
Background The red blood cell distribution width-to-albumin ratio (RAR) is a potential biomarker of systemic inflammation and nutritional status, but its association with mortality and early red blood cell (RBC) transfusion in sepsis remains unclear. Methods This retrospective cohort study included adults with Sepsis-3 in the MIMIC-IV database. The primary endpoint was 28-day all-cause mortality, and the secondary endpoint was RBC transfusion within 24h of ICU admission. LASSO-selected predictors were used to develop multiple machine-learning models. Temporal sensitivity analyses used pre-transfusion measurements for transfusion and a 24-h landmark framework for mortality. Results Among 25,518 patients, RAR was associated with 28-day mortality (OR=1.246, 95% CI: 1.223–1.270) and 24-h RBC transfusion (OR=1.208, 95% CI: 1.187–1.230; both P<0.001). XGBoost showed the highest discrimination for mortality (AUC=0.763) and transfusion (AUC=0.766). SHAP analysis identified RAR as an important contributor to model predictions. Conclusions RAR is a readily available biomarker associated with 28-day mortality and early RBC transfusion in sepsis and may support outcome-specific risk assessment.
Purpose To evaluate contrast sensitivity (CS) and its association with macular microvascular parameters measured by optical coherence tomography angiography (OCTA) and clinical severity in Parkinson's disease (PD). Methods Twenty-seven patients with PD (PwP) and 60 healthy controls (HC) underwent macular OCTA and CS testing with the Pelli–Robson chart. Vessel density (VD) and perfusion were quantified in central, inner, outer, and total macular regions.Clinical severity was evaluated using the MDS-UPDRS and Hoehn and Yahr (H&Y) staging. Results PwP showed reduced macular VD and perfusion across all regions and lower CS than HC (1.39 vs 1.53 log units, p=0.014). CS correlated negatively with MDSUPDRS (r=−0.498, p=0.008) and H&Y stage (r=−0.399, p=0.039). Outer macular perfusion showed the strongest correlation with MDS-UPDRS (r=−0.534, p=0.004, q=0.015). Within PwP, CS positively correlated with outer VD, outer perfusion, and total perfusion. In multivariable analysis, outer perfusion (β=0.382, p=0.036) and age (β=−0.476, p=0.007) were predictors of CS. Conclusions Reduced CS in PD correlated with outer macular perfusion and clinical severity, suggesting that it may serve as a non-invasive measure of retinal neurovascular dysfunction.
Older adults with serious illness follow heterogeneous trajectories shaped by multimorbidity, frailty, functional dependence, and competing risks. Consequently, a diagnosis or a single mortality estimate rarely answers the clinically relevant question of what care is needed and when. This narrative review critically examines prognostic markers and tools relevant to palliative and end-of-life care in older adults. Functional status, particularly when assessed serially, provides useful prognostic information; however, the Palliative Performance Scale and Eastern Cooperative Oncology Group performance status require cautious interpretation in older people because geriatric syndromes may not be captured. Frailty (assessed, with the Clinical Frailty Scale or the Frail-VIG index), cognitive impairment, dependence, and nutritional or inflammatory abnormalities add information but should not be treated as stand-alone treatment thresholds. We distinguish three groups of instruments: long-horizon geriatric mortality indices, short-horizon palliative survival models, and tools that prompt recognition of palliative-care needs. The Surprise Question is an intentionally unidimensional trigger rather than a stand-alone prognostic test; NECPAL and SPICT structure needs identification, with NECPAL 4.0 adding prognostic staging; and the Gold Standards Framework is a care model, not a diagnostic instrument. Prognostic information is most useful when combined with comprehensive geriatric assessment, repeated evaluation, transparent communication, and goal-concordant decisions.
BACKGROUND Evidence on factors associated with depressive symptoms in older Mexican adults remains dispersed. OBJECTIVE To synthesize sociodemographic, psychosocial, and functional factors associated with depressive symptoms among community-dwelling Mexican adults aged ≥60 years. METHODS Following PRISMA 2020, seven databases were searched for observational studies using validated depression instruments. Random-effects meta-analyses were conducted for factors with at least three comparable estimates; others were synthesized narratively. RESULTS Nineteen studies were included (six in the meta-analysis). Functional dependence in activities of daily living showed the strongest, most robust association with depressive symptoms (OR=2.26, 95% CI 1.95-2.63; I2=0%). Low educational attainment was directionally positive but imprecise (OR=2.47, 95% CI 0.42-14.50; I2=70%), and multimorbidity was inconclusive due to high heterogeneity (OR=1.77, 95% CI 0.77-4.07; I2=92%). Loneliness, elder abuse, food insecurity, poor self-rated health, and low social support showed suggestive associations but could not be pooled. CONCLUSIONS Functional dependence is the most consistent correlate of depressive symptoms in this population, supporting integrated depression screening and functional assessment. Longitudinal designs are needed to clarify causal pathways.