Fulfillment of Permanent Contraception among Patients with Cesarean Delivery in a Multi-Site Cohort

Abstract

Objectives:

This study aimed to assess the association between insurance type and permanent contraception fulfillment among those with cesarean deliveries. Additionally, we sought to examine modification by the scheduled status of the cesarean.

Study Design:

We used data from a multi-site cohort study of patients who delivered in 2018–2019 at Northwestern Memorial Hospital in Illinois, MetroHealth Medical System in Ohio, or University of Alabama at Birmingham in Alabama. All patients had permanent contraception as their contraceptive plan in their medical chart at the time of delivery. We used logistic regression to model the association between insurance type, scheduled status of cesarean and permanent contraception fulfillment by hospital discharge. The scheduled status of cesarean delivery was examined as an effect modifier.

Results:

Compared to patients with private insurance, those with Medicaid were less likely to have their desired permanent contraception procedure fulfilled by hospital discharge (89.3% vs 96.8%, p<0.001). After adjusting for covariates, patients with Medicaid had a lower odds of permanent contraception fulfillment by hospital discharge (OR: 0.41; 95% CI: 0.21, 0.77). This association was stronger among those who had unscheduled cesarean deliveries (OR: 0.29; 95% CI: 0.12, 0.74) than those with scheduled cesarean deliveries (OR: 0.77; 95% CI: 0.32, 1.88).

Conclusions for Practice:

Compared to patients with private insurance undergoing a cesarean delivery, those with Medicaid insurance were less likely to have their desired permanent contraception fulfilled. Physicians and hospitals must examine their practices surrounding Medicaid forms to ensure that patients have valid consent forms available at the time of delivery.

Introduction

There are a variety of well-documented barriers to fulfillment of postpartum permanent contraception (Bullington & Arora, 2022). At the physician and hospital levels, barriers include scheduling constraints, lack of available operating rooms, and poor reimbursement, among others. At the policy level, numerous studies have documented that patients with Medicaid insurance are less likely to achieve their desired permanent contraception procedure (K. S. Arora et al., 2023; Boardman et al., 2013; Hahn et al., 2019; Zite et al., 2005, 2006). This is, in part, due to the Medicaid sterilization policy; all patients who plan to undergo permanent contraception must sign a consent form and subsequently undergo a 30-day waiting period from the time the form is signed to the time the procedure is performed (Block-Abraham et al., 2015; Borrero et al., 2014). This policy applies to all permanent contraception procedures covered by Medicaid, including vasectomy, but does not apply to any other procedures. Because many of the barriers to permanent contraception are intertwined, isolating the impact of the Medicaid sterilization policy from other barriers to permanent contraception is difficult.

Delivery type is one of the strongest predictors of whether a patient has their desired permanent contraception procedure fulfilled(K. Arora et al., 2018; K. S. Arora et al., 2023). Studies have found that the adjusted odds of receiving permanent contraception within 42 days of delivery among patients who deliver via cesarean is between 15.5 and 64.7 times the odds of receiving permanent contraception among patients who deliver vaginally (K. Arora et al., 2018; K. S. Arora et al., 2023). This is likely because undergoing permanent contraception during cesarean delivery removes many physician- and hospital-level barriers, like scheduling and operating room availability. In fact, the American College of Obstetrics and Gynecology notes performing postpartum permanent contraception at the time of cesarean delivery may be “ideal” given “technical ease for the physician and convenience for the patient” (“Access to Postpartum Sterilization: ACOG Committee Opinion Summary, Number 827,” 2021). Thus, understanding if and how insurance type impacts fulfillment among those with cesarean delivery allows us to look further at the association between the Medicaid sterilization policy and permanent contraception fulfillment. One study conducted at MetroHealth Medical System in Ohio found that, among patients who delivered via cesarean, those with Medicaid were 15% less likely to receive permanent contraception at the time of delivery compared to those with private insurance (Morris et al., 2019).

However, it remains unclear whether this insurance-based disparity for patients desiring permanent contraception and undergoing cesarean delivery persists in a multi-center cohort. Furthermore, it is unknown whether the cesarean delivery was scheduled or unscheduled impacts fulfilment of desired permanent contraception. An unscheduled delivery may preclude fulfillment of postpartum permanent contraception, particularly if patients were not counseled on or previously considering permanent contraception or if the acuity of the unscheduled cesarean procedure prevents confirming a patient’s desire to proceed with permanent contraception during an emergent procedure. Patients with Medicaid insurance may be more impacted by unscheduled cesarean deliveries given the Medicaid sterilization consent form may not have been signed sufficiently in advance, especially if the initial clinical plan had been a vaginal delivery followed by an interval permanent contraception procedure. Therefore, differentially examining the association between insurance type and fulfillment of permanent contraception among those with scheduled and unscheduled cesareans allows for a closer look at policy-level impacts of the Medicaid consent process. Our goal was to explore fulfillment of postpartum permanent contraception among patients who undergo cesarean delivery in a multi-institution cohort. Among all patients with cesarean deliveries, we hypothesized that those with Medicaid would be less likely to have their desired permanent contraception procedure fulfilled compared to those with private insurance. We hypothesized that this difference would be larger for those with unscheduled compared to scheduled cesarean deliveries.

Methods and Materials

Sample

Data for this planned secondary analysis come from a retrospective cohort of patients who delivered $\ge$20 weeks gestation between January 1, 2018 and December 31, 2019 at one of four hospitals across the US: University of California in San Francisco (UCSF) in California, Northwestern Memorial Hospital (NMH) in Illinois, MetroHealth Medical System (MHS) in Ohio, or University of Alabama at Birmingham (UAB) in Alabama. As California requires a waiting period for all patients undergoing permanent contraception, regardless of insurance, we excluded patients for this secondary analysis who delivered at UCSF (N=86) from this analysis focusing on the impact of the federal Medicaid policy. Patients included in the cohort had permanent contraception as their documented contraceptive plan, which was defined as either the plan documented in the 1) patients’ delivery hospitalization discharge summary or 2) last inpatient progress note from their delivery hospitalization, if there was no contraceptive plan documented in the discharge summary. Patients were excluded from the cohort if they suffered peripartum mortality, had previously received permanent contraception but conceived using in vitro fertilization, or underwent a cesarean hysterectomy due to suspected placenta accreta spectrum. Full methodological details are available for the primary analysis which focused on insurance-based differences across the cohort, regardless of mode of delivery (K. S. Arora et al., 2023).

Relevant demographic and clinical information were abstracted from inpatient and outpatient electronic medical records, including type of delivery, insurance status, and fulfillment of permanent contraception at three timepoints (before delivery discharge, within six weeks of delivery, and within one year of delivery).

For this secondary analysis focused on cesarean delivery, patients who delivered vaginally (N=1,609) or for whom delivery type was unknown (N=5) were excluded. Because our objective focused on identifying the impact of policy-level barriers to permanent contraception among those who desired the procedure, we excluded patients who changed their mind (as documented in the medical record) about their desire for permanent contraception prior to delivery (N=94).

Variables of interest

Our primary outcome was fulfillment of postpartum permanent contraception prior to hospital discharge. We also examined permanent contraception fulfillment within six weeks of delivery and within one year of delivery. Among participants who did not have the procedure fulfilled by hospital discharge, we examined reasons for non-fulfillment, which were determined via electronic medical record and categorized by a trained research assistant.

Our primary exposures in this analysis were insurance type and scheduled status of cesarean. Patients who had more than one insurance plan including Medicaid were categorized as having Medicaid insurance, as the federal sterilization consent process and waiting period still apply. Scheduled status of cesarean was determined from the medical record and analyzed as either scheduled or unscheduled.

Covariates of interest included body mass index at admission, parity, maternal age, and gestational age at delivery. Covariates were chosen because they are factors that impact contraceptive decision-making and/or fulfillment of permanent contraception (K. S. Arora et al., 2023; Bullington & Arora, 2022; Robinet et al., 2023). All covariates were determined using electronic medical records. Body mass index (BMI) was included because there is variation in clinician comfort performing permanent contraception on patients with higher BMI (K. S. Arora et al., 2018). Parity at admission and maternal age at the time of delivery were included because both have been shown to influence clinician contraceptive counseling and patient contraceptive decision-making (Mosley et al., 2023). Finally, gestational age at the time of delivery was included because early preterm birth can influence patient decision-making and clinician comfort performing permanent contraception (K. S. Arora et al., 2018; Chen et al., 2022); gestational age was dichotomized at 32 weeks (<32 weeks, $\ge$32 weeks).

Medicaid sterilization form validity was defined as having a form available in the medical record with a waiting period of 30 days between the date the form was signed and delivery (or 72 hours in the case of delivery prior to 37 weeks gestation). Reasons for permanent contraception non-fulfillment by hospital discharge were abstracted by trained research assistants from the medical chart using a predetermined list of possible reasons for non-fulfillment developed based on existing literature (K. Arora et al., 2018). Some participants were categorized as having more than one reason for non-fulfillment, so percentages may not add to 100%.

Analytic Approach

We descriptively examined sociodemographic and clinical characteristics of the sample, stratified by insurance type, using Mann Whitney U tests and χ² tests for continuous and proportional variables, respectively. We then calculated the proportion of participants with Medicaid and private insurance who had their desired permanent contraception procedure fulfilled by discharge, six weeks after delivery, and one year after delivery, using χ² tests to determine differences. We further stratified by scheduled status of cesarean, calculating the proportion of participants who had the procedure fulfilled by delivery by insurance status among those with scheduled and unscheduled cesareans. We examined this both by individual site and across all sites, using χ² tests for significance.

We modeled the association between insurance type scheduled status of cesarean, and fulfillment of permanent contraception by hospital discharge using logistic regression. We examined effect measure modification by scheduled status of cesarean. We conducted a Breslow Day Test and Wald Chi-Square Test to examine whether there were differences in the association between insurance type and fulfillment by scheduled status of cesarean. In our models, we used an interaction term between insurance type and scheduled status of cesarean. We decided to present results overall and stratified by scheduled status of cesarean regardless of significance of statistical tests because of the substantive interest in cesarean scheduled status as an effect measure modifier. We therefore estimated odds ratios and 95% confidence intervals for the association between insurance type and fulfillment at delivery overall and stratified by scheduled status of cesarean. We adjusted for BMI, parity before delivery, maternal age at delivery, and gestational age at delivery. There was no significant collinearity across covariates. Throughout, we used two-tailed tests and evaluated statistical significance with an alpha of 0.05. Institutional review board approval was received from MHS on January 24, 2020 (#IRB20–00026).

Results

A total of 1,313 participants met inclusion criteria, of whom 815 (62.1%) had Medicaid insurance and 498 (38.0%) had private insurance (Figure 1). Compared to participant with private insurance, those with Medicaid insurance were, on average, more likely to be younger, have more than two children, be unmarried, and have delivered at less than 32 weeks’ gestation (Table 1). Participants with Medicaid versus private insurance were less likely to have had a scheduled cesarean (50.7% vs. 70.7%). Overall, 64% of those with Medicaid had valid Medicaid sterilization consent forms at time of delivery.

Figure 1.

Flow chart of multi-center analytic cohort

Table 1. Demographics and clinical characteristics of patients who desired permanent contraception and delivered via cesarean delivery, by insurance status, 2018–2019.

	Medicaid N=815		Private N=498		p-value
	N/median	%/IQR	N/median	%/IQR
Maternal age at delivery (years)	32.0	27.9, 36.0	36.0	32.0, 38.8	<0.001
Parity less than 2 at admission	184	22.58%	237	47.59%	<0.001
Gestational age <32 weeks	67	8.22%	15	3.01	<0.001
Married (%)	215	26.38%	398	79.92%	<0.001
Urgency of cesarean
Scheduled	413	50.67%	352	70.68%	<0.001
Unscheduled	396	48.59%	145	29.12%
Missing	6	0.74%	1	0.20%
BMI (kg/m2)	35.5	31.0, 41.2	33.6	29.2, 38.8	<0.001
Site
MetroHealth	244	29.94%	52	10.44%	<0.001
Northwestern	161	19.75%	288	57.83%
UAB	410	50.31%	158	31.73%

Compared to participants with private insurance, those with Medicaid insurance were less likely to have their desired permanent contraception procedure fulfilled by discharge (89.3% vs. 96.8%; p<0.001), within six weeks of delivery (89.5% vs. 97.0%; p<0.001), and within one year of delivery (90.8% vs. 97.0%; p<0.001). Notably, in the full sample, only 13 of the 103 (12.6%) participants who did not undergo surgery for permanent contraception during their cesarean delivery underwent the procedure within one year of delivery.

Table 2 shows fulfillment of permanent contraception by the time of hospital discharge by insurance type and scheduled status of cesarean for each site individually and pooled across all sites. In pooled analysis, insurance status was not associated with fulfillment of desired permanent contraception at time of scheduled cesarean delivery. However, those with Medicaid insurance who had unscheduled cesareans were significantly less likely to have their desired permanent contraception procedure fulfilled by the time of discharge compared to those with private insurance who had unscheduled cesareans (83.6% vs. 95.2%; p<0.001). Overall, rates of fulfillment for permanent contraception were lower for patients undergoing unscheduled versus scheduled cesarean deliveries.

Table 2. Fulfillment of postpartum permanent contraception at the time of delivery among those with cesarean deliveries by insurance status and scheduled status of cesarean, 2018–2019.

	Scheduled cesarean					Unscheduled cesarean
	Medicaid		Private insurance		p-value	Medicaid		Private insurance		p-value
	N	%	N	%		N	%	N	%
MetroHealth	108	93.10%	31	100%	0.13	79	61.72%	17	80.95%	0.09
Northwestern	104	94.55%	236	97.52%	0.15	41	80.39%	43	93.48%	0.06
UAB	181	96.79%	76	96.20%	0.81	211	97.24%	78	100.00%	0.14
Total	393	95.16%	343	97.44%	0.10	331	83.59%	138	95.17%	<0.001

In Table 3, we present univariate and multivariable analyses for the association between insurance type and fulfillment overall and among those with scheduled and unscheduled cesarean deliveries. The adjusted odds of permanent contraception fulfillment before hospital discharge among those with Medicaid insurance was 0.41 times the odds among those with private insurance (95% confidence interval (CI): 0.21, 0.77). For those with unscheduled cesarean deliveries, the odds of permanent contraception fulfillment by hospital discharge among those with Medicaid was 0.29 times the odds of fulfillment among those with private insurance (95% CI: 0.12, 0.74). For those with scheduled cesareans, the odds ratio comparing Medicaid to private insurance was 0.77 (95% CI: 0.32, 1.88). Results from the Breslow Day Test indicated that p=0.2 for differences in association between insurance type and fulfillment among strata of scheduled status of cesarean. The Chi-Square Test for the interaction term between insurance type and cesarean scheduled status had p=0.1. Graphical representation of potential effect measure modification by scheduled status is shown in Supplemental Figure 1.

Table 3. Univariate and multivariable analyses for the association between insurance status, scheduled status of cesarean, and fulfillment of permanent contraception before hospital discharge, 2018–2019.

	Unadjusted		Adjusted*
	OR	95% CI	OR	95% CI
Overall Medicaid **	0.28	0.16, 0.48	0.41	0.21, 0.77
Scheduled cesarean Medicaid **	0.52	0.23, 1.15	0.77	0.32, 1.88
Unscheduled cesarean Medicaid **	0.26	0.12, 0.58	0.29	0.12, 0.74

^*Adjusted for BMI, site, parity, maternal age at delivery, and gestation age at delivery

^**Referent – private insurance

Finally, we examined reasons for permanent contraception non-fulfillment at the time of hospital discharge among those with cesarean deliveries who did not undergo the procedure before hospital discharge (Table 4).. Of these 103 participants, 87 (84.5%) had Medicaid insurance. Of those with Medicaid, 72 (82.8%) did not have a valid Medicaid consent form at the time of delivery. Of the 72 participants with invalid Medicaid consent forms, 56 (77.8%) had unscheduled cesareans.

Table 4. Reasons for permanent contraception fulfillment among participants who did not achieve their desired permanent contraception procedure by delivery hospitalization discharge^*.

	Medicaid n=87	Private insurance n=16
Invalid Medicaid consent form	72 (82.8%)	0 (0.0%)
Financial reason	1 (1%)	0 (0.0%)
Provider barrier	32 (36.8%)	4 (25.0%)
Medical barrier	25 (28.7%)	11 (68.8%)
Planned to have an interval procedure	18 (20.7%)	2 (12.5%)
Decided on permanent contraception after delivery	4 (4.6%)	0 (0.0%)

^*Some participants had multiple reasons for non-fulfillment, so percentages may not add to 100%.

Discussion

In this secondary analysis of a multi-site cohort study, we examined differences in fulfillment of postpartum permanent contraception among those with cesarean delivery by insurance type and scheduled status of cesarean to examine the policy-level impact of the Medicaid sterilization consent process. Overall, we found that participants with Medicaid undergoing cesarean delivery were less likely to have their desired permanent contraception procedure fulfilled compared to those with private insurance. These associations were particularly apparent among the subgroup of people who had unscheduled cesarean deliveries. Cesarean delivery removes some physician and hospital-level barriers, like scheduling constraints and operating room availability, that impact prioritization and ease of scheduling surgery for permanent contraception. During cesarean deliveries, these barriers are removed for all patients, regardless of insurance type or scheduled status of cesarean. However, patients with Medicaid insurance who undergo unscheduled cesareans continue to face policy-level barriers to desired permanent contraception compared to their privately insured counterparts given that they are required to have a valid Medicaid consent form in order to receive the procedure.

This analysis builds on previous work that has examined the association between insurance type and permanent contraception fulfillment. Numerous papers, including the primary analysis from this multi-site cohort study (K. S. Arora et al., 2023), have reported that patients with Medicaid are less likely to have their desired permanent contraception procedure fulfilled compared to patients with private insurance (Hahn et al., 2019; Wolfe et al., 2017; Zite et al., 2005, 2006). Our prior work at MHS alone reported that 66% of people delivering via cesarean who desired but did not achieve permanent contraception did not have valid Medicaid consent forms at the time of delivery (Morris et al., 2019). The present study confirmed the single-site study’s findings in a multi-site cohort and also expanded upon them by examining the role of the scheduled status of a cesarean delivery in permanent contraception fulfillment by insurance type, finding that those with Medicaid and unscheduled, but not scheduled, cesarean deliveries had reduced odds of permanent contraception fulfillment compared to those with private insurance. Given differences when stratified by scheduled status of cesarean, the impact of physician practice and hospital-level processes on permanent contraception fulfillment are important. That is, processes that ensure Medicaid consent forms are signed early in the prenatal course help ensure that patients who have unscheduled cesareans can still receive their preferred method of contraception.

Future research is needed to more fully examine the nuances of the relationship between delivery type and permanent contraception fulfillment. For example, qualitative work that explores differences in clinician counseling regarding mode of delivery with documented desire for permanent contraception, as well as further study of how delivery type impacts patient contraceptive decision-making, would provide important insights. Additionally, research that explores clinician perceptions of whether unscheduled cesarean deliveries qualify as emergency abdominal surgery whereby the Medicaid waiting period can be reduced to 72 hours rather than 30 days would provide insights into the Medicaid policy as a barrier to fulfillment among those with unscheduled cesareans. Further, research that explores the implications of updating the Medicaid consent policy by reducing the length of the waiting period or updating the consent form would have broad policy implications.

The major strength of this study is that it examines a diverse, multi-center cohort of practices across the United States. The robust sample size and rigorous medical record abstraction homogenized across sites allowed for meaningful comparisons across groups. This study has several limitations. First, it is retrospective and thus limited by the quality of the medical records. Yet, using medical records allowed us to include all patients who desired permanent contraception at the included institutions, thereby mitigating selection bias or limitation solely to those patients who received surgery for permanent contraception. Second, loss to follow-up impacts our findings. However, given that the Medicaid sterilization consent form would need to be transferred or resigned at a new practice if care is transferred prior to fulfillment of desired permanent contraception, this likely biases our study findings towards the null. It is important to note that no states had extended Medicaid to 12-months postpartum during this study’s data collection period. While the impact on permanent contraception fulfillment during time of cesarean delivery is likely minimal, extended postpartum insurance coverage may increase interval outpatient fulfillment. Additionally, this study may be underpowered to assess effect measure modification due to small sample size in subgroups. However, given that no prior studies have examined associations between scheduled status of delivery type and permanent contraception fulfillment, this study provides important preliminary work on the topic.

Conclusions

This study builds on existing literature by examining the association between Medicaid and fulfillment of permanent contraception among those with cesarean delivery in a multi-site cohort and by exploring effect measure modification by scheduled status of cesarean delivery. Our study demonstrates that current Medicaid policy hampers a person’s ability to have their contraceptive decisions honored, especially during an unscheduled cesarean delivery. Until such policies are modified, it is imperative that physicians and hospitals analyze their practices surrounding permanent contraception to ensure that patients have valid consent forms at time of delivery and that physician- and hospital-level barriers are minimized for permanent contraception at time of cesarean delivery – particularly for unscheduled cesarean cases.

Significance:

What is already known about this subject?

Studies have shown that the Medicaid sterilization consent process are associated with decreased fulfillment of desired postpartum permanent contraception. Cesarean delivery is one of the strongest predictors of permanent contraception fulfillment.

What this study adds?

This study found that Medicaid is associated with decreased permanent contraception fulfillment in a subset of patients who had cesarean deliveries in a national multi-center cohort. These associations were particularly strong among patients who had unscheduled, compared to scheduled, cesarean deliveries.