C-reactive protein to albumin ratio outperforms Nutritional Risk Screening-2002 in predicting prolonged hospital stay in internal medicine patients

Aim: Malnutrition and inflammation are critical factors affecting the duration of hospitalization. Although the Nutritional Risk Screening 2002 (NRS-2002) is widely recommended for nutritional assessment, its predictive value for length of stay (LOS) remains controversial. The C-reactive protein to albumin ratio (CAR) has emerged as an objective inflammatory-nutritional marker. This study aimed to compare the prognostic performance of NRS-2002 and CAR in predicting prolonged hospitalization in an internal medicine cohort.Methods: This retrospective cohort study included 375 adult patients hospitalized in the internal medicine ward of a tertiary care hospital between April and October 2023. Prolonged LOS was defined as a hospitalization duration exceeding the median value (>9 days). The predictive performance of NRS-2002, albumin, CRP, and CAR was analyzed using univariate and multivariate binary logistic regression and receiver operating characteristic (ROC) curve analysis.Results: The median LOS was 9 days, and 48.8% of patients experienced prolonged hospitalization. In univariate analysis, CAR (8.5 [24.3] vs. 15 [36.9], p=0.004), albumin (3.3 [0.9] vs. 3.0 [0.8] g/dL, p=0.001) and CRP (25.9 [72.5] vs. 46.0 [104.8] mg/L, p=0.008) differed significantly between the normal and prolonged LOS groups. NRS-2002 scores did not differ between groups (3 [2] vs. 3 [2], p=0.617), and nutritional risk prevalence (NRS-2002 ≥3) was identical in both groups (68.3%). ROC analysis showed that CAR (area under the curve [AUC]=0.585) and albumin (AUC=0.604) had superior discriminative ability compared with NRS-2002 (AUC=0.515). In multivariate analysis, CAR remained an independent predictor (odds ratio [OR]=1.01, p=0.038), whereas NRS-2002 was not (p=0.598).Conclusion: CAR index outperforms the NRS-2002 screening tool in predicting prolonged hospital stay in internal medicine inpatients. While NRS-2002 remains essential for identifying patients in need of nutritional intervention, the integration of objective inflammatory-nutritional biomarkers such as CAR may improve risk stratification for hospitalization outcomes.