Novel Stratification of Mortality Risk by Kidney Disease Stage.

ABSTRACT

BACKGROUND: Chronic kidney disease (CKD) is a common disorder with a variable clinical course and it is associated with increased mortality. The Intermountain Risk Score (IMRS) is an electronic risk calculator that utilizes complete blood count (CBC) and basic metabolic panel (BMP) values to predict mortality in various healthcare populations. We hypothesized that IMRS would predict mortality in patients with CKD even with adjustment for serum phosphate and urinary albumin. METHODS: Three thousand eight hundred seventy-two patients with CKD classes IIIA-V had IMRS calculated retrospectively and survival analysis was performed investigating 1- and 5-year mortality. Kaplan-Meier survival curves were generated for predefined IMRS groups of low, medium and high risk for CKD patients overall and by sex and CKD stage. Serum phosphate and urinary albumin/creatinine ratios were modeled in multivariate Cox-proportional hazard models. Receiver operator characteristic curves were used to determine c-statistics for mortality. RESULTS: For all patients with CKD, mortality was significantly greater for those with medium- or high-risk compared to low-risk IMRS categories, among each CKD stage. Overall, IMRS was predictive of mortality at both 1 and 5 years, even when adjusted for CKD stage and predicted mortality more accurately than CKD stage alone. Albuminuria was not independently associated with mortality and serum phosphate weakly predicted mortality. CONCLUSION: IMRS is a strong predictor of mortality in patients with CKD and is robustly complementary to CKD stage in refining risk prediction. Given the universal availability and low cost of the CBC and BMP, IMRS may be of a substantial value in CKD risk assessment and management.