Comparison of renal replacement therapy and mortality using 1 versus 2 estimated glomerular filtration rates

Patients with chronic kidney disease (CKD) often progress to end-stage renal disease or die. Population managers may benefit from using the National Kidney Foundation's suggestion of measuring 2 estimated glomerular filtration rates (eGFRs) more than 90 days apart to determine whether a patient has CKD. This study examines the utility of using a single eGFR vs. 2 eGFRs for identifying CKD patients who are more likely to benefit from disease management programs. We conducted a retrospective study within a large administrative dataset. Patients with no history of CKD and a single eGFR <60 mL/min per 1.73 m(2) in a single year were compared to those with a second eGFR and followed for 4 years. A Cox regression model was created to examine the impact of common risk factors on renal replacement therapy (RRT) and mortality. Patients with 1 eGFR were less likely to require RRT (0.1% for 1 eGFR vs. 1.4% for 2 eGFRs, P value <0.0001) or to die (14.9% vs. 21.7%, P value <0.0001). The adjusted hazard ratio (2 eGFR vs. 1 eGFRs) among patients with multiple risk factors (ie, diabetes, hypertension, proteinuria) was 2.54 (95% confidence interval [CI] 0.54-11.98) for RRT and 8.92 (95% CI 1.05-75.63) for death. The adjusted hazard ratio among patients with no additional risk factors was 2.02 (95% CI 0.21-19.57) for RRT and 1.97 (95% CI 1.41-2.75) for death. The hazard ratio among patients with any risk factor for RRT or death was 4.90 (95% CI 1.48-16.26) and 1.25 (95%CI 0.97-1.61), respectively. The addition of a second eGFR to identify which patients with CKD are likely to require RRT or die decreases the overall number of patients identified while increasing the proportion who subsequently require RRT or die. The increased risk of RRT was particularly evident among patients with risk factors.