Health Care Use by Commercially Insured Postpartum and Nonpostpartum Women in the United States

Abstract

Objective:

To compare the type, frequency, and timing of health care use among commercially insured postpartum and nonpostpartum women.

Methods:

This retrospective cohort study used data from a large national commercial claims database. Women between 18 and 44 years of age who gave birth in 2016 (n=149,563), and women who were neither pregnant nor postpartum between 2015 and 2017 (n=2,048,831) (nonpostpartum) were included. We examined preventive, problem, and emergency department visits, and hospitalization among postpartum women during the early postpartum (< 21 days after childbirth), postpartum (21 – 60 days postpartum), and extended postpartum period (61 – 365 days after childbirth). Visits among nonpostpartum women were assessed during time periods of equivalent duration.

Results:

Almost 24% of postpartum women had a problem visit in the early postpartum period compared to 19.7% of nonpostpartum women (adjusted difference 4.8 percentage points [95% CI 4.6–5.0]). Approximately 3% of postpartum women had an early ED visit, more than double the percent among nonpostpartum women (adjusted difference 2.3 percentage points (95% CI, 2.2–2.4). Both problem and ED visits among postpartum women remained elevated relative to nonpostpartum women during the postpartum and extended postpartum periods. While postpartum women were more likely than nonpostpartum women to receive preventive care during the early and postpartum periods, only 43% of postpartum women had a preventive visit during the extended postpartum period, a rate 1.8 (95% CI, −2.1 to −1.5) percentage points lower than that of nonpostpartum women. Adjusted hospitalization rates among postpartum women in the early (0.8%), postpartum (0.3%), and extended postpartum (1.4%) periods were higher than those of nonpostpartum women (0.1%, 0.2%. 1.6% respectively).

Conclusions:

Commercially insured postpartum women use more health care than nonpostpartum women, including inpatient care. Differences are largest in the early postpartum period and persist beyond 60 days postpartum.

Precis

Commercially insured postpartum women use more health care than nonpostpartum women; differences are largest in the early postpartum period, and differences persist beyond 60 days postpartum.

Introduction

Women face unique pregnancy-related health challenges during the postpartum period including pain, heavy bleeding, breastfeeding issues, urinary incontinence, depression and hypertensive disorders of pregnancy.^1–3 In the United States, postpartum maternal morbidity has risen dramatically over time and an estimated one third of pregnancy-related deaths occur between one week and one year after childbirth.^4,5 Until recently, clinical guidelines only recommended a single postpartum follow-up visit six weeks after childbirth.⁶ In May 2018, the American College of Obstetricians and Gynecologists (ACOG) recommended that all women receive an early postpartum contact within three weeks and a comprehensive postpartum visit within three months postpartum. Further, guidelines recommended that postpartum women receive ongoing postpartum care, tailored to their individual needs, throughout the first three months after childbirth.⁶

Little evidence exists to guide decisions about how to reorganize follow-up care so that clinical practice conforms to new ACOG recommendations. Most of the literature on health care utilization postpartum has focused on readmission,^7,8 and on examining attendance rates for the routine 6-week preventive visit.^9–12 Beyond documenting rates of routine postpartum care, there is a paucity of evidence on outpatient health care seeking in the first weeks after childbirth. Further, existing studies that document outpatient care among postpartum women do not include a comparison group of women who are not in the postpartum period. Without a comparison to benchmark care relative to nonpostpartum women, it is not possible to determine whether rates of outpatient care reflect elevated health needs in the postpartum population or are instead consistent with general care use among reproductive-aged women.

Better evidence documenting postpartum women’s health care use, how it differs from that typically used by nonpregnant or postpartum women, and for how long this difference lasts, may help encourage new models of care organization and payment. In this study, we used a large national sample of commercial insurance claims to compare the frequency, timing and reason for health care use among postpartum women in the year after childbirth with health care received by a comparison group of reproductive-aged women who were neither pregnant nor postpartum.

Methods

We conducted a retrospective cohort study in which we used data from the Health Care Cost Institute (HCCI) national database that include inpatient, outpatient, physician and enrollment claims from over 50 million commercially insured individuals per year drawn from four major US payers.¹³ We constructed a sample of postpartum women by identifying all inpatient claims for childbirth in 2016 among women age 18 – 44. Claims for childbirth were identified using the following diagnosis-related group (DRG) codes: 765–768, 774, 775. To identify pre-pregnancy chronic conditions and track utilization for one year postpartum, we restricted our primary sample to births to women who were enrolled by the same HCCI-contributing payer for at least 21 months before (12 months pre-pregnancy plus nine months of pregnancy) and 12 months after birth. Our sample includes approximately 8% of all births to commercially insured women in the United States in 2016.¹⁴ Among women with more than one birth in 2016, only the first birth was retained in the sample.

To compare health care use among postpartum women with the care use that would be expected without pregnancy and childbirth, we created a comparison group composed of reproductive aged women (18–44 years) who were neither pregnant nor postpartum between 2015 and 2017. This comparison sample – the nonpostpartum group – was restricted to women enrolled by the same HCCI-contributing payer from 2015–2017, and who had no claim for childbirth, prenatal care or pregnancy care during these years. Differences in the mean characteristics of women in the two groups were testing using t-tests.

As the data were deidentified, this study was considered to be nonhuman subjects research by Brown University’s and by Columbia University Irving Medical Center’s Institutional Review Boards and was conducted between January 1, 2020 – February 1, 2021.

The primary outcomes were preventive visits, problem visits, emergency department (ED) visits and hospitalizations. To measure preventive and problem visits, we identified all outpatient and physician claims (excluding ED) for postpartum and nonpostpartum women. We considered a “visit” all outpatient and physician claims associated with an individual woman on a given date. Definitions for preventive and problem visits are described in the Appendixes 1–3.

The outcomes were measured during three time periods over a full year of follow-up. For postpartum women, we measured visits in the early postpartum (< 21 days childbirth), postpartum (21 – 60 days postpartum), and extended postpartum period (61 – 365 days after childbirth). We chose these time periods for consistently with ACOG’s revised visit schedule and with the HEDIS quality metric for postpartum care that was used during the time period of this study.¹⁵ For the comparison group, we created a random start date of follow-up by generating a random variable between 0 and 365 and adding this integer to January 1, 2016. These start dates correspond to the range of possible follow-up start dates among postpartum women who gave birth in 2016. As in the postpartum group, the early period included the first 20 days of follow-up, the postpartum period included days 21 – 60 of follow-up, and the extended period included days 61 – 365 of follow-up.

We created indicator variables for pregnancy complications if postpartum women had any inpatient, outpatient or physician claims for these conditions, as defined using ICD codes from previous literature,⁷ in the nine months before childbirth. Indicator variables for chronic disease were created if a woman had an inpatient, outpatient or physician claim corresponding to HCCI ICD code definitions for asthma, hypertension, and diabetes, and any ICD code for nonpsychotic mood and anxiety disorders in the year before pregnancy (assuming a nine-month gestational period) for postpartum women and in 2015 for the non-postpartum comparison group (ICD codes listed in Appendix 4).¹³

Based on these conditions, we defined three groups of women for subgroup analyses: 1) women with pregnancy complications (placenta previa, preterm labor, gestational diabetes, gestational hypertension, preeclampsia, multiple gestation, hemorrhage, infection, laceration, operative injury, and uterine rupture, 2) women with chronic disease (asthma, hypertension, and diabetes, and mood and anxiety disorders), and 3) women with cesarean birth.

We calculated the percent of women in both groups with any visits in each category (preventive, problem, ED, and hospitalization) during the early postpartum, postpartum, and extended postpartum periods. For each outcome and time period, we measured the difference in utilization using unadjusted and adjusted linear probability models with each outcome as the dependent variable and an indicator for postpartum women as the independent variable. Adjusted models included controls for age group (18–24, 25–34, 35–44), chronic conditions, zip-code level income, and region (Northwest, Midwest, South, West).

The coefficient on the indicator variable for postpartum women represents the percentage point difference in the outcome between postpartum women and nonpostpartum women. For example, for early postpartum problem visits, the coefficient measures the difference in the percent of postpartum women with an early postpartum problem visit and the percent of nonpostpartum women with a problem visit during a time period of equivalent duration. We chose to use linear probability models, and corresponding percentage point differences, instead of logistic regression because the magnitude of the coefficient can be directly interpreted as the difference in the rate of the outcome between groups.

To examine how the difference in care use between postpartum and nonpostpartum women changed over the follow-up period, we calculated and plotted the percent of women in each group with a preventive or problem visit by week of follow-up and the weekly mean difference (and 95% confidence interval) between the two groups.

To examine the timing and reasons for care in the early postpartum period, we plotted the day of first outpatient problem visit (including visits to the ED) in the first 20 days after birth among all postpartum women and by subgroup. We then classified the primary diagnosis for all problem visits in this period using ICD10 disease categories. When possible, we grouped disease categories into related health issues (e.g. N61 inflammatory issues of the breast, O91 infections of breast, and O92 other disorders of breast and disorders of lactation were all classified as “breast issues”).

Results

We identified 149,563 hospitalizations for childbirth in 2016 among women aged 18–44 with continuous enrollment in the year before and after pregnancy (the postpartum group). We also identified 2,048,831 women who were neither pregnant nor postpartum and who were enrolled continuously during these years (the nonpostpartum comparison group). State was missing from 474 (0.32%) and 58,388 (2.70%) postpartum and nonpostpartum women respectively, and zipcode was missing from 2,906 (1.94%) and 100,002 (4.63) postpartum and nonpostpartum women respectively.

While the majority of postpartum women (62%) were between 25 and 34 years of age, non-postpartum women were more evenly distributed between age groups (Table 1). Chronic disease prevalence among postpartum women ranged from 1.4% for diabetes to 10.0% for mood and anxiety disorders. All chronic disease conditions were more prevalent among nonpostpartum women. Average zip code level median-income was similar between groups ($66,959 among postpartum women and $65,052 among nonpostpartum women). Both groups had a similar regional representation with approximately 17% from the Northeast and West, approximately 42% from the South, and approximately 21% from the Midwest. Roughly 36% of postpartum women had a cesarean birth, slightly higher than the national average of 33% among commercially insured women.¹⁴

Table 1:

Characteristics of postpartum and nonpostpartum women of reproductive age

Variable	Postpartum Women N=149,563	Nonpostpartum Women N=2,048,831	P-value difference between groups
Age	n (%)	n (%)
18–24 years	17,116 (11.4)	552,711 (27.0)	<0.001
25–34 years	91,984 (61.5)	606,030 (29.6)	<0.001
35–44 years	40,463 (27.1)	890,090 (43.4)	<0.001
Presence of chronic conditions	n (%)	n (%)
Asthma	5,808 (3.9)	92,979 (4.5)	<0.001
Hypertension	4,183 (2.8)	107,444 (5.2)	<0.001
Diabetes	2,123 (1.4)	46,194 (2.3)	<0.001
Mood and anxiety disorders	14,913 (10.0)	276,346 (13.5)	<0.001
Zip code median income quartiles	Percentile	Percentile
25th percentile	$51,331	$49,844
50th percentile	$66,959	$65,052
75th percentile	$87,379	$85,154
Region	n (%)	n (%)
Northeast	25,942 (17.3)	361,057 (17.6)
Midwest	34,051 (22.8)	427,649 (20.9)	0.007
South	62,717 (41.9)	862,945 (42.1)	<0.001
West	26,379 (17.6)	339,601 (16.6)	0.161
Birth outcome	n (%)
Preterm labor	19,215 (12.8)	--
Cesarean birth	53,099 (35.5)	--

Note: Age refers to age during the month of delivery for postpartum women and during January 2016 for nonpostpartum comparison women. Hypertension did not include codes for pregnancy induced hypertension, and diabetes did not include ICD codes for gestational diabetes. Northeast included CT, ME, MA, NH, RI, VT, NJ, NY and PA. Midwest included IL, IN, MI, OH, WI, IA, KS, MN, MO, NE, ND, and SD. South included DE, DC, FL, GA, MD, NC, SC, VA, WV, AL, KY, MS, TN, AR, LA, OK, and TX. West included AZ, CO, ID, MT, NV, NM, UT, WY, AK, CA, HI, OR, and WA.

Table 2 compares health care use during one year of follow-up between postpartum and non-postpartum women. In the early postpartum period (<21 days after birth among postpartum women and during days 61 through 365 of follow-up among non-postpartum women), nearly a quarter of postpartum women (24%) had at least one outpatient problem visit and approximately three percent sought care in an ED (Table 2). While problem visits were common throughout the early postpartum period, the percent of postpartum women with a problem or ED visit was highest in the first (12.4%) and second (10.4%) week after childbirth (Figure 1). The most common reasons for early postpartum problem visits were respiratory issues (e.g. shortness of breath and chest pain), genitourinary issues (e.g. urinary tract infections), hypertension and breast or breastfeeding issues (Table 3).

Table 2:

Health care use during one year of follow-up among postpartum and nonpostpartum women of reproductive age

	All postpartum women N=149,563	Nonpostpartum women N=2,048,831	Difference (unadjusted) N=2,198,394	Difference (adjusted) N=2,198,394
	Percent	Percent	Coefficient (95% CI)	Coefficient (95% CI)
Early postpartum (<21 days)
Preventive visits	15.0	3.3	11.7*** (11.6–11.8)	11.6*** (11.5–11.7)
Problem visits	23.7	19.7	4.0*** (3.8–4.3)	4.8*** (4.6–5.0)
ED visits	3.2	1.0	2.2*** (2.1–2.3)	2.3*** (2.2–2.4)
Hospitalization	0.8	0.1	0.7*** (0.6–0.7)	0.7*** (0.7–0.7)
Postpartum (21–60 days)
Preventive visits	28.2	6.5	21.6*** (21.5–21.8)	21.4*** (21.3–21.5)
Problem visits	39.4	30.2	9.2*** (8.9–9.4)	10.2*** (9.9–10.4)
ED visits	2.0	1.9	0.1** (0.0–0.2)	0.3*** (0.2–0.4)
Hospitalization	0.3	0.2	0.0 (−0.0–0.0)	0.1*** (0.0–0.1)
Extended postpartum (day 61–365)
Preventive visits	42.5	42.7	−0.3* (−0.5,−0.0)	−1.8*** (−2.1,−1.5)
Problem visits	79.5	72.8	6.6*** (6.4–6.9)	6.5*** (6.2–6.7)
ED visits	11.2	11.1	0.2 (−0.0–0.3)	1.1*** (0.9–1.2)
Hospitalization	1.4	1.6	−0.2*** (−0.2,−0.1)	0.1** (0.0–0.2)

NOTE: Preventive visits were visits with any diagnosis code for a preventive office visit (99381–99397), encounter for general adult medical examination (Z00.00, Z00.01), well woman exam (Z01.411, Z01.419), or postpartum care (Z39.1, Z39.2). Problem visits were visits without any preventive codes. Adjusted regression models included age group (18–24 years, 25–34 years, and 34–44 years), chronic disease (diabetes, asthma, hypertension, mood or anxiety disorder, region (Northwest, West, South, Midwest), income and month when follow-up began.

^*p<0.05,

^**p<0.01

^***p<0.001

Figure 1:

Days to first outpatient visit (including emergency department) in the first 20 days after childbirth among postpartum women, Health Care Cost Institute births 2016.

Table 3:

Most common primary diagnosis group in postpartum outpatient problem and ED visits in the first 21 days after childbirth

Condition	Percent of total outpatient visits in the early postpartum period
Respiratory (shortness of breath, chest pain)	11.7
Genitourinary (dysuria, urinary tract infection)	11.4
Hypertension (hypertension, preeclampsia, eclampsia)	10.8
Breast or breastfeeding issue (disorders of lactation, breast inflammation)	9.5
Pain (pelvic, abdominal, back)	8.4
Mental health (anxiety disorder, postpartum depression)	6.6
Musculoskeletal (dysfunction of lumbar or thoracic region, soft tissue disorders)	5.9
Skin condition (rash, cellulitis, dermatitis)	5.2
Obstetric complications	5.1
Obstetric wound (cesarean birth wound, perineal obstetric wound)	4.2
Puerperal infection or sepsis	3.9
Cardiology (pulmonary embolism, peripartum cardiomyopathy)	3.8
Bleeding or hemorrhage	3.6
Pelvic organ disorders (inflammatory or noninflammatory)	3.3
Gastrointestinal	3.3
Percent of visits with none of the above listed conditions	3.4

Note: Percentages refer to the share of total outpatient visits in the early postpartum period associated with each of the listed conditions. The denominator used to calculate the percentages was 57,805 outpatient visits among postpartum women in the first 20 days after childbirth. For some visits, more than one primary diagnosis code was listed.

Compared to nonpostpartum women, health care visits of all types (preventive, problem, ED, and hospitalization) were higher among postpartum women during the early postpartum period. Fifteen and 23.7% of postpartum women had a preventive and problem visit, respectively, during this period, compared to 3.3% and 19.7% of nonpostpartum women during the same length of time (preventive visit adjusted difference 11.6 percentage points [95% CI, 11.5–11.7]; problem visit adjusted difference 4.8 percentage points [95% CI: 4.6–5.0]) (Table 2). In the early postpartum period, 3.2% of postpartum women had an ED visit compared to 1.0% of non-postpartum women (adjusted difference 2.3 percentage points [95% CI, 2.2–2.4]), and 0.8% of postpartum women were hospitalized compared to 0.1% of nonpostpartum women (adjusted difference 0.7 percentage points [95% CI, 0.7–0.7]) (Table 2).

In the postpartum period (during days 21–60 of follow-up), 28.2% and 39.4% of postpartum women had at least one preventive and one problem visit, respectively. As in the early period, both preventive and problem visits in the postpartum period were more common among postpartum women than nonpostpartum women (adjusted difference 21.4 percentage points [95% CI, 21.3–21.5] for preventive visits; adjusted difference 10.2 percentage points [95% CI, 9.9–10.4] for problem visits). In the postpartum period, the rate of ED visits and hospitalizations were 2% and 0.3% respectively for postpartum women and 1.9% and 0.2% respectively for nonpostpartum women. After adjusting for maternal characteristics, the percent of postpartum women with an ED visit in the postpartum period was 0.3 percentage points (95% CI, 0.2–0.4) greater and the percent of postpartum women with a hospitalization was n0.1 percentage points [95% CI, 0.0–0.1]). percentage points greater than among non-postpartum women.

While most postpartum women (80%) received problem-related health care in the extended postpartum period (between 61 days to 1 year of follow-up), only 43% received preventive care, including well-woman care (Table 2). The percent of postpartum women with a problem visit was higher than nonpostpartum women in the extended postpartum period (difference 6.5 percentage points [95% CI, 6.2–6.7]), while the percent with a preventive visit was lower (difference −1.8 percentage points [95% CI, −2.1, −1.5]). The percent of postpartum women with an ED visit in the extended postpartum period was 11.2%. In adjusted regression analysis, ED visits were more common among postpartum women in the extended postpartum period (difference 1.1 percentage points [95% CI, 0.9–1.2]) than in the nonpostpartum group. Though there were fewer hospitalizations among postpartum women compared to nonpostpartum women in unadjusted models, after adjusting for maternal characteristics the percent of postpartum women with an inpatient stay was 0.1 percentage points greater (95% CI, 0.0–0.2).

Examining the timing of preventive visits over the course of the follow-up year, the percent of postpartum women with a preventive visit was higher than the comparison group during weeks one through ten. After ten weeks, preventive visit rates were similar in both groups with approximately two percent of women seeking preventive care each week (Figure 2). For problem visits, the percent of postpartum women seeking care was elevated relative to the nonpostpartum group in most weeks during the early postpartum and postpartum periods. The difference in problem visit rates between the two groups remained statistically significant through 16 weeks postpartum (Figure 3).

Figure 2:

Weekly percentage of postpartum (A) and nonpostpartum (B) women (comparison group) with a preventive visit. Area to the left of the shaded area indicates early postpartum, shaded area indicates postpartum, area to the right of the shaded area indicates extended postpartum.

Figure 3:

Weekly percentage of postpartum (A) and nonpostpartum (B) women (comparison group) with a problem visit. Area to the left of the shaded area indicates early postpartum, shaded area indicates postpartum, area to the right of the shaded area indicates extended postpartum.

Compared with postpartum women overall, postpartum women with chronic disease, pregnancy complications and cesarean birth were more likely to receive health care of all types in the early postpartum period. Of the three sub-groups, health care use was highest among postpartum women with chronic disease among whom 35% had at least one outpatient problem visit, 5% had an ED visit, and 1.3% were hospitalized in the first 20 days after childbirth (Appendix 1). While health care use in the postpartum and extended postpartum periods was similar among postpartum women with pregnancy complications and cesarean birth, compared to postpartum women overall, postpartum women with chronic disease continued to be more likely to receive problem-related health care, ED visits, and be hospitalized than postpartum women overall through the full year after childbirth (Appendix 1). Comparisons of the prevalence of early, postpartum, and extended postpartum visits among postpartum and nonpostpartum women with chronic disease are presented in Appendix 2.

Discussion

Among a national sample of commercially insured women, we found that postpartum women were more likely to use health care of all types (preventive, problem-related, ED and hospitalization) during the first three weeks after childbirth compared to typical health care use among nonpostpartum women. These findings are consistent with previous studies that have documented elevated postpartum ED and hospitalization rates, and we contribute new evidence indicating that postpartum health care utilization rates are significantly higher than rates of health care utilization in the general population of reproductive age women when they are not pregnant or postpartum. Most notably, postpartum women were over three times more likely to have an ED visit and eight times more likely to be hospitalized than nonpostpartum women in the early postpartum period. Health care use of all types remained higher among postpartum women compared to nonpostpartum women during the postpartum period (21 to 60 days), though the magnitude of the difference was substantially reduced for ED visits and hospitalization compared to the early postpartum period. Further, we found that higher levels of problem visits extended beyond sixty days, remaining statistically significant through 16 weeks postpartum. Finally, we found that fewer than half (43%) of postpartum women received any preventive care between 61 and 365 days after childbirth, a lower rate than among nonpostpartum women.

The large share of women seeking care for health problems in the first three weeks after childbirth—including high rates of ED use and hospitalizations—demonstrates substantial health care needs during this time period. This evidence supports ACOG’s 2018 recommendation that all postpartum women receive care in-person or remotely within the first three weeks postpartum. Our findings demonstrate that most early problem visits occurred in the first two weeks postpartum, suggesting that contact as early as one or two weeks postpartum may benefit many women.

Obstetric care delivery models will have to adapt to meet these early postpartum needs. Telehealth and home visits are promising options to promote early and consistent health care contact and reduce known barriers to postpartum care seeking such as fatigue, lack of transportation, and childcare.^16,17 Health care professionals may also be able to reduce the escalation of common early postpartum problems identified in this study (e.g. respiratory problems, pain, urinary tract infections) with anticipatory postpartum counseling and care before hospital discharge such as ensuring that women have inhalers at home, developing a pain management plan, and screening for signs of urinary tract infection.

Additionally, our finding that postpartum women receive more care than nonpostpartum women until 16 weeks postpartum suggests that an expanded definition of the postpartum period for follow-up, for example the 12 weeks currently suggested in the ACOG 2018 guidelines, is warranted. It also suggests that policies that use a postpartum period ending at sixty days to define insurance eligibility (e.g. pregnancy-related Medicaid) or to structure reimbursement reforms (e.g. pay-for-performance schemes) are arbitrary in relation to the timing of women’s actual health needs and may result in missed opportunities to prevent maternal morbidity.

While earlier contact with postpartum women may result in earlier treatment and prevention of more severe outcomes, lack of payment for individual postpartum visits during this period may limit health care professional’s incentives to increase care. Prenatal, delivery and postpartum services are often reimbursed using a single global payment, which is currently only meant to cover a single postpartum visit.^18,19 Without separate reimbursement for additional postpartum visits before the single routine visit or an increase in the overall reimbursement rate coupled with requirements for earlier and more frequent postpartum care, health care professionals may not implement new care models that recommend enhanced postpartum follow-up. Policy makers would benefit from research examining whether alternative payment models increase early postpartum contacts.

This study has several limitations. First, claims data documents health care receipt not health care need, which would better inform clinical guidelines and reimbursement strategies for postpartum women. However, this issue would affect both postpartum and nonpostpartum groups and we do not expect it to bias our estimates of differences between groups. Second, the timing and occurrence of the routine postpartum visit may not be fully observable in claims data since the routine visit can be billed inside the global maternity payment. This could result in an underestimate of the percent of postpartum women experiencing health issues in the early and postpartum periods, which would bias the observed differences between the postpartum and nonpostpartum groups towards the null. Third, variables were not available to examine differences in health care use by maternal race-ethnicity. This is an important limitation due to persistent large racial disparities in postpartum morbidity and mortality.⁸ Finally, our data included only commercial claims from women with extended enrollment and our results may not generalize to women covered by Medicaid during and after pregnancy or women with less consistent enrollment in commercial plans. Further research on racial-ethnic and insurance-related disparities in postpartum care use could inform policy debates and health care reforms to improve access to postpartum care.

Current patterns of health care use among this large sample of commercially insured postpartum women suggest that existing models of postpartum care may not be responsive to the timing, frequency, and type of health care needed by women after birth. Health care payment and delivery reforms should focus on ensuring that postpartum women can access timely care for early postpartum problems and maintain access beyond 60 days postpartum to address ongoing health issues.