Feasibility and Acceptability of a Mobile App-Based TEAM-CBT (Testing Empathy Assessment Methods-Cognitive Behavioral Therapy) Intervention (Feeling Good) for Depression: Secondary Data Analysis.

BACKGROUND: The Feeling Good App is an automated stand-alone digital mobile mental health tool currently undergoing beta testing with the goal of providing evidence-informed self-help lessons and exercises to help individuals reduce depressive symptoms without guidance from a mental health provider. Users work through intensive basic training (IBT) and ongoing training models that provide education regarding cognitive behavioral therapy principles from a smartphone. OBJECTIVE: The key objective of this study was to perform a nonsponsored third-party academic assessment of an industry-generated data set; this data set focused on the safety, feasibility, and accessibility of a commercial automated digital mobile mental health app that was developed to reduce feelings associated with depression. METHODS: The Feeling Good App development team created a waitlist cohort crossover design and measured symptoms of depression and anxiety using the Patient Health Questionnaire-9, Generalized Anxiety Disorder-7, and an app-specific measure of negative feelings called the 7 Dimension Emotion Slider (7-DES). The waitlist cohort crossover design divided the participants into 2 groups, where 48.6% (141/290) of the participants were given immediate access to the apps, while 51.4% (149/290) were placed on a 2-week waitlist before being given access to the app. Data collected by the Feeling Good App development team were deidentified and provided to the authors of this paper for analysis through a nonsponsored university data use agreement. All quantitative data were analyzed using SPSS Statistics (version 28.0; IBM Corp). Descriptive statistics were calculated for demographic variables. Feasibility and acceptability were descriptively assessed. All participants included in the quantitative data were given access to the Feeling Good App; this study did not include a control group. RESULTS: In terms of safety, there was no statistically significant change in suicidality from preintervention to postintervention time points (t288=0.0; P>.99), and there was a statistically significant decrease in hopelessness from preintervention to postintervention time points (F289=30.16; P<.01). In terms of acceptability, 72.2% (166/230) of the users who started the initial 2-day IBT went on to complete it, while 34.8% (80/230) of the users who started IBT completed the entirety of the apps' 4-week protocol (150/230, 65.22% dropout rate over 4 weeks). CONCLUSIONS: This study is the first reported proof-of-concept evaluation of the Feeling Good App in terms of safety, feasibility, and statistical trends within the data set. It demonstrates a feasible and novel approach to industry and academic collaboration in the process of developing a digital mental health technology translated from an existing evidence-informed treatment. The results support the prototype app as safe for a select nonclinical population. The app had acceptable levels of engagement and dropouts throughout the intervention. Those who stay engaged showed reductions in symptom severity of depression warranting further investigation of the app's efficacy.

The development of the object sequencing imitation task to measure working memory in preschoolers.

We developed a new object sequencing imitation (OSI) task for preschoolers. We parameterized the task to test the effects of working memory load in 56 3- to 5-year-old children in a museum. We tested individual groups of 3-, 4-, and 5-year-olds on both "low" (2- to 4-step) and "high" (3- to 5-step) memory load sequences on two variants of the task. The experimenter demonstrated each sequence three times, and children were given 3 trials to imitate the sequence, receiving feedback from the experimenter following each trial. Children were tested on an object that was functionally the same but perceptually different. All preschoolers performed significantly above baseline. We found a significant load effect, showing that performance decreased with increasing steps per sequence. There was no effect of age on performance, showing that leveling of the task across age was successful. Overall, the newly developed OSI task can test age- and load-related changes in working memory in 3- to 5-year-old children. The new OSI task will have utility in longitudinal studies and in studies where multiple trials are needed such as neuroimaging.