Calibrating Readmission Risk Prediction Models for Determining Post-discharge Follow-up Timing.

Abstract

The soaring hospital readmission rates are straining the already limited financial resources in the US health system. Meanwhile, timely outpatient follow-up, an efficient and cost-effective intervention following hospital discharge, has been shown to reduce the readmission risk. However, the current and projected shortage of physicians in primary and specialty care poses a unique dilemma in transitional care planning: optimizing the utilization of post-discharge follow-up to reduce readmission rate while limiting the strain on the limited pool of outpatient physicians. The ideal solution would entail a strategy whereby patients at higher risk for readmission are stratified towards earlier outpatient follow-up and vice versa. This article explores the utility of Institution-specific readmission risk prediction algorithms for assessing patient population for diverse administrative, clinical and socioeconomic risk factors and further classifying the hospital’s patient population into high- and low-risk strata, so that appropriate risk-concordant timing of follow-up can be assigned at the time of hospital discharge, with earlier follow-up assigned to high readmission risk strata. This stratification shall help ensure judicious and equitable human resource allocation while simultaneously reducing hospital readmission rates.