility,” and “tubal.” We then reviewed each chart obtained in our electronic search to verify the request for sterilization and obtain missing data variables. We also consulted our electronic patient registration system and nonobstetrical patient records to complete missing data. We excluded women less than age 21 years. Because our population is predominately publicly insured women under the age of 21, they are ineligible for sterilization. We also excluded women who delivered at another institution and women who delivered before 30 weeks. The cutoff for gestational age was based on the risk of miscarriage, poor neonatal outcome, and inability to sign Medicaid-required sterilization consent forms. This study was approved by the institutional review board of the University of Illinois at Chicago.
Once women who desired postpartum sterilization during antepartum care were identified and exclusion criteria applied, we determined whether they eventually underwent sterilization. We then compared characteristics of women who underwent sterilization with characteristics of those who did not. Demographic variables examined included age, race/ethnicity, marital status, employment status, religion, body mass index (BMI), gravidity, and number of living children. Antenatal variables included insurance status, provider type (supervised resident physician, certified nurse midwife, maternal–fetal medicine specialist, general obstetrician gynecologist, or family medicine physician), trimester at initial prenatal care visit, trimester at request for sterilization, and number of hospital admissions before delivery. Intrapartum variables included mode of delivery and use of epidural analgesia. Data were analyzed using SAS 8.0 (SAS institute, Cary, NC). The Student t test was used to compare means. Univariate analyses of data were performed with the χ2 test or the Mantel-Haenszel test for trend. Variables that had P < .10 in univariate analysis or that were clinically relevant were eligible for inclusion in the initial multiple logistic regression model. Variables were removed from the regression model in a stepwise fashion, testing for significance sequentially with the likelihood ratio test. The final explanatory regression model includes all significantly associated variables and potential confounding variables. We report a generalized R2 to assess the variability in our model. Variables were considered significant if the Wald χ2 test statistic had a P value ≤ .05; 95% confidence intervals are reported. Observations with missing values were excluded from the model.
We reviewed 6,589 prenatal records and identified 799 unique records in which the patient requested sterilization (there were no women with more than one pregnancy). After excluding women less than age 21 years (n = 9), women who delivered at outside institutions or otherwise lacked delivery data (n = 52), and women with very premature deliveries (n = 26), a total of 712 subjects remained. The characteristics of these 712 women are presented in Table 1. Our population was predominately African American and Latino. Most were unmarried and unemployed. While over one third of our population identified themselves as nonreligious, Catholic and Baptist were the 2 most common religions in our population. Most women had 3 or more prior pregnancies and were insured by Medicaid (Table 1). The rate of cesarean delivery in this cohort was 21.3%, and 24.4% used epidural analgesia.
Of the 712 subjects who requested postpartum sterilization during their antepartum care, 327 (45.9%) women did not undergo the procedure. We compared demographic, antepartum, and intrapartum variables for women who received postpartum sterilization and those who did not. In univariate analysis, we found that those who did not undergo sterilization were significantly younger than those who did: mean age (standard deviation) 28.4 (5.3) years versus 30.5 (5.3) years (P < .001), respectively (data not shown). Examining age as a categorical variable demonstrated that women in the youngest age group (ages 21–25) were the least likely to undergo sterilization (66/166, 39.8%). Race was also significant, with Hispanics being more likely to undergo sterilization than African Americans (143/229, 62.4% compared with 216/434, 49.8%). Antepartum and intrapartum factors associated with undergoing sterilization included requesting sterilization in the third trimester (322/567, 56.8%) rather than the second (33/75, 44%) and cesarean delivery (161/204, 78.9%).
A multivariable logistic regression was constructed with the variables that were significant in univariate analysis as well as insurance status, parity, and use of epidural anesthesia. In this model, age, race, trimester of request and mode of delivery remained significant. We calculated a generalized R2 of 0.180. Body mass index was not included because of the large number of missing values. Our model was based on 612 observations (Table 2).
In this study, we determine the risk of and predictors for undergoing postpartum sterilization among women who expressed desire for this surgery during antepartum care. This research is important for a number of reasons. Data from the National Survey of Family Growth6 indicate that 38% of reproductive-aged women depend on sterilization for contraception. Women who desire sterilization but do not undergo the procedure have missed the opportunity to obtain a convenient and highly efficacious method of pregnancy prevention. Yet, we know little about who such women are. Have these women changed their minds, or do factors associated with the health care system prevent them from having desired surgery?
We found that only half of all women who, during antepartum care, request postpartum sterilization undergo this procedure. It is not surprising that some women who request sterilization do not follow through. Certainly women change their minds about desiring sterilization. Another possible explanation is that, because a large portion of our population is insured by Medicaid, consent for sterilization must be signed at least 30 days in advance of surgery. It is likely that some women will sign a consent form “just in case” they decide to have the surgery, only to change their minds at a later time.
In addition, we hoped to identify specific variables associated with not undergoing sterilization. The presence of such variables may indicate populations that are more likely to change their minds or are particularly vulnerable to issues related to the health care system. In our analysis, age, race, trimester of request, and vaginal delivery were predictors of not undergoing sterilization. The data we present cannot delineate the exact reasons why these risk factors were associated with failure to undergo a previously desired postpartum sterilization. Indeed, both personal decision making and system-related barriers may be factors in some of these findings. For example, age lends itself to both interpretations. Data from the National Survey of Family Growth demonstrate that women who have had difficulty using reversible contraception are more likely to undergo sterilization.7 A subanalysis of our data shows that women in our lowest age group (21–25 years) averaged 3.91 pregnancies (21 year olds had a mean of 3.15 pregnancies; 25 year olds a mean of 4.4 pregnancies), which suggests that these women have difficulty planning pregnancies. Thus, the request for sterilization may be a marker for women who are concerned about their ability to use reversible contraception rather than a true commitment to sterilization. Similarly, we know from the CREST study that young women are more likely to regret their sterilization decision.8 Thus, physicians may be less likely to sterilize a young woman and may actively dissuade her at the time of delivery.
Trimester of request may also relate to both personal decision making and the health care system. An analysis of data from the National Survey of Family Growth showed that reasons for choosing sterilization include not wanting more children and having medical problems.7 It is possible that women who choose sterilization for reasons of health may be ambivalent about sterilization. In our study, women who voiced their desire in the second trimester may have done so for reasons such as these, only to change their minds later on. Also, counseling early in pregnancy may give women more time to reverse their decision.
Our finding of racial differences in rates of sterilization is more difficult to interpret. One possibility is that many Latino patients served by our hospital have Medicaid insurance that ends shortly after pregnancy. These women may be more determined to undergo surgery in the immediate postpartum period. Our finding of a higher completion rate during cesarean delivery may reflect lack of motivation on the part of the physician or the patient to undertake another procedure, or it may be indicative of system-wide barriers to completing the surgery. Some of these possibilities have been raised in international studies and may also exist in the present population. Indeed, the health care system may even affect a woman's determination to undergo surgery. For example, because sterilization is an elective procedure, it can be delayed for more urgent operating room procedures. One can easily imagine an exhausted postpartum woman choosing to postpone her surgery if it were delayed. Dedicated operating rooms and anesthesia staff, scheduled surgery times, and preparing the patient for possible delays may improve rates of completing postpartum sterilization.
This study has a number of limitations. First, it is a retrospective study; thus, our data are limited by the quality of record keeping. Yet, our ability to collect data is greatly enhanced by the use of an electronic medical records system, and we completed 100% chart verification to improve accuracy. Second, these data are limited to a single hospital. It remains to be seen whether our findings are particular to our institution and patient population. Third, retrospective studies prevent one from understanding motivations; did women not have a desired surgery or did they change their mind? Our data prevent us from distinguishing between these 2 possibilities. Fourth, we may not have examined the most relevant risk factors for not undergoing sterilization. For example, despite reviewing anesthesia records and operating room schedules, we were unable to make a global assessment of operating room and anesthesia availability. Such information would be useful in teasing apart patient variables from system barriers. Finally, because of limited numbers and the uniformity of our patient population, we may lack the power to examine all variables adequately. A post hoc power analysis shows that we had 80% power to detect lower sterilization rates for women in our youngest age group. But we lacked power to examine other variables. For example, we had 50% power to detect a 25% difference in the risk of receiving sterilization compared with the risk of not receiving sterilization for women in our highest BMI quartile.
Despite its limitations, this study raises important questions regarding postpartum sterilization and points to areas for future research. Clearly, it will be important to determine which women who did not get their surgery still wanted it. Such instances mean that women are missing an opportunity for the most effective method of sterilization. We are currently reviewing hospital records to find out what reasons physicians and patients provide when sterilization is not done. It is also essential to learn more about women who change their minds regarding sterilization. This occurrence may be a marker for women who have concerns about their ability to use reversible contraception to prevent unintended pregnancy. Such women may be at particularly high risk of repeat pregnancy, especially if they have a history of unintended pregnancy. We are currently conducting a longitudinal study of postpartum women who expressed desire for sterilization but did not undergo the procedure for the purpose of examining factors such as rates of obtaining interval sterilization, regret, contraceptive compliance, and repeat pregnancy.
Ultimately, one might think that a woman who chooses sterilization does not require counseling regarding reversible contraceptive methods. Yet, our data suggest that providers should counsel all women who desire postpartum sterilization about a wide array of contraceptive methods, with the understanding that approximately half of these women may not undergo the sterilization procedure.
1. Peterson HB, Xia Z, Wilcox LS, Tylor LR, Trussel J. Pregnancy after tubal sterilization with bipolar electrocoagulation. Obstet Gynecol 1999;94:163–7.
2. Peterson HB, Xia Z, Hughes JM, Wilcox LS, Tylor LR, Trussel J. The risk of pregnancy after tubal sterilization: findings from the U.S. Collaborative Review of Sterilization. Am J Obstet Gynecol 1996;174:1161–8.
3. Goodman EJ, Dumas SD. The rate of successful reactivation of labor epidural catheters for postpartum tubal ligation surgery. Reg Anesth Pain Med 1998;23:258–61.
4. Verkuyl DA. Tubal ligation candidates who did not get their operation. Cent Afr J Med 1996;42:150–2.
5. Potter JE, Perpetuo IH, Berquo E, Hopkins K, Leal OF, de Carvalho Forminga MC, et al. Frustrated demand for postpartum female sterilization in Brazil. Contraception 2003;67:385–90.
6. Abma JC, Chandra A, Mosher WD, Peterson L, Piccinino LJ. Fertility, family planning, and women's health: new data from the 1995 National Survey of Family Growth. National Center for Health Statistics. Vital Health Stat 23 1997;19:1–114.
7. Chandra A. Surgical sterilization in the United States: prevalence and characteristics, 1965–95. National Center for Health Statistics. Vital Health Stat 23 1998;20:1–33.
8. Hillis S, Marchbanks PA, Tylor LR, Peterson HB. Poststerilization regret: findings from the United States Collaborative Review of Sterilization. Obstet Gynecol 1999;93:889–95.