Evaluation of the Padua Prediction Score ability to predict venous thromboembolism in Israeli non-surgical hospitalized patients using electronic medical records

Abstract

Venous thromboembolism (VTE) is considered a leading safety concern during hospitalization. The Padua Predication Score (PPS) is a risk model conceived to predict VTE among non-surgical hospitalized patients. The study aim was to evaluate the PPS ability to predict VTE in Israeli non-surgical hospitalized patients using data from electronic medical records. A single center, large-scale, historic cohort study of hospitalized non-surgical patients was conducted. Outcomes included clinically diagnosed symptomatic VTE events, bleeding events, and mortality during hospitalization and up to 90 days thereafter, and readmission up to 90 days after discharge. 5117 patient records were analyzed after screening and validation. 1120 (22%) patients were defined per PPS as high-risk, of which 277 (24.7%) were prophylactically treated. The low-risk group included 3997 (78%) patients. Prevalence of symptomatic VTE was low. Overall, 14 (0.27%) VTE events were diagnosed: 3 cases in the high-risk group (0.27%) and 11 (0.28%) in the low-risk group, with no significant difference, p = 0.768. Prophylactic treatment among the high-risk patients did not significantly improve VTE incidence: 1/277 (0.36%) treated vs. 2/843 (0.24%), p = 0.343. There was no significant difference between the study groups regarding the rates of bleeding, unexplained mortality or readmission. PPS was not found to be an efficient tool for identification of non-surgical hospitalized patients with high risk for clinically significant VTE.