To the best of our knowledge, no previous study using an mHealth application among children with asthma in a primary care setting in Thailand has been reported. The interactive Asmon application showed a trend to improve disease control especially in a specific group of children who had poorly controlled asthma at baseline. Acute exacerbation rate and the median dose of ICS used were not significantly different between those who did and those who did not use the application. The PAQLQ also was not impacted by the intervention. Even though the user’s satisfaction scores were promising, implementation of the application nevertheless presented a number of challenges, such as the user-friendliness of the application procedure and participants’ background factors.
In order to improve clinical outcomes among childhood asthma patients, healthcare providers have various choices in selecting different types of mHealth applications for special functions, characteristics, utility, and quality [28, 29]. Healthcare professionals’ perceptions were surveyed regarding creation of innovative electronic monitoring devices. This indicated that attractive features would be new visual evidence to enable clinical discussions with patients and engaging them by means of rewards or motivations. However, data confidentiality and cost-effectiveness need to be considered and addressed before employing these applications in daily practice [30]. From the users’ point of view, i.e., for children with asthma, aged 7‒12 years, and their guardians, desirable aspects of the application were receiving asthma knowledge in a fun way, including a comic; interactive activity between the child and caregiver; and a user-friendly design, such as the use of colors. Beside knowledge regarding disease management, universal design and enhanced elements, reminding functions, and clinical monitoring of the application are also needed [31]. The Asmon application had been cooperatively developed by pediatricians and an IT professional to focus on medication adherence through daily reminders, interactive activity with daily uploading of a video by users, which is relayed to the healthcare database, and employment of a reward system (cartoon applause animated gifs).
However, even though we followed behavioral theory and applied an interactive design for creating the Asmon application, there was little effect on the primary outcome regarding disease control. Notably, only children with poorly controlled asthma at enrollment had potentially impacted by using the application. This result was consistent with the study by Cook et al. [32] who followed guidelines and Scripps management pathways to prove the concept of their study. Participants aged 17–82 years with poorly controlled asthma were enrolled to test the usability and effectiveness of their application over a 4-month period. They found that the application improved asthma control, increased FEV1, decreased courses of systemic corticosteroids, and was considered highly satisfactory by users. Another study in youths, aged 12‒17 years, with persistent asthma, used a mobile asthma action plan (AAP) application that revealed a promising outcome (disease control). This result was demonstrated only in participants with uncontrolled asthma at baseline [33]. In contrast, a multi-center, randomized controlled trial, performed in primary care setting in the UK, with cost effectiveness analysis, using mobile phone-based interactive management showed that adolescents and adults with poorly controlled asthma did not improve asthma control or increase self-efficacy as compared with paper-based monitoring. This intervention did not show promising cost-effectiveness [34].
Acute asthma exacerbation, dose of ICS used, emergency visits or hospitalization, and quality of life are also important indicators for assessing benefits of intervention. Our study did not show any significant differences between the intervention and control groups. These findings were concordant with many previous studies [35, 36]. In data from the Cochrane Database Systematic Reviews 2014, 12 of 17 randomized controlled trials also showed non-promising results regarding clinical outcomes [37].
In our study, the lack of substantial changes in desirable clinical outcomes may have been due to the status of participants at the initial enrollment. The proportion of participants with well-controlled disease was high in both the intervention group (43.2%) and the control group (33.3%). The mean ACT scores and were also higher than the cut-off point (20) in both the intervention group (20.2) and the control group (21.7). In addition, the median PAQLQ scores at baseline in both groups were also high (intervention group, 157; control group, 159 from 161). These would leave little room for effective changes regarding implementation of an intervention. Lessons learned from our study were in accordance with the study of Kosse et al. [38] and Vasbinder et al. [35]. In addition, due to the Covid-19 pandemic during our research period, most participants/guardians had intentionally adapted their protection behaviors (mask-wearing, physical distancing, hand-washing, etc.), which may have resulted in less exposure to allergens or pollutants.
We were disappointed in participants’ adherence to our Asmon application. Based on our backup data, about one-third of the intervention group never used our tool, while 40.5% of those who had ever tried it, used it between one and seven times during a 6-month period. This was similar to the results of Lau et al. [39], Greenhalgh et al. [40], Christensen & Mackinnon [41], and Kosse et al. [42]. Even though the participants/guardians’ mean satisfaction score was promising (4.5/5.0), this may reflect a desire to appease the research team. Data from open-ended questionnaires reported that the application was perceived to be hard to use. Self-judged ICS discontinuation and change in guardians were additional main obstacles for regular engagement with our application. Furthermore, time and financial issues were also challenges. In terms of characteristics of participants who strongly engaged with an mHealth application in other studies were high adherence to mHealth applications at baseline [42], young and middle-aged users [43]. Monetary incentivization and advanced gamification may be needed to promote long-term engagement with applications in chronic diseases such as asthma [44, 45]. Eventually, users’ psychosocial‒behavioral needs and sociodemographic backgrounds are very significant factors for implementing a novel mHealth application in the real-world daily practices.
The limitations of our study were that this was a non-blinded, conveniently-randomized, small sample-sized, and short-duration (6 months) study. In addition, we did not actively monitor participants/guardians’ compliance in the initial weeks or directly track medical adherence, which may have resulted in the less-than-ideal engagement. Moreover, because our participants/guardians were predominantly in the low‒middle socioeconomic class, who benefited from a universal health coverage scheme and mostly resided in a rural area, which might limit generalizability to other childhood asthmatic populations. Due to the Covid-19 pandemic, we encountered a variety of challenges, such as a lack of peak flow meter assessment at the end of study, avoidance of pulmonary function test measurement, and a high loss to follow-up.
The strengths of our study included that this clinical trial was conducted in a real-world, primary care setting in a small town, rural area in Thailand, that there was randomization of participants/guardians and comparison with a control group, which is typically burdensome to enroll in an intervention study. We had comparable success in the initial enrollment.
Future research should focus on objective outcome measures and should address patients’ background characteristics to allow identification of those in whom it would be more suitable to implement mHealth applications.