This study focuses on the early identification and precision prevention of acute kidney injury (AKI) following interventional therapy in patients with liver cancer. A prospective, phased mixed-methods research design will be adopted. In the first phase, a prospective cohort study will be conducted to systematically collect clinical and nursing-related data. A Super Learner ensemble algorithm will be employed to develop and validate an AKI risk prediction model, enabling precise risk stratification of patients. In the second phase, a prospective randomized controlled trial will be implemented. Guided by the Health Belief Model (HBM), a risk-stratified proactive nursing intervention will be delivered, and its effectiveness will be compared with that of routine nursing care in reducing the incidence of AKI, improving adherence to hydration protocols, optimizing the management of nephrotoxic medications, and enhancing patients' self-management capabilities. By integrating intelligent risk prediction with proactive nursing intervention strategies, this study aims to establish a reproducible and scalable precision prevention and management model for AKI in liver cancer patients undergoing interventional therapy. The findings are expected to provide scientific evidence for early AKI warning and proactive nursing practice, thereby promoting the precision, intelligence, and standardization of oncologic interventional nursing care.

Inclusion Criteria:

• Completed eligibility screening in Phase I of the study
• Completed AKI risk assessment using the established prediction model
• Classified by the model as at risk for AKI (moderate- or high-risk patients prioritized; low-risk patients included only for exploratory/observational analysis)
• Conscious and able to communicate effectively
• Able to understand and comply with nursing interventions
• Expected hospital stay of at least 3 days
• Available for postoperative follow-up
• Provided written informed consent jointly with a primary caregiver

Exclusion Criteria:

• Development of AKI within 24 hours after interventional therapy according to KDIGO criteria
• Severe psychiatric disorders, cognitive impairment, or communication barriers preventing participation in HBM-based assessment or intervention
• Requirement for postoperative ICU admission and continuous life-support therapy
• Mandatory use of nephrotoxic medications that cannot be discontinued, substituted, or adjusted during the study period
• Participation in other clinical intervention studies that may affect renal function or nursing outcomes

Individual participant data (IPD) will not be publicly shared because the dataset contains sensitive clinical information that may pose a risk of participant re-identification despite de-identification procedures. In addition, data sharing is restricted by institutional ethics requirements and the informed consent provided by participants. Requests for access to de-identified data may be considered on a reasonable basis and subject to approval by the corresponding institution and ethics committee.