The respiratory geriatric severity index (RGSI): A multidimensional prognostic tool integrating geriatric syndromes to predict in-hospital mortality in older adults with respiratory diseases

Background and aims: Older adults hospitalized with respiratory diseases face high mortality risk, but existing prognostic models neglect geriatric vulnerability. We aimed to develop and internally validate the Respiratory Geriatric Severity Index (RGSI), integrating geriatric syndromes with clinical factors to predict in-hospital mortality. Methods: We retrospectively studied 728 patients aged ≥75 years admitted to a national respiratory referral center (2022–2024). Comprehensive geriatric assessment within 24 h included frailty, sarcopenia, disability, cognition, delirium, dysphagia, and nutritional measures. Independent mortality predictors were identified using multivariable logistic regression and weighted to construct the RGSI. Model discrimination and calibration were evaluated with bootstrap validation. Results: In-hospital mortality was 18.3% (133/728). Predictors included acute clinical factors (dyspnea, oxygen requirement) and geriatric syndromes (delirium, pressure ulcers, disability, frailty, sarcopenia, cognitive impairment, dysphagia, urinary incontinence, low handgrip strength, low BMI). The RGSI showed good discrimination (AUC 0.716, 95% CI 0.669–0.763) and calibration, stratifying mortality from 9.2% (low risk) to 38.1% (high risk). Conclusions: The RGSI incorporates multidimensional geriatric assessment into respiratory prognostication, demonstrating robust predictive performance. It offers a practical tool to support individualized care, timely palliative integration, and resource allocation for vulnerable older adults.