Table 3.1. Recommendations for telemonitoring, wearables, artificial intelligence, and machine learning in heart failure.

Recommendations	Class	LE	Comment	Table 2018	Ref.
Use of telemonitoring to manage patients with chronic HF.	IIa	A	NEW: Meta-analyses show reduction in mortality rates and in hospitalizations for HF.	New	29–32
Wearables as complementary tools in the diagnosis and treatment of patients with chronic or acute HF.	IIa	B	NEW: Several observational studies show the benefits of wearables use for HF patients.	New	33, 34
Artificial intelligence use in the diagnosis, prognostic assessment, or selection of patients who can most benefit from different therapies.	IIb	B	NEW: Observational studies indicate the benefits of using Machine Learning and Artificial Intelligence in the diagnosis and prognosis of HF.	New	35
Meta-analyses involving observational and randomized trials on invasive and noninvasive distance monitoring and support has found a positive impact on the prognosis for HF patients.29–32 Reductions in all-cause mortality may range from 19 to 31% with telemonitoring for HF patients, while the reduction in frequency of hospitalizations for HF ranges from 27 to 39%, especially for patients in functional class (FC) III/IV, according to the New York Heart Association (NYHA). Artificial intelligence has applications in HF, either for diagnosis, prognostic assessment, telemonitoring or selection of patients who can most benefit from various therapies.33,34 This is possible, for instance, in distinguishing phenotypes, assigning patients in different signature profiles; more accurate diagnosis of acute HF as compared to physicians; and also helping in referral for new or established therapies, such as additional analysis of baseline ECG to identify patients who would better respond to cardiac resynchronization therapy.35

FC: functional class; HF: heart failure.