Abstract: PO0760
Risk Score to Predict CKD Among Mexican Individuals with Diabetes Mellitus
Session Information
- Diabetic Kidney Disease: Clinical
November 04, 2021 | Location: On-Demand, Virtual Only
Abstract Time: 10:00 AM - 12:00 PM
Category: Diabetic Kidney Disease
- 602 Diabetic Kidney Disease: Clinical
Authors
- Raña, Alejandro, Harvard University T H Chan School of Public Health, Boston, Massachusetts, United States
- Lajous, Martín, Instituto Nacional de Salud Publica, Cuernavaca, Morelos, Mexico
- Denova, Edgar, Instituto Nacional de Salud Publica, Cuernavaca, Morelos, Mexico
- Chávez, Mildred Yazmin, Instituto de Seguridad y Servicios Sociales de los Trabajadores del Estado, Mexico City, Mexico City, Mexico
- López, Ruy, Centro Nacional de Programas Preventivos y Control de Enfermedades, Mexico City, Mexico City, Mexico
- Danaei, Goodarz, Harvard University T H Chan School of Public Health, Boston, Massachusetts, United States
Background
The two major causes of CKD are type 2 diabetes (T2D) & hypertension, which are responsible for up to two-thirds of the cases. More than half of patients in Mexico with incident ESRD have an underlying diagnosis of T2D. Some prediction models have been developed for the purposes of screening CKD & its progression. However, their generalizability to the Mexican population is not known, & few have been validated in different populations & rarely in LMIC. We aimed to develop & validate a lab and office-based risk prediction scores for CKD among Mexican patients with T2D
Methods
The prospective cohort consisted of 105,310 patients enrolled in the Integral Management of Diabetes by Stages program. 18,148 patients were randomly assigned to the training & testing sets on an 80-20 ratio. Logistic regression models were used to assess risk factors for CKD. A stepwise selection process was performed to determine the best predictive equations
Results
A total of 1,617 patients developed CKD (mean follow up of 1.1 years). Age, BMI, duration of T2D, treatment with insulin & oral hypoglycemics, treatment with nephroprotective agents, retinopathy & alcohol use were predictors in both models. Triglyceride levels & eGFR proved to be important predictors in the lab model. The lab score had a C statistic of 0.77 & a calibration slope of 1, whereas the office score had 0.67 & 0.89, respectively
Conclusion
These models can be used to identify individual patients with T2D who are at risk of developing CKD. This can facilitate early detection to intensify T2D treatment in a timely manner