The AHRQ Patient Safety Indicators (PSIs) are used for calculation of risk-adjusted postoperative rates for adverse events. The payers and quality consortiums are increasingly requiring public reporting of hospital performance on these metrics. We discuss processes designed to improve the accuracy and clinical utility of PSI reporting in practice.
The study was conducted at a 793-bed tertiary care academic medical center where PSI processes have been aggressively implemented to track patient safety events at discharge. A three-phased approach to improving administrative data quality was implemented. The initiative consisted of clinical review of all PSIs, documentation improvement, and provider outreach including active querying for patient safety events.
This multidisciplinary effort to develop a streamlined process for PSI calculation reduced the reporting of miscoded PSIs and increased the clinical utility of PSI monitoring. Over 4 quarters, 4 of 41 (10%) PSI-11 and 9 of 138 (7%) PSI-15 errors were identified on review of clinical documentation and appropriate adjustments were made.
A multidisciplinary, phased approach leveraging existing billing infrastructure for robust metric coding, ongoing clinical review, and frontline provider outreach is a novel and effective way to reduce the reporting of false-positive outcomes and improve the clinical utility of PSIs.
Department of Surgery (Dr Najjar), Department of Quality and Safety, (Drs Najjar and Kachalia and Mss Beloff and David-Kasdan), Department of Medicine (Drs Kachalia and Bates), and Department of Anesthesiology, Perioperative and Pain Medicine (Dr Urman), Brigham and Women's Hospital, Boston, Massachusetts; and Department of Anesthesia, Beth Israel Deaconess Medical Center (Dr Sutherland), Harvard Medical School (Drs Najjar, Kachalia, Sutherland, Bates, Urman), Boston, Massachusetts.
Correspondence: Richard D. Urman, MD, MBA, Department of Anesthesiology, Brigham and Women's Hospital, 75 Francis St, Boston, MA 02115 (firstname.lastname@example.org).
Internal departmental funds were used for this study.
All authors declare no conflict of interest.