Venous thromboembolism risk assessments on trauma patients has suboptimal interobserver reliability among inexperienced clinicians (fourth-year medical students)

BACKGROUND: Venous thromboembolic disease is a major cause of morbidity and mortality in hospitalized patients worldwide. The objective of this study is to compare interobserver reliability for qualitative and quantitative venous thromboembolism (VTE) risk assessments in hospitalized trauma patients. METHODS: We conducted a retrospective medical record review of 40 randomly selected trauma patients admitted to a 448-bed urban level-I trauma center from January 2013 to January 2014. Interclass correlation coefficient (ICC) was calculated based on a two-way mixed model. The sample was equally distributed between patients admitted to the floor and the intensive care unit (ICU). Eight fourth-year medical students performed the risk assessments by the medical record. Two forms for risk assessment were used: a qualitative screening and a quantitative screening. The composite of intraobserver and interobserver variabilities was determined. RESULTS: The ICC for qualitative VTE risk assessments was 0.845 and for quantitative VTE risk assessment was 0.628. CONCLUSION: To optimize accuracy of VTE risk stratification and appropriate prophylaxis, medical students and first-year residents should be formally trained to perform quantitative assessments.