Fragmentation of care for frequently hospitalized urban residents.

BACKGROUND: Fragmentation across sites of care may impede efficient healthcare delivery.
OBJECTIVES: The objectives of this study were to evaluate fragmentation of hospital care for chronically ill New York City (NYC) residents and its association with enrollment in the New York State (NYS) Medicaid program. RESEARCH
DESIGN: We conducted a cross-sectional study using the NYS Department of Health's Statewide Planning and Research Cooperative System discharge database. We identified 53,031 NYC residents admitted 3 or more times to acute care hospitals between 2000 and 2002 with the same principal diagnosis of a specific chronic illness (diabetes, sickle cell anemia, psychosis, substance abuse, cancer, gastrointestinal disease, chronic obstructive pulmonary disease/asthma, coronary artery disease, or congestive heart failure). We also evaluated a larger cohort of 225,421 patients with >or=3 admissions for a specific chronic illness coded as either the principal or a secondary diagnosis. A generalized logit model was used to examine the relationship between fragmentation and each patient's primary insurance adjusted for diagnosis and demographic characteristics. MEASURES: Fragmentation was characterized as high, moderate, or low based on the number of distinct hospitals a patient visited relative to the patient's total number of hospitalizations over the 3-year interval.
RESULTS: Among frequently hospitalized NYC residents with select chronic conditions, 17.1% experienced highly fragmented care. This rate was 9.9% for patients with commercial insurance, 24.4% for those with Medicaid, and 9.7% for those with Medicare. The unadjusted odds ratio describing high fragmentation of Medicaid enrollees compared with commercially insured patients was 3.82 (95% confidence interval [CI], 3.50-4.18) and, although attenuated, remained significant after adjustment for demographic characteristics (odds ratio, 1.33; 95% CI, 1.20-1.47). The strongest predictor of fragmentation was a diagnosis of psychosis (OR, 2.81; 95% CI, 2.43-3.25) or substance abuse (OR, 7.58; 95% CI, 6.55-8.77).
CONCLUSIONS: In NYC, Medicaid enrollment is associated with greater fragmentation of hospital care, but this is largely attributable to the preponderance of Medicaid enrollees with diagnoses of psychosis and substance abuse. Strategies to improve the efficiency of healthcare delivery should focus on patients with mental illness who are frequently admitted to general hospitals.