Deep learning based electrocardiographic screening for chronic kidney disease

Background Undiagnosed chronic kidney disease (CKD) is a common and usually asymptomatic disorder that causes a high burden of morbidity and early mortality worldwide. We developed a deep learning model for CKD screening from routinely acquired ECGs. Methods We collected data from a primary cohort with 111,370 patients which had 247,655 ECGs between 2005 and 2019. Using this data, we developed, trained, validated, and tested a deep learning model to predict whether an ECG was taken within one year of the patient receiving a CKD diagnosis. The model was additionally validated using an external cohort from another healthcare system which had 312,145 patients with 896,620 ECGs from between 2005 and 2018. Results Using 12-lead ECG waveforms, our deep learning algorithm achieved discrimination for CKD of any stage with an AUC of 0.77 (95% CI 0.76-0.77) in a held-out test set and an AUC of 0.71 (0.71-0.71) in the external cohort. Our 12-lead ECG-based model performance was consistent across the severity of CKD, with an AUC of 0.75 (0.0.74-0.77) for mild CKD, AUC of 0.76 (0.75-0.77) for moderate-severe CKD, and an AUC of 0.78 (0.77-0.79) for ESRD. In our internal health system with 1-lead ECG waveform data, our model achieved an AUC of 0.74 (0.74-0.75) in detecting any stage CKD. In the external cohort, our 1-lead ECG-based model achieved an AUC of 0.70 (0.70-0.70). In patients under 60 years old, our model achieved high performance in detecting any stage CKD with both 12-lead (AUC 0.84 [0.84-0.85]) and 1-lead ECG waveform (0.82 [0.81-0.83]). Conclusions Our deep learning algorithm was able to detect CKD using ECG waveforms, with particularly strong performance in younger patients and patients with more severe stages of CKD. Given the high global burden of undiagnosed CKD, further studies are warranted to evaluate the clinical utility of ECG-based CKD screening.