BACKGROUND/AIMS: Dynamic changes in estimated glomerular filtration rate (eGFR) predict death among patients with chronic kidney disease (CKD). Whether variability in serial eGFR measurements is associated with risk of end stage renal disease (ESRD) has not been reported. METHODS: We retrospectively analyzed the risk of ESRD as a function of eGFR variability (defined as the absolute value of the difference between the obtained clinical eGFR value at a given time and the eGFR value estimated by the linear regression line at the same time point) among a cohort of patients with Stage 3 CKD. The study population was comprised of adult primary care patients enrolled at Geisinger Clinic between January 1, 2004 and December 31, 2006, with Stage 3 CKD and a minimum of 4 serum creatinine results during this 3-year window, and without history of solid-organ transplant or metastatic cancer. Cohort members were followed through March 31, 2011 for ESRD (identified through linkage with the USRDS dataset of ESRD, or first outpatient eGFR < 15 ml/min/1.73 m2). A multivariate Cox proportional hazard model (adjusted for demographic factors, co-morbid conditions, medications, hospital-associated acute kidney injury, proteinuria, kidney function, and serum albumin, among other factors) was developed to test the association of eGFR variability with ESRD. RESULTS: 4,219 patients met study criteria. Those with greater eGFR variability were more likely to have diabetes, cardiovascular disease, and better baseline kidney function than those with lesser variability. 193 (4.6%) of the overall cohort developed ESRD during a median follow-up of 3.8 years, while 596 (14.1%) died prior to study end without ESRD. Results of the multivariate-adjusted Cox proportional hazard model showed that eGFR variability is not associated with ESRD (HR 1.00 for the highest-variability quartile, relative to the lowest; 95% CI 0.66 - 1.51). CONCLUSION: eGFR variability does not predict ESRD among patients with Stage 3 CKD.
BACKGROUND/AIMS: Dynamic changes in estimated glomerular filtration rate (eGFR) predict death among patients with chronic kidney disease (CKD). Whether variability in serial eGFR measurements is associated with risk of end stage renal disease (ESRD) has not been reported. METHODS: We retrospectively analyzed the risk of ESRD as a function of eGFR variability (defined as the absolute value of the difference between the obtained clinical eGFR value at a given time and the eGFR value estimated by the linear regression line at the same time point) among a cohort of patients with Stage 3 CKD. The study population was comprised of adult primary care patients enrolled at Geisinger Clinic between January 1, 2004 and December 31, 2006, with Stage 3 CKD and a minimum of 4 serum creatinine results during this 3-year window, and without history of solid-organ transplant or metastatic cancer. Cohort members were followed through March 31, 2011 for ESRD (identified through linkage with the USRDS dataset of ESRD, or first outpatient eGFR < 15 ml/min/1.73 m2). A multivariate Cox proportional hazard model (adjusted for demographic factors, co-morbid conditions, medications, hospital-associated acute kidney injury, proteinuria, kidney function, and serum albumin, among other factors) was developed to test the association of eGFR variability with ESRD. RESULTS: 4,219 patients met study criteria. Those with greater eGFR variability were more likely to have diabetes, cardiovascular disease, and better baseline kidney function than those with lesser variability. 193 (4.6%) of the overall cohort developed ESRD during a median follow-up of 3.8 years, while 596 (14.1%) died prior to study end without ESRD. Results of the multivariate-adjusted Cox proportional hazard model showed that eGFR variability is not associated with ESRD (HR 1.00 for the highest-variability quartile, relative to the lowest; 95% CI 0.66 - 1.51). CONCLUSION: eGFR variability does not predict ESRD among patients with Stage 3 CKD.
Authors: Carlos E Palant; Lakhmir S Chawla; Charles Faselis; Ping Li; Thomas L Pallone; Paul L Kimmel; Richard L Amdur Journal: Am J Physiol Renal Physiol Date: 2016-05-18