Kidney Failure Risk Scores Show Good Accuracy Worldwide

Although a calibration factor is sometimes needed, equations for predicting kidney failure risk developed in Canada perform well in widely varying world populations, concludes a study in The Journal of the American Medical Association.

Kidney failure risk equations developed and validated in Canada were further validated in 31 cohorts participating in the Chronic Kidney Disease Prognosis Consortium. Those cohorts included more than 720,000 participants with stage 3 to 5 CKD from 30 countries, with data collected from 1982 through 2014. New pooled risk equations were developed to compare with the original risk equations for prediction of kidney failure (dialysis treatment or kidney transplant). Two calibration factors were developed to address regional variations in risk.

The analysis included nearly 24,000 cases of kidney failure developing over a median four-year follow-up. The original Canadian equations showed very high discrimination of patients who developed kidney failure, with C statistics of 0.90 at two years and 0.88 at five years. Discrimination was also excellent in subgroups defined by age, race, and diabetic status, and was not further improved with the use of the pooled equations.

The Canadian risk equations showed good calibration in North American populations, but overestimated risk in some cohorts from other continents. With use of a calibration factor that lowered baseline risk by 32.9 percent at two years and 16.5 percent at five years, calibration improved in most non-North American cohorts.

Kidney failure risk equations can play an important role in targeting high-risk CKD patients for optimized nephrology care. The new study suggests that risk equations developed in a Canadian population accurately predict two- and five-year probability of kidney failure in international cohorts with differing characteristics. A calibration factor improves performance in some non-North American populations [Tangri N, et al. Multinational assessment of accuracy of equations for predicting risk of kidney failure: a meta-analysis. JAMA 2016; 315:164–174].