The worldwide prevalence of chronic kidney disease (CKD) is on the rise. The Global Burden of Disease study estimated that nearly 697 million individuals worldwide had reduced estimated glomerular filtration rate (eGFR) or increased albuminuria in 2016, an increase of 70% since 1990. During the same time period, the years of life lost to CKD increased by 53%; CKD is the 16th most common cause of years of life lost. The aging of the population and the increase in the prevalence of diabetes, hypertension, and obesity are the primary factors associated with the increase in CKD prevalence.
Identification of patients at risk for CKD may prevent the adverse health outcomes associated with the disease; in addition, management of patients with CKD may be hindered by lack of CKD awareness. Robert G. Nelson, MD, PhD, and colleagues conducted a study utilizing data from multinational cohorts to develop and evaluate risk prediction equations for CKD defined by reduced eGFR. Results were reported online in JAMA [doi:10.1001/jama2019/17379].
The study included data on 5,222,711 individuals in 34 multinational cohorts from the CKD Prognosis Consortium from 28 countries. Within cohorts, eligible participants were ≥18 years of age with an eGFR of less than 60 mL/min/1.73 m2 at baseline. Inclusion criteria were no prior end-stage kidney disease and at least one serum creatinine value recorded during follow-up. Most cohorts used self-report to define race/ethnicity. The data were collected from 1970 through 2017.
Of the total cohort, 15.0% (n=781,627) had diabetes. Mean age in the population without diabetes was 54 years and 38% were women. Among the population with diabetes, mean age was 62 years and 13% were women. The low percentage of women was due primarily to the Veterans Administration cohort, which was 97% male.
During a mean follow-up of 4.2 years, in the 4,441,084 participants without diabetes, there were 660,856 incident cases (14.9%) with an eGFR less than 60 mL/min/1.73 m2; 56.7% (n=374,513) were confirmed by subsequent eGFR measurements. In the cohort with diabetes, the percentage of incident cases of reduced eGFR was 40.1% (n=313,646) during a mean follow-up of 3.9 years; 67.7% of those (n=212,246) were confirmed by subsequent eGFR measurements.
In cohorts with and without diabetes, there was a significant association between incident eGFR less than 60 mL/min/1.73 m2 and older age, female sex, black race, hypertension, history of cardiovascular disease, lower eGFR values, and higher urine albumin:creatinine ratio. In the cohorts without diabetes, there was also a significant association between smoking and an incident eGFR less than 60 mL/min/1.73 m2. In cohorts with diabetes, elevated hemoglobin A1c and the presence and type of diabetes medication were significantly associated with an incident eGFR of less than 60 mL/min/1.73 m2.
In the cohorts without diabetes, the median C statistic for the 5-year predicted probability of all eGFR events of less than 60 mL/min/1.73 m2 was .0845; in the cohorts with diabetes, it was 0.801, reflecting good discrimination. For confirmed eGFR events of less than 60 mL/min/1.73 m2, the median C statistic was 0.869 in the cohorts without diabetes and 0.808 in the cohorts with diabetes.
Model calibration was assessed visually by plotting observed versus predicted risk per decile of predicted risk at 5 years in the cohorts with frequent measures of creatinine. In calibration analyses, nine of 13 study populations (69%) had a slope of observed to predicted risk between 0.80 and 1.25. Calibration was generally better for the eGFR of less than 60 mL/min/1.73 m2 end point than for the lower eGFR end points, where calibration was poor in some cohorts.
Discrimination was similar in 18 study populations in nine external validation cohorts (n=2,253,540); calibration showed that 16 of 18 (89%) had a slope of observed to predicted risk between 0.80 and 1.25.
The researchers cited some limitations to the study, including the absence of data on albuminuria in most cohorts of patients without diabetes, necessitating that a statistical patch derived from cohorts without diabetes, but with albuminuria data, be applied to the remaining cohorts in order to estimate how including albuminuria altered the models. The risk equations developed incorporated routinely collected demographic, clinical, and laboratory data, but not genotype data or newly identified biomarkers of early CKD. The equations developed were intended to identify individuals at increased risk of an intermediate health outcome; the equations did not identify risk of progression of CKD, cardiovascular events, or death. Calibration varied across setting, with particularly poor performance in some of the cohorts.
In conclusion, the researchers said, “Equations for predicting risk of incident chronic kidney disease were developed from more than five million individuals from 34 multinational cohorts and demonstrated high discrimination and variable calibration in diverse populations. Further study is needed to determine whether use of these equations to identify individuals at risk of developing chronic kidney disease will improve clinical care and patient outcomes.”
Takeaway Points
- Researchers conducted a study of data from multinational cohorts to develop and evaluate risk prediction equations for chronic kidney disease defined by reduced estimated glomerular filtration rate (eGFR).
- Equations for 5-year risk of reduced eGFR included age, sex, race/ethnicity, eGFR, history of cardiovascular disease, ever smoker, hypertension, body mass index, and albuminuria concentration. For patients with diabetes, the model also includes diabetes medication and hemoglobin A1c.
- The developed equations demonstrated high discrimination and variable calibration in diverse populations.
