Data Source, Setting, and Study Population

We used a previously described provincial administrative health and laboratory data repository—the Alberta Kidney Disease Network. ²² This is an established computerized repository of health data across Alberta that includes over 4.5 million adults. ²² We created a cohort of adults (18 years of age and older) with one or more outpatient serum creatinine and proteinuria measurements between January 1, 2017 and December 31, 2017 in Alberta, Canada. We used measures of estimated glomerular filtration rate (eGFR) and proteinuria (when available) to determine kidney function. We then categorized patients into 6 risk categories (including patients treated with dialysis) based on the Kidney Disease Improving Global Outcomes (KDIGO) guidelines. ²³ This included: low risk (G1A1, and G2A1), moderate risk (G1A2, G2A2, and G3aA1), high risk (G1A3, G2A3, G3aA2, G3bA1), very high risk (G3aA3, G3bA2, G3bA3, G4A1, G4A2, G4A3, G5-NDA1, G5-NDA2, G5-NDA3, G4, G5-ND), dialysis (G5-DA1, G5-DA2, G5-DA3, G5-D), and high risk with unmeasured proteinuria (G3a and G3b). Individuals with CKD were included in the final cohort if they had: (a) an eGFR measurement ≥60 mL/min/1.73 m² with measured proteinuria or (b) a series of 2 or more eGFR measurements <60 mL/min/1.73 m² spanning 90 or more days (with or without a proteinuria measurement). The index eGFR was defined by the first eGFR measurement <60 mL/min/1.73 m². Dialysis dependence was identified from the provincial dialysis registry. ²⁴ We excluded patients with only one eGFR <60 mL/min/1.73 m² and those with only an eGFR ≥60 mL/min/1.73 m² and unmeasured proteinuria to reduce the risk of misclassifying them as low, moderate, or high-risk CKD. Those with a prior kidney transplant were also excluded.