Facility Medicaid Payer Burden and Nonelective Admission for Chronic Limb-Threatening Ischemia

Abstract

Importance: Chronic limb-threatening ischemia (CLTI) is a major public health issue that requires considerable human and physical resources to provide optimal patient care. It is essential to characterize the disease severity and resource needs of patients with CLTI presenting to facilities of varying resource capacities.

Objective: To investigate the association between facility-level Medicaid payer proportions and the incidence of nonelective admissions among patients admitted for CLTI.

Design, setting, and participants: In this retrospective multicenter cohort study, 876 026 CLTI-related inpatient admissions at 8769 US facilities from January 1, 1998, through October 31, 2020, were identified in the National Inpatient Sample. Facilities were ranked into quintiles according to increasing Medicaid burden, defined as the annualized proportion of Medicaid patient discharges for all hospitalizations. Inpatient admissions for CLTI were identified using International Classification of Diseases codes for rest pain, foot ulcers, and gangrene. Patients younger than 18 years or older than 100 years were excluded, as were those with missing admission type. Statistical analysis was conducted from January to August 2024.

Exposure: Facility-level Medicaid burden quintiles.

Main outcomes and measures: Emergency and urgent admissions defined as nonelective admissions.

Results: The study included 876 026 CLTI-related admissions (mean [SD] patient age, 68.6 [14.5] years; 54.3% men). Increasing nonelective admission rates were associated with increasing facility Medicaid burden (low Medicaid burden, 59.7%; low-moderate Medicaid burden, 62.2%; moderate Medicaid burden, 63.6%; moderate-high Medicaid burden, 63.6%; and high Medicaid burden, 66.8%; P < .001). This trend persisted across all CLTI-related diagnoses (patients with rest pain: low Medicaid burden, 29.8%; high Medicaid burden, 36.1%; patients with lower-limb ulceration: low Medicaid burden, 63.5%; high Medicaid burden, 71.5%; and patients with gangrene: low Medicaid burden, 61.2%; high Medicaid burden, 67.4%; P < .001). In the adjusted model, odds of nonelective admission for CLTI indications increased progressively among facilities as Medicaid burden increased from low to high (adjusted odds ratio for low-moderate Medicaid burden, 1.05 [95% CI, 1.00-1.11]; P = .06; adjusted odds ratio for high Medicaid burden, 1.44 [95% CI, 1.36-1.52]; P < .001).

Conclusions and relevance: High Medicaid burden facilities were associated with increased nonelective admissions for CLTI. This highlights an important mismatch: that resource-constrained facilities are at greater odds of seeing more resource-intensive admissions. Facility-level patient cohort characteristics should be considered when planning for resource allocation to achieve equitable patient care.