The role of early venous duplex screening in improving patient safety indicator-12 (PSI-12) classification accuracy: a retrospective cohort study.

Abstract

Background: Venous thromboembolism (VTE) is a major cause of morbidity among surgical patients. Patient Safety Indicator 12 (PSI-12) identifies perioperative VTE events. This study evaluates whether early admission duplex ultrasound identifies VTEs associated with PSI-12 coding and assesses PSI-12 classification accuracy.

Methods: A single-center retrospective analysis was conducted on surgical patients with a venous duplex ultrasound within 48 h of admission at a large academic medical center (2013-2024). Two cohorts were analyzed: (1) all surgical patients with early duplex imaging, (2) PSI-12 positive elective surgical patients with early duplex imaging. The primary outcome was the prevalence of pre-existing DVT identified within 48 h among PSI-12 positive patients. Diagnostic accuracy of early abnormal duplex findings for predicting PSI-12 classification was assessed using sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).

Results: Among all 10,498 surgical patients, 10.6% had positive duplex findings for DVT. PSI-12 positive patients had double the prevalence compared to PSI-12 negatives (20.9% vs. 10.4%, p < 0.001). In PSI-12 positive elective cases, 10.7% of preoperative duplex studies identified a DVT. Misclassification, defined as PSI-12 positive admissions without imaging-confirmed DVT or pulmonary embolism, occurred in 8.7% of elective cases. Overall, 36.7% of PSI-12 positive patients had abnormal early duplex findings. Early duplex demonstrated limited sensitivity but high specificity (sensitivity 36.7%, specificity 81.7%, PPV 3.7%, NPV 98.5%).

Conclusions: Our findings suggest possible pre-existing disease rather than hospital-acquired VTE in PSI-12 positive patients. Incorporating admission duplex studies could improve PSI-12 classification accuracy, reduce costs, and enhance patient safety.