Development and Validation of an Algorithm to Identify Planned Readmissions From Claims Data

摘要: BACKGROUND It is desirable not to include planned readmissions in readmission measures because they represent deliberate, scheduled care. OBJECTIVES To develop an algorithm identify readmissions, describe its performance characteristics, and improvements. DESIGN Consensus-driven development chart review validation study at 7 acute-care hospitals 2 health systems. PATIENTS For development, all discharges qualifying for the publicly reported hospital-wide measure. For validation, same-hospital that were characterized by as planned, a random sampling of unplanned. MEASUREMENTS We calculated weighted sensitivity specificity, positive negative predictive values (version 2.1), compared gold standard review. RESULTS In consultation with 27 experts, we developed characterizes 7.8% planned. reviewed 634 readmissions. The was 45.1% overall, 50.9% large teaching centers 40.2% smaller community hospitals. specificity 95.9%, value 51.6%, 94.7%. We identified 4 minor changes improve performance. revised had 49.8% (57.1% hospitals), 96.5%, 58.7%, 94.5%. Positive poor most common potentially procedures: diagnostic cardiac catheterization (25%) procedures involving devices (33%). CONCLUSIONS An administrative claims-based feasible can facilitate public reporting primarily unplanned Journal Hospital Medicine 2015;10:670–677. © 2015 Society Medicine.