Predictive analytics support for complex chronic medical conditions: An experience-based co-design study of physician...

Area of Science:

• Health Informatics
• Clinical Decision Support Systems
• Chronic Disease Management

Background:

• Predictive technology in healthcare requires clinician input during development.
• Managing patients with multiple chronic conditions (Heart, Nephrology, Diabetes) presents complex care challenges.

Purpose of the Study:

• To explore physician managers' needs and preferences for predictive analytics in decision support for patients with Heart, Nephrology, and Diabetes (HND) conditions.
• To understand how predictive analytics can aid in managing complex comorbidities.

Main Methods:

• Qualitative study using an experience-based co-design model.
• Phases included patient mapping via observation and process mining, semi-structured interviews, and a co-design workshop.
• Data collected from physician managers at an integrated HND center.

Main Results:

• Complex care challenges arise from interacting pathophysiologies and multiple caregivers, hindering continuity.
• Physician managers see predictive analytics as valuable for decision support, risk calculation, and resource optimization.
• Key factors for adoption include simple visual interfaces, algorithm understanding, and addressing professional concerns.

Conclusions:

• Integrated practice units managing complex comorbidities can serve as ideal sites for piloting predictive technologies.
• Involving healthcare professionals early in the design process is crucial for effective implementation and adoption.
• Enhanced collaboration between healthcare and IT professionals, alongside user-centered design, is essential for successful predictive analytics integration.