Variations in Receipt of Contraceptives by Insurance Status and Race/Ethnicity

Discussion

Our primary finding was that insurance is associated with receipt of contraceptives among young women aged 18–29 and that this relationship varies by race/ethnicity. Specifically, non-Hispanic Black and Hispanic women with Medicaid had higher odds of receiving contraceptives, either LARC or non-LARC, compared to non-Hispanic White women with commercial insurance. In addition, Hispanic women with commercial or other insurance also had significantly higher odds of receiving LARC, compared to non-Hispanic White women with commercial insurance.

At 8.4%, the percent of women receiving LARC in our sample was lower than in the 2011–2015 pooled National Survey of Family Growth (NSFG), which reported 9.5% [23]. This is notable given that we included only women who had at least 1 primary care visit during the study period, reflecting active engagement with the health care system. Compared to the sample in the 2011–2015 NSFG, our sample had a higher percentage of women with commercial insurance and a lower percentage of women with either Medicaid or other insurance [23]. High out-of-pocket cost is a well-known barrier to LARC use [12, 19, 24]. Broecker and coauthors found that patients were 22% less likely to choose LARC for each $100 increase in out-of-pocket cost [17]. In North Carolina, the Medicaid Family Planning Program (BE SMART) covers LARC and other contraceptives without any copayments for eligible men and women with incomes less than or equal to 195% of the federal poverty level [25]. North Carolina Medicaid also began covering inpatient postpartum LARC insertion beginning in 2018. For women with other types of coverage, including those with commercial insurance and the uninsured, cost may continue to be a barrier to LARC use.

Our findings also indicate that use of different contraceptive methods varies by racial/ethnic group. While non-Hispanic Black women with Medicaid were more likely to receive LARC and non-LARC than no contraception when compared to non-Hispanic White women with commercial insurance, the percent of young Black women receiving either contraceptive type in our sample was less than 30%. Efforts to increase contraceptive use in this population need to address historical injustices that contribute to mistrust of health care professionals. Between 1929 and 1974, the North Carolina Eugenics Board authorized the forcible sterilization of over 7000 citizens, mostly in Mecklenburg County [26, 27]. The board gave the power to social workers who threatened poor Black women by withholding welfare benefits if they did not comply [27]. During this period many teenagers were put under sterilization surgery without formal consent from their legal guardians [26]. Although the Eugenics Board was disbanded in 1974, perspectives regarding contraception as a means of government population control remain [9]. Building trust between providers and patients, increasing women’s autonomy in contraceptive choice, and shared decision-making are all needed and under-studied, especially among Black women and women with low incomes.

Hispanic women in our sample across all insurance statuses had significantly higher odds of receiving LARC when compared to non-Hispanic White women with commercial insurance. Mecklenburg County, where the women in our sample resided, is a relatively new Latino immigrant gateway, where immigrants have moved into older suburbs and neighborhoods with a high percent of Latinos due to a booming economy and large number of jobs [28, 29]. Coffman and colleagues estimated that fewer Latino adults in Mecklenburg County had health insurance than reported in the US Census due to undercounts of immigrant and undocumented residents [30]. Lack of insurance coverage has been shown to influence use of prescription contraceptives in studies involving national samples and among Hispanic women [23]. However, 1 recent national study found that Hispanic women aged 18–45 with insurance coverage were not more likely to use prescription contraceptives during the period following the ACA when compared to those without insurance [31]. A national survey of 897 female respondents aged 20–29 found that, among Hispanic women, foreign-born individuals had lower awareness of contraceptives—including the IUD and vaginal ring—compared to their United States-born peers [32]. These data also showed that Hispanic women knew less about available contraceptive methods and side effects than White women. More acculturated immigrants tend to have earlier sexual debut and increased use of contraception [33, 34], while less-acculturated women may choose not to use LARC because of social stigma [12], less access to health care [10, 11], or distrust of the health care system [24]. With limited access to health care, foreign-born and less acculturated Hispanic women may have fewer opportunities to improve their knowledge of available contraception options and to cover the associated costs.

Provider bias may also play a role in how patients receive contraceptives. Researchers have reported that, without proper training, providers are uncomfortable talking about contraceptives with their patients and that some are unwilling to educate their patients about LARC because of the increased time required [19]. A study using national data from women aged 18–44 found no differences among racial/ethnic groups in terms of overall use of family planning services [35]. However, the study also found that Hispanic and Black women were more likely than White women to receive counseling for birth control and Hispanic women were more likely than White women to be counseled for sterilization [35]. On the other hand, Higgins and colleagues found that women, mostly White respondents, considered providers to be a trusted source of contraceptives and LARC [36]. However, several women in their study expressed that their preferences regarding contraceptive selection or removal were undervalued by their providers, especially when the conversation may lead to LARC removal. In addition, many participants, including all races, believed that providers tend to recommend LARC to “women of color, poor women, and women deemed uneducated or unintelligent by providers” [36]. Although studies have found mixed results regarding whether including training during residency helps with overall IUD insertion [37, 38], continuing education for physicians was found to be a strong predictor of LARC insertion [39]. Efforts to improve contraceptive rates need to a) consider ways to address cost, especially for postpartum LARC; b) create strategies to reduce physician bias [40] and misconceptions regarding promoting LARC; and c) provide culturally tailored patient education.

This study is the first to use real-world data to examine factors associated with contraceptive use in North Carolina. In addition, this study provides a unique opportunity to study how young women with different insurance status and racial/ethnic backgrounds use contraceptives. Yet, our approach is not without limitations. First, our data only captured prescription contraceptives ordered by a physician. Other nonprescription methods, such as condoms or vaginal contraceptive film, were not captured. Therefore, our results may underestimate the use of non-LARC contraceptives. Second, the data used in this study did not include information on potential confounding variables including parity, religion, level of acculturation, and socioeconomic status. We addressed the lack of individual data on socioeconomic status by including 4 area-level measures as proxies. Third, misclassification of insurance status is possible because of changes in insurance over the study period that are not reflected in our measure based on the most frequent payor type. However, we expect that misclassification would be nondifferential between groups of women studied. Fourth, although we excluded about one-fourth of the population due to lack of PCP visits and invalid addresses, patients who were excluded did not differ from the study sample with respect to age or race/ethnicity. Lastly, data were from a single health care system in the Southeastern United States and may not be generalizable to other regions.