Massachusetts Medicaid ACO Program May Have Improved Care Use And Quality For Pregnant And Postpartum Enrollees

Abstract

Value-based care models, such as Medicaid accountable care organizations (ACOs), have the potential to improve access to and quality of care for pregnant and postpartum Medicaid enrollees. We leveraged a natural experiment in Massachusetts to evaluate the effects of Medicaid ACOs on quality-of-care-sensitive measures and care use across the prenatal, delivery, and postpartum periods. Using all-payer claims data on Medicaid-covered live deliveries in Massachusetts, we used a difference-in-differences approach to compare measures before (the first quarter of 2016 through the fourth quarter of 2017) and after (the third quarter of 2018 through the fourth quarter of 2020) Medicaid ACO implementation among ACO and non-ACO patients. After three years of implementation, the Medicaid ACO was associated with statistically significant increases in the probability of a timely postpartum visit, postpartum depression screening, and number of all-cause office visits in the prenatal and postpartum periods, with no changes in severe maternal morbidity, preterm birth, postpartum glucose screening, or prenatal or postpartum emergency department visits. Changes in cesarean deliveries were inconclusive. Results suggest that implementing Medicaid ACOs in the thirty-eight states without them could improve maternal health care outpatient engagement, but alone it may be insufficient to improve maternal health outcomes.

EXHIBIT 2.. Association between Medicaid ACO implementation and quality of care-sensitive measures among pregnant and postpartum patients (2016–2020)

Notes: Left graph shows adjusted marginal effects before (q12016 – q12018) versus after (q22018 – q42020) Medicaid ACO implementation. Right graph shows difference-in-differences estimates. All results are adjusted for age, a vector of patient diagnoses, number of insurance enrollment days during pregnancy, multiple gestation (if applicable), parity number, patient zip code characteristics, and include fixed effects for delivery month, county, and delivery hospital. Difference-in-differences estimates are calculated as absolute probabilities, reported as percentage points. ACO is accountable care organization.

EXHIBIT 3.. Association between Medicaid ACO implementation and health care utilization among pregnant and postpartum patients: difference-in-differences results (2016–2020)

Notes: Results are adjusted for age, a vector of patient diagnoses, number of insurance enrollment days during pregnancy, multiple gestation (if applicable), parity number, patient zip code characteristics, and include fixed effects for delivery month, county, and delivery hospital. Estimates shown represent the difference-in-differences between Medicaid ACO vs Medicaid non-ACO before (q12016 – q12018) versus after (q22018 – q42020) Medicaid ACO implementation. Rate ratios (RR) and incident rate ratios (IRRs) are relative, where values >1.0 indicate a positive association between the Medicaid ACO and the outcome. ACO is accountable care organization. SMM is severe maternal morbidity. ED is emergency department.