Background:
The increasing use of digital health applications poses new challenges for decision makers in the evaluation of these applications. For the majority of available digital health applications there is little or no evidence of (medical, organisational or economic) benefit. Currently available assessment frameworks often do not cover all domains of a full HTA. At the same time, the thorough evaluation of a digital health application requires additional, technology-specific aspects such as software updates, questions of connectivity and compatibility, as well as digital data protection. This report aims to examine the applicability of the currently available assessment frameworks for selected digital health applications and to develop recommendations for Austria.

Methods:
A systematic literature search in four databases and a survey of INAHTA members was conducted to identify assessment frameworks. Six assessment frameworks were exemplarily selected and analysed. A manual search was performed to identify already reimbursed digital health applications, for eleven digital health applications the published studies were identified. A hand search was conducted to describe the strategies and regulations of several countries in dealing with digital health applications.

Results:
Suggestions for study designs were described in 4 of 6 analysed frameworks. A risk classification of the digital health application to be evaluated was suggested in 1 of 6 analysed frameworks. Aspects of artificial intelligence were considered in 1 of 6 analysed frameworks. Only for 2 of the 11 selected digital health applications the present study design met the evidence standards of the NICE framework. Belgium, France, the Netherlands, England and Germany have different strategies for the regulation and reimbursement of digital health applications. Recent developments in these countries show that there are efforts to find regulations at the national level.

Conclusion:
There is a great heterogeneity of assessment frameworks, especially with regard to risk classification and technology-specific aspects such as data protection and artificial intelligence. Only the Evidence Standards Framework of NICE offers a precise classification into defined risk classes with correspondingly required study designs. The published evidence of digital health applications show differences in the choice of study designs. Controlled studies with comparisons of the standard of care are not always available, but are more frequent in recent (ongoing) studies.