Admission and Mid-Stay MedisGroups Scores as Predictors of Hospital Charges and 30-Day Mortality

摘要

The report explores the ability of MedisGroups' mid-stay review scores and other data collected during the review to predict hospital resource use (charges and length of stay) and mortality. It builds on prior work which suggests that admission MedisGroups scores have only a modest ability to predict the costs of hospitalizations. During the mid-stay review, which occurs approximately one week into the hospitalization for most patients, data are collected which are similar to those gathered during the admission review. However, they are processed differently by the MedisGroups computer algorithm, to produce a three-level score: no morbidity, morbid, and major morbidity. Patients receive mid-stay reviews only if they remain in-hospital for the required length of time. The work evaluates the relationship between admission MedisGroups score alone, the mid-stay review score alone and the clinical trajectory and two outcome variables: resource use (hospital charges and length of stay) and mortality.