Analysis of current venous thromboembolism risk assessment tools in trauma patients treated with cast immobilisation

A.M. Perera, U. Watson

¹University Hospital of Wales, Cardiff, United Kingdom

Introduction: NICE guidelines state that every patient should be assessed for their VTE risk on admission to hospital. The aim of this study was to determine whether currently recommended risk assessment tools (Nygaard, Caprini, NICE and Plymouth) can correctly identify the patients at risk.

Methods: In a consecutive series of over 750 trauma patients treated with cast immobilisation 23 were found to have suffered a VTE. Their notes were retrospectively reviewed to discover how many had been assessed for their VTE risk on admission. Additionally, the 4 most current Risk Assessment Tools were used to retrospectively score the patients for their VTE risk to determine whether they would have been identified as at risk of sVTE, had the RAMs been used at the time. We also identified a matched group of patients in the same cohort who had not suffered a VTE and they were also retrospectively risk assessed.

Results: NICE (2010), Caprini (2001) and Nygaard (2009) identified 100% of the 750 patients as at risk of sVTE but had a specificity of 0% as only 23 went on to develop VTE. The Plymouth Score (2010) was more specific and identified 56.3% patients of the 23 confirmed VTEs as ‘at risk’. However it would not have recommended prophylaxis in the remaining 46.7& that did in fact go on to developed VTE.

Conclusion: The tools used in this study have no clinical utility in this patient group. Detailed evaluation of the different RAMs is required in order to improve their discriminatory power. A reliance on NICE, Caprini and Nygaard tools would have required all 750 patients in this group to have been treated with thromboprophylaxis and therefore lacked sensitivity. However the Plymouth Score would have failed to recommend thromboprophylaxis in half of the patients who eventually developed VTE.

Click thumbnail below to view poster / thumbnail: