Optimizing Electronic Medical Records for an Acute Decompensated Heart Failure Registry: A Mixed-Methods Analysis of...

Introduction Clinical registries are essential for monitoring heart failure (HF) care quality, yet their effectiveness depends on the availability of structured data within electronic medical records (EMRs). This study aimed to analyze the completeness of core data elements and identify systemic barriers to documentation for patients with acute decompensated heart failure (ADHF) at the National Cardiovascular Center Harapan Kita in Jakarta, Indonesia. Materials and methods A mixed-methods sequential explanatory design was employed. Phase one involved a retrospective quantitative analysis of 305 EMRs of patients with ADHF admitted between January 2024 and January 2025. Data were extracted across 82 core variables harmonized with American College of Cardiology/American Heart Association (ACC/AHA) and European Society of Cardiology (ESC) EuroHeart standards. Phase two consisted of in-depth interviews with eight key informants, including cardiologists, residents, nurses, medical records staff and the head of the hospital information system, to explore the root causes of documentation gaps. Results The aggregate data completeness was 77.2%. However, disparities existed between the domains. Documentation for inpatient therapy was near-universal (99.5%), whereas medical history documentation was low (40.4%). Critical registry variables such as New York Heart Association (NYHA) functional class, discharge weight, and National Identity Number (NIK) had 0% structured completeness. However, qualitative findings revealed these parameters were frequently present in unstructured free-text narratives, indicating a data capture gap rather than a clinical one. This was driven by a preference for Stevenson profiles (wet/dry) in acute settings and terminological mismatches. The analysis confirmed that documentation habits are primarily shaped by reimbursement incentives and technical EMR limitations rather than clinical oversight. Conclusions While the institution possesses a strong foundation for administrative data, the current EMR usage prioritizes billing over clinical relevance. Developing a functional, impactful clinical registry requires transitioning from free-text narratives to structured data entry, implementing mandatory fields for high-value clinical and prognostic markers, and automating data synchronization between EMR modules. Establishing this high-quality data infrastructure is critical for future applications in predictive analytics or even artificial intelligence, enabling precision medicine approaches tailored specifically to the unique HF demographic in Southeast Asia, including Indonesia.