CHATBOT-DELIVERED PSYCHOTHERAPY FOR HELPING DEPRESSIVE AND ANXIOUS ADULTS: A SYSTEMATIC REVIEW AND META-ANALYSIS OF RANDOMISED CONTROLLED TRIALS

Abstract

Background: Depression and anxiety are major causes of disability worldwide. Psychotherapy has been widely adopted to treat depression and anxiety. Chatbot technology is an efficient and cost-effective platform to deliver psychotherapy for the improvement of depressive symptoms. Objectives: (1) To synthesize the available evidence that examine the effectiveness of chatbotdelivered psychotherapy in improving depressive symptoms among adults with depression and anxiety; (2) evaluate the preferred features for the design of chatbot-delivered psychotherapy. Methods: A systematic search of eight databases (EMBASE, PubMed, CINAHL, Cochrane, PsycINFO, SCOPUS, IEEExplore and ProQuest) was administered from commencement till 31 December 2019. Cochrane risk of bias tool was utilized for individual trial quality assessment. GRADE system was used to assess the overall quality of evidence. Meta-analysis was conducted using RevMan 5.3 software. Assessment of heterogeneity was done using ??2 and I2 tests. Overall effect was determined using z-statistics at significance level p < 0.05. Results: A total of 5,180 records were identified. 11 RCTs were included for meta-analysis involving 1,099 depressive and anxious adults across four countries. Meta-analysis revealed chatbot-delivered psychotherapy significantly improved depressive symptoms (z= 8.42, p < 0.001) of moderate to large effect size and quality of life (QoL) (z= 2.24, p = 0.03) with a small to moderate effect size, versus the control group. Subgroup analyses revealed chatbot embodiment, combination of input and output formats, unstructured content delivery, problem solving therapy and offline platforms are preferred features of chatbot-delivered psychotherapy. Conclusion: Chatbot-delivered psychotherapy are effective in improving depressive symptoms and QoL among adults with depression and anxiety. Chatbot features preferred include embodiment, combined forms of input and output, unstructured content, problem solving therapy and offline platforms. Implications: Chatbot-delivered psychotherapy can be adopted in healthcare institutions as an alternative treatment for depression and anxiety. Future trials should adhere to the guidelines recommended by the Consolidated Standards of Reporting Trials Statement in the conducting and reporting of trials to ensure reliability of findings.