JAMA Health Forum. 2026 Jul 2;7(7):e262226. doi: 10.1001/jamahealthforum.2026.2226.
ABSTRACT
IMPORTANCE: Forty-three state Medicaid programs have bed-hold policies that reimburse nursing homes (NHs) to reserve a resident’s bed during a hospitalization or other temporary absence. Although bed-hold policies are intended to increase continuity of care, some evidence suggests they may incentivize NHs to hospitalize residents; however, these data are outdated, and the impact of bed-hold policies on resident outcomes remains understudied.
OBJECTIVE: To examine the association between Medicaid bed-hold policies and hospitalizations and emergency department (ED) visits among NH residents receiving long-term care.
DESIGN, SETTING, AND PARTICIPANTS: This cohort study of long-stay NH residents in the US used claims data for a 20% national sample of Medicare fee-for-service beneficiaries from 2013 through 2018. Adjusted comparisons between states that did and did not adopt bed-hold policies were conducted using a staggered difference-in-differences design. Sensitivity analyses were conducted that (1) used composite outcomes combining death with hospitalization and ED visits, (2) restricted the sample to dually eligible residents, (3) excluded hospital-based NHs, (4) restricted the sample to NHs that were present for the entire study period, (5) restricted the sample to for-profit NHs, and (6) used alternate approaches to adjust standard errors. Data were analyzed between April and November 2025.
EXPOSURES: State adoption of a Medicaid bed-hold policy.
MAIN OUTCOMES AND MEASURES: Hospitalizations and ED visits in a given month were the 2 primary outcomes.
RESULTS: The sample included 219 719 long-stay NH residents (mean [SD] age, 80.2 [12.6] years; 67.2% female) in 22 states and the District of Columbia, of which 16 states adopted bed-hold policies. There were no significant changes in differences in resident and NH characteristics between those in states that did and did not adopt the policies from the pre- to the post-adoption period. Adjusted estimates did not show statistically significant changes in the likelihood of hospitalization (-0.1 percentage points [pp]; 95% CI, -0.3 to 0.1 pp; P = .41) or an ED visit (0.0 pp; 95% CI, -0.1 to 0.2 pp; P = .68) associated with adoption of bed-hold policies. Results of sensitivity analyses were consistent with the primary analyses.
CONCLUSIONS AND RELEVANCE: This cohort study found that states’ adoption of Medicaid bed-hold policies was not associated with hospitalizations or ED visits among long-stay NH residents, suggesting that the goal of bed-hold policies may not be undermined by increases in these events.
PMID:42467449 | DOI:10.1001/jamahealthforum.2026.2226
