Failure risk equation overestimates progression to kidney failure in older CKD patients
Click Here to Manage Email Alerts
A retrospective cohort study in Kidney International Reports found the Kidney Failure Risk Equation overestimated progression from chronic kidney disease to kidney failure among older patients.
“Despite widespread external validation, concerns have been raised about the use of risk prediction models, such as the [Kidney Failure Risk Equation (KFRE)], where death prior to kidney failure is treated as a censoring event rather than a competing event,” Gregory L. Hundemer, MD, MPH, of the division of nephrology, Ottawa Hospital Research Institute and clinical epidemiology program at University of Ottawa, and colleagues wrote. “Not accounting for the competing risk of death has been shown to result in an overestimation of the risk of kidney failure, particularly over longer periods of follow-up.”
Thus, Hundemer and associates analyzed data from patients with advanced CKD admitted to the Ottawa Hospital Multi-Care Kidney Clinic between Jan. 1, 2010, and Dec. 31, 2020. The data of 1,701 and 1,078 patients (62% men; 74% white) were used for 2-year and 5-year KFRE analysis, respectively. For 5-year KFRE assessment, only patients first seen between 2010 and 2015 were included.
Patients younger than 60 years of age comprised 30% of the study population, while 24% of patients were aged 60 to 69 years, 27% were aged 70 to 79 years and 19% were aged 80 years or older.
The median baseline KFRE score was 41% for the 2-year cohort and 81% for the 5-year cohort, both of with decreased with age.
In the youngest age categories of both cohorts, there was no statistically significant difference between the predicted and actual risk of kidney failure. Patients in the 2-year cohort who were at least 80 years of age had an overestimated risk, with an absolute difference of 7.6% and relative difference of 22.8%. Patients in the 5-year cohort between 70 and 79 years of age and older than 80 years of age had an overestimated risk as well, with absolute differences of 10% and 24.7% and relative differences of 14.5% and 40.4%, respectively.
In the 5-year cohort, KFRE underestimated kidney failure risk in patients younger than 60 years of age.
Kaplan-Meier analysis showed an overestimation of cumulative incidence kidney failure when accounting for death as a censoring event rather than a competing risk. Hundemer and colleagues advised nephrologists to be aware of this KFRE limitation while planning advanced care for older patients with advanced CKD.
While risk overestimation may be preferable to underestimation, it still may impact patient decisions on invasive procedures and allocation of kidney failure resources in countries such as Canada, they wrote.
The study was strengthened by its large continuous sample, availability of KFRE scores for all patients throughout the study period and minimized loss of follow-up. It was limited by its single-center design and predominantly white population, higher incidence of diabetes mellitus than in previous study populations and more restricted sample size for 5-year analysis.
Collapse
Click Here to Manage Email Alerts