C-Reactive Protein/Albumin Ratio vs. Prognostic Nutritional Index as the Best Predictor of Early Mortality in Hospitalized Elderly Patients, Regardless of Admitting Diagnosis

Background: Among elderly hospitalized individuals, factors like malnutrition and systemic inflammation significantly influence clinical outcomes such as length of hos-pital stay (LOS), mortality, and readmission risk. The C-reactive protein to albumin (CRP/Alb) ratio serves as a valuable prognostic indicator, reflecting both inflammatory and nutritional status. Additionally, the Prognostic Nutritional Index (PNI) offers a straightforward method to assess both the nutritional state and mortality risk in older patients. Objective: This study aimed to evaluate the prognostic utility of both the PNI and CRP/Alb ratio in forecasting hospital mortality at 7- and 30-days post-admission. We also sought to compare their predictive performance within a large cohort of elderly patients, irrespective of their admitting diagnosis. Methods: We conducted a retro-spective observational study, analyzing data from 2776 patients, aged 65 years or older, who were admitted to the Internal Medicine and Aging Department at the "Policlinico Riuniti" University Hospital in Foggia, Italy, between 2019 and 2025. Upon admission, serum C-reactive protein (CRP), albumin levels, and Total Lymphocyte Count (TLC) were recorded. To determine the predictive capacity of the CRP/Alb ratio and PNI for in-hospital mortality at 7 and 30 days, we performed ROC curve analysis, Cox regres-sion, and Kaplan–Meier survival analyses. Furthermore, the Positive Predictive Value (PPV) for mortality was calculated for both the CRP/Alb ratio and PNI. Likelihood-ratio test was also performed to compare the mortality PPVs of mortality of the CRP/Alb ratio and the PNI, at 30 and 7 days, both for all patients and for re-hospitalized patients. Results: In-hospital mortality occurred in 444 patients (16%). Our findings indicated that deceased patients exhibited notably higher CRP/Alb ratios and lower PNI values compared to survivors (p < 0.001). Both the CRP/Alb ratio and PNI demonstrated a ro-bust correlation with early mortality, particularly within the initial 7 days (AUC = 0.89 and AUC = 0.87, respectively). A CRP/Alb ratio exceeding 8 independently predicted 30-day mortality (HR = 3.07; 95% CI: 2.33–4.03) and 7-day mortality (HR = 10.46; 95% CI: 6.32–17.32). Similarly, a PNI below 38 was an independent predictor for 30-day mortality (HR = 3.54; 95% CIs = 2.70–4.64) and 7-day mortality (HR = 8.84; 95% CIs = 5.67–13.77). When comparing the PPVs, PNI emerged as a more refined mortality predictor than the CRP/Alb ratio among non-rehospitalized patients, regardless of LOS (p < 0.001). For re-hospitalized patients, PNI was a better predictor only for those with a LOS under 7 days, whereas the CRP/Alb ratio was superior for those with a LOS of 7 days or more. Conclusion: Among non-rehospitalized patients, PNI offers a more precise mortality prediction, irrespective of the length of stay. Among re-hospitalized patients, the pre-dictive superiority of PNI or CRP/Alb ratio varies with the length of stay. Anyway, both CRP/Alb ratio and PNI are simple, cost-effective biomarker that may support early risk stratification and guide targeted interventions in geriatric care.