Evaluation of the Maryland Total Cost of Care Model: Quantitative Only Report for the Model’s First Three Years (2019 to 2021)

Evaluation of the Maryland Total Cost of Care Model: Quantitative Only Report for the Model’s First Three Years (2019 to 2021)

Published: Dec 20, 2022
Publisher: Mathematica
Associated Project

Evaluating Accountability for Statewide Health Cost and Quality Outcomes: The Maryland Total Cost of Care Model

Time frame: 2019-2025

Prepared for:

U.S. Department of Health and Human Services, Center for Medicare & Medicaid Innovation

Key Findings

We find that the Maryland Model had the following favorable effects during the first three years of the MD TCOC period (2019 to 2021):

  • It substantially reduced rates of all-cause acute care hospital admissions (by 16.1 percent).
  • It moderately reduced total Medicare fee-for-service (FFS) spending (Part A and Part B) by 2.5 percent. The Maryland Model increased non-hospital spending (by 2.7 percent) but reduced hospital spending by more (6.6 percent), leading to a $781 million reduction in total spending.
  • It improved several quality-of-care measures, including reducing potentially preventable admissions (by 16.1 percent), reducing the likelihood of an unplanned readmission to the hospital (9.5 percent), and increasing timely follow-up after hospital discharge (2.5 percent).
  • For most outcomes, the impacts were larger and more favorable during the MD TCOC period than they were at the end of the previous model’s period (2017–2018), indicating further improvement. The larger, more favorable effects during the MD TCOC period might be due to (1) the growing influence of global budgets that began in 2014; (2) the broader accountability, incentives, and supports that Maryland and CMS introduced in 2019 and that continue to evolve; and (3) the synergies between the two.

Research Objective

The Maryland Total Cost of Care (MDTCOC) Model is one of the first payment and delivery reforms that holds a state accountable for reducing the total cost of medical care while improving quality of care. MDTCOC is a joint state and Centers for Medicare & Medicaid Services (CMS) initiative that builds on its predecessor, the Maryland All-Payer Model (MDAPM, 2014-2018), which created all-payer global budgets for Maryland hospitals. MDTCOC continues hospital global budgets and extends transformation beyond hospital walls by expanding statewide accountability for cost and quality outcomes and broadening the incentives and supports to providers, including new investments in primary care. This study estimates the impact of MDTCOC on spending, utilization, quality, and population health in the Model’s first three years (2019-2021).

Study Design

We used a regression adjusted difference-in-differences strategy, matching Maryland geographic regions to a comparison group drawn from the rest of the country. Because global budgets, started in 2014, form the foundation of the TCOC model, we used a 2011-2013 baseline. We estimated impacts of a single, evolving Maryland Model each year from 2014-2021 relative to outcomes that would have occurred without any changes that were made starting in 2014. We contrast the average effects during the MDTCOC period (2019-2021) with the end of the MDAPM period (2017-2018).

Population Studied

Medicare FFS beneficiaries in Maryland (~800,000/year) and the matched comparison regions (~8M/year).

Principal Findings

On average from 2019-2021, the Model significantly reduced all-cause hospital admissions (16.1%), potentially preventable admissions (16.1%), and the likelihood of an unplanned readmission (9.5%). The Model also decreased total Medicare spending by 2.5%, with large decreases in hospital spending (6.6%) that were partially offset by increases in non-hospital spending (2.7%). For most outcomes, impacts during the MDTCOC period were more favorable than they were at the end of the MDAPM period (2017-2018), including hospital admissions (6.1 percentage points [pp] lower) and total spending (1.5pp lower). While the average effects on total spending in 2019-2021 were favorable, the model increased non-hospital spending substantially in 2021 (by 5.5%), diminishing total savings that year (p<0.01 for all impacts reported).


On average, MDTCOC had significant, mostly favorable effects on utilization, spending, and quality-of-care outcomes from 2019-2021. In most cases, these effects were larger than they were at the end of MDAPM period (2017-2018), suggesting continued improvement. Such improvement could be the result of growing effects of hospital global budgets, new Model elements added in 2019, or synergies between the two. Recent increases in non-hospital spending, if they continued to grow, could present a risk to the Model’s goal of reducing the total cost of care.

Implications for Policy or Practice

As states consider ways to improve value in health care MDTCOC shows that hospital global budgets coupled with incentives to engage other providers can constrain cost growth, prevent the need for some acute care, and shift acute care to lower-cost settings. However, this model also hinges on widespread participation and a state regulatory body able to set all-payer hospital budgets statewide. States aiming to replicate aspects of the MD TCOC model’s success will need to assess which model elements they can realistically incorporate.

How do you apply evidence?

Take our quick four-question survey to help us curate evidence and insights that serve you.

Take our survey