Abstract
Aims
Heart failure is an escalating burden on global health care systems. Modernizing heart failure care is inevitable, with eHealth products poised to play an important role. However, eHealth devices that can initiate and adjust heart failure medication are currently lacking. Consequently, this study aimed to develop an artificial intelligence-based decision engine to provide guideline-based recommendations for disease-modifying medication in heart failure patients.
Methods and Results
We developed the decision engine by converting the ESC heart failure guidelines into Business Process Model and Notation, a visual modeling language suitable for developing complex decision engines. A safety evaluation, based on clinical parameters, was conducted to ascertain the system’s applicability to specific cases. The decision engine renders specific decisions concerning disease- modifying therapy for heart failure patients. We defined 72 virtual heart failure patient scenarios, encompassing a broad spectrum of baseline characteristics and background medication. All recommendations offered by the engine were evaluated by an independent heart failure specialist. All but three recommendations (94%) were identical to the treatment decisions by the heart failure specialist and all (100%) were in line with the 2021 ESC heart failure guidelines.
Conclusion
The decision engine offers guideline-based recommendations for disease-modifying therapy, positioning it as a tool to enhance self-care among heart failure patients. To validate our results, the decision engine is being prospectively tested in real-world patients in a multicenter clinical trial (NCT04699253).
Heart failure is an escalating burden on global health care systems. Modernizing heart failure care is inevitable, with eHealth products poised to play an important role. However, eHealth devices that can initiate and adjust heart failure medication are currently lacking. Consequently, this study aimed to develop an artificial intelligence-based decision engine to provide guideline-based recommendations for disease-modifying medication in heart failure patients.
Methods and Results
We developed the decision engine by converting the ESC heart failure guidelines into Business Process Model and Notation, a visual modeling language suitable for developing complex decision engines. A safety evaluation, based on clinical parameters, was conducted to ascertain the system’s applicability to specific cases. The decision engine renders specific decisions concerning disease- modifying therapy for heart failure patients. We defined 72 virtual heart failure patient scenarios, encompassing a broad spectrum of baseline characteristics and background medication. All recommendations offered by the engine were evaluated by an independent heart failure specialist. All but three recommendations (94%) were identical to the treatment decisions by the heart failure specialist and all (100%) were in line with the 2021 ESC heart failure guidelines.
Conclusion
The decision engine offers guideline-based recommendations for disease-modifying therapy, positioning it as a tool to enhance self-care among heart failure patients. To validate our results, the decision engine is being prospectively tested in real-world patients in a multicenter clinical trial (NCT04699253).
| Original language | English |
|---|---|
| Pages (from-to) | 285-288 |
| Number of pages | 4 |
| Journal | European Heart Journal - Digital Health |
| Early online date | 1 Dec 2023 |
| DOIs | |
| Publication status | Published online - 1 Dec 2023 |
Data Access Statement
The data underlying this article will be shared upon reasonable request to the corresponding author.Keywords
- Heart failure
- Artificial intelligence
- Decision support systems
- Clinical
- Guideline
