Comparison of 30-day readmission and emergency department revisit rates among homeless patients at teaching versus non-teaching hospitals

Abstract

Providing quality healthcare for homeless patients is a major public health challenge, and some hospitals may be better at treating homeless patients than others. However, whether the quality of care that homeless patients receive differs by the teaching status of hospitals remains unclear. Using statewide databases that include all hospital admissions and emergency department (ED) visits in four states (Florida, Massachusetts, Maryland, and New York) in 2014, we compared 30-day readmission and ED revisit rates for homeless and non-homeless patients discharged from teaching hospitals versus non-teaching hospitals, after adjusting for patient and hospital characteristics. Among 3,438,538 patients (median age [IQR]: 63 [49-77] years) analyzed, 132,025 (4%) were homeless patients. Overall, homeless patients had a higher readmission rate (28.3% vs. 17.7%; average marginal effects [AME], 10.5%; 95% confidence interval [CI], 8.2%-12.9%; p<0.001) and a higher ED revisit rate (37.6% vs. 23.9%; AME, 13.7%; 95%CI, 10.9%-16.6%; p<0.001) than non-homeless patients. Patients from teaching hospitals had similar readmission rate (18.2% vs. 18.3%; AME, -0.1%; 95%CI, -0.8%-0.5%; p=0.69) and slightly lower ED revisit rate than those from non-teaching hospitals (24.1% vs. 25.2%; AME, -1.1%; 95%CI, -1.9% to -0.3%; p<0.01). When we focus on joint effects of homelessness and hospital teaching status, we found that homeless patients treated at teaching hospitals had lower rates of 30-day readmission (AME, -5.8%; 95%CI, -9.7% to -1.8%; p<0.01) and ED revisit (AME, -9.3%; 95%CI, -13.1% to -5.5%; p<0.001) compared to those treated at non-teaching hospitals. For non-homeless patients, in contrast, we found no evidence that rates of hospital readmission (AME, 0%, 95%CI, -0.1%-0.1%; p=0.94) or ED revisit (AME, -0.9%; 95%CI, -1.7% to -0.1%; p=0.02) differ between teaching and non-teaching hospitals. These findings suggest the healthcare settings in which homeless patients receive care have important implications for their patient outcomes.