Geographic disparities in hospital readmissions: a retrospective cohort study among patients with chronic disease in rural China.

Background: Frequent hospital readmissions place a significant burden on patients, families, and society. Many high-income countries have implemented financial incentives to reduce readmissions. In China, readmission metrics have also been introduced as part of the performance evaluation for secondary hospitals. However, the understanding of hospital readmissions, particularly in rural and remote areas of China, remains limited. This study aims to analyze geographic disparities in hospital readmissions among high-need patients.

Methods: This retrospective cohort study used anonymized hospital discharge data from January 1, 2017, to December 31, 2021, from three public secondary county hospitals. We included rural patients aged 15 and older with hypertension or type 2 diabetes. The outcomes were 30-day, 90-day, and annual readmissions. The explanatory variable was the travel distance to county hospitals, calculated based on the longitude and latitude of registered addresses. Covariates included patient demographics (gender, age, marital status, and ethnicity); health status (Charlson comorbidity score, types of chronic diseases, surgery, and length of stay); and other factors (health insurance and admitted departments). We first reported unweighted readmissions stratified by travel distances (< 40 km versus ≥ 40 km). Multiple logistic regression models were then used to examine the relationship between travel distances and readmissions.

Results: The 30-day, 90-day and annual readmission rates for hypertension or type 2 diabetes were 8.5%, 19.1%, and 39.7%, respectively. Patients living far away were more vulnerable - older (aged 65 and older 59.1% versus 58.5%, P < 0.001), predominantly minorities (Minority 55.6% versus 29.4%, P < 0.001), and having more hypertension and diabetes-related complications, as well as undergoing more surgeries (surgery 29.4% versus 23.3%, P < 0.001) compared to those living nearby. After adjusting covariates and weights, patients living 40 km away had 11% decrease in the odds of being readmitted within 30 days (OR = 0.89, 95%CI = 0.83-0.96), 10% decrease in the odds of 90-day readmissions (OR = 0.90, 95%CI = 0.85-0.94), and 13% decrease in the odds of annual readmissions (OR = 0.87, 95%CI = 0.84-0.91) compared to those living within 40 km.

Conclusion: We found significant geographic disparities in hospital readmissions among high-need patients. Patients living farther from hospitals had significantly lower odds of readmissions. Readmission rates reflect patients' healthcare utilization patterns in rural and remote areas. Policymakers should address the geographic access barriers and be cautious when using readmission rates as a measure of hospital performance.