A first step towards translating evidence into practice: heart failure in a community practice-based research network

Area of Science:

• Primary Care Medicine
• Health Informatics
• Cardiology

Background:

• Electronic Health Records (EHRs) are increasingly used in primary care.
• Accurate patient problem lists within EHRs are essential for effective disease management.
• Left ventricular dysfunction (LVD) and heart failure (HF) require timely diagnosis and treatment.

Purpose of the Study:

• To assess the accuracy of heart failure diagnoses listed in an EHR.
• To determine the validity of EHR problem list entries for identifying patients with left ventricular dysfunction in a primary care setting.

Main Methods:

• A cross-sectional study was conducted across nine primary care clinics within the Providence Research Network.
• A database query identified patients with a heart failure diagnosis on their problem list.
• Left ventricular ejection fraction (LVEF) data were cross-referenced using echocardiography databases and manual chart reviews.

Main Results:

• Out of 205,755 active patients, 1,731 had a heart failure (HF) problem list entry.
• The sensitivity of the EHR problem list for identifying HF was low (43.9% for LVEF ≤55% and 54.4% for LVEF ≤40%).
• The positive predictive value was not explicitly stated but implied to be poor given the low sensitivity.

Conclusions:

• The validity of heart failure diagnoses on EHR problem lists in primary care is poor.
• Reconciling EHR data with other sources can improve problem list accuracy.
• Inaccurate EHR problem lists may lead to underprescribing of essential therapies for conditions like left ventricular dysfunction.