• Center on Health Equity and Access
  • Clinical
  • Health Care Cost
  • Health Care Delivery
  • Insurance
  • Policy
  • Technology
  • Value-Based Care

Risk Prediction Model Effective in Determining CKD Risk in Patients With Diabetes


Patients with type 2 diabetes can determine their risk of chronic kidney disease (CKD) by using a risk prediction model.

A study published in Diabetes, Obesity and Metabolism found that a risk prediction model developed in a separate study was able to predict the risk of chronic kidney disease (CKD) in participants from Germany and Austria with type 2 diabetes.

The leading cause of end-stage kidney disease has been found to be CKD in patients with type 2 diabetes, with the prevalence of CKD stage 5 predicted to increase by 3.2% per year between 2012 and 2025. However, identification of patients at risk of kidney disease is usually late and only diagnosed when complications develop. This study aimed to assess whether a risk prediction model that considers 9 risk factors was able to effectively assess the risk of CKD in patients with type 2 diabetes.

Data were obtained from the German/Austrian Diabetes Prospective Follow-up database, which collects data for patients with diabetes from centers in Germany, Austria, Switzerland, and Luxembourg. Patients were included if they had type 2 diabetes, were aged 39 to 75 years, had an estimated glomerular filtration rate (eGFR) of greater than or equal to 60 mL/min/1.73 m2, and had normo-albuminuria.

The first physician-patient contact at the treating physician was deemed baseline. Physician diagnosis, medication use, and/or laboratory tests were used to determine type 2 diabetes status. Patients who had type 1 diabetes, other diabetes types, or had incomplete baseline data were excluded.

The 9 factors included in the risk prediction model were: age, body mass index (BMI), smoking status, diabetic retinopathy status, hemoglobin A1C (HbA1c), systolic blood pressure (SBP), high-density lipoprotein cholesterol, triglycerides, and urinary albumin-to-creatinine ratio.

There were 10,922 patients that were included in this study, with a mean age of 61 years and a population that was made up of 44.4% of women. A total of 16.3% of patients were smokers and 9.1% had diabetic retinopathy. All patients were non-abluminuric at baseline.

Patients in the study tended to be older (61 vs 55 years), had a shorter duration of diabetes (6 vs 7 years), higher BMI (31.7 vs 26.6 kg/m2), were less likely to be smokers (16.3% vs 41.3%), had lower HbA1c (6.9% vs 8.5%), and had lower prevalence of diabetic retinopathy (9.1% vs 24.7%) compared with the validation cohort.

A total of 4084 patients (37.4%) developed CKD after a median follow-up of 59 months compared with 6838 (62.6%) who had not. Patients who had been diagnosed with CKD while they had diabetes were older (63 vs 59 years), less often smokers (13.4% vs 17.9%), were more often female (47.0% vs 42.8%), and had longer diabetes duration (7 vs 6 years) compared with patients without CKD. Patients who developed CKD also had a lower eGFR at baseline (83.9 vs 92.4 mL/min/1.73 m2).

A higher total risk prediction score at baseline was found in patients who developed CKD (17.2 vs 15.7 points) compared with the validation group. The risk scores for age (mean 4.8 vs 4.0 points), SBP (mean 2.8 vs 2.5 points), and smoking (0.5 vs 0.7 points) were found to have substantial differences between those with and without CKD.

All patients were separated into 4 risk groups of low risk (score of less than 12), moderate risk (score from 12 to 15.5), high risk (score from 16 to 26.5), and very high risk (score from 27 to 37). A total of 15.0%, 20.1%, 27.7%, and 40.2% of patients in the 4 risk groups developed CKD, respectively. Increasing risk score also correlated with an increasing cumulative risk of incident CKD.

There were some limitations to this study. There was a potential discrepancy in estimating eGFR between this study and the validation cohort found in a separate study. The effect of drug treatment that was initiated between baseline and follow-up could not be described.

The researchers concluded that the risk prediction model for CKD in patients with type 2 diabetes was able to “achieve moderate discrimination and good calibration in a German/Austrian [type 2 diabetes] population.”


Kress S, Bramlage P, Holl RW, et al. Validation of a risk prediction model for early chronic kidney disease in patients with type 2 diabetes: data from the German/Austrian DPV registry. Diabetetes Obes Metab. Published online November 29, 2022. doi:10.1111/dom.14925

Related Videos
Chase D. Hendrickson, MD, MPH
Steven Coca, MD, MS, Icahn School of Medicine, Mount Sinai
Matthew Crowley, MD, MHS, associate professor of medicine, Duke University School of Medicine.
Susan Spratt, MD, senior medical director, Duke Population Health Management Office, associate professor of medicine, division of Endocrinology, Metabolism, and Nutrition,
Stephen Nicholls, MD, Monash University and Victorian Heart Hospital
Amal Agarwal, DO, MBA
Dr Robert Groves
Dr Robert Groves
Jeremy Wigginton, MD
Related Content
© 2023 MJH Life Sciences
All rights reserved.