In this study, the investigators aim to develop a risk prediction model for acute kidney injury (AKI) in hospitalized patients using the calcineurin inhibitor tacrolimus. This will be achieved by mining electronic medical record data and employing explainable deep learning methods. The model will provide clinical decision support for timely intervention and treatment. Compared to traditional machine learning models, deep neural networks can extract more nuanced features from complex medical data and perform more precise pattern recognition, thereby enhancing prediction accuracy and reliability. By constructing a predictive tool based on explainable deep learning models, the investigators will better assess the association between the use of calcineurin inhibitors and AKI, explore targeted prevention strategies, and offer more precise predictions and intervention guidance to clinicians. Additionally, this research has significant socio-economic benefits and application potential. By reducing the incidence of AKI, the investigators can lower patient hospitalization duration and re-treatment costs, conserve medical resources, and improve patient quality of life. Preventive healthcare not only alleviates the physical and psychological burden on patients but also reduces the strain on the healthcare system, enhances healthcare efficiency, and promotes the rational allocation of medical resources.

Inclusion Criteria:

• Use of tacrolimus during hospitalization, with standardized therapeutic drug monitoring
• Age of 18 years or older at the time of admission
• Length of hospital stay ≥ hours
• At least two serum creatinine level tests conducted during the hospital stay

Exclusion Criteria:

• Stage 5 chronic kidney disease prior to admission
• Incomplete clinical data
• Serum creatinine levels consistently below 40 mmol/L during hospitalization

This project is a multicenter study involving hospitalized patients who received tacrolimus treatment and underwent therapeutic drug monitoring at three medical centers from January 2020 to December 2023: Shandong University First Affiliated Hospital, Binzhou Medical University Affiliated Hospital, and Jinan First People's Hospital.

The diagnosis and staging of acute kidney injury (AKI) in this study follow the relevant diagnostic criteria outlined in the 2012 KDIGO Clinical Practice Guidelines for AKI. AKI is defined by meeting at least one of the following conditions: (1) an increase in serum creatinine (SCr) > 26.5 mmol/L (>0.3 mg/dL) within 48 hours, or an increase in SCr to >1.5 to 1.9 times the baseline value within a continuous 7-day period; (2) a urine output of <0.5 mL/(kg·h) for 6 to 12 hours.