Prevalence of preoperative Deep Venous Thrombosis (DVT) following elderly intertrochanteric fractures and development of a risk prediction model

Abstract

BackgroundThis study aimed to investigate the prevalence of preoperative deep venous thrombosis (DVT) following intertrochanteric fractures in the elderly and identify the associated factors, based on which a risk prediction model was developed.MethodThis was a retrospective single-center study of elderly patients presenting with intertrochanteric fractures between our institution between January 2017 and December 2020. Patients' duplex ultrasound (DUS) or venography results were retrieved to evaluate whether they had a preoperative deep venous thrombosis (DVT) of bilateral extremities, whereby patients were dichotomized. Various variables of interest on demographics, comorbidities, injury and biomarkers were extracted and their relationship between DVT were investigated. Statistically significant variables tested in multivariate logistics regression analyses were used to develop a risk prediction model.ResultsThere were 855 patients eligible to be included in this study, and 105 were found to have preoperative DVT, with a prevalence rate of 12.3%. Ten factors were tested as significantly different and 2 marginally significant between DVT and non-DVT groups in the univariate analyses, but only 6 demonstrated the independent effect on DVT occurrence, including history of a VTE event (OR, 4.43; 95%CI, 2.04 to 9.62), time from injury to DVT screening (OR, 1.19; 95%CI, 1.13 to 1.25), BMI (OR, 1.11; 95%CI, 1.04–1.18), peripheral vascular disease (OR, 2.66; 95%CI, 1.10 to 6.40), reduced albumin (2.35; 95%CI, 1.48 to 3.71) and D-Dimer > 1.0 mg/L(OR, 1.90; 95%CI, 1.13 to 3.20). The DVT risk model showed an AUC of 0.780 (95%CI, 0.731 to 0.829), with a sensitivity of 0.667 and a specificity of 0.777.ConclusionDespite without a so high prevalence rate of DVT in a general population with intertrochanteric fracture, particular attention should be paid to those involved in the associated risk factors above. The risk prediction model exhibited the improved specificity, but its validity required further studies to verify.