Developing a User-Centered Digital Clinical Decision Support App for Evidence-Based Medication Recommendations for...

Area of Science:

• Health Informatics
• Clinical Decision Support
• Human-Computer Interaction

Background:

• Bridging the gap between evidence-based guidelines and clinical practice remains a challenge.
• Clinical decision support systems (CDSSs) aim to aid clinicians but often suffer from poor usability, leading to workflow issues and alert fatigue.

Purpose of the Study:

• To validate a user-centered design (UCD) process for developing a standards-based medication recommender for type 2 diabetes mellitus (T2DM).
• To evaluate the prototype app's effectiveness, efficiency, and user satisfaction in a simulated clinical setting.

Main Methods:

• Iterative user testing with 8 clinical leaders and 2-8 prescribers per round informed app development.
• A validation study with 43 participants compared a prototype app against a traditional electronic health record (EHR).
• Measures included adherence to guidelines, task completion time, perceived cognitive load (NASA-TLX), and usability (SUS).

Main Results:

• The prototype app significantly increased adherence to evidence-based guidelines compared to the EHR (96.2% vs. 72.0%).
• Medication ordering time decreased with the app in the second period (107.7s vs. 130.7s).
• System Usability Scale (SUS) scores indicated good usability, ranging from 68-86, suggesting provider satisfaction.

Conclusions:

• A UCD process can yield a point-of-care medication recommender app that enhances adherence to guidelines, such as those from the American Diabetes Association.
• The T2DM app shows potential for improving provider efficiency in medication ordering with practice.
• The app achieved good usability and provider satisfaction scores.