New risk prediction models have been developed that may make it easier to identify individuals at increased risk for chronic kidney disease (CKD), investigators reported at the American Society of Nephrology’s Kidney Week 2019, November 5-10, Washington, DC.
The models, developed based on analysis of >5 million individuals worldwide, facilitated prediction of the 5-year probability of developing estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m2 in adults across age ranges and ethnicities. The study was led by Josef Coresh, MD, PhD, Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland.
Although further study is needed to determine whether use of the equations to identify patients would lead to improved outcomes, early identification of persons at increased CKD risk could improve clinical care by facilitating surveillance and improved management of underlying health conditions, said authors in the full study published in JAMA to coincide with presentation of the results at the meeting.
To develop the equations, researchers performed an individual-level analysis of data based on 34 multinational cohorts in the CKD Prognosis Consortium from 1970 through 2017, including >5.2 million individuals in 28 countries. They sought to identify individuals at increased 5-year risk of CKD, defined as an eGFR <60 mL/min/1.73 m.2
Clinical variables were often differentially available by diabetes status, so separate models were developed for participants with and without diabetes.
Modeling was based on a range of readily available factors including age, race, sex, body mass index, eGFR levels, albuminuria concentration, cardiovascular disease, high blood pressure, and smoking status. For diabetes patients, it was based on diabetes, A1c, and diabetes medications.
The resulting equations demonstrated high discrimination with a median C statistic for 5-year predicted probability of eGFR of 0.801 for cohorts with diabetes and 0.845 in those without diabetes. Calibration was variable, with a slope of observed-to-predicted risk between 0.80 and 1.25 seen in nearly 70% of study populations.
In a related editorial also published in JAMA, Sri Lekha Tummalapalli, MD, MBA, and Michelle M. Estrella, MD, MHS, both from the Kidney Health Research Collaborative, University of California, San Francisco, said that with validated equations for risk prediction now available, the challenge will be to evaluate approaches that individualize care based on calculated risk.
“An accurate tool that identifies persons at elevated risk of developing CKD may facilitate efforts to modify relevant risk factors and could be used to determine the frequency of kidney health monitoring among persons at risk of CKD,” said Tumalpalli and Estrella.
Reference: Nelson RG, Grams ME, Ballew SH, et al. Development of risk prediction equations for incident chronic kidney disease. JAMA. 2019 Nov 8. doi: 10.1001/jama.2019.17379.