Table 3.

Description of the control groups, procedures, and findings in studies on self-monitoring, treatment, and automatically generated data collected via smartphones in children, adolescents, and young adults with psychiatric disorders (N=15).

Author, year of publication	Control group	Baseline assessment	Follow-up	Adherence	Findings
Bucci et al, 2018 [13]	24 participants received Actissist plus TAUa, 12 received ClinTouch plus TAU	Demographics; PANASb; PSYRATSc; CDSSd; Global Assessment of Functioning Scale; Personal and Social Performance Scale; Empowerment Rating Scale; EQ-5D-5Le; Timeline Followback, Medication Adherence Rating Scale	Clinical assessment at 12 weeks and 22 weeks	Data points completed (>33% data entries): 75% (Actissist) and 50% (ClinTouch)	Actissist was feasible, acceptable (90% recommend Actissist), and safe (0 serious adverse events), with high levels of user satisfaction. Treatment effects were large on negative symptoms, general psychotic symptoms, and mood. The addition of Actissist conferred benefit at posttreatment assessment over routine symptom-monitoring and TAU.
Cao et al, 2020 [21]	N/Af	Mini International Neuropsychiatric Interview to confirm diagnosis of MDDg, PHQ-9h, HAM-Di, and HAM-Aj	Biweekly clinical assessment	79.0%	Significant correlation between the self-evaluated mood averaged over a 2-week period and the biweekly psychometric scores from PHQ-9, HAM-D, and HAM-A (0.45≤|r|≤0.63; P=.009, P=.01, and P=.003, respectively). The daily steps taken, SMS frequency, and average call duration were also highly correlated with clinical scores (0.44≤|r|≤0.72; all P<.05). By combining self-evaluations and smartphone sensor data, they could predict the PHQ-9 score with an accuracy of 88%.
Dennis et al, 2015 [24]	N/A	GAIN-Q3k	2 visits a week to complete survey and provide urine sample	Participants completed 89% EMAsl; 18 participants completed over 90% of the EMAs	EMA observations were classified into 3 risk groups: “Current Use” (3%), “Unrecognized Risk” (42%), or “Recognized Risk” (55%). Unrecognized Risk (50%) and Current Use (96%) groups reported significantly higher rates of use in the next week compared with the Recognized Risk group’s use in the following week (31%). Drug use following an EMA that was accessed was lower compared to that when EMA was not accessed (32% vs 43%).
Jones et al, 2018 [22]	N/A	Caregivers completed aberrant behavior checklist, child behavior checklist, PANAS, visual analog scale	Clinic visits in weeks 1, 4, and 8.	Not described	Over 8 weeks, caregivers reported improvements in their child’s mood, irritability, and disruptive behaviors during TAU.
Kennard et al, 2018 [25]	34 patients received As Safe As Possible app+ TAU, 32 received TAU	PHQ, SIQ-JHSVm, CSSRSn, youth self-report scale, CRAFTo	At weeks 4, 12, and 24	70.6% used the app at least once. Participants rated their mood at a median of 19 times	There were no treatment effects on suicidal ideation. Participants reported high satisfaction with both the intervention and the app.
Kolar et al, 2016 [17]	Healthy controls	ChEDEp; Symptom Checklist 90	Not described	1030 completed the observations entered	Participants with anorexia nervosa showed higher mean and maximum levels of aversive tension. Reported food intake was associated with higher levels of aversive tension in the anorexia nervosa group, whereas reported school or sport-related events were not linked to specific states of aversive tension. After food intake, subsequent increases of aversive tension were diminished, and decreases of aversive tension were induced in adolescents with anorexia nervosa.
Kumar et al, 2018 [14]	N/A	GFqsocial, GFrole; BPRSr; Clinical Global Impression Severity; assorted questionnaires	At the end of project with repeat of clinical assessment	Daily survey completion rate was 41% and weekly survey completion rate was 39%	27 of 41 (66%) participants with early psychosis who completed the study and 11 of 13 (85%) treatment providers who responded to satisfaction surveys reported they would continue to use the app as part of treatment services.
Lim et al, 2020 [16]	N/A	SCID-5s, Positive and Negative Syndrome Scale, CDSS, Social Skills Performance Assessment, National Adult Reading Test, UCLA-LSt, SIASu, Scales of Psychological Well-being	After treatment and 3-month follow up	Participants on average completed 95.47% of the +Connect (mean 40.10 days, SD 3.04)	Data indicate preliminary evidence that +Connect may reduce loneliness, with scores from preintervention (mean 52.58, SD 8.47) to postintervention (mean 48.10, SD 10.38) and at 3 months after the intervention (mean 42.89, SD 7.04) on UCLA-LS
Lim et al, 2019 [27]	11 lonely students without mental health conditions	SCID-5-Research version, UCLA-LS, SIAS, Centre for Epidemiological Studies-Depression	Posttreatment and 3-months follow up	Social anxiety group: 84.66%; control group: 90.26%	The UCLA-LS and straightforwardly worded SIAS scores decreased in a linear trend from baseline to months after the intervention for the case group. There were higher acceptability ratings across different ratings in a nonclinical lonely student group compared with those with social anxiety disorder.
Neumayr et al, 2019 [19]	50% were randomized to receive intervention	EDE-Qv, BMI, BDI-IIw	8 weeks and 6 months	Mean of 231 logs during the 8-week intervention.	There were postintervention nonsignificant small to moderate between‐group effect sizes favoring the intervention group regarding BMI (d=–0.24 [–0.90, 0.41]) and eating disorder symptoms. At 6‐month follow‐up, there were no differences between the intervention group and control group
Niendam et al, 2018 [15]	N/A	BPRS	Not described	Weekly survey completion: mean 77.3%; Daily survey completion: mean 69.0%.	Weekly survey positive symptoms were significantly associated with BPRS-positive symptoms (β=.56, SE=0.10; P<.001).
Seidel et al, 2016 [18]	Healthy controls, normal weight, no history of eating disorders	Structured Interview for Anorexic and Bulimic Syndromes for Experts; Eating Disorder Inventory; BDI; Perseverative Thinking Questionnaire; State-Trait Anxiety Inventory; body weight	Not described	Participants answered 84.19% of their prompts compared to 75.73% in the control group	Momentary negative affect is positively associated with a higher amount of disorder-related rumination in participants (P<.001).
Shrier and Spalding, 2017 [20]	N/A	BDI-II	Not described	Not described	Participants expressed positive opinions about the ecological momentary intervention. They desired a high degree of personalization of the message quality, style, and voice.
Stallard et al, 2018 [26]	N/A	MFQx; Revised Children’s Anxiety and Depression Scale; Strengths and Difficulties Questionnaire; Safety-assessment; Self-Harming information	After 2 weeks (postfamiliarization) and 12 weeks (after use)	Not described	73% of those who had recently self-harmed reported reductions in self-harm after using BlueIce for 12 weeks. There was a statistically significant mean difference of 4.91 (P=.04) on postuse symptoms of depression (MFQ) and 13.53 (P=.001) on symptoms of anxiety (Revised Child Anxiety and Depression Scale), which was evident across all anxiety subscales.
Sweidan et al, 2019 [23]	N/A	Not described	Not described	Not described	A detailed survey filled out by 100 parents and teachers after testing showed encouraging results

^aTAU: treatment as usual.

^bPANAS: Positive Affective and Negative Affective Scale.

^cPSYRATS: Psychotic Symptom Rating Scale.

^dCDSS: Calgary Depression Scale for Schizophrenia.

^eEQ-5D-5L: 5-level health status and health-related quality of life.

^fN/A: not applicable.

^gMDD: major depressive disorder.

^hPHQ-9: patient health questionnaire-9.

ⁱHAM-D: Hamilton Depression Rating scale.

^jHAM-A: Hamilton Anxiety Rating scale.

^kGAIN-Q3: Global Appraisal of Individual Needs-Quick 3.

^lEMA: ecological momentary assessment.

^mSIQ-JHSV: Suicidal Ideation Questionnaire-Junior High School Version.

ⁿCSSRS: Columbia Suicide Severity Rating Scale.

^oCRAFT: Car, Relax, Alone, Forget, Friends, and Trouble questionnaire.

^pChEDE: child eating disorder examination.

^qGF: global functioning.

^rBPRS: Brief Psychiatric Rating Scale.

^sSCID-5: Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders.

^tUCLA-LS: University of California, Los Angeles Loneliness Scale score.

^uSIAS: Social Interaction Anxiety Scale.

^vEDE-Q: Eating Disorder Examination-Questionnaire.

^wBDI-II: Beck Depression Inventory-II.

^xMFQ: mood and feeling questionnaire.