Deep vein thrombosis and validation of the Caprini risk assessment model in Chinese orthopaedic trauma patients: a multi-center retrospective cohort study enrolling 34,893 patients

Abstract

Abstract Background The risk of venous thromboembolism among orthopaedic trauma patients is high, but prevalence of deep vein thrombosis (DVT) remains unknown. In addition, the Caprini risk assessment model (RAM) score in orthopaedic trauma patients is undetermined in previous research. This study is aimed to determine the incidence of DVT and then validate the Caprini RAM in orthopaedic trauma patients. Methods This is a retrospective cohort study enrolling orthopaedic trauma inpatients from seven tertiary and secondary hospitals during a 3-year period (from April 1, 2018 through April 30, 2021). Caprini RAM scores were assessed by experienced nurses on admission. The patients with suspected DVT were verified through duplex ultrasonography by qualified radiologists, and then prospectively followed once a year after discharge. Results In total, 34,893 patients were enrolled in our study. The Caprini RAM identified 45.7% of patients at low risk (Caprini score 0–2), 25.9% at medium risk (3–4), and 28.3% at high risk (5–6), highest risk (7–8), and superhigh risk (> 8). Patients with Caprini score > 5 were likely to be older, female, and with longer length of hospital stay. Moreover, 8695 patients had received ultrasonography to detect DVT. The prevalence of DVT was determined to be 19.0% [95% confidence interval (CI) 18.2–19.9%], which significantly increased with Caprini score. The area under curve of the Caprini RAM for DVT was 0.77 (95% CI 0.76–0.78) with a threshold of 4.5. Furthermore, 6108 patients who had received ultrasonography completed the follow-up. DVT patients had a hazard ratio of 1.75 (95% CI 1.11–2.76; P = 0.005) in the mortality, compared to non-DVT ones. Caprini scores were significantly associated with increase in the mortality [odds ratio (OR) 1.14; 95% CI 1.07–1.21; P < 0.001]; DVT remained an independent effect (OR 1.5; 95% CI 1.02–2.26; P = 0.042). Conclusions The Caprini RAM may be valid in Chinese orthopaedic trauma patients. Prevalence of DVT and higher Caprini score were significantly associated with increased all-cause mortality among orthopaedic trauma patients after discharge. Further study is warranted to explore the causes of higher mortality in patients with DVT.