Abstract 147: Do Billing Data Accurately Describe Causes Of Readmission After PCI? An Analysis Of A Large PCI Readmission Database.
Introduction: Early readmission after PCI can be a useful performance measure, particularly if the reasons for readmission are related to patient management (e.g. angina, stent thrombosis or bleeding). Prior studies on readmission have used principal billing discharge diagnosis, but the validity of billing codes for this purpose is unclear.
Methods: PCI patients readmitted within 30 days of discharge at the Massachusetts General Hospital (January 2007-December 2011) and Brigham and Women’s Hospital (June 2009-December 2011), were identified, and their medical records reviewed by a cardiologist. For each readmission, the principal billing discharge diagnosis of the readmission was compared to the primary diagnosis of the readmission as determined by the chart review. The accuracy of billing discharge diagnoses for readmissions due to chest pain or other symptoms concerning for angina and vascular access complications was assessed.
Results: Of 9081 patients undergoing PCI and surviving to hospital discharge, 1011 (11.1%) were readmitted to the index hospital within 30 days. After excluding repeat readmissions, 894 readmissions were reviewed and of those, 754 (84.3%) could be matched to billing diagnoses. For each reason for readmission, corresponding billing diagnoses were diverse (Table 1).
The sensitivity and specificity of billing code 414.01 for a readmission for chest pain or other symptoms concerning for angina were 0.36 and 0.85, respectively. The sensitivity and specificity of billing code 997.2 (Peripheral vascular complications, not elsewhere classified) for a bleeding or vascular complication of PCI were 0.28 and 0.997, respectively.
Conclusions: The diversity of ICD-9 codes does not allow straightforward categorization of principal diagnoses of readmission from billing data. More complex algorithms need to be developed and validated to reliably derive reasons for readmission after PCI from billing data.
Author Disclosures: J.H. Wasfy: None. C. O'Brien: None. J.B. Strom: None. A. Zai: None. J. Luttrell: None. J.A. Spertus: B. Research Grant; Significant; American College of Cardiology. F. Ownership Interest; Significant; Health Outcomes Sciences. L. Mauri: B. Research Grant; Significant; Abbott, Boston Scientific, Cordis, Medtronic, Eli Lilly, Daiichi Sankyo, Bristol Myers Squibb, Sanofi-Aventis. G. Consultant/Advisory Board; Significant; Cordis, Medtronic. S.T. Normand: None. R.W. Yeh: None.
This research has received full or partial funding support from the American Heart Association, Founders Affiliate (Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Rhode Island, Vermont).
- © 2014 by American Heart Association, Inc.