Inter-rater reliability in the evaluation of a thrombosis risk assessment tool

Introduction: Venous thromboembolism (VTE) is a health concern that is common among hospitalized patients. There are more than 200,000 new VTE events that occur annually. The majority of VTE cases are asymptomatic, and 70% to 80% of fatal pulmonary embolisms (PE) in the hospital are not predictable. The ENDORSE study evaluated more than 35,000 patients and reported that only 58% of surgical and 39% of medical patients at risk for VTE received American College of Chest Physicians (ACCP) recommended VTE prophylaxis.¹ Nebraska Methodist Hospital evaluated and revised the thrombosis risk assessment tool published by Motykie et al³ for use with the institution's medical and surgical patients. Objectives: The purpose of this study was to assess the reliability of the thrombosis risk assessment tool. Secondary end points were to determine if patients were treated with the appropriate therapy based on their total risk score, to assess if therapy was appropriately adjusted after pharmacist intervention, and to determine the incidence of VTE. Methods: Three evaluators completed a thrombosis risk assessment of 100 patients chosen randomly upon hospital admission during a 1-month period. Pearson's correlation coefficients were determined to assess reliability. Patient medication profiles were evaluated for the proper prophylaxis regimen and interventions were made as needed. Quality control provided data on the incidence of thromboembolic events. Results: The thrombosis risk assessment tool was shown to have good clinical reliability. Seventy percent of patients were treated appropriately. Interventions were generally accepted after pharmacist intervention unless patients were soon to be discharged. The incidence of deep vein thrombosis in surgical and medical patients and pulmonary embolism in surgical patients was decreased after implementation of the thrombosis risk assessment tool. However, the impact of the thrombosis risk assessment tool on the incidence of pulmonary embolism in the medical patients could not be determined.