Statement of the problem: Not all readmissions are preventable, but some may be prevented through the application of proven standards of care. Policymakers are increasingly focusing on this care dimension as a potential quality measure that can be linked to payment. Despite such significant potential impact, no research has examined the characteristics of and the extent to which these types of readmissions occur across categories of rural hospitals or by diagnoses of rural patient populations.

Project goals: The purpose of this project is to: 1) Measure geographic variation in potentially preventable readmission rates across categories of rural hospitals and compared to urban hospitals, adjusting for hospital and patient characteristics; 2) Assess how hospital and patient attributes affect potentially preventable hospital readmission rates; and 3) Examine the relationship between patient discharge locations (e.g., home, nursing home, home health) and subsequent potentially preventable hospital readmissions.

Methods: Potentially preventable readmission rates will be determined using the 3-M algorithm software. Readmissions for which there is no possible clinical relation to the initial admission, those clearly not related to improvement opportunities in hospital care, and those occurring more than 15 days after the initial hospitalization will not be considered in the rates. For each of the four diagnoses (heart failure, pneumonia, acute myocardial infarction, and chronic obstructive pulmonary disease), a separate logistic model will be estimated, controlling for hospital location/payment type category. Patient and hospital attributes will be used as predictors of potentially preventable readmissions. Relative risks and rates of readmissions for each hospital location/payment type category will be estimated and adjusted for by hospital volume.

Anticipated publications or products: The products of this project will include a final report, a policy brief, an article that will be submitted to a peer-reviewed journal, presentation abstracts that will be submitted to state, regional and national conferences and a possible fact sheet.

Publications

• Geographic Differences in Potentially Preventable Readmission Rates in Rural and Urban Hospitals (Final Report)
  Upper Midwest Rural Health Research Center
  Date: 11/2010
  Potentially preventable readmissions (PPRs) among Medicare patients are examples of inefficiencies in the healthcare system. This study estimated PPRs in three types of acute care hospitals: urban prospective payment system (PPS), rural PPS, and Critical Access Hospitals.
• Potentially Preventable Readmissions in Rural Hospitals
  Policy Brief
  Upper Midwest Rural Health Research Center
  Date: 08/2009
  Using 3-M algorithm software and Medicare inpatient claims data from five states, this brief examines potentially preventable readmission rates for rural and urban hospitals and discusses the rural implications of policy initiatives to reduce readmission rates.