Insurance Churn and Postpartum Health among Texas Women with Births Covered by Medicaid/CHIP

Abstract

Background:

Insurance churn (changes in coverage) after childbirth is common in the U.S., particularly in states that have not expanded Medicaid. Although insurance churn may have lasting consequences for healthcare access, most research has focused on the initial weeks following a birth.

Methods:

We analyzed data from a cohort study of postpartum Texans with pregnancies covered by public insurance (n=1,489). Women were recruited shortly after childbirth from 8 hospitals in 6 cities, completing a baseline survey in the hospital and follow-up surveys at 3, 6, and 12 months. We assessed insurance trajectories, healthcare utilization, and health indicators over the 12 months following childbirth. We also conducted a content analysis of women’s descriptions of postpartum health concerns.

Results:

A majority of participants (64%) became uninsured within 3 months of the birth and remained uninsured for the duration of the study; 88% were uninsured at some point in the year following the birth. At three months postpartum, 17% rated their health as fair or poor, and 13% reported a negative change in their health after the 3-month survey. Women’s open-ended responses described financial hardships and other difficulties accessing care for postpartum health issues, which included acute and ongoing conditions, undiagnosed concerns, pregnancy and reproductive health, mental health, and weight/lifestyle concerns.

Conclusions:

Insurance churn was common among postpartum women with births covered by Medicaid/CHIP and prevented many women from receiving healthcare. To improve postpartum health and reduce maternal mortality and morbidity, states should work to stabilize insurance coverage for women with low incomes.

Introduction

In the United States (U.S.), instability in insurance coverage, or insurance churn, is common in the months preceding and following pregnancy, particularly for women living on low incomes (D’Angelo et al., 2015; Daw et al., 2017; Ranji et al., 2019). During pregnancy, women with low incomes who may otherwise be uninsured can gain coverage through pregnancy Medicaid—which has more generous income limits than standard Medicaid in most states—or through the Children’s Health Insurance Program (CHIP) Unborn Child Option,¹ which covers pregnant women who are ineligible for Medicaid due to immigration status (Green et al., 2016). However, CHIP coverage for postpartum care is limited and not available in all states, and pregnancy Medicaid coverage expires after 60 days postpartum.² Postpartum coverage after 60 days depends on eligibility for standard Medicaid, which varies by state (Daw et al., 2021; Green et al., 2016; Ranji et al., 2019). Thus, women whose pregnancies are covered by Medicaid or CHIP are at risk of becoming uninsured early in the postpartum period (Daw et al., 2017, 2021). Stable insurance is vital to the diagnosis and management of chronic conditions, such as hypertension and diabetes (Daw & Sommers, 2019), and inconsistent coverage between pregnancies may contribute to a cycle of poor interconception health leading to higher-risk pregnancies and additional health problems (Daw et al., 2017; Daw & Sommers, 2019).

Medicaid expansion under the Affordable Care Act has reduced insurance churn among reproductive-aged women and increased healthcare utilization in the postpartum period (Daw et al., 2021; Daw & Sommers, 2019; Dunlop et al., 2020; Gordon et al., 2020). However, residents of expansion states are still vulnerable to insurance churn if their incomes exceed 138% of the federal poverty level (FPL), the eligibility threshold for nonelderly adults.³ Moreover, women living in the 12 states that have not expanded Medicaid (Kaiser Family Foundation, 2021c) remain at higher risk of losing coverage after childbirth and experiencing adverse health outcomes associated with insurance loss. This is particularly concerning given that women with low incomes and women of color are disproportionately affected by insurance churn and by serious maternal morbidity and maternal mortality (D’Angelo et al., 2015; Joseph et al., 2021; Liese et al., 2019; Louis et al., 2015; Maternal Mortality and Morbidity Task Force, 2020; Petersen et al., 2019).

Texas is an important setting to examine postpartum women’s health and experiences of insurance churn in the months following childbirth, given its large population of reproductive-age women and its high rates of maternal morbidity and mortality (MacDorman et al., 2018). The state has the largest share of women age 19–64 without insurance (23%) in the U.S. (Kaiser Family Foundation, 2019), as well as large racial/ethnic disparities in the percent uninsured: 14% of non-Hispanic White women, compared to 18% of non-Hispanic Black women and 37% of Hispanic women (U.S. Census Bureau 2019).⁴ As of June 2021, Texas has not expanded Medicaid and Medicaid eligibility for non-pregnant people is limited to those with incomes equivalent to <17% of the Federal Poverty Level (FPL) (Kaiser Family Foundation, 2021b). Uninsured women earning below 200% FPL qualify for family planning services and reproductive healthcare through Healthy Texas Women (HTW), a Medicaid waiver program; Healthy Texas Women Plus, an expansion program implemented in 2020, covers an additional subset of health services (e.g., screening and treatment for hypertension, postpartum depression, and substance use disorders) for postpartum women with low incomes in the 12 months after childbirth. Postpartum care for undocumented women is extremely limited: HTW and HTW Plus are restricted to U.S. citizens and legal residents, and CHIP Perinatal (the Texas implementation of the CHIP Unborn Child Option) covers only two postpartum visits. Other state-funded programs that do not restrict services based on immigration status have historically been underfunded (Carpenter et al., 2021), further limiting the scope of care available to individuals who do not qualify for Medicaid.

Typical sources of state-level data on pregnant and postpartum women’s health, such as the Pregnancy Risk Assessment Monitoring System (PRAMS), offer limited insight into the consequences of insurance churn because they are cross-sectional and do not collect information on women’s health beyond the early postpartum period. Longitudinal assessment of postpartum women’s experiences with insurance churn is important as states consider policy strategies to address maternal and morbidity, such as extending Medicaid coverage to 12 months following a birth. In this analysis, we use data from a prospective cohort study of postpartum women in Texas to examine insurance churn and maternal health in the 12 months following a birth. After describing longitudinal patterns of insurance coverage and several maternal health indicators, we conduct a content analysis of open-ended survey responses in order to characterize postpartum women’s salient health concerns.

Methods

Data

We conducted a secondary analysis of survey data from the Texas Postpartum Contraception Study. This longitudinal study recruited women who gave birth between 2014 and 2016 at one of eight participating hospitals in six cities throughout Texas. Women between the ages of 18 and 44 were eligible if they gave birth to a singleton infant that did not require extended care in a neonatal intensive care unit, had a birth that was not covered by private insurance, lived in Texas within the catchment area of the hospital where they gave birth, spoke English or Spanish, and did not plan to have another child within the next two years. Of the 1,825 women who met these criteria, 1,700 (93%) agreed to participate. After providing informed consent, women completed a baseline survey in person with the assistance of a trained interviewer before leaving the hospital. Women completed follow-up surveys via telephone at 3, 6, 12, 18, and 24 months following childbirth. Women who did not complete a follow-up survey were contacted to participate in the next survey; those who missed two consecutive surveys were considered lost to follow-up. This study was approved by the institutional review boards of all participating hospitals and the University of Texas at Austin.

This analysis uses data from the baseline survey and follow-up surveys conducted 3, 6, and 12 months following childbirth. Data collection for the 12-month surveys concluded in 2017, prior to the implementation of Healthy Texas Women Plus. Our sample includes 1,489 women who had pregnancies covered by Medicaid or CHIP and completed at least one follow-up survey. This analysis excludes 59 women who were uninsured at the time when they gave birth.

Measures

At the baseline survey, women reported their age, parity, race/ethnicity, and nativity. We constructed a composite measure of race/ethnicity and nativity with the following categories: Hispanic, born in U.S; Hispanic, born outside U.S.; non-Hispanic Black; non-Hispanic White; and non-Hispanic Other. The 3-month survey captured receipt of government assistance.

Outcomes of interest included insurance coverage, healthcare utilization during the postpartum period, and indicators of women’s healthcare needs in the 12 months following childbirth.

Insurance was captured at each survey wave. At the time of the birth, all respondents indicated having public insurance (i.e., Medicaid, CHIP Perinatal, or Emergency Medicaid.) In later surveys, women were coded as publicly insured, privately insured (including Tricare), or uninsured. Women without private or public insurance who reported enrollment in the Healthy Texas Women program (or its predecessor program)⁵, enrollment in county programs, or clinic discounts were considered uninsured because these programs do not provide comprehensive healthcare coverage (i.e., broad coverage for a wide range of acute and chronic health conditions, encompassing both diagnosis and treatment.)

Measures of healthcare utilization in the postpartum period included completion of a postpartum visit (captured at 3 months, or at 6 months if no 3-month survey), whether women ever wanted to have a health condition or illness checked by a doctor in the year following the birth (measured at 12 months), and whether women who wanted to see a doctor during this time had done so (measured at 12 months).

In order to assess women’s probable need for healthcare during this same period, we compiled several indicators of postpartum health and illness. Self-rated health (captured at 3 months) provided a holistic assessment of women’s health early in the postpartum period (excellent/very good/good/fair/poor). Women selecting “fair” or “poor” received an open-ended survey question prompting them to explain their response. In the 6- and 12-month surveys, women were asked whether they had experienced any change in their health since the previous survey; women indicating any change in their health received an open-ended survey question in which they were asked to describe the change(s). Responses to these open-ended survey questions provided insight into the range of health issues that women experienced in the postpartum period.

We also constructed indicators of specific chronic conditions linked to maternal mortality and severe maternal morbidity (Admon et al., 2017; Maternal Mortality and Morbidity Task Force, 2020). Hypertension and diabetes in the 12 months after childbirth were composite measures based on several items in the baseline and follow-up surveys. We considered women to have hypertension if they ever reported current high blood pressure, use of blood pressure medication, or gestational hypertension that continued after the birth. We coded women as having diabetes if they reported having diabetes before their pregnancy or diabetes that was diagnosed in pregnancy and persisted beyond three months postpartum. Since maternal mortality and morbidity are associated with obesity (Maternal Mortality and Morbidity Task Force, 2020; Nelson et al., 2018), we computed BMI (kg/m²) at each survey wave using height (measured at baseline) and weight (captured in each follow-up survey.)

Analysis

Quantitative analyses were conducted in Stata 16 unless otherwise noted. First, we calculated the percentage with Medicaid/CHIP, private insurance, and no insurance at each survey wave. We used the R package TraMineR to identify the most common sequences of insurance status during the 12 months after childbirth. Next, we calculated the percentage of women who completed a postpartum visit. We also calculated the distribution of self-rated health 3 months after the birth, the percentage who reported a negative change in their health, the distribution of BMI 12 months after the birth, and the prevalence of hypertension and diabetes. In light of known disparities in hypertension and diabetes among American adults (Centers for Disease Control and Prevention, 2020; Ostchega et al., 2020), we conducted an exploratory analysis of racial/ethnic differences. We also computed the percentage of women who wanted to have a concern or illness checked by a doctor during the postpartum period and the percentage of these women who reported visiting a doctor about their health concern.

Finally, we conducted a content analysis of women’s open-ended survey responses (n=445) describing their health. The responses collected at 3 months postpartum explained why women had rated their health as fair or poor; at 6 and 12 months postpartum, these responses described changes in women’s health since the previous survey. Open-ended responses were coded through an iterative process. Responses provided in Spanish were translated by a bilingual author. Two authors read all responses and proposed a preliminary set of codes. The resulting categories and coding decisions were refined through discussion among all authors. Responses that described multiple health concerns or changes could be assigned multiple codes. All final coding decisions were reviewed by the full group of authors. This content analysis was conducted in Microsoft Excel.

Results

Just under half of respondents (46%) were between 18 and 24 years old; another 26% were 25 to 29 years old and 28% were 30 years or older. Hispanic women born in the U.S. made up 36% of the sample and Hispanic women born outside the U.S. made up 44%. Non-Hispanic Black women accounted for 14% of the sample and the remaining 6% of women were Non-Hispanic White, Asian, Native American, or some other race/ethnicity. One quarter (26%) of participants had just given birth to their first child, 32% had two children, and 43% had three or more children.

Three months after the birth, 83% lived in a household participating in the Supplemental Nutrition Program for Women, Infants, and Children (WIC), 58% lived in a household participating in the Supplemental Nutritional Assistance Program (SNAP), 2% lived in a household receiving Temporary Assistance for Needy Families (TANF), and 91% of women had a child in their household enrolled in the Children’s Health Insurance Plan (CHIP).

Health insurance and healthcare utilization

Among women who had Medicaid/CHIP coverage at the time of the birth, 77% were uninsured at 3 months postpartum (Figure 1). The share of women without insurance remained stable at around three-quarters for the rest of the study period. Overall, 86% of women were uninsured at some point during this time, and only 12% of women held private insurance at any time in the year after childbirth.

Figure 1:

Insurance coverage in the 12 months following childbirth (n=1,489)

The most common trajectory of insurance coverage during study period, experienced by 59.3% of women, was to have a birth covered by public insurance, become uninsured by 3 months postpartum, and remain uninsured for the remainder of the year (Figure 2). The second-most common trajectory, experienced by 5.8% of women, was to remain on public insurance for the entire 12 months.

Figure 2:

Most common insurance trajectories in the 12 months after childbirth (n=1,489)

Six months after the birth, 17.1% of women had not had a postpartum visit. Twelve months after the birth, 14.9% of women reported having a condition or illness in the previous year that they had wanted to have checked by a doctor. Of these women, 53.5% reported seeing a doctor.

Health indictors

In this sample, 8% of women reported that they had hypertension and 1% of women reported that they had diabetes. Non-Hispanic Black women were nearly three times as likely to have hypertension (18%) than US-born Hispanic women (6%), foreign-born Hispanic women (6%), or women in other racial/ethnic groups (6%) (p<0.001). Before the index pregnancy, 24% of women had a BMI in the overweight category and 30% had a BMI in the obese category; 12 months after the birth, 26% had a BMI in the overweight category and 37% had a BMI in the obese category.

Although the majority of women rated their health as excellent, very good, or good 3 months after the birth, 17% considered their health to be fair or poor. Half of non-Hispanic Black women (50%) described their health as excellent or very good, compared to 40% of foreign-born Hispanic women and 38% of US-born Hispanic women; only 14% of non-Hispanic Black women described their health as fair or poor, versus 16% of foreign-born Hispanic women and 19% of US-born Hispanic women (p<0.05). Among women who had insurance at 3 months postpartum, 47% rated their health as excellent or very good, compared to 40% of women who had recently lost their insurance (p=0.063). Overall, 13% of women reported a negative change in their health in either the 6-month survey or the 12-month survey.

Open-ended health descriptions

Postpartum women’s open-ended survey responses (n=445) reflected a variety of health concerns experienced in the postpartum period as well as obstacles that limited access to care. We identified seven categories of health responses: ongoing conditions, undiagnosed concerns, acute conditions, mental health, weight/lifestyle, pregnancy, and non-pregnancy reproductive health. Regardless of their specific health concerns, women described navigating a fragmented healthcare system on their own with little support and varying levels of success. Moreover, women’s challenges managing their health often intersected with other financial and material hardships. Our analysis below focuses on health concerns not related to pregnancy or reproductive health (see Table 4 for additional examples drawn from all seven categories.)

Responses about ongoing conditions (n=118) described formally diagnosed conditions requiring ongoing care or management by a healthcare provider. This category included severe, potentially life-threatening conditions such as cancer and cardiomyopathy, as well as conditions requiring long-term management, such as diabetes, hypertension, anemia, and chronic pain. For instance, a 36-year-old woman said, “My blood pressure is high and my heart beats a lot and there are other things, too. And because I don’t have money, I can’t pay or anything, you know, as an immigrant” (foreign-born Hispanic, uninsured shortly after childbirth). A woman suffering from migraines described material hardships that complicated the management of her symptoms, explaining that she was “taking Advil because of my migraines and I don’t have money to buy better medication… there’s cracks in the house and mold starting to come out and I told [the landlord], but he has done nothing about it. Roaches and other kinds of animals are starting to come in through the cracks. We can’t break the contract because if we do, we lose the [financial assistance]” (28 years old, US-born Hispanic, lost and regained public insurance). Some respondents reported multiple ongoing conditions, such as a 33-year-old woman who explained, “I have issues with anemia, still working on the carcinoma and doing treatments for that, and I have cardiomyopathy” (US-born Hispanic, uninsured after 3 months). A 37-year-old woman said, “I am diabetic and I have high blood pressure and my diabetes is starting to affect the retina to my eye” (US-born Hispanic, consistent public insurance).

Undiagnosed concerns (n=122) were symptoms that were not connected to a formal diagnosis. This heterogeneous category included symptoms such as dizziness, headaches, fatigue, or episodes of unexplained nausea and vomiting. One woman explained, “There are times when I throw up every day for a week and then it goes away and it comes back and I don’t know why” (26 years old, foreign-born Hispanic, uninsured shortly after childbirth). Sometimes, these concerns remained undiagnosed because symptoms were non-specific or tests results were inconclusive. This was true for a 21-year-old woman who said, “Every day since I gave birth iťs just really hard to get through the day. Constant headaches, going to and from the ER and they can’t find out what’s wrong with me. There hasnť been a day that I've felt good since I had [the baby]” (US-born Hispanic, uninsured after 3 months). More frequently, women indicated that they had not visited a healthcare provider about their symptoms. A woman whose public insurance was set to expire said, “lately I have been sick with migraines, fever, vomiting, and lately I've been feeling very weak. I made an appointment but they canť see me until October when I donť have insurance. Same with [the local hospital]” (20 years old, foreign-born Hispanic, uninsured after 3 months). Some women’s symptoms suggested that potentially serious medical conditions were going undetected. For instance, an uninsured woman described “pain in my chest and arm sometimes, and I don’t know if it’s cholesterol or something” (34 years old, foreign-born Hispanic, uninsured shortly after childbirth). Another woman commented, “my arms get numb—especially the left one gets completely numb, sometimes my left leg as well” (24 years old, US-born Hispanic, consistent public insurance).

Acute conditions (n=27) included short-term conditions with a formal diagnosis that may require medical attention, but not ongoing care or medical management. These included conditions such as broken bones, appendicitis, and pneumonia. The most common acute condition was gallbladder pain, including severe cases requiring gallbladder removal. For example, one woman explained that “apparently the hormones from getting pregnant caused me to get gallstones, and now I have to get my gallbladder taken out and I have to watch what I eat or I'm in severe pain. My Medicaid just ran out so I have to either get another insurance or find the money to get my gallbladder removed” (29 years old, non-Hispanic White, uninsured shortly after childbirth).

Mental health responses (n=22) included descriptions of mental health diagnoses or symptoms that women attributed to depression, anxiety, or panic attacks. One woman described feeling “like I can’t handle things and I feel depressed and I try to fight it off” (30, US-born Hispanic, uninsured shortly after childbirth). Another woman said, “Sometimes I stress out a lot and that gives me headaches, dizziness, and a lot of body aches. Sometimes I have panic attacks and I have chest pain and find it hard to breathe” (27 years old, foreign-born Hispanic, uninsured shortly after childbirth).

Finally, responses were classified as weight/lifestyle changes (n=166) when women described changes in their weight⁶ or health behaviors (e.g., smoking cessation, exercise). Most responses in this category related to weight gain during or after pregnancy, successful or unsuccessful attempts to lose weight during the postpartum period, and body image. A woman concerned about her weight explained, “I am trying to work on it. It’s just with three kids and I have no car for myself, I can’t go work out” (23 years old, US-born Hispanic, lost public insurance and later obtained private insurance). Another woman said, “I feel like I lost track of my…well everything, eating healthy, everything I used to do since I had my kid” (21 years old, US-born Hispanic, uninsured shortly after childbirth).

Discussion

Insurance churn was nearly universal within our sample of postpartum women with births covered by Medicaid/CHIP: long periods without coverage were typical and only 14% of women were continuously insured throughout the 12 months after childbirth. This finding corroborates and expands on prior studies showing high risk of insurance loss soon after a birth, particularly among non-White women in the South whose births were covered by public insurance (Daw et al., 2017, 2020). Since we captured 12 months of insurance and health data following childbirth, we were able to demonstrate that most women who lost their insurance early in the postpartum period did not regain coverage within the year. Over the same period, many women had health conditions requiring short-term or long-term medical care, and others described symptoms that could indicate serious underlying conditions.

Moreover, our results indicated that conditions that were identified during pregnancy or developed in the months following childbirth often went untreated. Among women who wished to have a health concern or illness checked by doctor during the study period, only about half were able to visit a provider. In open-ended responses, many women volunteered information about their obstacles to receiving care, such as the prohibitive cost of healthcare and difficulty obtaining appointments before their temporary pregnancy coverage ended, and described attempts to self-manage their symptoms when they did not receive medical care they needed.

Inaccessible or inadequate healthcare in the year after childbirth may have serious consequences. Although women’s health concerns varied in type and reported severity, they included conditions such as diabetes, hypertension, and cardiomyopathy, which rank among the top causes of maternal death in Texas (Maternal Mortality and Morbidity Task Force, 2018, 2020). Moreover, the higher prevalence of self-reported hypertension among non-Hispanic Black women in our sample was a troubling echo of racial/ethnic disparities in maternal outcomes (Louis et al., 2015). If an end result of insurance churn is that women with lower incomes living with chronic conditions only have access to treatment when pregnant, postpartum women may experience severe (and preventable) complications.

Implications for Policy and Practice

Medicaid expansion has reduced perinatal insurance churn in participating states (Daw et al., 2020) and would likely have similar benefits if implemented in the remaining 12 non-expansion states. Extending pregnancy Medicaid coverage beyond 60 days would benefit residents of non-expansion states, as well as residents of expansion states who do not qualify for standard Medicaid because their incomes exceed 138% FPL. Stabilizing coverage in the postpartum period would ensure that women can receive needed healthcare and would likely be cost-effective (Eckert, 2020). Proposals to extend pregnancy Medicaid have been endorsed by the American Medical Association, the American College of Obstetricians and Gynecologists, and Maternal Mortality Review Committees in Texas and other states (Eckert, 2020; Maternal Mortality and Morbidity Task Force, 2020).

The 2021 American Rescue Plan Act (ARPA) creates new fiscal incentives to expand Medicaid for states that have not yet done so, including Texas. The American Rescue Plan Act also permits states to extend pregnancy Medicaid to cover a full year after a birth by filing a state plan amendment, eliminating the need to obtain a Section 1115 waiver from CMS in order to receive federal matching funds (Daw et al., 2021; Musumeci, 2021; Ranji et al., 2021). This option will become available to states in 2022, amid significant bipartisan interest in extending pregnancy Medicaid: as of June 2021, 3 states have obtained waivers to extend postpartum coverage, 3 states have similar waivers pending, and 5 additional states have enacted legislation to extend Medicaid by applying for a waiver or filing a state plan amendment (Kaiser Family Foundation, 2021d). California has implemented an extension of pregnancy Medicaid using state funds, although this coverage is limited to women with a diagnosed mental health condition (Daw et al., 2021).

During the recently concluded 2021 session of the Texas State Legislature, bills to expand Medicaid did not receive a hearing (Harper, 2021). Although the legislature considered a proposal to extend pregnancy Medicaid to cover 12 months after childbirth, the final version approved by the legislature extended coverage to 6 months following childbirth. This bill was signed into law in June 2021. We expect that this extension of pregnancy Medicaid will reduce insurance churn early in the postpartum period and help women to obtain care for the wide range of health concerns described in our data. However, our results suggest that without an expansion of Medicaid or a longer extension of pregnancy Medicaid, there is likely to be substantial insurance churn at 6 months postpartum and unmet need for healthcare in subsequent months. Some of these needs may be addressed by the recent Healthy Texas Women Plus expansion, which covers a subset of conditions associated with maternal mortality and morbidity (e.g., hypertension, diabetes, substance use, and postpartum depression) for women with low incomes in the 12 months following childbirth. The impact of HTW Plus on postpartum health has not yet been assessed, given the recent implementation of the program in September 2020, and its impact will likely depend on the level of participation by patients and providers. Recent programmatic changes to HTW, including the discontinuation of auto-enrollment from pregnancy Medicaid and eligibility verification based on participation in other means-tested programs, could reduce participation among those who are eligible. There are also concern about the availability of providers to offer the specialist care newly covered by HTW Plus (Lerma et al., 2021).

Additional interventions are needed to improve healthcare access for women who cannot receive Medicaid or enroll in Healthy Texas Women Plus due to their immigration status. Given the exclusion of undocumented immigrants from Medicaid, ⁷ the American Rescue Plan Act does not incentivize the expansion of coverage to this population. However, states may address the health needs of immigrants and undocumented people by electing to cover their healthcare costs using state revenue (Green et al., 2016; Kelley & Tipirneni, 2018). For instance, 8 states use their own revenue to provide coverage for low-income undocumented children (Brooks et al., 2021). While the expenditures would not be eligible for federal matching funds, Texas could use state revenue to expand postpartum coverage for recent immigrants and undocumented women.

Health systems may improve maternal health outcomes through proactive communication with patients throughout the postpartum period. For instance, mobile applications could facilitate conversations between patients and providers about their health concerns and streamline the process of seeking care. Healthcare organizations can reduce barriers to care for immigrant women by ensuring the availability of interpretation services, conducting community outreach in patients’ own languages, and engaging organizations embedded in immigrant communities. Finally, healthcare providers already participate in state programs (such as the Texas Home Visiting program, which provides voluntary home visits to pregnant women and families with young children) that could be leveraged to ensure that postpartum women are connected to resources.

Limitations and Strengths

Although the Texas Postpartum Contraceptive Study collected longitudinal health information, it was not designed to assess women’s health risks and did not include a full medical history. Participants may have been experiencing health problems that we did not detect, including additional risk factors for maternal mortality and morbidity. Open-ended survey responses were collected from a subset of postpartum women who reported poor health or negative health changes; while these data provide rich description of health concerns and obstacles to receiving care, they do not indicate the prevalence of specific conditions at the population level. In particular, we caution against inferring low need for perinatal mental health services from the rarity of survey responses explicitly mentioning mental health. Mental health conditions such as postpartum depression are a leading cause of pregnancy-related death in Texas and disproportionately affect women of color (Maternal Mortality and Morbidity Task Force, 2020; Muzik, 2021; Rich-Edwards et al., 2006). We attribute the low number of responses about mental health in our study to the lack of a specific survey question about mental health, underdiagnosis of mental health conditions, and underreporting of mental health symptoms due to stigma. Finally, since our study examined the experiences of a majority-Hispanic sample recruited from eight urban hospitals in Texas, our conclusions are not necessarily generalizable to all postpartum women in Texas and in other states that have not expanded Medicaid. For instance, our sample included relatively few non-Hispanic Black women. Additional research is needed to capture the impact of postpartum insurance churn within other populations.

Nevertheless, our results are drawn from a large, multi-site sample of women with several risk factors for postpartum insurance churn, including low incomes, births covered by public insurance, and residence in a state that has not expanded Medicaid. Another strength of our analysis was the longitudinal measurement of insurance coverage and health conditions well beyond the early postpartum period, when most data on maternal health (e.g., PRAMS) are collected. Thus, we were able to describe postpartum women’s healthcare access, utilization, and needs over an entire year following a birth. Finally, recruitment immediately following a birth facilitated participation by women whose subsequent insurance loss reduced their access to healthcare providers—women who are likely underrepresented in clinic-based studies of postpartum health.

Conclusions

Women are at risk of insurance churn in the months following childbirth, particularly in states that have not expanded Medicaid. By examining 12 months of survey data collected from postpartum women in Texas whose births were covered by Medicaid/CHIP, we found that people experiencing insurance churn have healthcare needs that are not being adequately addressed. The limitations of existing public programs leave postpartum women in Texas with lower incomes (especially immigrant women) vulnerable to long-term health problems, poor preconception health, and higher-risk future pregnancies. A more inclusive and comprehensive safety net is required in order to improve maternal outcomes among women with low incomes and women of color.

Table 1:

Sample Characteristics (n=1,489 Postpartum Women with Births Covered by Medicaid/CHIP)

	n	%
Age
18–24 years	684	45.9
25–29 years	389	26.1
30 + years	416	27.9
Race/ethnicity and nativity
Hispanic, born in US	541	36.3
Hispanic, born outside US	657	44.1
Non-Hispanic Black	202	13.6
Non-Hispanic White	71	4.8
Non-Hispanic Other	18	1.2
Parity
1	381	25.6
2	474	31.8
3+	634	42.6
Had a postpartum visit
Yes	1172	82.9
No	241	17.1
Household participates in Supplemental Nutrition Program for Women, Infants, and Children (WIC)
Yes	1222	82.1
No	267	17.9
Household participates in Supplemental Nutritional Assistance Program (SNAP)
Yes	868	58.3
No	621	41.7
Household receives Temporary Assistance for Needy Families (TANF)
Yes	30	2.0
No	1459	98.0
Child(ren) in household covered by Children’s Health Insurance Program (CHIP)
Yes	1362	91.5
No	127	8.5

Table 2:

Health Status in the 12 Months after childbirth (n=1,489 Postpartum Women with Births Covered by Medicaid/CHIP)

	n	%
Self-rated health 3 months after childbirth
Very good/Excellent	581	41.3
Good	590	41.9
Fair/Poor	237	16.8
Reported a negative health change in 12 months after childbirtha	192	13.4
BMI 12 months after childbirthb
Underweight	23	1.7
Normal	334	25.0
Overweight	353	26.5
Obese	490	36.7
Missing BMI	134	10.0
Chronic conditions in 12 months after childbirthc
Hypertension	118	7.9
Diabetes	19	1.3

^an=1,434 women who completed either the 6-month or 12-month survey

^bn=1,334 women who completed 12-month survey

^cExcludes cases reported during pregnancy only

Table 3:

Typology of Open-Ended Health Responses and Sample Quotations (n=445 responses)

Ongoing Conditions (n=118):

• “My blood pressure has been off the roof lately. I’ve been trying to take my medicine and go to the gym for it.” (29, non-Hispanic Black, consistent public insurance)

• “I have problems with my stomach and I have irritable bowel syndrome, but I don’t have the medicine now since my Medicaid ran out.” (30, US-born Hispanic, uninsured shortly after childbirth)

• “I was diagnosed with thyroid cancer 3 months into my pregnancy and we’re starting treatment in a week.” (18, Middle Eastern, consistent public insurance)

• “...I just try to stay in a calm frame because I have epileptic seizures...” (34, non-Hispanic Black, consistent public insurance)

Undiagnosed Concerns (n=122):

• “When I sit down, I can’t really breathe well.” (19, foreign-born Hispanic, uninsured shortly after childbirth)

• “Weight loss, dizziness, sometimes I feel dizzy, weak, my hair is falling out, headaches.” (21, US-born Hispanic, uninsured after 3 months)

• “...lately I have been sick with migraines, fever, vomiting, and lately I’ve been feeling very weak. I made an appointment but they can’t see me until October when I don’t have insurance. Same with [the local hospital].” (20, foreign-born Hispanic, uninsured after 3 months)

• “Sometimes I have a lot of pain in my body, a bad headache, no desire to eat, sometimes lack of sleep, I sleepwalk during the day.” (25, foreign-born Hispanic, uninsured shortly after childbirth)

Reproductive Health, Non-Pregnancy (n=36):

• “Weight gain and bloating and spotting and headaches as side effects from the birth control.” (28, US-born Hispanic, private insurance shortly after childbirth)

• “I had a cyst that ruptured on my ovary about 3 months ago now...I haven’t really had any problems since then, but my menstrual cramps have worsened a lot.” (23, non-Hispanic Black, uninsured shortly after childbirth)

Acute Conditions (n=27):

• “I had gallbladder pain. They had to operate on me.” (25, foreign-born Hispanic, gained and lost private insurance)

• “I was hospitalized with an infection two weeks ago.” (39, foreign-born Hispanic, uninsured shortly after childbirth)

Mental Health (n=22):

• “I’ve had three months of postpartum depression and with the medication I’m taking now I am doing better. They gave me an injection to calm my nerves and I’m feeling better.” (28, US-born Hispanic, public insurance until 12 months, then uninsured)

• “I have anxiety, real bad anxiety.” (24, US-born Hispanic, consistent public insurance)

Pregnancy (n=20):

• “Besides the ectopic pregnancy. I lost a lot of blood and had to take some iron pills just enough so they didn’t have to give me a blood infusion. I was in the hospital for a week.” (28, non-Hispanic Black, uninsured shortly after childbirth)

• “The infections — mastitis. I’ve gone to the hospital. I even went to the WIC office and they try to help me through the process, told me to continue pumping regularly. I just [use] remedies at home, herbs, anything I can do to get better. This last time I had it, it was really bad. I passed out at work and everything. It’s gotten better. So now I go and take a 15-minute break every 3 hours and I pump because this last time my breast was really bad. It felt like I had a tumor; it was horrible. I tried to apply for ObamaCare, but they told me I can’t make enough money.” (30, US-born Hispanic, uninsured shortly after childbirth)

Weight and Lifestyle (n=166):

• “Losing a little bit of weight because I’m exercising more.” (24, US-born Hispanic, uninsured after 3 months)

• “I’m a little overweight and I probably don’t eat nearly as well as I should. I’m also probably not as active as I could be, plus I don’t drink enough water.” (37, US-born Hispanic, uninsured shortly after childbirth)

Author Biographies

Elizabeth Ela, MA PhD completed this work during a postdoctoral fellow with the Texas Policy Evaluation Project at the University of Texas at Austin. She studies contraceptive use, reproductive health, and maternal health.

Elsa Vizcarra, BS is a Research Associate at the Population Research Center at the University of Texas at Austin. Her research interests focus on access to and experiences with reproductive health services, including contraception and abortion.

Lauren Thaxton, MD MBA MSBS is an Obstetrician-Gynecologist with subspecialty in Complex Family Planning. She is also an Assistant Professor at the Dell Medical School Department of Women’s Health and studies novel, patient-centered interventions for reproductive health.

Kari White, PhD MPH is an Associate Professor at the Steve Hicks School of Social Work and Department of Sociology at the University of Texas at Austin. She studies the effect of policies on family planning service delivery and women’s access to reproductive health care.