Abstract
The study aimed to assess the predictive significance of various inflammatory biomarkers among critically ill elderly patients with osteoporotic hip fracture (OHF). The research identified critically ill elderly patients with OHF through the MIMIC-IV database. Nine systemic inflammatory indicators, formed by diverse combinations of neutrophils, lymphocytes, monocytes, and platelets, were evaluated for their associations with 30-day all-cause mortality after ICU admission. An examination of the receiver operating characteristic curve was conducted to determine the strongest predictive biomarker. In addition, the least absolute shrinkage and selection operator (LASSO) regression was applied to identify potential influencing factors. Furthermore, Cox proportional hazards regression and restricted cubic spline analyses were further applied to evaluate the association between the optimal biomarker and time-to-event outcomes. Seven inflammatory markers demonstrated significant predictive value for 30-day all-cause mortality, among which the platelet-to-lymphocyte ratio (PLR) exhibited the largest area under the curve (AUC = 0.783). Subsequently, individuals were categorized into quartiles according to PLR. Multivariate Cox proportional hazards models demonstrated a significant association between elevated PLR levels and 30-day all-cause mortality. After adjustment for variables selected by LASSO regression, elevated PLR remained an independent prognostic factor for mortality (adjusted hazard ratio, 1.43; 95% confidence interval, 1.24-1.65; P < 0.001). Additionally, restricted cubic spline analysis further demonstrated a consistently increasing risk of all-cause mortality with rising PLR levels. This study demonstrates that PLR is an independent predictor of 30-day all-cause mortality in critically ill elderly patients with osteoporotic hip fracture. Patients with a PLR ≥ 188.64 exhibit a significantly increased non-linear risk of 30-day all-cause mortality. Furthermore, patients with a PLR ≥ 302.11 can be classified as high-risk individuals. PLR represents an inexpensive and readily accessible tool for rapidly identifying high-risk patients and optimizing individualized treatment strategies in ICU clinical practice.