Back to search

Boston University

Evaluation of a Massachusetts Medicaid housing support program in a safety-net health system: implementation and health care utilization outcomes

Abstract

dc:description.abstract

Housing-related needs, including homelessness and housing instability, are prevalent and increasing. In 2024 in the United States, an estimated 771,480 people were homeless on a given night, an 18% rise from 2023. Additionally, in 2023, 17% of households spent more than half of their income on housing or lived in severely inadequate conditions. Housing-related needs are linked to mortality, poor health, and costly acute health care utilization and exacerbated among people with substance use disorders. To improve patient outcomes and reduce costs, health care entities, including state Medicaid programs and local health systems, have adopted initiatives to identify and address patients’ housing-related needs. Yet such initiatives vary widely in their design, including supports offered and eligibility criteria. More research is needed to understand how setting-specific contextual factors like population and health system characteristics influence program implementation and effectiveness. Housing support program evaluations have faced several limitations, including challenges with identifying individuals’ housing-related needs, constructing comparison groups, and linking implementation factors to outcomes. This dissertation used qualitative and quantitative methods to evaluate the implementation and health care utilization outcomes of a Massachusetts Medicaid housing support program, MassHealth Flexible Services Program for Housing (FSP-H), within the largest safety-net health system in New England. In Chapter One, I provided an overview of the literature on housing as a health-related social need, an in-depth description of the health system setting and FSP-H, and the conceptual framework supporting my analyses. In Chapter Two, I validated a pragmatic algorithm that combined six structured data elements from the electronic health record to identify patient housing-related needs compared against two reference standards. With all data elements combined, algorithm sensitivity ranged from 62%–81%, and specificity was 97%. Health-related social need screening yielded the highest single-element sensitivity but remained limited: 57%–62% of those with housing-related needs were detected via screening. Address information and diagnostic codes had low sensitivity. Findings demonstrate the value of integrated clinical and administrative data and highlight areas for improvement including changes to health-related social need screening and documentation. In Chapter Three, I conducted in-depth interviews with key informant stakeholders involved in the health system’s administration and delivery of FSP-H. I aimed to understand facilitators and barriers to FSP-H implementation over time within the health system, with considerations for patients with substance use disorders. Key informants strongly valued FSP-H, noting the importance of formal contracted partnerships with external housing organizations which shifted case management responsibilities from clinical staff to housing experts. However, they also described challenges, including tension in patient prioritization (i.e., trade-offs between offering services broadly to those with housing-related needs, or focusing on highest-acuity patients). Frequent changes to FSP-H, including evolving eligibility criteria, were sometimes perceived as top-down or misaligned with local realities. Results highlight the acceptability of Medicaid-based housing—health partnerships, as well as the need for evidence-informed program design and evaluation and transparent bidirectional communication about program decision-making between state Medicaid agencies, program administrators, and on-the-ground stakeholders. In Chapter Four, I linked FSP-H programmatic records, electronic health records, and administrative claims data to evaluate the association between FSP-H receipt and health care utilization and cost outcomes. I used a staggered difference-in-differences approach to compare monthly inpatient and emergency department utilization and total cost of care in FSP-H recipients versus a comparison group of individuals with housing-related needs who were ineligible for FSP-H due to their insurance plan (identified via the Chapter Two algorithm). FSP-H was not significantly associated with utilization but was associated with decreased monthly total cost of care, which approached statistical significance ($634 decrease, p=0.053). Significant declines in monthly inpatient utilization were observed among individuals with substance use disorders and those with frailty, and in monthly total cost of care among those with frailty. Results suggest FSP-H may be associated with reduced health care utilization and costs, particularly in high-risk clinical groups. Findings offer early evidence for state Medicaid plans implementing or iterating housing support programs. Findings from this dissertation can inform health system practice and Medicaid health-related need-focused waiver policy, with insights applicable to similar interventions in other settings and to future research.

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy
Level thesis:degree_level
doctoral
Discipline thesis:degree_discipline
Health Services Research
Grantor
Boston University
Year dc:date.issued
2026

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • McCann, Nicole Cristina
Advisor dc:contributor.advisor
  • Shafer, Paul R.
ORCID dc:identifier.orcid

Subjects

dc:subject × 2

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/2144/52946

Chain of custody

source
Harvested from
Boston University
Base URL
open.bu.edu/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

McCann, Nicole Cristina. Evaluation of a Massachusetts Medicaid housing support program in a safety-net health system: implementation and health care utilization outcomes. doctoral thesis, Boston University, 2026. https://hdl.handle.net/2144/52946