Time Series AI Model for Acute Kidney Injury Detection Based on a Multicenter Distributed Research Network: Development and Verification Study

Background: Acute kidney injury (AKI) is a marker of clinical deterioration and renal toxicity. While there are many studies offering prediction models for the early detection of AKI, those predicting AKI occurrence using distributed research network (DRN)-based time series data are rare. Objective: In this study, we aimed to detect the early occurrence of AKI by applying the interpretable LSTM-based model on a hospital EHR-based time series in patients who took nephrotoxic drugs using a DRN Methods: We conducted a multi-institutional retrospective cohort study of data from six hospitals using a DRN. For each institution, a patient-based dataset was constructed using five drugs for AKI, and the interpretable multi-variable long short-term memory (IMV-LSTM) model was used for training. This study employed propensity score matching to mitigate differences in demographics and clinical characteristics. Additionally, the temporal attention values of the AKI prediction model's contribution variables were demonstrated for each institution and drug, with differences in highly important feature distributions between the case and control data confirmed using one-way analysis of variance. Results: This study analyzed 8,643 and 31,012 patients with and without AKI, respectively, across six hospitals. When analyzing the distribution of AKI onset, vancomycin showed an earlier onset (median: 12 days), and acyclovir was the slowest compared to the other drugs (median: 23 days). Our temporal deep learning model for AKI prediction performed well for most drugs. Acyclovir had the highest average area under the receiver operating characteristic curve score per drug (0.94), followed by acetaminophen (0.93), vancomycin (0.92), naproxen (0.90), and celecoxib (0.89). Based on the temporal attention values of the variables in the AKI prediction model, verified lymphocytes and calcium had the highest attention, whereas lymphocytes, albumin, and hemoglobin tended to decrease over time, and urine pH and prothrombin time tended to increase. Conclusions: Early surveillance of AKI outbreaks can be achieved by applying the IMV-LSTM based on time series data through hospital electronic health records (EHR)-based DRNs. This approach can help identify risk factors and enable early detection of adverse drug reactions when prescribing drugs that cause renal toxicity before AKI occurs.