JMIR Publications

ABSTRACT

Background:

The desire for healthcare organizations to reduce the cost of chronic care and to prevent disease from occurring to begin with, has coincided with the development of new technology that is revolutionizing digital health. Numerous health-oriented mobile phone applications (referred to as mHealth apps) have been developed and are available for download into smartphones. These mHealth apps serve a wide range of functions. There are apps that monitor data to treat or avoid chronic illness; apps for managing daily activities and diet; apps promoting healthy choices for people who want to maintain and improve their overall health, and many others.

Objective:

While it is generally recognized that mHealth apps have a significant potential for promoting public health, little research has been done to determine user preferences for such apps. The objective of our study was to tests the major product qualities of mHealth apps, asking if users seek interaction with a live physician, or are they willing to rely on artificial intelligence to analyze data from their app. Next, the research presented here tests how judgments of aspects such as instrumentality, aesthetics, and symbolic value influence product preference. Understanding what users want in their mHealth apps can help increase their acceptability and encourage healthy lifestyles. The contribution of this paper is its focus on user preferences which may help in the design of mHealth apps to better address peoples’ needs—thus encouraging a wide, frequent, and effective use of such tools which promote public health.

Methods:

A total of 347 respondents volunteered to rate three models of mHealth apps that included commonly used mHealth features, health-indicators, and social-oriented features according to sixteen items measuring instrumentality, aesthetics, and symbolism. Models were rated after reading one of two different scenarios. In one scenario the data collected by the app was analyzed by a physician. In another scenario an automatic analysis of an AI algorithm analyzed the data.

Results:

As shown by the responses, the involvement of a human physician in the application had a significant effect (p < .05) on the perceived instrumentality of the model that presented only mHealth features. The model that included commonly used mHealth features, health-indicators, and social-oriented features was rated as significantly more aesthetic when there was a human physician analyzing the data compared to when an AI algorithm was suggested to do the analysis.

Conclusions:

Generally, judgments of models where there was a “human touch” behind the analysis (a human physician was assumed to analyze the data) were judged more positively. The presence of a human caregiver is important for users, due to an increased sense of connectedness to a knowledgeable caring healthcare giver.