Inter-observer reliability of a risk assessment model for venous thromboembolism in acutely-ill medical hospitalized patients: Results from a prospective cohort study

Objectives To analyze the inter-observer reliability of risk for venous thromboembolism (VTE) in a population of adult acutely-ill medical patients. Methods In this prospective cohort study, we collected risk factors and risk classification for VTE using RAM IMPROVE7. Kappa statistics was used to evaluate inter-observer reliability between lead clinicians and trained researchers. We evaluated occurrence of VTE in patients with mismatched classification. Results We included 2,380 patients, median age 70 years (interquartile range [IQR], 58-79), 56.2% female. Adjusted Kappa for VTE risk factors ranged from substantial (0.64, 95% confidence interval [CI], 0.61-0.67) for immobilization, to almost perfect (0.98; 95% CI 0.97-0.99) for thrombophilia; risk classification was 0.64 (95% CI 0.60-0.67). Divergent risk classification occurred in 434 patients (18.2%) of whom seven (1.6%) developed VTE. Conclusion Despite substantial to almost perfect reliability between observers for risk factors and risk classification, lead clinicians tended to underestimate the risk for VTE.

Automated summary

In a population of adult acutely-ill medical patients, there was substantial to almost perfect inter-observer reliability in assessing risk factors and risk classification for venous thromboembolism (VTE). However, lead clinicians tended to underestimate the risk for VTE despite this high reliability.