OBJECTIVES: Frailty is frequently present in patients admitted to ICU and associated with worse outcomes. Standard severity-of-illness scores (SOISs) do not account for frailty. We investigated the impact of frailty burden (proportion of frail patients per ICU) on performance indicators estimated by Simplified Acute Physiology Score (SAPS) 3 and Mortality Probability Model (MPM)0-III. DESIGN: Retrospective cohort study using prospectively collected data. SETTING: One hundred fifty-nine ICUs in Brazil and Uruguay. PATIENTS: Two hundred forty-two thousand one hundred forty-one patients admitted in 2022-2023. Frailty was defined as a Modified Frailty Index (MFI) of 3 or higher. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Outcomes were hospital mortality and ICU length of stay. Performance indicators were standardized mortality rate (SMR) and standardized resource use (SRU). SAPS 3 and MPM0-III were recalibrated, and SRU parameters reestimated. Associations between frailty burden and SMR or SRU were examined with scatter plots, Spearman correlations, and efficiency matrices. The median frailty burden was 23.0% (interquartile range IQR, 16.8-31.7%); median hospital mortality was 14.8% (IQR, 9.1-25.2%). Although higher frailty burden ICUs showed greater mortality, no consistent trend emerged. Frailty burden was not significantly correlated with performance indicators: MPM0-III (SMR: r = -0.123 95% CI, -0.262 to 0.035 and SRU: r = -0.091 95% CI, -0.252 to 0.081) and SAPS 3 (SMR: r = -0.105 95% CI, -0.258 to 0.044 and SRU: r = -0.074 95% CI, -0.236 to 0.096). Recalibrating models with the addition of MFI did not improve estimates. These results were consistent in practically all sensitivity analyses. ICUs with varying frailty burdens were evenly distributed across efficiency matrix quadrants. CONCLUSIONS: Frailty burden had no significant impact on ICU benchmarking with SAPS 3 or MPM0-III, suggesting limited relevance for performance comparisons.
Bastos et al. (Mon,) studied this question.