Comparison of Cost-Effectiveness Between Digital Health Interventions and Pharmacotherapy for Depression: Systematic Review (Preprint)

<sec> <title>BACKGROUND</title> Owing to the unique characteristics of digital health interventions (DHIs), a tailored approach to economic evaluation is needed—one that is distinct from that used for pharmacotherapy. However, the absence of clear guidelines in this area is a substantial gap in the evaluation framework. </sec> <sec> <title>OBJECTIVE</title> This study aims to systematically review and compare the economic evaluation literature on DHIs and pharmacotherapy for the treatment of depression. </sec> <sec> <title>METHODS</title> We searched for articles published between January 2013 and October 2023 in Ovid MEDLINE, Embase, Cochrane Library, and PsycINFO databases. Studies were eligible if they evaluated DHIs or pharmacotherapies for depression and reported economic outcomes. We extracted data on the study characteristics, input parameters, and economic evaluation modeling components. Chi-square tests were used to analyze the frequency of various components across intervention types. A qualitative comparison was performed to assess the costs, effects, and modeling aspects of each intervention. The Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist was used to evaluate the quality of the selected studies. </sec> <sec> <title>RESULTS</title> A total of 42 articles were included, of which 23 (23/42, 55%) focused on DHIs and 19 (19/42, 45%) on pharmacotherapy. Cost-utility analysis was used more frequently in pharmacotherapy (16/19, 84%) than in DHIs (12/23, 52%), with a significant difference between the 2 intervention types (&lt;i&gt;P&lt;/i&gt;=.01). Similarly, the types of comparators differed significantly, with DHIs more often being compared to usual care (12/23, 52%) or waitlist controls (5/23, 22%) and pharmacotherapy studies mainly involving active controls (17/19, 89%; &lt;i&gt;P&lt;/i&gt;&amp;lt;.001). In addition, pharmacotherapy was more likely to be used in model-based studies (13/19, 68%), whereas DHIs predominantly relied on trial-based studies (17/23, 74%; &lt;i&gt;P&lt;/i&gt;=.006). Although not statistically significant (&lt;i&gt;P&lt;/i&gt;=.28), a notable trend was observed: the payer perspective was most commonly applied in pharmacotherapy studies (10/19, 53%), compared with approximately 30% (7/23) in DHIs. Furthermore, studies with a time horizon exceeding 12 months were more common for pharmacotherapy (5/19, 26%) than for the DHIs (3/23, 13%). Assessment using the CHEERS checklist indicated that pharmacotherapy studies generally had higher reporting quality compared with the quality of DHI studies in areas such as study parameters, comparators, time horizon, and discount rate. </sec> <sec> <title>CONCLUSIONS</title> Compared with pharmacotherapy, DHIs involved a higher proportion of trial-based studies reporting short-term outcomes and studies with ambiguously defined cost items. This underscores the need for improved measurement and modeling to accurately capture the costs and effectiveness of DHIs. </sec> <sec> <title>CLINICALTRIAL</title> PROSPERO CRD42023471565; https://www.crd.york.ac.kr/PROSPERO/wiew/CRD42023471565 </sec>