Preoperative lung ultrasound score and multifactorial predictors of 1-year mortality in geriatric hip fracture patients with pneumonia: a retrospective cohort study
Preoperative pneumonia severity independently predicts postoperative mortality in elderly hip fracture patients, yet accurate assessment remains challenging. This study evaluated lung ultrasound score (LUS) for predicting 1-year mortality and compared its predictive performance with traditional risk scores. We retrospectively reviewed 839 elderly (≥65 years) hip fracture patients treated between August 2019 and August 2023. After applying inclusion/exclusion criteria, 92 patients with preoperative pneumonia were enrolled. Pneumonia was diagnosed based on clinical symptoms and chest X-ray findings, with severity assessed using the CURB-65 scoring system. All patients underwent lung ultrasound to evaluate aeration patterns across six regions of each lung, with scores assigned based on the degree of aeration loss (0–3 per region). Baseline data, including demographic characteristics, ASA scores, laboratory results, and medical history, were collected. Surgical procedures included hemiarthroplasty, total hip arthroplasty, or intramedullary nail fixation, depending on fracture type. Cox regression and ROC curve analyses were performed to evaluate the association between candidate risk factors and 1-year mortality, and to compare the diagnostic efficacy of LUS, ASA, and CURB-65. Of 92 enrolled patients, 26 (28.3%) died within one year. ASA and CURB-65 scores were significantly higher in the mortality group ( P =0.027 and P =0.008, respectively). Postoperative acute respiratory failure was significantly more frequent in the mortality group (50% vs. 13.6%, P <0.001). LUS was the strongest predictor of 1-year mortality (AUC=0.86, 95% CI: 0.79-0.94), outperforming CURB-65 (AUC=0.67) and ASA (AUC=0.63). The optimal LUS cut-off was 13.5 (sensitivity=1.000, specificity=0.667). Binary logistic regression confirmed LUS≥14 as the only independent predictor of 1-year mortality (OR=108.45, 95% CI: 3.30-3567.00, p =0.009). LUS outperforms ASA and CURB-65 in predicting 1-year mortality in geriatric hip fracture patients with preoperative pneumonia, though combining LUS with clinical scores optimizes prognostic evaluation.