Obesity is a considerable problem in populations with serious mental illness (SMI), in part due to medication-induced weight gain and social disadvantage. Individuals with SMI often have cognitive deficits that need to be considered when delivering interventions. In-person interventions for weight management that are tailored for these cognitive deficits have been developed and found in multiple trials to result in lower weight. Unfortunately, the impact of these interventions is limited by low rates of utilization. Similar to the general population, underserved populations increasingly use smartphones for communication and internet access, which allows for more convenient engagement and retention in a weight management program. Evidence-based interventions have been delivered via mobile technologies for the general population, yet there has been almost no effort to do the same for the population with SMI.
1) Develop CoachToFit, a weight management system that is tailored to meet the needs of individuals with SMI and delivers MOVE! via a smartphone app with support from peer wellness coaches; 2) Study the acceptability and usability of CoachToFit in Veterans with SMI; and 3) Evaluate changes in self-efficacy, motivation, and readiness among CoachToFit users. Although evaluating acceptability and usability is the primary goal for this study, we will also explore preliminary evidence for the efficacy of CoachToFit.
App development utilized user-centered and agile development processes. Participants contributed to the design and evaluation of CoachToFit through participation in focus groups, in-lab usability trials, and a 30-day experiential usability trial. Participants ("users") were Veterans with SMI who were overweight who owned a phone running Android OS or iOS (iPhone). Data were collected from patient assessments and the smartphone app.
Focus groups of overweight Veterans with SMI (n=6) were presented with the CoachToFit screen mockups for feedback. Feedback was used to design and name the CoachToFit app. A peer coach was hired who provided input on the coaching dashboard where CoachToFit user data was visualized to guide coaching. Additional functionality was added to the app to benefit users; including integrating the app with an external Bluetooth activity tracker watch and Bluetooth scale for weight monitoring.
With a working version of CoachToFit, in-lab usability testing with overweight Veterans with SMI (n=10) was completed. Cohort demographics were 10 males; aged 61.7+8.6; 5 diagnosed with schizoaffective disorder, 4 with bipolar disorder, and 1 with schizophrenia; 9 had Android phones and 1 had an iPhone. In-lab usability indicated that more explicit directional aids were necessary and the graphing of data (weight and steps) needed further simplification, and there were particular problems with Bluetooth scale usability. There were very positive reactions to the activity tracker watch, and the educational modules and goal choices within the modules. Further refinement was made to CoachToFit. In sum, CoachToFit has 30 modules providing education on nutrition (n=15 modules) and exercise (n=15 modules). CoachToFit allows tracking of efforts to meet the personalized nutrition and exercise goals set at the end of each module, tracking and visualization of weight and step counts over time, and review of completed modules. A dashboard allows a peer coach to see an individual's progress for weekly supportive coaching by phone.
With a revised version of CoachToFit and the coaching dashboard, we enrolled 18 Veterans with SMI to carry the app for 30 days. Cohort demographics were 16 males, 2 females; aged 57.8+10.7; 5 diagnosed with schizoaffective disorder, 7 with bipolar disorder, 5 with schizophrenia, and 1 with recurrent major depressive disorder; 13 had Android phones, 5 had iPhones. The mean body mass index of the sample was 32.2+3.7 (obese). This cohort had an average PROMIS Global Physical Health T-score of 41.8 and average PROMIS Global Mental Health T-score of 42, meaning this cohort at baseline was one standard deviation worse (less physically, less mentally healthy) than the general population.
Overall, early results indicate strong acceptability and usability of CoachToFit in Veterans with SMI. The majority of the sample agreed that the they would like to use CoachToFit often, thought CoachToFit was easy to use, felt the app and the watch and scale worked well together, felt confident using CoachToFit, found it not complicated, not cumbersome, and would not need technical assistance to use it. The majority reported being satisfied with how easy it was to use, found the education modules easy to understand and informative. There was strong agreement across the cohort that they would recommend CoachToFit to a friend, would like to keep using it themselves, and felt it was made for people like them. Additionally, the majority felt comfortable that their information was collected by the app and would be happy for their clinical team to know of their progress with CoachToFit, although a few had concerns about the privacy of their data. There was consistent agreement that the weekly peer coaching was valuable.
Early results provided information about self-efficacy, motivation, and readiness among CoachToFit users. Self-efficacy and motivation for making changes around diet and exercise was high at baseline, so much so that there was little room for change. This indicates that a different measurement of self-efficacy and motivation should be explored for a future larger test of CoachToFit. Readiness to make a change in nutrition was improved in half of the sample, but there was limited movement in readiness to make a change in exercise. Given this was only a 30 day trial, users only received a maximum of 8 (4 nutrition; 4 exercise) of 30 modules; in a full trial allowing time for all 30 modules we expect change in readiness for both nutrition and exercise.
Although not powered for examination of the efficacy of the program in 30 days, we explored weight changes. Several of the early users lost weight in 30 days; a range of 2-16 pounds lost.
CoachToFit can deliver an evidence-based weight management treatment that is specifically designed for the population with SMI, who report it has high acceptability and usability. Early results also indicate it is efficacious and a larger trial is warranted.
- Cohen AN, Rocchio R. Developing and Initial Testing of a Mobile App Delivering an Evidence-based Weight Management Program to Veterans with Schizophrenia: CoachToFit. Spotlight on Mental Health Centers of Excellence [Cyberseminar]. VA HSR&D. 2019 Feb 11.