Effects of a Smartphone-Based Self-management Intervention for Individuals With Bipolar Disorder on Relapse, Symptom Burden, and Quality of Life

Abstract

ImportanceBipolar disorder–specific psychotherapy combined with pharmacotherapy improves relapse risk, symptom burden, and quality of life, but psychotherapy is not easily accessible.ObjectiveTo determine if a smartphone-based self-management intervention (LiveWell) can assist individuals with bipolar disorder to maintain wellness.Design, Setting, and ParticipantsAn assessor-blind randomized clinical trial enrolled participants from March 20, 2017, to April 25, 2019, with 48-week follow-up ending on April 10, 2020. Participants were randomly assigned to usual care or usual care plus the smartphone intervention stratified by relapse risk based on initial clinical status (low risk: asymptomatic recovery; high risk: continued symptomatic, prodromal, recovering, symptomatic recovery). Participants with bipolar disorder I were recruited from clinics in the Chicago and Minneapolis-Saint Paul areas. Data were analyzed from June 19, 2020, to May 25, 2022.InterventionsThe smartphone-based self-management intervention consisted of an application (app), coach, and website. Over 16 weeks, participants had a coach visit followed by 6 phone calls, and they completed daily and weekly app check-ins. The app provided adaptive feedback and information for developing a personalized wellness plan, the coach provided support, and the website provided summary data and alerts.Main Outcomes and MeasuresThe primary outcome was time to relapse. Secondary outcomes were percentage-time symptomatic, symptom severity, and quality of life.ResultsOf the 205 randomized participants (mean [SD] age, 42 [12] years; 125 female individuals [61%]; 5 Asian [2%], 21 Black [10%], 13 Hispanic or Latino [6%], 7 multiracial [3%], 170 White [83%], 2 unknown race [1%]), 81 (40%) were randomly assigned to usual care, and 124 (60%) were randomly assigned to usual care plus the smartphone intervention. This clinical trial did not detect a reduction in relapse risk for the smartphone intervention (hazard ratio [HR], 0.65; 95% CI, 0.39-1.09; log-rank P = .08). However, decreased relapse was observed for low-risk individuals (HR, 0.32; 95% CI, 0.12-0.88; log-rank P = .02) but not high-risk individuals (HR, 0.86; 95% CI, 0.47-1.57; log-rank P = .62). Reduced manic symptom severity was observed for low-risk individuals (mean [SE] difference, −1.4 [0.4]; P = .001) but not for high-risk individuals (mean [SE] difference, 0 [0.3]; P = .95). The smartphone-based self-management intervention decreased depressive symptom severity (mean [SE] difference, −0.80 [0.34]; P = .02) and improved relational quality of life (mean [SE] difference, 1.03 [0.45]; P = .02) but did not decrease percentage-time symptomatic (mean [SE] difference, −5.6 [4.3]; P = .20).Conclusions and RelevanceThis randomized clinical trial of a smartphone-based self-management intervention did not detect a significant improvement in the primary outcome of time to relapse. However, a significant decrease in relapse risk was observed for individuals in asymptomatic recovery. In addition, the intervention decreased depressive symptom severity and improved relational quality of life. These findings warrant further work to optimize the smartphone intervention and confirm that the intervention decreases relapse risk for individuals in asymptomatic recovery.Trial RegistrationClinicalTrials.gov Identifier: NCT03088462

Additive effects of adjunctive app-based interventions for mental disorders - A systematic review and meta-analysis of randomised controlled trials

“…No difference was found for symptoms of depression regarding primary and secondary outcomes ( g = 0.03; 95 % CI -0.07 to 0.13, p = 0.578; I 2 = 13.9 %; 95 % CI 0.0 to 48; PI -0.19 to 0.25; k = 22 ( Bastiaansen et al, 2020 ; Bell et al, 2020 ; Boettcher et al, 2018 ; Farren et al, 2022 ; Faurholt-Jepsen et al, 2015 ; Faurholt-Jepsen et al, 2020 ; Faurholt-Jepsen et al, 2021 ; Goulding et al, 2023 ; Keeler et al, 2022 ; Law et al, 2023 ; Lewis et al, 2020 ; Liu et al, 2023 ; Mackintosh et al, 2017 ; Mantani et al, 2017 ; O'Toole et al, 2019 ; Raevuori et al, 2021 ; Schmädeke and Bischoff, 2015 ; Schnall et al, 2022 ; Tønning et al, 2021 ; Wallace et al, 2022 ); see Appendix F for forest plot Fig. F.1).…”

“…F.4) with low heterogeneity ( I 2 = 0.0 %; 95 % CI 0.0 to 84.7; PI -0.20 to 0.59). When studies with a low risk of bias were pooled, the difference was no longer significant ( g = 0.18; 95 % CI, -0.06 to 0.42; k = 3 ( Faurholt-Jepsen et al, 2015 ; Faurholt-Jepsen et al, 2020 ; Goulding et al, 2023 )).…”

“…For bipolar disorder, a significant reduction of manic symptoms was observed across four comparisons ( Faurholt-Jepsen et al, 2015 ; Faurholt-Jepsen et al, 2020 ; Faurholt-Jepsen et al, 2021 ; Goulding et al, 2023 ) in favour of the intervention group ( g = 0.20; 95 % CI 0.02 to 0.38, p = 0.031; see Appendix F for forest plot Fig. F.4) with low heterogeneity ( I 2 = 0.0 %; 95 % CI 0.0 to 84.7; PI -0.20 to 0.59).…”

“…The reported completion rate of the primary intervention programme in the intervention and control group ranged between 33.39 % ( Asayut et al, 2022 ) and 98.33 % ( Aharonovich et al, 2017 ), and no significant differences between the intervention and the control group were reported in 11 comparisons ( g = 0.01; 95 % CI -0.13 to 0.15, p = 0.852; I 2 = 0.0 %; 95 % CI 0.0 to 60.2; PI -0.13 to 0.19). The reported completion rate of the adjunctive app-based intervention programme ranged between 22 % ( Goulding et al, 2023 ) to 93 % ( Faurholt-Jepsen et al, 2015 ), from which five completed less and 13 completed more than half of the intervention programme.…”

“…However, this contrast is not solely due to the broader scope of the other meta-analysis; three out of four studies were identical. The distinctions arise from our inclusion of a more recent study by Goulding and colleagues ( Goulding et al, 2023 ), which contributed to a substantial single effect size that significantly impacted the overall effect. It is noteworthy that the other meta-analysis pooled pre-post change scores.…”