collaborative, individual-level analysis compared longitudinal outcomes across the International Network of Chronic Kidney Disease (iNETCKD) cohorts.

Rates of chronic kidney disease (CKD) progression, end stage kidney disease (ESKD), all-cause mortality, and cardiovascular (CVD) events among individuals with CKD vary widely across countries. Well-characterized demographic, comorbidity, and laboratory markers captured for prospective cohorts may explain, in part, such differences. To investigate whether core characteristics of individuals with CKD explain differences in rates of outcomes, we conducted an individual-level analysis of eight studies that are part of iNET-CKD, an international network of CKD cohort studies. Overall, the rate of CKD progression was 40 events/1000 person-year (95% confidence interval 39 - 41), 28 (27 - 29) for ESKD, 41 (40 - 42) for death, and 29 (28 - 30) for CVD events. However, standardized rates were highly heterogeneous across studies (over 92.5%). Interactions by study group on the association between baseline characteristics and outcomes were then identified. For example, the adjusted hazard ratio for CKD progression was 0.44 (95% confidence interval 0.35 - 0.56) for women vs. men among the Japanese (CKD-JAC), while it was 0.66 (0.59 - 0.75) among the Uruguayan (NRHP). The adjusted hazard ratio for ESKD was 2.02 (95% CI 1.88 - 2.17) per 10 units lower baseline eGFR among Americans (CRIC), while it was 3.01 (2.57 - 3.53) among Canadians (CanPREDDICT) (significant interaction for comparisons across all studies). The risks of CKD progression, ESKD, death, and CVD vary across countries even after accounting for the distributions of age, sex, comorbidities, and laboratory markers. Thus, our findings support the need for a better understanding of specific factors in different populations that explain this variation.