Variations in Incidence and Prognosis of Stage 4 CKD Among Adults Identified Using Different Algorithms: A Population-Based Cohort Study

Abstract

Rationale & ObjectiveClinical guidelines define chronic kidney disease (CKD) as abnormalities of kidney structure or function for >3 months. Assessment of the duration criterion may be implemented in different ways, potentially impacting estimates of disease incidence or prevalence in the population, individual diagnosis, and treatment decisions, especially for more severe cases. We investigated differences in incidence and prognosis of CKD G4 identified by one of four algorithms. Study DesignPopulation-based cohort study in Alberta, Canada. Setting & ParticipantsResidents >18 years old with incident CKD G4 (eGFR 15-29 mL/min/1.73 m2) diagnosed between April 1, 2015 and March 31, 2018, based on administrative and laboratory data. ExposuresFour outpatient eGFR-based algorithms, increasing in stringency, for defining cohorts with CKD G4 were evaluated: (1) a single test, (2) first eGFR <30 mL/min/1.73 m2 and a second eGFR 15-29 mL/min/1.73 m2 >90 days apart (two tests), (3) ≥2 eGFR measurements <30 mL/min/1.73 m2 sustained for >90 days (qualifying period) and the last eGFR in the qualifying period of 15-29 mL/min/1.73 m2 (relaxed sustained), (4) ≥2 consecutive measurements 15-29 mL/min/1.73 m2 for >90 days (rigorous sustained). OutcomesTime to the earliest of death, eGFR improvement (a sustained increase in eGFR to ≥30 mL/min/1.73 m2 for >90 days and >25% increase from the index eGFR), or kidney failure. Analytical ApproachFor each of the 4 cohorts, incidence rates and event-specific cumulative incidence functions at 1 year from cohort entry were estimated. ResultsThe incidence rates of CKD G4 decreased as algorithms became more stringent, from 190.7 (single test) to 79.9 (rigorous sustained) per 100,000 person-years. The two cohorts based on sustained reductions in eGFR were of comparable size and 1-year event-specific probabilities. The two cohorts based on a single test and a two-test sequence were larger and experienced higher probabilities of eGFR improvement. LimitationsA short follow-up period of 1 year and a predominantly White population. ConclusionsThe use of more stringent algorithms for defining CKD G4 results in substantially lower estimates of disease incidence, the identification of a group with a lower probability of eGFR improvement, and a higher risk of kidney failure. These findings can inform implementation decisions of disease definitions in clinical reporting systems and research studies.