Establishment and verification of postoperative 1-year mortality risk prediction model in elderly patients undergoing hip fracture surgery / 中华麻醉学杂志

ABSTRACT

Objective:To establish a prediction model of postoperative 1-year mortality risk in elderly patients undergoing hip fracture surgery and verify its efficacy.Methods:Patients of both sexes, aged ≥65 yr, of American Society of Anesthesiologists physical status Ⅰ-Ⅳ, who underwent an operation for traumatic hip fracture in the Second Affiliated Hospital of Wenzhou Medical University from January 2017 to December 2018, were enrolled and randomly assigned to model group and verification group in a ratio of 3∶1.The demographic characteristics, clinical data and results such as laboratory examinations were collected.In model group, the logistic regression analysis was used to recognize the independent risk factors for 1-year mortality after procedure, and the prediction model was established.In verification group, the prediction efficacy was analyzed using the receiver operating characteristic curve, and the degree of fitting was evaluated by Hosmer-Lemeshow goodness-of-fit test.Results:Multivariate logistic analysis indicated that age ≥84 yr, Charlson comorbidity index ≥2 points, Braden score on admission to hospital ≤16 points, preoperative urea nitrogen ≥8.8 mmol/L and postoperative albumin ≤ 29.6 g/L were the independent risk factors for 1-year mortality after hip fracture surgery in elderly patients ( P<0.05). The prediction model was established based on the risk factors mentioned above.The area under receiver operating characteristic curve was 0.870, and the sensitivity and specificity were 82.8% and 80.0%, respectively.The prediction model showed good fitting ( χ2=4.672, P=0.700). Conclusion:Age ≥84 yr, Charlson comorbidity index ≥2 points, Braden score on admission to hospital≤16 points, preoperative urea nitrogen ≥8.8 mmol/L and postoperative albumin ≤ 29.6 g/L are the independent risk factors for 1-year mortality after hip fracture surgery in elderly patients, and the prediction model established based on the above indicators has good efficacy.