Introduction
Hospitalized advanced heart failure (HF) patients are at high risk for malnutrition and death. The Nutritional Risk Index (NRI) is a simple, well-validated tool for identifying patients at risk for nutrition-related complications. We hypothesized that in advanced HF patients from the ESCAPE (Evaluation Study of Congestive Heart Failure and Pulmonary Artery Catheterization Effectiveness) trial, the NRI would improve risk discrimination for 6-month all-cause mortality.
Methods
We analyzed the 160 ESCAPE index admission survivors with complete follow-up and NRI data, calculated as follows: NRI = (1.519 × discharge serum albumin, g/dL) + {41.7 × discharge weight (kg)/ideal body weight (IBW; kg)}; as in previous studies, if discharge weight > IBW, this ratio was set = 1. The previously developed ESCAPE mortality model includes age, 6-minute walk distance, CPR/mechanical ventilation, discharge beta-blocker prescription and diuretic dose, and discharge serum sodium, blood urea nitrogen, and B-type natriuretic peptide levels. We used Cox proportional hazards modeling for the outcome of 6-month all-cause mortality.
Results
30/160 patients died within 6 months of hospital discharge. The median NRI was 96 (IQR 91-102), reflecting mild-to-moderate nutritional risk. The NRI independently predicted 6-month mortality, with adjusted HR 0.60 (95% CI 0.39-0.93), p=.02) per 10 units, and increased Harrell's c index from 0.74 to 0.76 when added to the ESCAPE model. Body mass index and NRI at hospital admission did not predict 6-month mortality. The discharge NRI was most helpful in patients with high (≥ 20%) predicted mortality by the ESCAPE model, where observed 6-month mortality was 38% in patients with NRI < 100 and 14% in those with NRI >100 (p=0.04).
Conclusions
The NRI is a simple tool that can improve mortality risk stratification at hospital discharge in hospitalized patients with advanced HF.