JMIR Publications

ABSTRACT

Background:

The management of heart failure (HF) requires complex, data-driven decision-making. Although electronic health record (EHR) systems and clinical decision support (CDS) tools can streamline access to essential clinical information, it remains unclear which EHR elements and tools cardiologists and general medicine physicians prioritize when caring for HF patients.

Objective:

This study aims to identify these elements and tools to improve the user interface design of future EHR applications.

Methods:

This study used a user-centered design research approach to understand physician workflows and decision-making needs in HF care. A cross-sectional online survey was administered to 302 physicians, comprising 150 cardiologists (including 15 heart failure specialists) and 152 general medicine physicians. Respondents reported their use of EHR variables (e.g., medication lists, lab results, diagnostic tests, problem lists, clinical notes) for decision-making in HF care, as well as their time spent in the EHR before, during, and after patient visits along with their use of predictive models and patient reported outcome questionnaire. Descriptive analyses, chi-square tests, and t-tests were conducted to compare groups, with statistical significance set at p < 0.05.

Results:

A total of 302 healthcare providers participated in the survey, nearly evenly split between cardiologists (49.7%, 150/302) and general medicine physicians (50.3%, 152/302). Both groups consistently relied on medication lists, vital signs, lab results, diagnostic tests, problem lists, and clinical notes for HF decision-making. Cardiologists placed greater emphasis on diagnostic tests for inpatient HF care (4.66 ± 0.50 vs. 4.44 ± 0.64, P = .012) and outpatient HF care (4.67 ± 0.55 vs. 4.35 ± 0.71, P < .001). In contrast, general medicine physicians relied more on problem lists for inpatient HF care (4.63 ± 0.58 vs. 4.43 ± 0.72, P = .034), with no significant difference in the outpatient setting (p > 0.05). Both groups underutilized standardized questionnaires and predictive models, with only 20.1% (29/144) of cardiologists and 4.5% (6/133) of general medicine physicians using standardized questionnaires (P < .001)

Conclusions:

Both physician groups depend on medication lists, lab results, diagnostic tests, and problem lists. Cardiologists prioritize diagnostic tests, whereas general medicine physicians more often use problem lists. Low use of questionnaires and predictive models highlights the need for better integration of these tools. Future EHR design interface should tailor functionalities to accommodate these differing priorities and optimize HF care.