Postpartum expansion of Emergency Medicaid is associated with increased receipt of recommended glycemic screening and care

Abstract

Background

Oregon expanded Emergency Medicaid coverage to 60 days of postpartum care in 2018, facilitating ongoing care for conditions such as gestational diabetes.

Methods

We linked Medicaid claims and birth certificates from 2010 to 2019 in Oregon and South Carolina, which did not expand postpartum care. We used a difference-in-difference design to measure the effects of postpartum care coverage among Emergency Medicaid recipients with gestational diabetes. Primary outcomes were receipt of recommended glucose tolerance testing and new diagnosis of Type 2 diabetes.

Results

Our sample included 2,270 live births among a predominantly multiparous, Latina population. Postpartum coverage was associated with a significant increase in receipt of a recommended glucose tolerance test (23.1 percentage points, 95% CI 16.9 to 29.3) and in diagnosis of Type 2 diabetes (4.6 percentage points, 95% CI 3.3 to 65.9).

Discussion

Expansion of postpartum coverage increased recommended screenings and care among Emergency Medicaid enrollees with pregnancies complicated by gestational diabetes.

Introduction

High quality obstetric care provides evidence-based screenings and treatment to optimize maternal health from the preconceptual period through postpartum.(1-5) Postpartum follow up is critical for those diagnosed with chronic conditions during pregnancy.(6) Gestational and pre-gestational diabetes are becoming increasingly common among reproductive age individuals and both are frequently diagnosed during pregnancy and the postpartum period.(7) Among individuals diagnosed with gestational diabetes, approximately 70% will develop Type 2 diabetes in their lifetime; postpartum follow-up and screening is needed to monitor and mitigate this lifelong risk.(8-13) Timely intervention for those with impaired glucose tolerance can also prevent development of Type 2 diabetes mellitus.(13)

National Medicaid policy restricts perinatal care coverage for a population at high risk for diabetes: undocumented immigrants.(7,14) In the United States (US), the obstetric immigrant population is majority Latina (73-91.4%).(15,16) Latinas are at increased risk for diabetes both during and outside of pregnancy. However, many are unable to access comprehensive Medicaid coverage.(17,18) Medicaid is the largest payer for obstetric care in the US and covers comprehensive prenatal and postpartum care.(19-21) Federal law requires that Medicaid recipients be citizens or permanent residents with greater than five years of residence. Full Medicaid benefits are not extended to people who do not meet the citizenship requirements, even if they would qualify on the basis of income.(22) Instead, these women may only receive obstetric delivery coverage through Emergency Medicaid, a federal safety net program covering acute, emergent events and obstetric admissions: no prenatal or postpartum care is provided.(22)

While federal Emergency Medicaid does not cover prenatal care, states can extend coverage using state funds or the Children’s Health Insurance Program Unborn Child clause.(23) Currently, 18 states, including Oregon, cover prenatal care for immigrants who do not qualify for full Medicaid benefits, while 32 states, including South Carolina, do not.(22,23) Previous research showed that Oregon’s expansion of prenatal care to Emergency Medicaid recipients resulted in a 31.6 % percentage point increase in receipt of anti-diabetic medications during pregnancy (95% CI 28.9% to 34.2%).(16) In 2018, Oregon became one of a handful of states that expanded postpartum care for 60 days for Emergency Medicaid recipients.(24) South Carolina has not expanded postpartum coverage for the immigrant population.

When a new diagnosis of diabetes is made during pregnancy, postpartum intervention is critical for long term health.(9-12) The risk of developing Type 2 diabetes for individuals with a history of gestational diabetes is significantly elevated. Type 2 diabetes is associated with significant morbidity, reduced quality of life, and increased health system costs.(8,25) For those with gestational diabetes, both the American College of Obstetricians and Gynecologists and American Diabetes Association recommend a 2-hour glucose tolerance test (GTT) between 4 and 12 weeks postpartum to evaluate for Type 2 diabetes.(18) A minority of people with a history of gestational diabetes receive screening, even among those with postpartum insurance coverage.(13) Marked racial disparities exist in the risk of developing diabetes and the receipt of a timely diagnosis.(7,9-12) Specifically, Latina women of reproductive age are at increased risk of diabetes, an estimated 48.1% of which is undiagnosed.(7)

The objective of this study was to determine the association of Oregon’s 2018 policy expanding access to postpartum care for immigrant women with improved receipt of recommended postpartum screenings and care for women with gestational diabetes. We leverage nine years of linked birth certificate and Medicaid claims data from Oregon and South Carolina to evaluate the association of postpartum coverage with receipt of a postpartum glucose tolerance test (GTT) and new diagnosis of Type 2 diabetes.

Materials and Methods

Data Source and Study Population

We conducted a retrospective cohort study using linked birth certificate data and Medicaid claims from Oregon and South Carolina. Our study period was from January 1, 2016 to December 31, 2019 which allowed for 16 months of data before the policy change and 20 months after the policy change. Our study sample included births occurring between January 1, 2016, to October 31, 2019 to allow for 60 days of postpartum follow-up for all births. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.(26) OHSU’s Institutional Review Board approved the study.

Our study population included individuals diagnosed with gestational diabetes, aged 12 to 44, with births between 23 and 44 weeks’ gestation. We identified patients with gestational diabetes from the birth certificate records. We restricted our study population to births among Emergency Medicaid recipients (Supplemental Figure 1). We used Emergency Medicaid as a measure of immigration status; the program’s enrollment is restricted to non-citizens who meet the financial eligibility requirements for full Medicaid benefits.(15,27)

Our primary outcomes were receipt postpartum of the recommended GTT to evaluate for Type 2 diabetes, and a new diagnosis of Type 2 diabetes. Secondary outcomes included attendance at a postpartum visit, receipt of postpartum contraception, and hospital readmission. We used International Classification of Disease 10^th edition (ICD-10) diagnosis and procedure codes, as well as pharmacy claims to identify our outcomes of interest (Supplemental Table 1). We examined all outcomes within the first 60 days postpartum to align with Oregon’s policy change.(24) Beginning April 1, 2018, Oregon provided full Medicaid benefits for 60 days postpartum to the Emergency Medicaid population that previously qualified during pregnancy.(24) Notably, this is distinct from recently publicized 12-month Medicaid postpartum coverage for citizen Medicaid recipients made possible through the American Rescue Plan of 2021.(28) Throughout this paper, we reference Oregon’s program for Emergency Medicaid recipients as postpartum expansion. South Carolina did not cover postpartum care for Emergency Medicaid recipients during the study period. Our independent variable was postpartum coverage for the Emergency Medicaid population.

We used a previously published algorithm to identify births and dates of delivery.(15) We calculated the difference in days between the date of delivery and the date of the first postpartum claim to determine the timing of our study outcomes. We used diagnosis codes to identify claims for postpartum GTT, new diagnosis of Type 2 diabetes, and postpartum visits (see Supplemental Table 1). We calculated the difference in days between the date of delivery and the date of the first claim to determine the timing of the postpartum test or visit. We used procedure and National Drug Codes (NDC) associated with receipt of contraception to identify claims for postpartum contraception (see Supplemental Table 1). We identified claims for emergency room admittance and inpatient hospital claims, and classified individuals with either type of claim within 60 days of delivery as having a hospital readmission.

We abstracted demographic and clinical information from the birth certificate files and claims data. We included the demographic variables of maternal age, multiparity, race/ethnicity, county of residence (metropolitan, non-metropolitan, missing), state, maternal tobacco use, and pre-pregnancy body mass index (BMI). We included self-reported information on race and ethnicity from the birth certificate files due to well described differences in rates of diabetes by racial and ethnic group.(29-31) We included the clinical variables of chronic hypertension, and gestational hypertensive disorders, preterm birth (births less than 37 weeks gestation), current mode of delivery, and prior cesarean delivery.

Statistical Analyses

To estimate the effect of Oregon’s expansion of postpartum care, we utilized a difference-in-difference design. This design allowed us to compare changes in our outcomes from the pre-policy period (January 1^st, 2016 to March 31^st 2018) to the post-policy period (April 1^st, 2018 to October 31^st 2019) in Oregon and South Carolina. This approach relies on the assumption that the time trends in outcomes were the same in the two states in the pre-policy period (i.e. parallel pre-trends).(32) We tested for parallel pre-trends and did not find any significant difference in our outcomes prior to the policy change (Supplemental Table 2).

We selected covariates for models based on clinical or reported associations with our outcomes.(18) In all models, we adjusted for maternal age, pre-pregnancy BMI, rural location, cesarean birth, and preterm birth. Standard errors were clustered at the county level.

Latinas are considered high-risk for both gestational diabetes and make up the majority of Emergency Medicaid recipients though South Carolina has a more ethnically diverse Emergency Medicaid population than Oregon.(17,18,27,33) We therefore conducted sub-analyses that restricted the study population to only Latinas enrolled in Emergency Medicaid to better understand how these policies affect the population at greatest risk.

For all of our difference-in-difference models, we conducted 2-sided tests with an alpha level of 0.05. We utilized R version 4.0.3 to conduct our analyses.

Results

Our study sample included 2,269 live births among 2,152 individuals (Table 1). More births in our sample occurred in the Oregon cohort than the South Carolina comparison group (69.5% vs 30.5%). The majority of all births were to individuals who were multiparous (86.4%) with no history of a cesarean delivery (76.9%). Most births were delivered vaginally (64.7%). Our sample had a mean pre-pregnancy BMI of 30.0 (SD = 6.0).

Table 1:

Demographics of births covered by Emergency Medicaid to individuals with gestational diabetes in Oregon and South Carolina, 2016 – 2019 (n= 2,272)

Characteristic	No. (%)a
	Oregon (N = 1,579, 69.5%)	South Carolina (N = 693, 30.5%)
Maternal age at birth, years (Mean (SD))	33.1 (5.5)	32.4 (5.4)
Multiparous	1,363 (86.3)	601 (86.7)
Race/Ethnicity
White	52 (3.3)	24 (3.7)
Black	22 (1.4)	13 (2.0)
Latina	1,343 (85.1)	327 (50.1)
Asian	98 (6.2)	27 (4.1)
American Indian/Alaska Native	0 (0.0)	1 (0.1)
Native Hawaiian/Pacific Islander	50 (3.2)	2 (0.3)
Other	10 (0.6)	257 (39.4)
Unknown	4 (0.3)	2 (0.3)
County of residence
Metropolitan	1,353 (85.7)	549 (79.2)
Non-Metropolitan	194 (12.3)	138 (19.9)
Missing	32 (2.0)	6 (0.9)
Maternal tobacco	16 (1.0)	2 (0.3)
BMI (Mean (SD))	30.0 (6.0)	29.9 (6.4)
Adequate prenatal care	1,440 (92.6)	604 (87.4)
Pre-pregnancy hypertension	43 (2.7)	25 (3.6)
Gestational hypertension	141 (8.9)	93 (13.4)
Preterm birth	187 (11.8)	104 (15.0)
Cesarean delivery	550 (34.8)	253 (36.5)
Prior cesarean delivery	368 (23.3)	158 (22.8)

^aIndividual variable denominators differ depending on missingness.

Individuals giving birth in Oregon were more likely to be older, (33.1 years (SD = 5.5) vs 32.4 years (SD = 5.4) in South Carolina), Latina (85.1% vs 50.1% in South Carolina), living in metropolitan counties (85.8% vs 79.2% in South Carolina), and to deliver at term (88.1% vs 85.0% in South Carolina).

Before the policy change, only 3.6% of Emergency Medicaid enrollees in Oregon received the recommended postpartum GTT within 60 days of delivery. After the policy, 26.2% of Emergency Medicaid enrollees in Oregon received the recommended postpartum GTT within 60 days of delivery. In our adjusted DID model, the policy was associated with an increase in GTT/blood sugar testing of 23 percentage points (95% CI 16.6 to 29.4, P = <.001) (Table 2, Figure 1).

Table 2:

Changes in primary and secondary outcomes among the Emergency Medicaid enrollees with gestational diabetes following the expansion of postpartum coverage, 2016-2019, 2,272

	Treatment		Comparison		Difference-in- Difference Estimate
	(Oregon)		(South Carolina)
	Pre- Policy	Post- Policy	Pre- Policy	Post- Policy	Adjusted Difference
	N = 969 No. (%)	N = 610 No. (%)	N = 432 No. (%)	N = 261 No. (%)	% (95% CI)
Primary Outcomes
GTT/Blood Sugar Testing	35 (3.6)	159 (26.1)	6 (1.4)	3 (1.1)	23.0 (16.0 - 29.9)
Type 2 Diabetes Mellitus Diagnosis	4 (0.4)	27 (4.4)	2 (0.5)	0 (0.0)	4.6 (3.3 - 6.0)
Secondary Outcomes
Postpartum Visit Attendance	83 (8.6)	365 (59.8)	76 (17.6)	55 (21.1)	47.3 (38.4- 56.2)
Postpartum Contraception	125 (12.9)	281 (46.1)	11 (2.5)	21 (8.0)	27.2 (18.9 - 35.7)
Hospital Readmission	7 (0.7)	19 (3.1)	4 (0.9)	0 (0.0)	3.2 (1.3 – 5.1)

Figure 1: Adjusted trend estimates of recommended postpartum Glucose Tolerance Test Screening for Type 2 Diabetes among Emergency Medicaid recipients with gestational diabetes in Oregon and South Carolina, 2016-2019, N = 2,270.

Completion of glucose tolerance testing among Emergency Medicaid recipients following implementation of postpartum coverage. Event time figure depicts implementation at time 0, and demonstrates an associated and sustained increase in testing following implementation.

We observed a similar increase in new diagnoses of Type 2 diabetes postpartum. Pre-policy, 0.4% of Emergency Medicaid recipients in Oregon with gestational diabetes were diagnosed with Type 2 diabetes within 60 days postpartum. After the policy, 4.4% of Emergency Medicaid enrollees in Oregon with gestational diabetes received a diagnosis. The policy was associated with an increase in overall diagnoses of 4.6 percentage points in our adjusted model (95% CI 3.2 to 6.0, P = <.001). (Table 2, Figure 2).

Figure 2: Adjusted trend estimates of new diagnosis of Type 2 Diabetes Mellitus postpartum among Emergency Medicaid recipients with gestational diabetes in Oregon and South Carolina, 2016-2019, N = 2,270.

Diagnosis of Type 2 Diabetes Mellitus among Emergency Medicaid recipients in the postpartum period following implementation of postpartum coverage. Event time figure depicts implementation at time 0, and demonstrates an associated and sustained increase in diagnosis following implementation.

Oregon’s policy was also associated with changes in postpartum visit attendance, receipt of contraception, and hospital readmission for individuals with either gestational diabetes. Prior to the policy change, 8.6% of individuals with Emergency Medicaid in Oregon attended a postpartum visit within 60 days of delivery. We observed an increase in postpartum visit attendance of 47.2 percentage points following the expansion of postpartum care (95% CI 38.5 to 55.8, P = <.001, Table 2). We observed a similar trend with postpartum contraceptive use: pre-policy, only 12.9% of Emergency Medicaid enrollees in Oregon received contraception. The policy change was associated with an increase in the receipt of contraception of 27.1 percentage points (95% CI 18.6 to 35.5, P = <0.001). Similarly, prior to the policy, only 0.7% of Emergency Medicaid recipients in Oregon were readmitted to the hospital postpartum for ongoing care. The policy change was associated with an increase in hospital readmission of 3.2 percentage points (95% CI 1.6 to 4.9, P = <0.001).

A similar pattern was found within our Latina sub-analysis (see Supplemental Table 3 for Latina demographics). Among the subpopulation of Latina individuals with gestational diabetes, only 3.6% received GTT/blood sugar testing prior to the policy. In our adjusted difference-in-difference model, the policy was associated with a 25.4 percentage point increase in receipt of a postpartum GTT (95% CI 18.7 to 32.2, Table 3). Before the policy change, 0.4% of Latina individuals with gestational diabetes received a postpartum diagnosis of Type 2 diabetes. Postpartum coverage was associated with a 5.4 percentage point increase in detection of new diagnoses of Type 2 diabetes (95% CI 3.1 to 7.7, Table 3).

Table 3:

Changes in primary and secondary outcomes among Latina Emergency Medicaid enrollees with gestational diabetes following the expansion of postpartum coverage, 2016-2019, N = 1,670

	Treatment		Comparison		Difference-in- Difference Estimate
	(Oregon)		(South Carolina)
	Pre- Policy	Post- Policy	Pre- Policy	Post- Policy	Adjusted Difference
	N = 824 No. (%)	N = 519 No. (%)	N = 223 No. (%)	N = 104 No. (%)	% (95% CI)
Primary Outcomes
GTT/Blood Sugar Testing	30 (3.6)	141 (27.2)	5 (2.2)	1 (1.0)	25.4 (18.7 - 32.2)
Type 2 Diabetes Mellitus Diagnosis	3 (0.4)	24 (4.6)	2 (0.9)	0 (0.0)	5.4 (3.1 - 7.7)
Secondary Outcomes
Postpartum Visit Attendance	70 (8.5)	318 (61.3)	40 (17.9)	20 (19.2)	52.0 (39.4 - 64.5)
Postpartum Contraception	98 (11.9)	242 (46.6)	9 (4.0)	4 (3.8)	35.6 (26.1 - 45.0)
Hospital Readmission	7 (0.8)	17 (3.3)	2 (0.9)	0 (0.0)	3.2 (0.8 - 5.5)

Discussion

Expansion of postpartum coverage to the Emergency Medicaid population was associated with rapid and significant increases in recommended postpartum care and detection of new diagnoses of Type 2 diabetes, specifically. Expanded Medicaid coverage to include 60 days of postpartum care was further associated with receipt of recommended services, including a postpartum visit, a glucose tolerance test, and the use of postpartum contraception.

While highlighting important positive health effects of a policy initiative expanding coverage for immigrants, these results also emphasize a missed opportunity in current national reforms aimed at extending postpartum coverage. States now have the option of expanding postpartum Medicaid eligibility for 12 months from the previous 60-day limit.(34) However, federal funds cannot be used to cover unauthorized immigrants under this provision and may miss out on some of the potential benefits of extended access to care.(34)

Timely diagnosis and management of Type 2 diabetes has significant implications for the lifelong health of people capable of pregnancy. Early intervention can delay or prevent severe complications of diabetes, including hypertension, renal disease, retinopathy.(25) Among the general population, postpartum follow-up for diabetes screening ranges has been reported as only 33.7% to 67% following up postpartum.(35-37) Prior to the policy change, only 3.6% of the Emergency Medicaid population received the recommended GTT postpartum; following the policy change screening rates improved to approximate that of the general Medicaid population.(35,36) While this represents a marked improvement, it also demonstrates a need to address other barriers to screening for this and the general Medicaid population. Ongoing going access to often costly medication and medical care will be improved by the 12-month postpartum expansion but not for those Emergency Medicaid recipients who remain in a coverage gap.

When postpartum care became a covered benefit in Oregon, we observed a 47.3 percentage points increase in attendance at a postpartum visit and a 23 percentage points increase in receipt of a postpartum glucose tolerance test among Emergency Medicaid recipients with a pregnancy complicated by diabetes. This policy change brought Emergency Medicaid enrollees to similar levels of receipt of postpartum care as other publicly-insured populations, suggesting that expanded insurance coverage is a key strategy to mitigate reproductive health disparities among the immigrant population.(13) Similarly, we identified a 27.2 percentage points increase in the use of postpartum contraception when postpartum care was a covered benefit. Contraception is a critical strategy to promote maternal and child health by ensuring that individuals living with diabetes have optimal interpregnancy intervals and excellent blood sugar control prior to conception.

For women planning subsequent pregnancies, postpartum contraception facilitates optimal interpregnancy intervals, reducing maternal risk for obesity and worsening glycemic control.(38) Achieving euglycemia prior to a subsequent pregnancy is essential to prevent fetal embryopathy, stillbirth, and maternal complications.(17) However, significant racial and ethnic disparities exist in the detection and management of diabetes.(7,39) In the general obstetric population, adherence to recommended postpartum glucose testing recommendations is low, and research to improve follow-up is ongoing. Multiple barriers such as transportation, insurance coverage, education, lack of childcare, and emotional stress may make it more difficult for women with diabetes to attend a postpartum visit.(40,41) These are compounded in settings offering the federal minimum standard of care for Emergency Medicaid—no access to funded postpartum care, contraception or postpartum glucose tolerance testing.

Our study should be interpreted with the following limitations in mind. Administrative claims data may be subject to errors in coding. We addressed this limitation by using two different data sources, Medicaid claims and birth certificate data, which allowed us to corroborate health outcomes and improve the demographic information available. We used data from two states, Oregon and South Carolina, which may limit our generalizability to other areas. State-specific contexts affect policy interpretation, implementation and care availability including, for example, ethnic background among Emergency Medicaid recipients, geographic distribution of providers accepting Medicaid, and political sentiments toward immigrants. These factors may influence results. We were not able to capture subsequent births to women who moved out of state or switched to a private payor or who may have received care through charity programs, safety net clinics, or federally qualified health centers. Our study is also limited by systematic under detection of pre-gestational diabetes among the Emergency Medicaid population and of gestational diabetes among South Carolinians lacking access to prenatal care; our estimates likely are lower than the true prevalence of gestational diabetes. We examined gestational diabetes and pre-gestational diabetes separately and together in an attempt to address this limitation.

Conclusions

We found that Oregon’s policy change to expand coverage to the Emergency Medicaid population for 60 days postpartum swiftly eradicated existing disparities in recommended postpartum care for women with pregnancies complicated by diabetes. Coverage of 60 days of postpartum care was associated with marked improvements in evidence-based care and diagnosis of new cases of Type 2 diabetes. To ensure ongoing optimal management of chronic diseases, including diabetes, health insurance coverage outside of pregnancy is needed.