In June 2024, the first author (GJ) conducted a literature search in Google Scholar for digital, integrated interventions for MDD and SUD. Eligible articles included one of the terms in Table 1 in the title, abstract, keywords, or article text. The reviewer also screened the reference list of all eligible articles, plus a search of all articles that had cited the included papers.

A report of a digital intervention trial of MDD and SUD treatment was included if (1) the article reported on an intervention intended to improve substance use and depression simultaneously, and encompassed various severity levels from mild symptoms to diagnostic thresholds; this included participants diagnosed with MDD and SUD, as well as those without confirmed diagnoses; (2) the treatment described was grounded in CBT, MI, or MET; (3) the treatment was delivered entirely or partially via digital modality (computer, smartphone, internet, and text message) outside of the traditional F2F therapy setting; and (4) the article was written in English as a peer-reviewed journal article. Studies were excluded if they (1) targeted only depression or only substance use, (2) were published protocols without preliminary outcomes or findings, (3) did not specify theoretical underpinnings or an evidence-based rationale for the treatment, or (4) were based on approaches other than CBT, MI, or MET.

For each study included in the review (Multimedia Appendix 1), we extracted and summarized inclusion and exclusion criteria, study design, measures used to assess MDD, SUD, treatment adherence, and other outcomes the digital modality used, as well as study duration.

Twenty-five treatment strategies (Table 2) were coded in July and August 2024. The content analysis coded any mention of used treatment strategies. The intensity of treatment strategies, specifically the frequency of their presentation within the interventions, was not systematically assessed. In addition, the authors did not independently verify whether these strategies were actually implemented beyond the descriptions provided in the articles and associated supplemental materials. In total, three authors (GJ, QG, and HA) coded all reports (n=14) using an open-coding approach [33]. Initially, we conducted a line-by-line coding of each original article, allowing treatment strategies to emerge directly from the data without using a predefined codebook. This inductive process was followed by team discussions to compare our codes to existing evidence-based treatment strategies [13,17,34-37]. Any disagreements in coding were resolved through consensus among the researchers to confirm that the final coding was consistent with and reflective of the included studies [38].

The number, frequency, and percentage of treatment strategies used in the included interventions were summarized using R (version 4.3.2; R Core Team). Plots summarizing the frequencies and overlap of treatment strategies were generated using the “ggplot2” [39] and “pheatmap” [40] packages in R.

Of the 14 studies that met inclusion criteria (Multimedia Appendix 1) [41-55], half of the studies (n=7, 50%) cited a protocol paper with an in-depth description of the research methodology, including participant recruitment procedures, data collection methods, and statistical analysis plans [56-62].

This study aimed to review integrated digital interventions that support individuals with co-occurring MDD and SUD, focusing on study design and characteristics, demographic inclusivity, and therapeutic strategies. Our findings indicate that these interventions often exhibit significant methodological variability, use inconsistent diagnostic criteria, lack demographic diversity, and implement treatment strategies inconsistently. Addressing these issues is critical for making digital integrated interventions more effective and widely applicable for diverse populations with different levels of comorbidity.

Comorbid MDD and SUD vary widely in severity and affect individuals across diverse racial, ethnic, financial, and educational backgrounds. A main finding of our review was the considerable variation in inclusion and exclusion criteria across studies. Most studies targeted individuals with mild to moderate depression and hazardous substance use, with few requiring clinically validated diagnoses. This focus on mild cases and alcohol use is consistent with current F2F treatments, which primarily address depression or anxiety alongside alcohol misuse [18,76,77]. Emerging interest in cannabis misuse [78,79] is beginning to address a broader scope, but significant gaps remain regarding the effectiveness of digital interventions for other substances. While F2F treatments have tested efficacy in clinically diagnosed MDD and SUD [80], the current focus of existing digital interventions on individuals with subclinical or mild-to-moderate symptom severity raises questions about their applicability for those on the more severe end of the comorbidity spectrum.

Demographic data in these studies were generally limited to age and gender, with few studies reporting on racial or socioeconomic diversity. Diverse sample recruitment is critical for advancing equitable mental health care [81]. Known disparities in MDD and SUD treatment access, where some racial groups face unique barriers such as language differences, cultural mistrust, and lack of social support, demonstrate the need for detailed demographic reporting [82]. Without such data, particularly for socially vulnerable communities, assessing whether these interventions are culturally relevant, effective, or capable of providing necessary access is challenging.

Finally, most studies targeted adult populations, with a limited focus on young adults. This is a significant gap, as young adulthood is a formative period where substance use and mood symptoms have the potential to escalate [83,84]. Early intervention for young adults could promote healthier coping mechanisms, resilience, and overall well-being, potentially thwarting the onset of more severe comorbidity later in life and reducing health care burden [85]. Programs for young adults should be tailored to reach individuals at risk or showing early signs of depression or substance use, even if they have not yet been clinically diagnosed.

Future research should expand inclusion criteria to encompass a broader range of MDD and SUD severity, including clinically validated diagnoses, to evaluate these interventions across diverse clinical profiles. It should also expand beyond alcohol to other substances to address gaps in integrated treatments for varying substance use types. Finally, improved demographic reporting is essential to ensure these interventions are accessible, equitable, and effective across diverse racial, ethnic, and socioeconomic backgrounds.

Our review found significant methodological variability across studies, including differences in study design, treatment length, therapist involvement, follow-up periods, outcome measures, and adherence reporting. Study designs varied, with some comparing interventions to active treatments, while others only used assessment or psychoeducation as controls. Treatment duration ranged from a single session to 24 weeks, and therapist involvement also differed, with some interventions being guided and others self-guided. Follow-up periods were inconsistent, from no follow-up to 12 months. Outcome measures varied widely, with depression assessed using tools like the BDI-II and PHQ-9 and substance use measured by tools such as the Timeline Follow-Back with monitoring periods from 7 to 90 days. Adherence reporting was inconsistent as well, with some studies tracking total modules completed and others measuring time spent on the intervention.

Standardizing methodological components would support the identification of effective strategies across diverse populations and clinical settings. For instance, brief smartphone-based interventions may be ideal for community settings or ongoing recovery support, whereas longer, more intensive computer-based programs may better serve individuals requiring intensive, structured care. Standardization would allow for clearer comparisons across studies and more precise matching of intervention tools to treatment needs and stages.

The reviewed interventions consistently included evidence-based strategies grounded in CBT, MI, and MET, though fewer than half provided detailed session-by-session descriptions. The number of strategies used varied widely, with web-based interventions including the most (3 to 19 strategies), followed by smartphone apps, computer-based programs, and supportive text messaging, which used the fewest.

The most frequently used strategies across digital modalities were self-monitoring, psychoeducation, coping skills, feedback, and assessment. About half of the studies incorporated activity scheduling, decisional balance, goal setting, and reminders, while less common strategies included time management, relaxation, interpersonal effectiveness, and assertiveness training. These findings align with existing literature on CBT for MDD and SUD. For instance, a study of CBT experts rating effective components of CBT for depressed adults emphasized the importance of psychoeducation, feedback, and activity scheduling and monitoring [86], reflecting the presence of these strategies in our findings. Similarly, a meta-review of effective interventions for alcohol use disorder identifies feedback, goal setting, and self-monitoring as commonly used strategies, further supporting the frequent use of these strategies in the included studies [87]. However, strategies such as assertiveness training, drug refusal skills, and contingency management, important for SUD treatment [88-90], were less frequently used in the digital interventions we analyzed, highlighting areas for further exploration, especially in addressing the substance use aspect of these integrated treatments.

To our knowledge, this study is the first to map the overlap of integrated digital strategies for MDD and SUD. The heatmap (Figure 2) shows that strategies like psychoeducation often overlap with assessment, goal setting, feedback, and coping skills, covering core elements of ICBT [16]. However, the map also reveals a “scattered” implementation pattern across different modalities. Although CBT is a protocolized treatment, the map demonstrates that there is no standardized approach to integrating these evidence-based strategies, potentially contributing to the variability in treatment outcomes. Some critical strategies for MDD and SUD, such as activity planning (for depression) [91] and drug refusal skills (for substance use) [87], appear less frequently despite being widely recognized as effective in the literature. While future research should explore innovative ways to adapt these more challenging strategies to a digital format, it’s important to acknowledge that some may not fully translate, presenting a persistent limitation of digital approaches. To address this challenge, integrated hybrid models that combine digital interventions with in-person components may offer a more effective solution [92,93].

The scattered distribution also suggests that some interventions may over-rely on strategies that lend themselves well to digital formats (eg, psychoeducation and self-monitoring) while neglecting others that, although potentially more effective, are harder to adapt digitally (eg, practicing drug refusal skills, interpersonal effectiveness, or roleplaying). The nature of some modalities, such as text messaging, inherently limits the strategies that can be implemented. For instance, the simplicity and brevity required for text messaging make it difficult to include more complex strategies, which may require interactive or intensive engagement, such as practicing communication skills or cognitive restructuring. Researchers should recognize that different digital modalities may be suited to different purposes; for example, text messaging may be better suited for ongoing recovery support, while smartphones or internet-based platforms might be more appropriate for delivering more detailed, longer-duration interventions. This understanding should guide the design and evaluation of future digitally integrated interventions, ensuring that each modality is used for its strengths rather than attempting to incorporate all treatment strategies, which could overwhelm the capabilities of certain modalities.

The variability in strategy implementation within digital treatments for MDD and SUD identifies areas for advancement in research and clinical practice. While reflecting the flexibility of digital interventions, these inconsistencies make it difficult to understand whether the issues lie within the digital format itself or the strategies themselves. Future research should focus on testing specific strategies within defined digital modalities to clarify their impact. Comparative studies could, for instance, assess whether strategies like psychoeducation, self-monitoring, and activity planning are more effective on particular platforms (eg, smartphone apps) for specific demographics, such as young adults or those with severe SUD. Identifying the most impactful strategies will help tailor interventions to better meet patient needs based on comorbidity severity.

This study reviewed all available digitally integrated treatments for MDD and SUD rather than limiting to only one kind of digital modality. The results offer a broad picture of the present status of digital treatments and their treatment strategies. This review will guide future research and development of these tools by encouraging developers and researchers to be more transparent about the evidence-based strategies incorporated in these interventions, contributing to the effectiveness or lack thereof. This review is limited in that it serves as a descriptive analysis, primarily focused on observing and cataloging the characteristics of the individual treatments and identifying frequencies of treatment strategies used in integrated digital interventions for MDD and SUD. While it provides an idea of the strategies used across different modalities, this study lacks the methodological grounding necessary to draw definitive statistical connections between specific strategies and their effectiveness. As a result, we cannot confidently infer which strategies are most closely associated with successful treatment outcomes or which combinations might be more effective in different contexts. Furthermore, the description of intervention characteristics was based on whether the study provided session-specific details about the implemented evidence-based strategies delivered during the intervention, which may have limited the precision of our treatment coding. Therefore, this study’s findings should be viewed as preliminary, observing trends that warrant further investigation.

Digital integrated treatments for MDD and SUD have the potential to expand access to evidence-based treatments for individuals with comorbid MDD and SUD. Existing digital integrated treatments offer a wide range of treatment strategies. However, methodological variability in research to date makes summarizing efficacy challenging at this time and highlights study reporting details that will be important for future replicability and generalizability of this literature.