Patient safety indicators (PSI) were developed for hospitals to screen for healthcare-associated adverse events. PSIs are believed to be preventable and have become a part of major pay-for-performance programs. PSI-12 captures perioperative venous thromboembolism (VTE), which contributes to morbidity and mortality of hospitalized patients. We aimed to evaluate PSI-12 events at our institution to identify areas for improvement of perioperative VTE prevention.
We identified PSI-12 events from June 2015 to June 2017 using the Agency for Healthcare Research and Quality software version 5. Events were reviewed using our electronic medical record to identify further details of each event.
A total of 154 perioperative VTE cases were analyzed in the 2-year period. Pulmonary embolism (PE) occurred in 62.9% of cases, deep venous thrombosis (DVT) in 24%, and concurrent DVT/PE in 12.9%. The mean age of patients was 56 years old. Deficiencies in guideline-appropriate prophylaxis were identified in only 17 (11%) of cases. Unfractionated heparin was used in 61 cases, enoxaparin in 31 cases, and nine events occurred on therapeutic anticoagulation. Mechanical prophylaxis was used in 51 cases because of bleeding risk, thrombocytopenia, and/or liver associated coagulopathy. Four events occurred prior to the index procedure, with another eight cases occurring intraoperatively, or on the day of the procedure.
PSI-12 has several limitations in identifying quality of care issues in perioperative VTE. While it may be useful as a screening tool, further research for improvements are needed if it will remain one of the key measures in pay-for-performance.

Author