Mobile apps and conversational agents are becoming popular in healthcare by improving usability and accessibility. To reveal their clinical benefits, we analyzed conversational agent usage characteristics of asthma patients within a mobile app. 

Asthma survey results and conversational agent usage data from the users of an asthma-specific mobile app (Albert Health Assistant) were analyzed using descriptive statistics. The relationships between the groups were evaluated using chi-square test. The distribution of continuous variables was checked using Shapiro-Wilk test.

Among the 254 users who completed the survey, 42.9% self-reported getting an asthma diagnosis within the last year. Most (57.1%) expected the app to be helpful for better asthma management. Also, 42.1% were not using any inhaler for asthma treatment. Fifty-three users (20.9%) interacted with the conversational agent 360 times, with a mean success rate of 88%. Voice-based interaction was higher than text-based interactions (6.78 and 2.48 times per user). While patients used both interfaces similarly (51.7% vs. 48.3%) for getting information about asthma; they primarily interacted by speaking to use functions of the application, such as setting a medication reminder or keeping an asthma diary, 72.2% vs. 27.8%, respectively (p<0.001). They also used voice interface more in other contexts out of asthma(92.1% vs. 7.9%, p<0.001).

The results highlight the potential of mobile apps and conversational agents to assist patients in managing their asthma. The higher frequency of voice-based interactions over text-based interactions indicates that patients may find it more convenient and efficient to interact with the application using voice.