How accurately are we coding readmission diagnoses after total joint arthroplasty?

Abstract

Readmission rates have been cited as an important quality measure in the Affordable Care Act. Accordingly, understanding and accurately tracking the causes for readmission will be increasingly important. We queried an electronic database for all patients who underwent primary THA or TKA at our institution from 2006 through 2010. We identified those readmitted within 90 days of surgery and analyzed 87 random de-identified medical records. We then assigned a clinical diagnosis for each readmission, which was then compared with the coder-derived diagnosis by ICD-9 code. The overall 90-day readmission rate was 7.9%. We identified 22 of 87 patients for whom there was disagreement (25.3%, 95% CI=16.6-35.8%). The most common were procedure-related complications. Coded diagnoses frequently did not correlate with the physician-derived diagnoses. The unverified use of coded readmission diagnoses in calculating quality measures may not be clinically relevant.