Implementation and Validation of Decision-Aid Tool to Reduce Unnecessary Testing Before Thoracic Surgery.

Abstract

Background: The aim of this study was to implement and prospectively validate a previously published decision-aid tool to guide ordering of preoperative type and screen (preT&S) tests.

Study design: In this interrupted time-series quasi-experimental study, we implemented a decision-aid tool for patients undergoing elective thoracic surgery at a single academic institution. Data were collected 6 months before and prospectively after implementation. The tool, a previously published nomogram, predicts the need for a preT&S using age, BMI, planned operation, approach, and preoperative hemoglobin. We excluded patients who had previous transfusions, neoadjuvant therapy, redo operations, and/or inpatient consults. We validated the tool using multivariable logistic regression, regression discontinuity, c-index, sensitivity, predictive values, and cost savings.

Results: One hundred seventy-seven consecutive patients met the inclusion criteria. Eighty-eight were after implementation and 89 patients were before implementation. No differences were observed between the groups in terms of age, sex, BMI, comorbidities, approach, or preoperative hemoglobin (all p > 0.05). Overall transfusion rates were similar (6.8% vs 6.7%; p > 0.99); however, the rate of ordering preT&S was reduced significantly (94.4% vs 60.2%; p < 0.001). At our institution, the decision-aid tool resulted in cost savings of $25,048 over 6 months alone. With a c-index of 0.977, our validation demonstrated 100% sensitivity, 90.3% specificity, and 100% negative predictive value.

Conclusions: Implementation and validation of the preT&S nomogram proved feasible, accurate, and resulted in reducing unnecessary testing and costs before elective noncardiac thoracic surgery. Wider implementation has the potential for substantial cost savings.