Identifying Patients with Heart Failure Eligible for Guideline-Directed Medical Therapy

Area of Science:

• Cardiology
• Health Informatics
• Population Health Management

Background:

• Many heart failure patients do not receive guideline-recommended therapy.
• Manual review of electronic health records (EHRs) by clinical staff is a significant barrier for population health management (PHM) programs aiming to improve medication adherence.
• Automated tools are needed to streamline the identification of eligible patients for guideline-directed care.

Purpose of the Study:

• To develop and evaluate a rule-based system (RBS) for automatically identifying patients with heart failure (HF) eligible for guideline-directed therapy.
• To assess the performance and identify areas for improvement in an automated system designed to support PHM programs.

Main Methods:

• A rule-based system (RBS) was developed to parse EHRs and identify potentially eligible HF patients.
• The RBS was deployed in a PHM program, running every other month to flag patients for manual screening.
• System performance was evaluated, including an error analysis of false-positive cases.

Main Results:

• The RBS identified approximately 4200 patients per execution from a cohort of 161,000 patients with echocardiograms.
• Manual screening of 5460 patients identified 1754 as truly eligible, resulting in a 32.1% accuracy rate.
• Over 38% of false positives were attributed to inaccuracies in determining symptomatic HF and patient medication history.

Conclusions:

• The RBS offers a systematic approach to enrich patient populations for guideline-directed therapy eligibility.
• Integrating clinical note processing could significantly improve the system's accuracy and performance.
• Implementing automated tools for guideline-directed care presents practical challenges that require further solutions.