560. Medicare Claims Successfully Track Surgical Site Infections Following Vascular Surgery in U.S. Hospitals
Session: Oral Abstract Session: Advancing Public Reporting and Surveillance of HAIs
Thursday, October 3, 2013: 2:30 PM
Room: The Moscone Center: 200-212
Background: Vascular surgery patients are at high risk for surgical site infection (SSI) due to comorbid conditions. Billing codes suggestive of SSI were previously shown to improve SSI detection following vascular surgery. We sought to validate this approach in a national sample of U.S. hospitals.

Methods: We conducted a retrospective cohort study of fee-for-service Medicare patients ≥65 years old who underwent vascular surgery in U.S. hospitals from 2005 through 2008. Hospitals were ranked using predicted rates of ICD-9 codes suggestive of SSI within 60 days of surgery, from generalized linear mixed models with individual-level adjustment for age, gender, and comorbidities. These rates account for clustering within hospital and the number of surgeries performed. For hospitals ranked in the best and worst-performing deciles, we obtained medical records on a random sample of patients for chart-validation of infection. We then calculated the risk-adjusted odds of developing a chart-confirmed SSI in the worst- versus best-decile hospitals.

Results: Among 203,023 eligible Medicare patients who underwent vascular surgery at 2,512 U.S. hospitals, there was a 2.5 times higher odds of developing a chart-confirmed SSI in a worst-versus-best decile hospital (95% CI 2.0-3.1). Of 726 patient charts flagged by an SSI code and reviewed, 46% met CDC/NHSN definitions for SSI, including 21% with a deep or organ/space (D/OS) infection. Every 2.2 charts reviewed identified an SSI and every 4.8 charts reviewed identified a D/OS SSI. Across all U.S. hospitals, 17% of patients were flagged. The chart-confirmed SSI rate in best decile hospitals was 3.3% (1.4% for D/OS) compared with 8.2% for worst decile hospitals (3.8% for D/OS).

Conclusion: In a national study, billing codes suggestive of SSI were highly efficient in identifying charts for identification of SSI events, thus providing a standardized approach to SSI surveillance within and across U.S. hospitals. In an era of increasing public reporting and value-based purchasing, claims-based surveillance is an effective approach to improve case identification and data validation.

Michael S. Calderwood, MD, MPH1,2, Ken Kleinman, ScD1, Dale W. Bratzler, DO, MPH3, Rebecca E. Kaganov, BA1, Christina B. Bruce, BA1, Elizabeth C. Balaconis, BA1, Claire Canning, MA1, Richard Platt, MD MS1, Susan S. Huang, MD, MPH, FIDSA4 and for the CDC Prevention Epicenter Program, (1)Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, MA, (2)Division of Infectious Diseases, Brigham and Women's Hospital, Boston, MA, (3)College of Public Health, University of Oklahoma Health Sciences Center, Oklahoma City, OK, (4)Division of Infectious Diseases and Health Policy Research Institute, University of California Irvine School of Medicine, Irvine, CA


M. S. Calderwood, None

K. Kleinman, None

D. W. Bratzler, None

R. E. Kaganov, None

C. B. Bruce, None

E. C. Balaconis, None

C. Canning, None

R. Platt, None

S. S. Huang, None

Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 2nd with the exception of research findings presented at the IDWeek press conferences.