Data from more than 5 million individuals has been used to develop an equation for predicting the risk of incident chronic kidney disease (CKD) in people with or without diabetes, according to a presentation at Kidney Week 2019, sponsored by the American Society of Nephrology.
In a paper published simultaneously online in, researchers reported the outcome of an individual-level data analysis of 34 multinational cohorts involving 5,222,711 individuals – including 781,627 with diabetes – from 28 countries as part of the .
“An equation for kidney failure risk may help improve care for patients with established CKD, but relatively little work has been performed to develop predictive tools to identify those at increased risk of developing CKD – defined by reduced eGFR [estimated glomerular filtration rate], despite the high lifetime risk of CKD – which is estimated to be 59.1% in the United States,” wrote, from the National Institute of Diabetes and Digestive and Kidney Diseases in Phoenix and colleagues.
Over a mean follow-up of 4 years, 15% of individuals without diabetes and 40% of individuals with diabetes developed incident chronic kidney disease, defined as an eGFR below 60 mL/min per 1.73m2.
The key risk factors were older age, female sex, black race, hypertension, history of cardiovascular disease, lower eGFR values, and higher urine albumin to creatinine ratio. Smoking was also significantly associated with reduced eGFR but only in cohorts without diabetes. In cohorts with diabetes, elevated hemoglobin A1c and the presence and type of diabetes medication were also significantly associated with reduced eGFR.
Using this information, the researchers developed a prediction model built from weighted-average hazard ratios and validated it in nine external validation cohorts of 18 study populations involving a total of 2,253,540 individuals. They found that in 16 of the 18 study populations, the slope of observed to predicted risk ranged from 0.80 to 1.25.
Moreover, in the cohorts without diabetes, the risk equations had a median C-statistic for the 5-year predicted probability of 0.845 (interquartile range, 0.789-0.890) and of 0.801 (IQR, 0.750-0.819) in the cohorts with diabetes, the investigators reported.
“Several models have been developed for estimating the risk of prevalent and incident CKD and end-stage kidney disease, but even those with good discriminative performance have not always performed well for cohorts of people outside the original derivation cohort,” the authors wrote. They argued that their model “demonstrated high discrimination and variable calibration in diverse populations.”
However, they stressed that further study was needed to determine if use of the equations would actually lead to improvements in clinical care and patient outcomes. In an accompanying editorial,, and , of the at the University of California, San Francisco, said the study and its focus on primary, rather than secondary, prevention of kidney disease is a critical step toward reducing the burden of that disease, especially given that an estimated 37 million people in the United States have chronic kidney disease.
It is also important, they added, that primary prevention of kidney disease is tailored to the individual patient’s risk because risk prediction and screening strategies are unlikely to improve outcomes if they are not paired with effective individualized interventions, such as lifestyle modification or management of blood pressure.
These risk equations could be more holistic by integrating the prediction of both elevated albuminuria and reduced eGFR because more than 40% of individuals with chronic kidney disease have increased albuminuria without reduced eGFR, they noted (JAMA. 2019 Nov 8.).
The study and CKD Prognosis Consortium were supported by the U.S. National Kidney Foundation and the National Institute of Diabetes and Digestive and Kidney Diseases. One author was supported by a grant from the German Research Foundation. Nine authors declared grants, consultancies, and other support from the private sector and research organizations. No other conflicts of interest were declared. Dr. Tummalapalli and Dr. Estrella reported no conflicts of interest.
SOURCE: Nelson R et al. JAMA. 2019 Nov 8. .