Unintended pregnancies in the United States incur an estimated $21 billion in health care expenditures annually and place long-term financial burdens on women and families.1,2 A key strategy in reducing unintended pregnancy rates is to increase use of the most effective contraceptive methods. Long-acting reversible contraceptives, including intrauterine devices (IUDs), are associated with high patient satisfaction, fewer unintended pregnancies, and greater cost-effectiveness from many perspectives compared with other reversible contraceptive methods.3–6 High upfront cost is one of several known barriers to IUD use.7–11
The reproductive health community was hopeful that the elimination of out-of-pocket cost under the contraceptive mandate portion of the Affordable Care Act (ACA) would result in more women choosing highly effective contraceptives. Indeed, early studies of the contraceptive mandate have consistently shown that fewer women have out-of-pocket costs for contraception.12–14 However, contemporary studies have not consistently demonstrated that eliminating cost sharing resulted in increased long-acting reversible contraceptive use.9,15 One plausible explanation is that women's response to the mandate might depend on their baseline out-of-pocket cost. Prior studies have not considered baseline cost and have examined women enrolled in employer-based health plans, which often had generous contraceptive coverage even before mandated coverage.12 Our objective was to evaluate whether women enrolled in health plans with high out-of-pocket cost before the ACA's contraceptive mandate responded differently to the elimination of out-of-pocket costs than those in plans with lower baseline cost. We hypothesized that women with higher baseline out-of-pocket cost would be more likely than their peers to respond to the elimination of these costs.
MATERIALS AND METHODS
We conducted a cross-sectional pre–post analysis at the plan level using deidentified data from The Clinformatics Data Mart Database to examine changes in cost sharing and service utilization among commercially insured women between 2009 and 2014. This database represents more than 140 million unique members among nonretired employees and their dependents enrolled in commercial health insurance plans sponsored by large or medium-sized U.S.-based employers. It includes monthly enrollment; inpatient and outpatient claims; and complete patient payment and charge information, including patient copayment, deductible, and coinsurance amounts as well as standardized costs. This study was deemed exempt by the University of Michigan Medical School institutional review board. This study was deemed exempt because it involved the use of existing data without participant identifying information.
Our initial sample included women aged 15–45 years with prescription drug or medical coverage who had a prescription or medical claim for an IUD insertion during 2009–2014 and at least 12 months of continuous plan enrollment. We identified services using Current Procedural Terminology codes, Healthcare Common Procedure Coding System procedure codes, or both (Table 1). To estimate patient total out-of-pocket cost, we summed the copayment, coinsurance, and deductible payments for an IUD device and placement. We did not include other costs associated with the office visit because noncontraceptive services might be provided at the same time as contraception. Costs were calculated per calendar year for each patient who used IUD services and were used to calculate the plan average. All payments and cost estimates were adjusted for inflation to 2014 dollars using the U.S. Bureau of Labor Statistics Medical Consumer Price Index.
We further restricted our sample to patients in plans used for IUD services in both 2009 and 2014. Patient information was then aggregated by plan and year. To test our hypothesis that response to cost-sharing elimination depends on baseline out-of-pocket cost, we examined plan rates of IUD placement during 1 year in the precontraceptive mandate period (2009) and 1 year in the postcontraceptive mandate period (2014). These years were chosen to span the period surrounding the ACA's contraceptive mandate, implemented August 1, 2012 (Fig. 1).
Our primary outcome of interest was changes in IUD insertions before and after mandated coverage. For this analysis, we limited our sample to 1,930 unique plans used for IUD insertions in both 2009 and 2014 (Table 2). Plans were grouped into three categories for each year based on the plan's mean patient out-of-pocket cost: no out-of-pocket cost, low out-of-pocket cost, and high out-of-pocket cost. The 75th percentile for costs among the full sample of patients in 2009 was set as the cutoff for high out-of-pocket costs informed by the results from a sensitivity analysis of patient out-of-pocket cost. Low out-of-pocket cost was defined as greater than $0 but less than the 75th percentile, and high out-of-pocket cost was defined as greater than the 75th percentile. We further examined patient out-of-pocket cost by determining the proportion of cost represented by cost-sharing structure including deductible payments, copayments, and coinsurance. Figure 2 depicts the adjusted 75th percentile of patient out-of-pocket cost by cost-sharing structure and year.
Plan groups were then compared in 2009 and 2014 and were further categorized based on change in out-of-pocket cost over time. Nine categories represent each of the possible patterns over time. For each plan type in 2009 (no, low, or high out-of-pocket cost), three categories were created based on change from 2009 to 2014: one for no change and two for change to either of the other cost levels. For example, one group is defined by plans with high out-of-pocket cost in 2009 that changed to having low out-of-pocket cost in 2014 (Table 3).
Patient out-of-pocket cost by year stratified by plan category was summarized using descriptive statistics. A generalized estimating equation was used to model the average proportion of women within a plan with IUD utilization in 2009 and 2014. The model accounted for the dependence resulting from repeated measures per plan and was fit with a Gaussian distribution, an identity link function, and an exchangeable correlation structure. Robust standard errors were obtained to produce valid standard error estimates even if the within-plan correlation does not match the model-specified correlation structure. The primary variables of interest were the plan category, the claim year, and the interaction between plan type and claim year. Additional plan time-varying controls included the plan proportions of those within 400% of the federal poverty line (low income), the plan proportion of those with a race identified as Caucasian or white, and the plan proportion of those aged 41 years or older. Model-based marginal effects were obtained and represent the average predicted level of IUD service utilization within a plan by plan categorization (Table 4). A P value of .05 was considered significant. All analyses were performed using Stata 14.
We identified 1,420,954 women between the ages of 15 and 45 years with at least 12 months of continuous enrollment in 167,964 employer-based health plans (Table 2). Of these plans, our analytic sample consisted of 1,930 plans that were used for IUD services in both 2009 and 2014. These plans included 543,499 women, the majority of whom were Caucasian or white (67%) and with higher income (greater than 400% of the federal poverty level, 58%) (Table 2).
Across all plans and plan categories that used new IUD devices in 2009 and 2014, there was a small but significant increase in average plan-level IUD insertions (12.5%, 95% CI 11.6–13.4% vs 13.8%, 95% CI 13.0–14.7%, P<.001). By categorizing plans based on their out-of-pocket cost pattern over time, we examined this overall change in more detail and found differences between categories in average plan utilization over time. Of the 1,930 plans that were used for IUD insertions in both 2009 and 2014, 49.7% of plans went from low out-of-pocket cost to no out-of-pocket cost, 15.3% of plans went from high out-of-pocket cost to no out-of-pocket cost, 14.5% of plans had no out-of-pocket cost in both years, and 0.8% of plans had high out-of-pocket cost in both years. Each of the remaining combinations represented fewer than 2% of plans and thus were omitted for clarity. Two percent of plans continued to have out-of-pocket costs after the contraceptive mandate in 2014, which is consistent with the number of plans that continued to have grandfathered status at that time.16 In models predicting the probability of IUD insertion by plan category, while controlling for race and poverty level, plans that changed from high cost to no cost increased their use for IUD insertions at the highest rate (3.6%), whereas use among plans that remained no cost did not significantly change (Fig. 3). After controlling for plan levels of poverty along with racial and age compositions, plans that went from high out-of-pocket cost in 2009 to no out-of-pocket cost in 2014 saw a significantly higher average increase in the average level of plan IUD utilization over time (2.4%, 95% CI 0.04–4.5%) compared with plans with no out-of-pocket cost in both 2009 and 2014 (−1.0%, 95% CI −3.3 to 1.4%, P=.02).
Among all plans with out-of-pocket costs for IUD placement, the 75th percentile of out-of-pocket cost in 2009 was $368. By 2014, this number dropped to $0 (Fig. 2). Before the implementation of mandated coverage in 2012, deductible payments appeared to account for much of the observed out-of-pocket cost above the 75th percentile.
Our examination of the effect of the contraceptive mandate of the ACA on a national sample of women in employer-based health plans indicates that full coverage of contraception has successfully led to an increase in the proportion of patients with no out-of-pocket cost for IUDs. We found that elimination of out-of-pocket cost for IUD placement occurred quickly after mandated coverage. Our findings suggest that baseline levels of out-of-pocket cost may help identify groups with unmet demand for IUD placement before mandated first-dollar coverage. Those with the highest out-of-pocket cost at baseline (who are often in high-deductible health plans) saw the largest decreases in out-of-pocket costs and also demonstrated the steepest increases in IUD insertions after elimination of costs. This finding has important implications for the conduct and interpretation of studies on contraception coverage under health care reform.
Some prior studies reported that the implementation of the ACA has had little to no effect on use of IUDs.11,12 Our study adds more nuanced information on the effect of cost sharing on IUD utilization by accounting for baseline out-of-pocket cost. By grouping plans based on out-of-pocket cost before and after mandated coverage, this analysis demonstrates that 1) most commercially insured women already had low or zero out-of-pocket cost and 2) the elimination of out-of-pocket cost for the smaller group of women in plans with high baseline costs was associated with a greater increase in IUD placement. This finding aligns with the large body of existing literature before the ACA, suggesting a strong inverse association between out-of-pocket cost and IUD use.9,10,17 Prior studies on the effect of the ACA on IUD insertion did not account for the fact that most women in these types of health plans already had low out-of-pocket cost and likely little unmet demand as a result of costs. Based on our findings, we recommend that these studies of service utilization after the elimination of out-of-pocket costs should account for baseline costs.18
Despite mandated coverage of contraception, IUD utilization remains low for patients seeking contraception. Beyond reducing out-of-pocket cost, it is important to consider other factors that may represent barriers to IUD insertion such as availability of same-day IUD insertion at outpatient clinics and a sufficient supply of trained health care providers.19–22 Additionally, our findings should be interpreted in light of limitations common to large employer-sponsored insurance claims databases. The sample is drawn from a pool of women with employer-sponsored coverage, which represents approximately 66% of U.S. women; findings may therefore not be generalizable to women with public insurance, no insurance, or marketplace plans. These data rely on coding generated for billing insurance and thus may not capture every IUD insertion (We would not identify IUD insertions that occurred out of plan or if the plan offered no coverage). Conversely, the accuracy of a paid claim for IUD insertion is high. Furthermore, plans used in this analysis had to be used in both 2009 and 2014 for IUD insertion, which limits the number of plans available to be analyzed and could overrepresent plans with above average utilization. Finally, it is also worth remembering that IUD placement and contraceptive use are only intermediate outcomes to the health outcome of interest—unintended pregnancy. Given the well-demonstrated strong association between use of highly effective contraceptives such as IUDs and prevention of unintended pregnancy, IUD insertion is a well-accepted proximate determinant of unintended pregnancy.
This study has several strengths. First, this analysis demonstrates a novel approach to examining the effect of mandated full coverage of contraception use by demonstrating the importance of accounting for baseline costs. Our findings may also have implications for studies that seek to evaluate the effect of out-of-pocket cost changes on women's use of other preventive health services.
Aggregate analyses of utilization may miss changes in important subgroups of women because of large variations in out-of-pocket cost at baseline and subsequent variation in unmet demand. Moving forward, studies that examine responses to changes in patient out-of-pocket cost over time should include baseline cost in their analyses.
1. Sonfield A, Kost K. Public costs from unintended pregnancies and the role of public insurance programs in paying for pregnancy-related care: national and state estimates for 2010. New York (NY): Guttmacher Institute; 2015.
2. Institute of Medicine (US) Committee on Unintended Pregnancy. Brown SS, Eisenberg L, editors. The best intentions: unintended pregnancy and the well-being of children and families. Washington, DC: National Academies Press; 1995.
3. Peipert JF, Zhao Q, Allsworth JE, Petrosky E, Madden T, Eisenberg D, et al. Continuation and satisfaction of reversible contraception. Obstet Gynecol 2011;117:1105–13.
4. Trussell J, Henry N, Hassan F, Prezioso A, Law A, Filonenko A. Burden of unintended pregnancy in the United States: potential savings with increased use of long-acting reversible contraception. Contraception 2013;87:154–61.
5. Trussell J, Lalla AM, Doan QV, Reyes E, Pinto L, Gricar J. Cost effectiveness of contraceptives in the United States. Contraception 2009;79:5–14.
6. Russo JA, Miller E, Gold MA. Myths and misconceptions about long-acting reversible contraception (LARC). J Adolesc Health 2013;52(suppl):S14–21.
7. Harper CC, Blum M, de Bocanegra HT, Darney PD, Speidel JJ, Policar M, et al. Challenges in translating evidence to practice: the provision of intrauterine contraception. Obstet Gynecol 2008;111:1359–69.
8. Forrest JD. U.S. women's perceptions of and attitudes about the IUD. Obstet Gynecol Surv 1996;51(suppl):S30–4.
9. Secura GM, Allsworth JE, Madden T, Mullersman JL, Peipert JF. The Contraceptive CHOICE Project: reducing barriers to long-acting reversible contraception. Am J Obstet Gynecol 2010;203:115.e1–7.
10. Dusetzina SB, Dalton VK, Chernew ME, Pace LE, Bowden G, Fendrick AM. Cost of contraceptive methods to privately insured women in the United States. Womens Health Issues 2013;23:e69–71.
11. Postlethwaite D, Trussell J, Zoolakis A, Shabear R, Petitti D. A comparison of contraceptive procurement pre- and post-benefit change. Contraception 2007;76:360–5.
12. Pace LE, Dusetzina SB, Keating NL. Early impact of the Affordable Care Act on uptake of long-acting reversible contraceptive methods. Med Care 2016;54:811–7.
13. Becker NV, Polsky D. Women saw large decrease in out-of-pocket spending for contraceptives after ACA mandate removed cost sharing. Health Aff (Millwood) 2015;34:1204–11.
14. Finer LB, Zolna MR. Declines in unintended pregnancy in the United States, 2008-2011. N Engl J Med 2016;374:843–52.
15. Carlin CS, Fertig AR, Dowd BE. Affordable Care Act's mandate eliminating contraceptive cost sharing influenced choices of women with employer coverage. Health Aff (Millwood) 2016;35:1608–15.
16. Kaiser Family Foundation. Women's health insurance coverage. Available at: http://kff.org/womens-health-policy/fact-sheet/womens-health-insurance-coverage-fact-sheet/
. Retrieved December 17, 2017.
17. Gariepy AM, Simon EJ, Patel DA, Creinin MD, Schwarz EB. The impact of out-of-pocket expense on IUD utilization among women with private insurance. Contraception 2011;84:e39–42.
18. Bearak JM, Finer LB, Jerman J, Kavanaugh ML. Changes in out-of-pocket costs for hormonal IUDs after implementation of the Affordable Care Act: an analysis of insurance benefit inquiries. Contraception 2016;93:139–44.
19. Centers for Disease Control and Prevention (CDC). Contraceptive methods available to patients of office-based physicians and title X clinics—United States, 2009-2010. MMWR Morb Mortal Wkly Rep 2011;60:1–4.
20. Biggs MA, Harper CC, Malvin J, Brindis CD. Factors influencing the provision of long-acting reversible contraception in California. Obstet Gynecol 2014;123:593–602.
21. Harper CC, Stratton L, Raine TR, Thompson K, Henderson JT, Blum M, et al. Counseling and provision of long-acting reversible contraception in the US: national survey of nurse practitioners. Prev Med 2013;57:883–8.
22. Harper CC, Henderson JT, Raine TR, Goodman S, Darney PD, Thompson KM, et al. Evidence-based IUD practice: family physicians and obstetrician-gynecologists. Fam Med 2012;44:637–45.