Blog

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

When HSBC unveiled a $13.6bn deal to take full control of its local Hong Kong lender on Thursday, chief executive Georges Elhedery was clear: after a year spent in retreat from high-profile businesses such as investment banking, the bank was back on the front foot.

“This is an investment for growth . . . for the medium to long term in what is a leading local bank in Hong Kong,” Elhedery said shortly after the announcement of the offer for the 37 per cent of Hang Seng Bank owned by other investors.

The bid is Elhedery’s first big strategic move that involves spending money rather than cutting back since he took charge in September last year.

Elhedery’s first term in office has been marked by his restructuring of Europe’s largest bank, with a cost-cutting plan that involved pulling the bank out of retail markets such as France and Argentina, and scrapping the HSBC investment bank in Europe and the US.

Another element of his plan however was to simplify HSBC’s structure into one that emphasised its two core geographical markets of the UK and Hong Kong. HSBC had US$9.1bn in pre-tax profit in Hong Kong in 2024 — 28 per cent of the total for the group — against $6.6bn in the UK.

Buying out minority investors in the Hong Kong bank has been welcomed by analysts as a long-awaited, common-sense simplification of the business in HSBC’s core market.

HSBC first bought a controlling stake in Hang Seng Bank, a local retail bank with a strong brand, in 1965 as a banking crisis hit the then-British territory — a deal that turned it into a dominant local player and ranks among HSBC’s most strategically significant transactions.

“Hong Kong has long been HSBC Holdings’ most profitable home market. We view the proposed transaction as a strategic redeployment of the substantial excess capital it is generating,” S&P analysts said on Friday.

Using HSBC’s excess capital to wholly privatise Hang Seng should boost the group’s capital ratios by ending the so-called “minority-interest deduction” — the accounting adjustment to HSBC’s cash buffer that represents the fact that Hang Seng was not fully owned by the European bank.

“The ability to be able to scale investments across both brands across the international network will be enhanced through this alignment,” Elhedery said on Thursday. “And it is more value generative for our shareholders than a share buyback.”

Investors’ reaction has been less positive: HSBC’s shares closed the week more than 5 per cent lower, weighed down in part by the bank’s decision to hold off new buybacks until at least the middle of 2026.

And after months of rumblings about Hang Seng’s exposure to Hong Kong property, there are questions about whether HSBC is making a billion-dollar bet on one of its twin home markets — or bailing out a troubled subsidiary.

Hang Seng is heavily exposed to the Hong Kong economy. It claims “close to 4mn customers”, almost all in Hong Kong — a city of 8mn. Most of its business comes from retail banking and lending to small and medium-sized companies — but it is particularly exposed to smaller Hong Kong real estate developers that are now under pressure.

China’s property bubble burst in 2021, imperilling some of the world’s largest developers. Hong Kong’s property market has followed suit, hit by rising interest rates, weaker demand and a loss of confidence in the territory following the anti-National Security Law protests and strict Covid-era lockdowns.

Hang Seng’s profits have fallen this year as interest rates have come down and demand for loans has declined — and the bank has ratcheted up its expected credit losses to reflect higher risk particularly on loans to Hong Kong’s real estate developers.

Pre-tax profits at Hang Seng dropped 28 per cent in the first half of 2025 compared with a year earlier to HK$8.1bn, and its non-performing loan ratio was 6.7 per cent: higher than even during the Asian financial crisis of the late 1990s.

Hang Seng’s difficulties also showed up in HSBC’s group results, which include Hang Seng. By the end of June, 73 per cent of its Hong Kong commercial real estate loans were marked as either impaired or as having increased credit risk, the Financial Times reported in August.

Two days after that report, Hong Kong Monetary Authority chief executive Eddie Yue weighed in, insisting that the city’s financial system remained strong.

“A recent media report highlighted the risks associated with CRE loans, with a particular focus on the accounting of banks’ ‘expected credit losses’,” Yue wrote in a blog post. “Hong Kong’s banking system is well-capitalised and has sufficient provisions and good financial strength to withstand market volatilities.”

But HSBC has nonetheless installed a new chief executive at Hang Seng, announcing last month that group veteran Luanne Lim would take over.

Some analysts said that even if this deal was not a bailout of Hang Seng, taking full control of the bank would allow HSBC to better handle any fallout from the property crisis.

“[The crisis] is HSBC’s responsibility, they need to take responsibility for it”, said Michael Makdad, an analyst at Morningstar. “If it were a choice between spinning off Hang Seng and taking 100 per cent control, then that is what matches the strategy and they have capital to do it.”

Rivals in the industry praised the move by HSBC to consolidate its operations in Hong Kong and get a handle on its exposure to struggling property investments.

“This has been a long-term goal for HSBC and now it is more politically possible,” said one former financial executive.

“Now it’s an easier time to gain control. This gets you the deposit base and in terms of dealing with the property market, it allows you to manage . . . without minority friction.”

Introduction

The global issue of illicit drug use has worsened, with 292 million users in 2022, a 20% increase over the past decade []. Cannabis is the most widely used illicit drug (228 million), followed by opioids (60 million), cocaine (23 million), and others []. Illicit drug users face various psychological and physiological problems, including mental disorders, cognitive deficits, cardiovascular dysfunction, and blood-borne infections. The social burden is also high, due to links with crime, violence, and sexual abuse []. Treatment is urgently needed, but globally, only about 10% of users receive treatment, a decline since 2015 [].

Traditional face-to-face psychosocial treatments remain important for illicit drug users but often fail to meet the needs of most patients due to time, location, and social stigma []. The COVID-19 pandemic accelerated the development of telehealth [] and pushed digital interventions from early simple interactions to more complex forms []. Modern digital interventions can provide multiple interaction methods via smart devices, such as apps, websites, email, text messages, video, audio, and computer programs. They overcome the limitations of traditional treatments and are valued for their flexibility and cost-effectiveness [-], better meeting personalized needs and improving treatment engagement []. Meta-analyses show that digital interventions are effective across different populations of illicit drug users [-].

However, dropout rates are particularly prominent in digital interventions [-]. Meta-analyses indicate that about one-third of individuals with substance use disorders fail to complete treatment [] and only 48% of early dropouts seek help again [], significantly increasing the risk of adverse outcomes [,]. Methodologically, the relatively high dropout rate limits the completeness of research findings, affecting the validity of results and the interpretation of treatment effects []. To improve the accuracy, this study clearly distinguishes three key concepts: engagement refers to behavioral involvement during use []; adherence reflects the alignment between actual behavior and intervention expectations []; while the dropout rate in this study is strictly defined as participants leaving, being lost to follow-up, or stopping participation before the outcome assessment for any reason. This conceptual clarification both distinguishes commonly confused terms and provides a methodological basis for enhancing the effectiveness of digital interventions, with important clinical implications.

Although the dropout rate is an important outcome indicator of intervention efficacy [], few studies have examined dropout rates among illicit drug users in digital interventions. A meta-analysis published in 2017 was the first to evaluate internet-based interventions in reducing illicit substance use after treatment and follow-up, but dropout rate was not the focus []. Moreover, existing research lacks systematic examination of clinical factors and intervention design, as well as dynamic assessment of dropout patterns at different time points [-], directly limiting the optimization of targeted intervention strategies.

Based on current research, this study aims to address the gap in dropout rate research in digital interventions. The study compared average dropout rates between the digital intervention and control groups to assess treatment retention under different experimental conditions. It also analyzed how variables at posttreatment and the longest follow-up time points affected dropout rates in the intervention group to support personalized intervention design for different research stages. These findings are important for advancing academic research and expanding clinical applications [].

Methods

Protocol Registration

This study strictly adheres to the guidelines of the Cochrane Handbook for Interventions [] and is reported according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines [] (the complete PRISMA checklist is available in ). The research protocol has been registered in the PROSPERO system: CRD42024534389.

Search Strategy

To comprehensively and systematically collect relevant literature, this study searched five major databases up to August 27, 2025, including Web of Science, PubMed, PsycINFO, Embase, and the Cochrane Controlled Trials Register. The search strategy combined controlled vocabulary (eg, MeSH terms) and free-text keywords using Boolean operators (“AND” and “OR”). The main search terms included the following: (“digital intervention” OR “internet intervention” OR “e-health” OR “m-health”) AND (“drug abuse” OR “substance use disorder” OR “illicit drugs”) AND (“psychotherapy” OR “psychoeducation” OR “psychodynamic”) AND (“randomized controlled trial” OR “single blind procedure” OR “random sample”). The complete search strategy for each database is provided in .

Inclusion and Exclusion Criteria

Inclusion criteria were as follows: (1) Individuals aged 18 years and above with illicit drug use behavior. Illicit drugs refer to controlled substances used for nonmedical or nonscientific purposes, including but not limited to cannabis, cocaine, amphetamines, and opioids []. (2) Digital psychosocial intervention is the primary treatment. Operationally defined as structured psychological intervention primarily delivered through digital platforms, including mobile applications, web-based programs, or digital communication tools, with or without minimal human support. (3) The article must report sample size and dropout rates. (4) Randomized controlled trials. Exclusion criteria were as follows: (1) treatment involving only face-to-face therapy. (2) mixed samples with insufficient proportion of illicit drug users (less than 80%) or without independent subgroup data (eg, alcohol and tobacco users). (3) non-English studies. (4) unpublished reports, study protocols, meta-analyses, reviews, doctoral theses, or other gray literature.

To ensure the accuracy of literature screening, a dual-screening process was adopted. First, two researchers independently screened the titles and abstracts of retrieved literature to exclude those clearly not meeting inclusion criteria. Subsequently, the full texts of the literature were reviewed for further evaluation. Finally, manual searches were conducted on the reference lists of included studies and related reviews to identify additional studies meeting inclusion criteria. Any disagreements were resolved through discussion.

Select Variables and Data Extraction

Outcome Variable

This study uses the dropout rate from randomized controlled trials as the primary outcome measure. Considering that the influencing factors at different treatment stages may vary [,-], the dropout rate data of the intervention and control groups at the end of treatment and at the longest follow-up time were extracted separately.

Moderator Variables

Previous studies have explored the factors influencing dropout among illegal drug users [], but due to differences in confounding variable control methods and insufficient understanding of the complexity of predictive factors, the results have been inconsistent []. Withdrawal from treatment is a dynamic process, and its mechanisms involve complex interactions of multiple factors []. It is difficult to fully explain the complexity of single-variable analysis []. Therefore, this study refers to previous research [] and selects multidimensional variables (): (1) Demographic characteristics of participants: most studies emphasize the role of patient-related variables in predicting dropout [,], and investigating individual differences (such as age, gender, race, digital literacy, etc.) is crucial for developing treatment interventions for specific populations [16]. (2) Baseline clinical characteristics of participants, including the type of illegal drug use, medication patterns, frequency of use, duration of use, and comorbid conditions. Different drugs may have differentiated effects on dropout rates due to their unique pharmacological mechanisms and withdrawal characteristics []. Additionally, the presence of comorbid mental disorders may exacerbate the likelihood of treatment interruption [], which also needs to be considered. (3) Therapist characteristics: the therapeutic orientation and experience level of therapists may be related to patient adherence []. Compared to busy clinic staff, full-time therapists are more likely to invest time and effort to retain and reengage patients who have discontinued treatment [32]. (4) Treatment characteristics: referring to the framework proposed by Derubeis et al [], which focuses on all factors that improve treatment and particularly on the relationship between treatment factors and outcomes. For example, this study extracted personalized feedback, real-time interaction, and therapeutic alliance. The optimization of these modifiable operational variables can directly enhance intervention effectiveness and improve patient treatment adherence [].

Data Extraction

Two researchers independently extracted data using a predesigned data extraction form. Disagreements were resolved through discussion or consultation with a third researcher. This data extraction form has been piloted in some studies and adjusted according to the recommendations and structured framework of the GRADE manual. For articles that met the inclusion criteria but lacked important data, we contacted the corresponding author via email, and studies that could not provide sufficient data to calculate effect sizes were excluded.

Quality Assessment

To assess the bias risk of the included studies, two researchers independently scored each study in five aspects using the revised Cochrane Risk of Bias tool ROB 2.0 []: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. Any disagreements were resolved through discussion.

Statistical Analysis and Software

We used Comprehensive Meta-Analysis software (CMA 4.0) to synthesize dropout rates across studies []. For each trial, dropout counts and total sample sizes were extracted separately for the intervention and control groups, from which group-specific dropout proportions were calculated. To stabilize variances and account for the bounded nature of proportions, these proportions were transformed into logit event rates with corresponding standard errors, which served as the primary effect size metric. Pooled estimates were calculated separately for intervention and control groups and subsequently back-transformed into raw proportions and expressed as percentages for interpretability, an approach that has been widely applied in meta-analyses of proportion-type outcomes []. Subsequently, between-study heterogeneity was examined using the Q statistic [,] and quantified with the I² statistic []. Given the significant heterogeneity among included studies in outcome measures and moderators [,], all analyses were conducted under a random-effects model []. Publication bias was assessed using funnel plots, Egger’s, Duval and Tweedie’s trim and fill, and Classic fail-safe N tests [], while sensitivity analyses were conducted to evaluate the robustness of the results. To explore potential influencing factors, meta-regression and subgroup analyses were further employed to examine the association between moderators in the intervention group and dropout rate.

Results

Characteristics of the Included Studies

After screening relevant articles based on predefined inclusion and exclusion criteria, a total of 41 studies were finally included (see ), involving 9693 participants with an age range of 19 to 50 years. The selection characteristics of the included studies are shown in . The studies included 82 intervention groups, with a total of 48 dropout rate data points, including 18 posttreatment dropout rates and 30 follow-up dropout rates, showing different data results between the two measurement points.

‎

Risk of Bias Assessment

The risk of bias in the included studies was assessed using the Cochrane Risk of Bias tool (ROB 2.0). Detailed results and percentage plots are presented in . The results showed that approximately 90% of the included studies had a low risk in terms of the randomization process (D1), measurement of the outcome (D4), and selection of the reported result (D5). Approximately 55% of the included studies had some concerns about deviations from intended intervention (D2). About 50% of the studies had a high risk of missing outcome data (D3), which is a key focus of our research.

Meta-Analysis Results

Posttreatment

An analysis of 18 studies was conducted using a random-effects model. The main effect results () showed that the mean dropout rate in the intervention group was 22% (95% CI 0.13‐0.36), lower than that in the control group of 26% (95% CI 0.16‐0.39) []. However, heterogeneity testing indicated high variability among the studies (Q=396.18, df=17, P<.001; I²=96%). Further analysis revealed that the variance of the true effect size reached 2.02 (logit units) with a standard deviation of 1.42 (logit units).

‎

Meta-regression and subgroup analysis revealed that this extreme variability was primarily due to four variables among three categories (): (1) Participant demographic characteristics: The proportion with employment rate showed a weak positive correlation with dropout rate (OR 1.04, 95% CI 1.00‐1.07; P=.03). (2) Participant clinical characteristics: Participants with baseline clinical diagnoses showed a significant positive correlation with dropout rate (odds ratio [OR] 1.03, 95% CI 1.01‐1.06; P=.01). The dropout rate for those using cocaine as the baseline primary medication (OR 1.96, 95% CI 0.31‐12.57; P=.48) was significantly higher than that for those using cannabis and opioid medications. (3) Intervention characteristics: Intervention frequency showed a significant negative correlation with dropout rate (OR 0.77, 95% CI 0.60‐0.99; P=.04). The other 27 factors showed no significant correlation with dropout rate.

The funnel plot showed some studies beyond the expected range (), suggesting the presence of studies with extreme dropout rates. Combined with Egger’s test results (P<.001), this further confirmed the presence of publication bias. After trimming the 5 missing studies on the right side, the effect size was adjusted from 22% to 33%, still not crossing the clinical threshold. Further leave-one-out analysis showed that 366 unpublished studies would need to be included to make the current result statistically insignificant. Overall, the results indicate that despite publication bias, the adjusted effect size did not exceed the clinical threshold and the leave-one-out number was high, supporting the stability of the study conclusions. Sensitivity analysis also showed () that removing any single study would not change the overall trend.

‎

‎

The Longest Follow-Up

Follow-up analysis of the intervention group was based on 30 studies, with an average dropout rate of 28.2% (95% CI 0.19‐0.39) (), while the rate in the control group was 27.8% (95% CI 0.20‐0.37). However, heterogeneity testing again indicated high variability among the studies (Q=1293.13, df=29, P=.000, I²=98%). Further analysis revealed that the variance of the true effect size reached 1.79 (logit units) with a standard deviation of 1.34 (logit units).

‎

Meta-regression analysis and subgroup analysis () revealed that this extreme variability is primarily due to 4 variables among three types of characteristics: (1) participant characteristics: dropout rate showed a negative correlation with single status (OR 0.95, 95% CI 0.91‐0.99; P=.01); (2) clinical characteristics: significantly positive correlation with baseline medication frequency (OR 1.18, 95% CI 1.05‐1.32; P=.004); (3) intervention characteristics: participants recruited via website showed a positive correlation with dropout rate (OR 5.74, 95% CI 1.85‐17.76; P=.002), while participants recruited via campus showed a negative correlation with dropout rate (OR 0.28, 95% CI 0.12‐0.66; P=.003); The association between the degree of digitalization and dropout rates varied depending on whether studies with unreported digitalization status (not reported [NR] group) were included. When all studies, including the NR group, were analyzed, the overall model reached statistical significance (Q=28.13, df=2, P<.001), with the NR group showing a strongly significant negative effect (OR 0.16, 95% CI 0.06‐0.41; P<.001). However, when the NR group was excluded and only studies explicitly reporting “fully digital” or “partially digital” were considered, the results were not statistically significant (Q=0.24, P=.62). The other 32 factors showed no significant correlation with dropout rate.

The funnel plot showed some studies beyond the expected range (see ). Combined with Egger test results (P=.023), publication bias was further confirmed. After trimming the six missing studies on the right side, the effect size changed from 28% to 37% after correction, without crossing the clinical threshold. Further leave-one-out sensitivity analysis showed that 1244 unpublished studies would need to be included to make the current results statistically insignificant, supporting the stability of the research conclusion. Meanwhile, sensitivity analysis (see ) indicated that the results of this study were robust and not dependent on individual studies.

‎

‎

Discussion

Principal Findings

This meta-analysis systematically evaluated the treatment retention effect of digital psychosocial interventions among adult illicit drug users. The pooled dropout rate was 22%, slightly lower than the approximately 30% reported for face-to-face psychosocial interventions [], suggesting potential advantages of digital formats for treatment retention. Nevertheless, the substantial heterogeneity across studies limits the generalizability of these findings. Dropout rates also varied across settings and populations. For instance, adults with co-occurring severe mental disorders and substance use had an average dropout of 27% [], whereas clinical samples of opioid users showed rates as high as 41% []. Beyond dropout, adherence constitutes another key indicator of engagement, with evidence showing that participants completed, on average, 60% of digital intervention modules, and only about half finished the full program []. Taken together, these results underscore the importance of considering both dropout and adherence when evaluating intervention effectiveness. Building on this, our moderator analyses further revealed complex interactive effects. To ensure clarity, we retained the classification system established during data extraction, presenting results separately across four major categories of characteristics as well as between short-term and longest intervention stages.

At the posttreatment stage, dropout was significantly influenced by participants’ demographic, intervention, and clinical characteristics. Regarding demographics, unemployment did not predict dropout, whereas higher employment was unexpectedly associated with greater attrition. This suggests that unstable or high-intensity work may interfere with regular participation. In addition, the short-term income from employment may reduce some patients’ motivation for treatment, especially when symptoms temporarily improve, leading them to discontinue prematurely due to “feeling better” []. For intervention characteristics, intervention frequency showed a negative correlation with dropout, indicating that more frequent contact may help consolidate behavior change, strengthen the therapeutic alliance, and enhance commitment [-]. Future studies should explore the optimal intervention frequency under different conditions [], balancing treatment intensity with patient burden [].

The results of baseline clinical characteristics indicated that both baseline clinical diagnosis and baseline cocaine use were significantly positively associated with dropout rates. Specifically, patients with a clear baseline diagnosis were at greater risk of dropout due to challenges such as dependency, withdrawal symptoms, and impaired cognitive or emotional functioning []. For this population, the integration of adjunctive pharmacological or behavioral therapies is recommended to reduce dropout []. Furthermore, consistent with previous findings [], participants with baseline cocaine use were more likely to discontinue treatment. Cocaine use disorder is often closely linked to impulsive behavior and diminished adherence []. These substance-specific risks highlight the importance of developing differentiated intervention strategies tailored to distinct types of substance use in future research []. Nevertheless, the small sample size of drug-use subgroups (k≤5) remains a limitation, which could be addressed through multi-institutional collaborations to expand subgroup samples.

During the longest follow-up, dropout was significantly influenced by demographic, clinical, and intervention characteristics. In demographics, a higher proportion of single participants was linked to lower dropout. This may be related to reduced drug exposure in family environments [-]. In addition, single participants with low social support were more likely to continue seeking health information online. Future research could involve non–drug-using significant others in monitoring the intervention process and integrate peer support modules []. In clinical characteristics, participants with higher baseline drug use frequency faced markedly greater dropout risk. This finding is consistent with recent studies []. For this high-risk group, we recommend the implementation of multistage intensive intervention programs [], together with the development of immediate-response modules (eg, crisis management tools, real-time consultation functions) to reduce early dropout [].

In terms of intervention characteristics, participants recruited through websites exhibited higher dropout rates, whereas those recruited from campus showed lower dropout rates. This may be explained by the lack of intensive treatment services typically provided in clinical settings, as well as the relative stability of campus environments [,]. Based on this finding, we recommend adopting a mixed online–offline recruitment strategy []. In addition, intervention content should be optimized for online recruits [], including simplifying operational procedures, providing regular reminders, and offering personalized feedback. The study also analyzed the association between the degree of digitalization and dropout rates. During data processing, studies that did not report their digitalization status (23/30, 77%) were categorized separately as a “Not reported” group for analysis rather than being directly excluded. The analysis revealed a significant association: compared to the nonsignificant negative correlation between fully digital interventions and dropout rates, interventions with unreported digitalization status showed a significant negative correlation, while non-fully digital interventions demonstrated a significant positive correlation with dropout rates. However, the reliability of these subgroup comparisons is constrained by the prevalent issue of poorly reported data. When we excluded the “Not reported” studies and repeated the analysis, no significant differences were found between fully digital and partially digital interventions. This suggests that the initial findings were likely confounded by nonrandom reporting bias rather than reflecting true effects, making definitive evaluation difficult. Therefore, these results primarily highlight the urgent need for future research to standardize the reporting of specific digital intervention details in order to more reliably explore the role of digitalization degree and human support in improving retention rates [].

Research Significance

This study systematically evaluated the dropout rate and its predictive factors among adult illicit drug users in digital psychological interventions, thereby addressing a critical research gap in the field. Unlike previous studies that primarily focused on demographic characteristics, this analysis incorporated multidimensional predictive variables—including clinical features, therapist-related factors, and intervention characteristics—to establish a more systematic theoretical framework. The identification of eight key predictive factors provides valuable insights for personalized interventions, guiding the development of tailored digital tools for patients at high risk of dropout. Optimization strategies derived from this evidence are expected to substantially reduce dropout rates and enhance intervention effectiveness [].

Limitations and Future Research

This study has several limitations. First, few of the included trials provided detailed information on software quality or reasons for dropout, which limited our ability to assess the reasons why participants stopped treatment []. Future studies could combine machine learning methods to predict dropout risk [] and use participant-centered questionnaires to collect data on perceived barriers. Previous research [-] emphasized common reasons for dropout, including technical difficulties, lack of engagement, and perceived ineffectiveness of the intervention. Collaboration with software engineers may help optimize the digital experience and reduce technical-related attrition []. Additionally, methodological improvements, such as combining intention-to-treat analysis with run-in phase dropout screening [,], may provide more refined methods for managing early dropout.

Second, most of the digital interventions included in the studies adopted limited forms, such as videos, virtual characters, or text messages, and lacked interactive features. Incorporating gamification elements may enhance user engagement [], especially when personalized to individual preferences [,]. Emerging evidence suggests that well-designed therapeutic video games can improve cognitive and mental health outcomes [], even inducing neurobiological changes, including alterations in white matter microstructure [-].

Finally, many studies did not clearly report key methodological details, such as the degree of digitalization or level of human support. Although we conducted analyses including and excluding the “Not reported” category, the lack of such information led to inconsistent findings, preventing definitive conclusions regarding the impact of digitalization on dropout rates. Future studies should standardize reporting of intervention details, including digitalization and human support, to better understand active components and optimize strategies [,]. Another limitation is the high heterogeneity in the meta-analysis (I²>90%), which may reduce robustness. Despite sensitivity and moderator analyses, some variability remained unexplained, suggesting pooled effects may not apply equally across interventions, populations, or outcomes. Future research should adopt rigorous methodologies, including detailed reporting, preregistration, data sharing, and large-scale RCTs. Individual participant data meta-analyses can further clarify subgroup effects and sources of heterogeneity, improving generalizability [].

Conclusion

In summary, this meta-analysis systematically examined dropout rates and their predictive factors in digital psychosocial interventions for adult illicit drug users, aiming to provide a comprehensive picture of the research landscape in this field. The results indicate that both short-term and long-term adherence to interventions are characterized by considerable complexity. In the short term, dropout rates were primarily associated with employment status, baseline clinical diagnoses, baseline primary substance use, and intervention frequency. Over longer follow-up periods, marital status, baseline drug use frequency, and recruitment source emerged as key predictors. These findings suggest the need for further investigation into factors that contradict common assumptions or remain insufficiently reported in the literature, as well as greater standardization in the design, measurement, and reporting of randomized controlled trials to improve research quality. Moreover, more attention should be given to tailoring interventions for specific populations, particularly through the design of intervention functions and modules. Continued exploration in these areas will contribute to better supporting patients’ long-term recovery.