Preoperative prediction of non-home discharge: a strategy to reduce resource use after cardiac surgery.

Abstract

Patients requiring discharge to a continuing care facility after cardiac surgery (non-home discharge) frequently have prolonged hospital stays while arrangements are made for posthospital care. We hypothesized that preoperatively identifying patients likely to require non-home discharge would allow earlier discharge planning, shorten length of stay, and thereby reduce resource use. This study sought to develop a validated tool for preoperative planning of non-home discharge. From October 2008 to December 2009, 4,243 patients were discharged alive after cardiac surgery at Cleveland Clinic. Of these, 4,031 resided in the 48 contiguous states or Alaska and formed the study cohort. Logistic regression analysis of non-home discharge was performed using preoperative data generally readily available at admission. A subsequent group of 2,005 patients discharged alive from December 2009 to July 2010 was used to validate this model. Eighteen percent of patients had non-home discharge, which was predictable from data readily available at admission for cardiac surgery (C-statistic 0.88 for model development, 0.87 for model validation). The strongest predictors included intra-aortic balloon pumping (odds ratio [OR] 7.5; 95% confidence interval [CI] 1.7 to 32), emergency status (OR 3.7; CI 2.1 to 6.5), older age (p < 0.001), longer preoperative stays (p < 0.001), poor nutritional state (p < 0.001), a number of comorbidities, and descending thoracic aorta procedures (OR 4.3; 95% CI 2.5 to 7.4). Non-home discharge can be easily predicted using data obtained during routine preoperative evaluation of cardiac surgical patients. We expect that early identification of patients at high risk for non-home discharge will allow for more intensive, personalized discharge planning, and will reduce wasted days and resource use.