Abstract
BACKGROUND: Frailty, characterized by diminished physiological reserve, and high mechanical power are well-established risk factors for mortality in critically ill patients. However, the potential synergistic impact of these two factors on clinical outcomes remains insufficiently explored. METHODS: Baseline data from critically ill patients were extracted from the MIMIC-IV and eICU-CRD databases. Frailty was evaluated using the Laboratory-based Frailty Index (FI-Lab), and mechanical ventilation intensity was quantified by mechanical power (MP). Generalized linear mixed models (GLMMs) and logistic regression analyses were employed to assess the associations between FI-Lab, MP, their interaction, and hospital mortality. RESULTS: A total of 17,881 critically ill patients were included-13,740 from MIMIC-IV and 4,141 from eICU. Both MP (MIMIC-IV: OR 1.126; 95% CI 1.093-1.160; eICU: OR 1.165; 95% CI 1.094-1.241) and FI-Lab (MIMIC-IV: OR 1.169; 95% CI 1.158-1.179; eICU: OR 1.240; 95% CI 1.215-1.266) showed significant associations with hospital mortality in both cohorts. The effect estimates for MP were consistent across databases (I(2) = 0%; Cochran's Q test, P = 0.34), whereas FI-Lab results demonstrated substantial heterogeneity (I(2) = 96.3% [89.9%-98.7%]; Cochran's Q test, P < 0.001). To account for this variability, a GLMM was used to pool estimates across datasets. Each 5 J/min increase in MP was associated with a 7.7% higher risk of hospital mortality (OR 1.077; 95% CI 1.045-1.109), while each 0.05-unit increase in FI-Lab corresponded to an 18% increased risk (OR 1.180; 95% CI 1.171-1.190). Importantly, a statistically significant interaction was observed between MP and FI-Lab (OR 1.011; 95% CI 1.009-1.013), indicating that their combined effect exceeds the sum of individual effects. CONCLUSIONS: FI-Lab and mechanical power are independently associated with hospital mortality, and their interaction further amplifies the risk. These findings suggest that integrating frailty assessment with mechanical power monitoring may enhance risk stratification in critically ill patients.