Abstract Background and Aims nutritional status clearly has a great impact on the prognosis of maintenance hemodialysis patients. Therefore, its management should be a priority and risk screening frequent and easily implemented, based on the biochemical and clinical routine parameters already available. Many tools fit these simple criteria, namely simple Protein Energy Wasting score (sPEW), Geriatric Nutritional Risk Index (GNRI) and Creatinine Index (Cr Index). These scores are associated with a high mortality and morbidity risk in hemodialysis (HD) patients. The objective of this study was to assess the performance of these tools regarding the estimation of all-cause mortality, in a 45-months follow-up of a large patient cohort. Method Historical cohort study of HD pts from 25 outpatient clinics. sPEW, GNRI and Cr Index were estimated. Kaplan-Meier estimator and univariable Cox regression models to analyze time until death were used. To compare survival curves the log-rank test or Tarone test were used, as appropriate. The level of significance α = .05 was considered. All data were analyzed using SPSS 22.0 (IBM Corp. Released 2013. IBM SPSS Statistics for Windows. Armonk, NY, USA: IBM Corp). Results We analyzed 2322 pts, 59% males, 31.7% diabetic, with a median age of 70 years (P25 = 60, P75 = 79) followed up for a maximum of 45-month (P25 = 31; P75 = 45). All-cause mortality was observed in 778 pts (33.5%). To assess the mortality risk, the exposures GNRI and CR Index, were discretized using quartiles. GNRI The median was 106.6 (P25 = 99.4, P75 = 114.2). The log-rank test results showed a significantly lower survival for patients in GNRI Q1 category (GNRI ≤ 99.4). A p-value <0.001 was obtained when comparing patients in Q4, Q3 and Q2 with patients in Q1. The univariable Cox regression model showed that patients in Q1 had a 2-fold increased risk of death, when compared with Q4 (HR = 2.1, 95% CI: 1.7-2.6, p<0.001). Creatinine Index The median was 12.648 (P25 = 11.908, P75 = 13.406). The log-rank test for the equality of survival functions for the different levels was considered statistically significant and showed a significantly lower survival for patients in CR Index Q1 category, when compared with the remaining categories (p<0.001). The univariable Cox regression model showed that comparatively with Q4, patients in Q1 had a 5-fold increased risk of death (HR = 4.8, 95% CI: 3.7-6.0, p<0.001), patients in Q2 a 3-fold increased risk of death (HR = 3.1, 95% CI:2.4– 3.9, p<0.001), and patients in Q3 a 2-fold increased risk of death (HR = 1.9, 95% CI: 1.4 – 2.4, p<0.001). sPEW The frequency of each score from 0 to 4 was 306, 380, 111, 1369 and 156, respectively. The log-rank test for the equality of survival functions corresponding to the different levels showed a significantly higher survival for patients with a sPEW score of 4 when comparing with the remaining levels (p<0.001). The univariable Cox regression model showed that comparatively with a score of 4, patients with a: score of 0 had an 8-fold increased risk of death (HR = 7.7, 95% CI: 4.5-13.3, p<0.001), score of 1 a 6-fold increase risk of death (HR = 5.6, 95% CI: 3.3– 9.7, p<0.001), score of 2 a 4-fold increase risk of death (HR = 4.1, 95% CI: 2.2– 7.7, p<0.001), and a score of 3, almost a 4-fold increased risk of death (HR = 3.7, 95% CI: 2.2– 6.3, p<0.001). Conclusion In this exploratory analysis, the three tools showed a significant association with mortality during follow-up. These tools, if adequately validated in future studies, may select patients for further intervention to modify the outcome.
Read full abstract