FPG Score: A Rapid Admission-Based Tool for Predicting In-Hospital Mortality in Elderly Hip Fracture Patients

Hip fractures in elderly patients are a major public health concern, associated with high morbidity and mortality. Early identification of high-risk patients is crucial to guide clinical decision-making, optimize resource allocation, and improve outcomes. However, existing risk prediction models, such as the Nottingham Hip Fracture Score (NHFS) and the Charlson Comorbidity Index (CCI), require laboratory or postoperative data, delaying risk stratification. This study aims to develop and validate the FPG score, a novel and simplified tool for predicting intrahospital mortality in elderly patients undergoing surgery for proximal femur fractures, using only admission data available at triage.

This single-center, observational cohort study was conducted in two phases: a retrospective derivation phase (2015-2019) and a prospective validation phase (2020-2022). Patients aged ≥ 65 years with proximal femur fractures (AO 31A, 31B) undergoing surgical treatment were included. Exclusions involved pathological, periprosthetic, and femoral head fractures (31C). Data on demographics, comorbidities, vital signs, and laboratory values were collected at Emergency Unit triage. The primary outcome was intrahospital mortality. Univariate and multivariate logistic regression identified predictors, and ROC analysis assessed the FPG score's predictive performance, with AUC, sensitivity, and specificity evaluated using SPSS v25 and MedCalc v18.

In the retrospective phase, 1984 patients (median age: 83.5 years, 28.7% male) were analyzed, with an observed intrahospital mortality of 3.8% (77 patients). The FPG score demonstrated an AUC of 0.79, outperforming NHFS and CCI. A score > 2 was associated with a > 50% mortality risk, with 61% sensitivity and 80% specificity. In the validation cohort (752 patients, 4.8% mortality), the FPG score maintained strong predictive performance (AUC = 0.751).

The FPG score provides a rapid, objective, and clinically applicable tool for mortality risk assessment in elderly patients with hip fractures, allowing for immediate triage-based decision-making. Unlike NHFS and CCI, it does not require laboratory or post-admission data, making it particularly useful in emergency settings. Its integration into clinical practice may enhance patient management, improve resource allocation, and facilitate early intervention. While the score has been validated in a single-center study, further multicenter validation is needed to confirm its broader applicability. Future research should explore the integration of frailty indices and laboratory markers to refine its predictive accuracy.