Medicaid Maternity Care Alternative Payment Models Analysis

With the increase in maternal mortality rates in 2021 (32.9 deaths per 100,000 live births versus 23.8 in 2020), health care providers, hospital systems, and policymakers are eager to find ways to improve outcomes in maternal care. Alternative Payment Models (APMs) provide different methods of payment structure for providers while encouraging patient-centered care, improved maternal and infant outcomes, and effective spending.

 Three common APMs for maternity care are Pay-for-Performance (P4P), Pregnancy Medical Homes (PMH), and Episodes of Care (EOC). The Medicaid and CHIP Payment and Access Commission (MACPAC) contracted with RTI International to conduct in-depth case studies of approaches in five states (Arkansas, Connecticut, Colorado, North Carolina, and Tennessee), implementing APMs to improve maternity care in Medicaid.  

The RTI team reviewed available documents on each state’s APM models and conducted interviews with individuals in each state (including state Medicaid and MCH personnel, providers, managed care organizations) and with national level entities including the Centers for Medicare & Medicaid Services, beneficiary groups, and national experts. The final case studies provide an understanding of design, implementation, and impacts of these models within Medicaid programs. RTI developed case studies for:

•    Arkansas (Episodes of Care)
•    Colorado (Episodes of Care & Pay-for-Performance)
•    Connecticut (Pay-for-Performance)
•    North Carolina (Pregnancy Medical Home), and 
•    Tennessee (Episodes of Care).

Model design parameters (duration, market and provider participation, patient population, and exclusions) varied by state. Since none of the states’ APMs had been formally evaluated prior to this policy analysis, RTI’s analysis provided details on each states’ specific payment components, methodology, and model effects. 

Under maternity Pay-for-Performance (P4P) models, hospitals, health care practices, physicians, and certified nurse midwives can receive additional payments for meeting certain clinical outcomes (such as a patient having at least one postpartum visit occurring within 21 days of delivery) and quality metric thresholds (lower rates of caesarean deliveries) during their care for pregnant people. 

• The Connecticut P4P maternity care model (established in 2013) offers bonus payments to eligible health care providers (specific providers providing obstetric care) participating in Medicaid. Payments are given for providing the first prenatal visit within two weeks of a positive pregnancy test and providing blood pressure cuffs to pregnant people at risk for hypertension. These bonus payments may vary in amount each year based upon funding availability and are often made in the form of annual lump sum bonuses. Participation by providers is voluntary. Medicaid is the sole payer.
• Colorado’s P4P model began in 2018. In Colorado, payments are made to participating hospitals based on their performance on quality measures in four categories (maternal health, perinatal care, patient safety, and patient experience). Points are assigned for each quality measure and summed and divided into five tiers to allow for larger rewards in higher tiers. There are no financial penalties for low performance. Participation is voluntary, and Medicaid is the sole payer in this model.

Episodes of Care (EOC) models control costs by financially rewarding providers with low costs (upside risk) and penalizing providers who incur high costs for providing the same routine pregnancy care (downside risk). In most EOC models, providers must meet financial benchmarks and pre-defined goals for select quality measures to avoid financial penalties and receive incentives. The threshold for acceptable cost ranges and quality measures are set by the state Medicaid agency and/or its managed care organizations. EOC payments and penalties are often rendered annually at the practice level for overall performance. RTI reviewed EOC models in Arkansas, Colorado, and Tennessee. In all three models, payments were tied to costs of caring for patients and clinical processes (e.g., screening for Group B streptococcus).

• In Colorado’s EOC model (established in 2020), Medicaid is the sole payer. Pregnant beneficiaries (including those with high-risk pregnancies) seeing participating providers could be included. Under the model, the accountable provider is held financially accountable for the episode’s cost and quality. In the first year of participation, the provider receives credit for reporting on a set of quality measures and in subsequent years, they are subject to gainsharing and risk-sharing payments based on cost and quality.
• Arkansas’ EOC model (established in 2012, sunset in 2021) required mandatory participation from Medicaid providers and allowed voluntary commercial insurer participation. The model specified that beneficiaries must be non-high-risk who delivered a live birth. Arkansas’s payments were not tied to clinical outcomes. Arkansas’s EOC model found a decrease in total spending by 3.8% across all episodes included in the state demonstration and relative to neighboring states. 
• Tennessee’s EOC model began in 2014 and required mandatory participation from Medicaid providers and allowed voluntary commercial insurer participation in their APMs. The model specified that the beneficiaries be low-to-moderate risk pregnant beneficiaries. A 2012 study of Tennessee reported cost savings of $632 per perinatal episodes in 2018. 

Pregnancy Medical Homes (PMH) seek to provide high-quality, coordinated care specific to a patient’s health needs during the prenatal and early postpartum period via a per-person care coordination payment to providers. Through coordinated networks of specialty health and social service providers, PMHs aim to address patient’s high-risk health factors, such as hypertension and diabetes, alongside social factors, such as housing and food insecurity. PMH coordination payments are paid to providers on a per-person basis alongside other billable services. This model can also be layered with other APMs, as in North Carolina where PMH is paired with a decades-old maternity bundled payment arrangement. In bundled payment arrangements, one comprehensive payment covers all of the care that was delivered during an episode of care (e.g., a pregnancy).  

• North Carolina’s PMH was established in 2011 and is still operational. Medicaid was the sole payer for the PMH; commercial plans were not eligible to participate in this APM. Medicaid beneficiaries who are pregnant and postpartum (up to 60 days) are eligible to participate in the program. There were no exclusions for beneficiaries based on health conditions. This PMH used lump sum payments to incentivize specific administrative practices (completion of a risk assessment screening and a postpartum visit) and outcomes (vaginal deliveries) in their payment arrangement. Additional required reported measures (not tied to payment) included initiation of prenatal care in the first trimester, patients who received tobacco cessation counseling, and patients who received a comprehensive postpartum visit 14-60 days after giving birth.  

APMs allow for states to incorporate the enhancement of patient centered outcomes into their payment structure. It’s important to recognize that monitoring quality metrics is helpful in terms of process, but ultimately the “proof” of goals met should be improved maternal health outcomes.

RTI’s analysis of models in these five states provides the groundwork for future outcome-based evaluations to be conducted. Importantly, patient centered metrics must yield results and not merely become to-do lists for providers and health care systems. To wholly address the specific shortcomings in maternal health outcomes, these models must account for health and social needs, as well as social determinants of health.