1. Development and Validation of a Methodology to Reduce Mortality Using the Veterans Affairs Surgical Quality Improvement Program Risk Calculator.
- Author
-
Keller, Deborah S., Kroll, Donald, Papaconstantinou, Harry T., and Ellis, C. Neal
- Subjects
- *
SURGERY , *TERTIARY care , *CARDIAC surgery , *MEDICAL centers , *MEDICAL decision making , *MANAGEMENT , *CLINICAL medicine , *COMPARATIVE studies , *DATABASES , *HEALTH status indicators , *LONGITUDINAL method , *VETERANS , *RESEARCH methodology , *MEDICAL cooperation , *MEDICAL referrals , *POSTOPERATIVE care , *QUALITY assurance , *RESEARCH , *RISK assessment , *ELECTIVE surgery , *VETERANS' hospitals , *EVALUATION research , *KEY performance indicators (Management) , *RETROSPECTIVE studies , *RECEIVER operating characteristic curves - Abstract
Background: To identify patients with a high risk of 30-day mortality after elective surgery, who may benefit from referral for tertiary care, an institution-specific process using the Veterans Affairs Surgical Quality Improvement Program (VASQIP) Risk Calculator was developed. The goal was to develop and validate the methodology. Our hypothesis was that the process could optimize referrals and reduce mortality.Study Design: A VASQIP risk score was calculated for all patients undergoing elective noncardiac surgery at a single Veterans Affairs (VA) facility. After statistical analysis, a VASQIP risk score of 3.3% predicted mortality was selected as the institutional threshold for referral to a tertiary care center. The model predicted that 16% of patients would require referral, and 30-day mortality would be reduced by 73% at the referring institution. The main outcomes measures were the actual vs predicted referrals and mortality rates at the referring and receiving facilities.Results: The validation included 565 patients; 90 (16%) had VASQIP risk scores greater than 3.3% and were identified for referral; 60 consented. In these patients, there were 16 (27%) predicted mortalities, but only 4 actual deaths (p = 0.007) at the receiving institution. When referral was not indicated, the model predicted 4 mortalities (1%), but no actual deaths (p = 0.1241).Conclusions: These data validate this methodology to identify patients for referral to a higher level of care, reducing mortality at the referring institutions and significantly improving patient outcomes. This methodology can help guide decisions on referrals and optimize patient care. Further application and studies are warranted. [ABSTRACT FROM AUTHOR]- Published
- 2017
- Full Text
- View/download PDF