Mobile health interventions for substance use reduction: a systematic review

Review question(s)

What are the effectiveness, delivery methods, engagement strategies, challenges, and moderating factors of mHealth interventions for substance use reduction among students in low- and middle-income countries (LMICs)?

Eligibility criteria: Inclusion criteria

Types of studies

The review focused on randomized controlled trials (RCTs), quasi-experimental designs, and qualitative studies conducted in LMICs to evaluate mHealth interventions for substance use reduction among students. RCTs provided robust quantitative evidence on intervention effectiveness, while quasi-experimental studies expanded the evidence base by capturing real-world applications in resource-limited settings, enhancing external validity. Qualitative studies offered deeper insights into students’ experiences, engagement strategies, cultural considerations, and barriers to success, aligning with the review’s focus on user satisfaction and contextual fit. Together, these study designs enabled a comprehensive assessment of intervention effectiveness, contextual influences, and user experiences, addressing the research objectives thoroughly and providing a well-rounded understanding of mHealth interventions in LMICs.

Types of participants

The review focused on studies that included students aged 16 to 25 years residing in LMICs.

Types of intervention(s)

The review focused on studies that reported on the use of mHealth interventions for substance use reduction specifically targeting young adult students in LMICs.

Types of comparators

The review included comparator studies involving students where participants received no intervention, only assessment, or standard care.

Outcomes

This review examined mHealth interventions designed to reduce substance use among young adult students in low- and middle-income countries (LMICs), evaluating their effectiveness, delivery methods, engagement strategies, and overall impact.

The review highlighted key elements that contribute to the success of mHealth programs, such as increasing self-awareness, building coping skills, fostering behavior change, and providing access to support networks. It also discussed the challenges and limitations of these interventions, while emphasizing mHealth’s potential as a valuable tool for addressing substance use in this population. Notably, an automated mHealth supplement designed to improve substance use behaviors was highlighted as a promising approach.

To assess the effectiveness of these interventions, studies utilized validated self-report tools such as the Alcohol Use Disorders Identification Test (AUDIT) and self-administered surveys on alcohol and smoking frequency. The AUDIT, a 10-item questionnaire developed by the World Health Organization, identifies hazardous drinking behaviors by assessing alcohol consumption, dependence symptoms, and associated problems. With strong validity and reliability (Cronbach’s alpha values ranging from 0.80 to 0.90), the AUDIT (https://www.who.int/publications/i/item/WHO-MSD-MSB-01.6a) is an effective tool for identifying harmful drinking patterns among students in LMICs.

Self-administered surveys, which assess the frequency of substance use within specific timeframes, also demonstrated strong face and content validity when questions were clear and culturally appropriate. Despite the potential for recall bias, these surveys are practical in LMIC settings, enabling efficient tracking of substance use patterns.

Together, these validated instruments support the conclusion that mHealth interventions can effectively reduce substance use among young adult students in LMICs, enhancing the credibility of the review’s findings.

Information sources

The databases searched included PubMed, Psych INFO, SCOPUS, Google, and Google Scholar.

Search strategy

The search strategy aimed to identify published and unpublished studies from 1991 onward, aligning with the rise of text messaging technology. A three-step approach was used: First, a limited initial search was conducted on PubMed, followed by an analysis of the text words in the titles and abstracts, as well as the index terms used for the identified articles. Next, a second search was carried out across all included databases using the identified keywords and index terms. Finally, the reference lists of all identified reports and articles were examined for additional studies. Only English-language studies were included, potentially limiting the review by excluding non-English research. Keywords covered students, mHealth, mobile interventions, text messaging, smartphones, and various substance use terms (e.g., alcohol, tobacco, cannabis). This comprehensive strategy ensured a thorough identification of relevant studies while acknowledging potential language-based limitations (see extended data: Appendix I: Search strategy).

Selection process

We conducted searches for technical reports and manually reviewed the reference sections of the resulting articles. Studies were selected, if they included young adults reporting substance use, incorporated a mHealth intervention component, and included an outcome measure. Two reviewers independently screened each record and all retrieved reports.

Data extraction

In line with the review questions and objectives, data from full-text articles were summarized, extracted, and manually organized into a table format. Two authors (MTG and KHA) independently conducted the data extraction. We gathered information on the primary author(s), the country or countries of origin, study design, age group of the population, study setting, sample size, effectiveness of mHealth interventions, type of intervention, target substance, student characteristics (age, gender, socioeconomic status), the intervention characteristics, follow-up duration, outcome assessment, frequency of mHealth interventions, study aim, key findings, purpose of the mHealth technology, and the platform used.

Study risk of bias assessment

Eligible papers selected for retrieval were assessed for methodological validity by two independent reviewers (MTG and KHA) prior to their inclusion in the review. Standardized critical appraisal tools were utilized: the Cochrane Effective Practice and Organization of Care (EPOC) guidelines for randomized controlled trials and quasi-experimental studies,²⁴ and the JBI Critical Appraisal Checklist for qualitative studies.²⁵ If needed, the authors of the original papers were contacted for clarification on relevant issues. Any disagreements between the reviewers were resolved through discussion or by involving a third reviewer (FA or DA).

Synthesis methods

The findings of the systematic review were synthesized narratively, around key themes, detailing aspects such as study design, author, country, participant demographics, sample size, intervention type, target substance, effectiveness, follow-up duration, and outcomes. It also explores mHealth platforms, usage frequency, and study aims, providing a comprehensive overview of the field. By summarizing methodologies and key findings, this narrative offers valuable insights into the current landscape of mHealth interventions for substance use reduction in LMICs, highlighting their potential and challenges.

Reporting bias assessment

The review used two critical appraisal tools to evaluate bias risk: the Cochrane EPOC checklist for randomized and quasi-experimental studies, and the JBI checklist for qualitative evidence. The Cochrane EPOC checklist assesses nine items, including sequence generation, allocation concealment, baseline comparability, incomplete data, intervention awareness, contamination, selective reporting, and other biases. The JBI checklist evaluates ten items, such as methodology alignment, data collection and analysis, interpretation, researcher influence, participant representation, ethical approval, and conclusions. These tools systematically assessed study rigor and reliability, strengthening the evidence base evaluation and ensuring a robust analysis of mHealth interventions for substance use reduction in LMICs.

Study selections and extractions

A total of 153 abstracts were identified from the databases PubMed, PsycINFO, SCOPUS, and Google Scholar. After removing 48 duplicates, 105 studies remained for title and abstract screening. Following the exclusion of 41 studies, 64 articles were retrieved, though 10 were unavailable. Fifty-four articles underwent full-text assessment, resulting in the exclusion of 49. Ultimately, five studies met the inclusion criteria and were included in the final analysis, all of which were published in English ( Figure 1).

Figure 1. PRISMA Flowchart describing the selection of studies reviewing existing literature on mHealth for substance use reduction in young adult students in the LMIC.

Study characteristics

Of the five studies reviewed, three utilized a randomized controlled trial (RCT) design,^26–28 one employed a quasi-experimental design,²⁹ and one was a qualitative study.³⁰ The sample sizes in these studies varied considerably, ranging from as few as 20 participants³⁰ to as many as 772,²⁶ with other studies involving 273,²⁷ 130,²⁹ and 179²⁸ participants. The sample size is a critical factor for ensuring statistical power and the reliability of results. Smaller sample sizes can limit the ability to detect significant effects, increasing the risk of Type II errors and compromising the generalizability of the findings. Future research should incorporate power analyses to determine appropriate sample sizes that align with the expected effect sizes and research outcomes, thereby ensuring more accurate and generalizable conclusions. Replication studies with larger samples are also encouraged to confirm results and enhance the robustness of the evidence.

The studies in the review used a range of sampling methods, including purposive, random, simple random, random selection, and cluster randomization, each of which influences the results and their generalizability. While purposive sampling targets specific groups, it carries the risk of selection bias, which can limit generalizability. Combining purposive sampling with random sampling within subgroups may improve the representativeness of the sample. Random and simple random sampling methods, although robust, can present logistical challenges and may not fully capture demographic variations. Stratified random sampling, which divides the population into subgroups, offers better representation of key demographics, improving generalizability. Cluster randomization, while practical in large-scale studies, may increase sampling error and reduce generalizability due to similarities within clusters. To address this, increasing diversity within clusters or employing multi-stage sampling can help mitigate these issues. Future studies should consider using combined sampling methods, such as stratified or multi-stage sampling, to balance feasibility with representativeness. Additionally, collecting demographic data to adjust for biases and ensuring diversity within clusters can further enhance the validity and generalizability of the findings. Although each sampling method has its limitations, strategically combining methods and adjusting for biases can significantly improve the robustness and generalizability of research outcomes.

Psychometric properties

The table 1 (refer extended data) presents key instruments used in these studies to assess outcomes, detailing their psychometric properties, including validity and reliability. By highlighting these characteristics, the measures’ dependability in tracking changes in alcohol and smoking behaviors is emphasized. These robust tools strengthen findings, supporting conclusions on mHealth interventions’ effectiveness for substance use (Refer extended data: Table 1).

Outcome assessment

All studies in the review used self-reported data for outcomes. One study recruited participants via purposive sampling based on Alcohol Use Disorder Identification Test (AUDIT) scores ≥8 to explore instant messaging (IM) apps for alcohol intervention.³⁰ Another compared IM and text messaging for alcohol reduction, using AUDIT for measurement.²⁶ A third evaluated a smoking cessation app, using the Heaviness of Smoking Index (HSI) to compare intervention and control groups.²⁷ Another assessed an educational intervention for smoking prevention, measuring outcomes via the Health Belief Model Questionnaire.²⁹ The final study examined text messaging for smoking cessation, using the Attitudes Towards Smoking Scale (ATS) and Fagerström Test for Nicotine Dependence (FTND) to compare groups.²⁸

Behavioral change techniques

This review examines behavioral change techniques (BCTs) in interventions, focusing on strategies rooted in theories like the Health Belief Model (HBM), Social Cognitive Theory (SCT), and Trans theoretical Model (TTM). Qualitative insights reveal that theory-based techniques—self-monitoring, goal-setting, risk awareness, and social support—enhance motivation and reinforce positive behaviors. Participants valued social and interactive elements, underscoring the importance of tailored, supportive approaches in mHealth interventions for youth and high-risk behaviors (Refer extended data: Table 1).

MHealth delivery methods

Several studies provided participants with a standardized alcohol brief intervention (ABI) at baseline, delivered face-to-face or via video conferencing by research nurses. The ABI included personalized feedback based on AUDIT risk levels and a 12-page booklet on alcohol’s health impacts. One study offered three months of chat-based instant messaging support for alcohol reduction to the intervention group, while the control group received SMS messages on general health topics.²⁶ Another evaluated the “Quit with US” app for young adult smokers, with assessments at baseline and 12 weeks.²⁷ A third study delivered a three-month educational intervention via Telegram, focusing on smoking harms and physical activity benefits for the experimental group, while the control group received no intervention.²⁹ Lastly, a study assessed a 12-week text messaging-based smoking cessation intervention among Chinese vocational school students.²⁸

The primary mHealth delivery methods included mobile apps, SMS messaging, and social media-based interventions, each offering unique advantages. SMS remains widely used due to its simplicity and affordability, while instant messaging apps are increasingly popular among younger demographics for their interactive features. These methods provide flexible, tailored approaches to address substance use in diverse technological environments (Refer to extended data: Table 2).

Engagement strategies in mHealth interventions

The review explored mHealth strategies targeting substance use among students in LMICs, focusing on smoking and alcohol interventions. One study examined instant messaging (IM) apps for alcohol reduction among university students, finding high acceptability and perceived utility for reducing drinking behaviors.³⁰ A follow-up study demonstrated the effectiveness of a mobile chat-based alcohol brief intervention (ABI) in reducing alcohol consumption.²⁶ For smoking cessation, the “Quit with US” program combined with pharmacist counseling significantly improved abstinence rates among light smokers.²⁷ Another study used the Health Belief Model (HBM) and health literacy concepts in a Telegram-based educational intervention, effectively promoting smoking prevention behaviors.²⁹ Lastly, a text messaging intervention increased self-reported smoking abstinence and reduced cigarette consumption.²⁸ Collectively, these studies highlight the potential of diverse mHealth approaches, including IM apps, mobile chat, educational content, and SMS, in reducing substance use among students in LMICs. The review underscores the importance of tailored, engaging strategies to sustain participation and reduce dropout rates (refer to extended data: Table 2).

Impact of mHealth interventions

The review examined mHealth interventions targeting substance use among students in LMICs, focusing on alcohol use and cigarette smoking. Studies utilized tools like instant messaging apps (WhatsApp, WeChat), smartphone applications, and text messaging. For alcohol reduction, instant messaging platforms demonstrated high acceptability and effectiveness in lowering AUDIT scores.^26,30 In smoking cessation, the “Quit with US” app improved smoking behaviors and attitudes,²⁷ while a Telegram-based intervention using the Health Belief Model (HBM) promoted sustained behavioral changes.²⁹ A text-messaging intervention also increased willingness to quit and reduced smoking levels among Chinese students.²⁸ These findings highlight the effectiveness of mHealth tools in fostering healthier behaviors and reducing substance use among students in LMICs, emphasizing their adaptability and accessibility in resource-limited settings (Refer to extended data: Table 3).

Methodological and contextual challenges

The review identified methodological and contextual challenges across studies. The first study lacked long-term follow-up data, relied on self-reported outcomes, and excluded non-smartphone users, raising concerns about representativeness.³⁰ The second study faced issues with social desirability bias, unclear intervention mechanisms, and incomplete follow-up despite high retention.²⁶ The third study struggled with app adherence, short duration, and a narrow participant pool limited to cigarette users.²⁷ The fourth study encountered difficulties measuring health literacy and generalizing findings due to educational gaps and a narrowly defined group.²⁹ The fifth study relied heavily on self-reported data, experienced follow-up loss, and had limited scalability due to small sample size and mobile phone access requirements.²⁸ Common challenges included reliance on self-reports, representativeness issues, short study durations, and scalability concerns, impacting the validity and applicability of findings in LMICs (Refer to extended data: Table 3).

Factors on the effectiveness of mHealth interventions

Table 4 (refer to extended data) summarizes studies on mHealth interventions targeting alcohol and smoking reduction among specific student populations. The first study used instant messaging (IM) apps for university students in Hong Kong aged 18+, requiring smartphone ownership and monthly alcohol use.³⁰ A follow-up study evaluated mobile chat-based IM for students with AUDIT scores ≥8, delivering 26 push messages on a tapering schedule.²⁶ The third study targeted young adult smokers (18–24 years) using the “Quit with US” app, requiring daily engagement.²⁷ Another study used Telegram for smoking prevention among university students, delivering weekly HBM-based interventions.²⁹ The final study employed daily text messaging for vocational students (16–19 years) to support smoking cessation.²⁸ These studies highlight tailored mHealth strategies and frequency schedules for diverse student demographics, addressing alcohol and smoking behaviors effectively (Refer to extended data: Table 4).

The intervention characteristics

The first study explored university students’ perceptions of instant messaging (IM) apps for alcohol reduction in Hong Kong.³⁰ A follow-up study evaluated a 12-week alcohol brief intervention (ABI) supplemented with mobile chat-based IM, tapering from three sessions weekly to one by the third month.²⁶ The third study assessed the “Quit with US” app for young adult smokers in Thailand over 12 weeks, involving daily app use.²⁷ Another study used Telegram to deliver weekly smoking prevention interventions to Iranian university students, integrating the HBM and HL.²⁹ The final study employed a 12-week daily text-messaging intervention for smoking cessation among Chinese vocational students.²⁸ These studies demonstrate the effectiveness of mobile and digital platforms in addressing alcohol and smoking challenges across diverse student populations in LMICs (Refer to extended data: Table 4).

Generalizability

Country

The review included studies from four upper-middle-income countries (UMICs)—Hong Kong, China, Iran, and Thailand—all in Asia, reflecting a regional focus on mHealth interventions. While these studies offer valuable insights, their applicability to low-income countries (LICs) is limited due to differences in socio-cultural factors, healthcare access, education, and public health resources. UMICs often share some economic and infrastructural similarities with LICs but differ significantly in substance use patterns, social attitudes, and intervention effectiveness. To enhance relevance, future research should prioritize low-cost, adaptable, and culturally tailored interventions that address health literacy and mental health challenges in LICs. Simplifying mHealth tools and engaging local stakeholders can improve scalability and effectiveness in resource-constrained settings, ensuring broader applicability and impact.

Population

The reviewed studies primarily focused on students aged 18 and older, with most targeting undergraduates aged 18–24.^26,27,30 Some studies specifically examined second- or third-year undergraduates²⁹ or vocational school smokers aged 16–19,²⁸ highlighting a focus on young adults in higher education. The analysis also explored cultural, socio-economic, and gender factors shaping substance use behaviors and intervention effectiveness. Cultural norms, such as associating substance use with masculinity in regions like Hong Kong, Thailand, Iran, and China, and socio-economic disparities, including financial stress and limited access to resources, were key considerations. Gender-responsive interventions, addressing stigma for females and masculinity norms for males, were emphasized to improve outcomes. Digital platforms were noted as a viable solution for accessible, affordable support.

The health implications and social contexts

Table 5 (refer to extended data), highlights the health risks of alcohol and tobacco use among youth and the social factors driving substance use. Peer influence plays a crucial role in initiation and continuation. The findings stress the need for age-specific, culturally adapted interventions, especially in environments where substance use is socially accepted, to support public health efforts.

The characteristics of the Study setting

The study involved participants from diverse higher education institutions across Asia, including Chinese university students in Hong Kong,^26,30 undergraduates from five universities in Chiang Mai, Thailand,²⁷ and students from Shahid Beheshti University of Medical Sciences in Iran.²⁹ Vocational school students were also included,²⁸ ensuring broad representation. Institutional policies, such as strict anti-substance rules in Hong Kong and Iran, contrasted with lenient smoking policies in Thailand, influencing intervention effectiveness. Student demographics, cultural acceptance of substance use, and resource availability—like high mobile penetration in Hong Kong—shaped intervention outcomes. These factors highlight the need for context-specific approaches to improve intervention adaptability globally (Refer to extended data: Table 5).

Aim of the included studies

The primary studies examined the effectiveness of mHealth interventions for alcohol and smoking cessation, with varied objectives. One study assessed university students’ perceptions of using instant messaging (IM) apps for alcohol reduction in Hong Kong, where high alcohol exposure and peer drinking are prevalent.³⁰ Another evaluated chat-based IM support for alcohol reduction among Hong Kong students at risk of alcohol use disorder.²⁶ A third tested the “Quit with US” app for smoking cessation among young adults.²⁷ Extended the Health Belief Model with Health Literacy to assess smoking prevention interventions,²⁹ while²⁸ evaluated text-messaging-based cessation for Chinese vocational students. These studies highlight the diverse applications of mHealth interventions across populations and contexts. This variety of research aims demonstrates the evolving understanding of mHealth interventions for substance use across different populations and contexts.

Results of individual studies

The primary studies revealed key insights into mHealth interventions for substance use reduction. One study found IM apps highly acceptable and engaging for alcohol reduction, suggesting their potential as alternatives to text-based programs.³⁰ Another demonstrated that mobile chat-based messaging, combined with ABI, effectively reduced alcohol consumption among at-risk university students in Hong Kong.²⁶ The “Quit with US” app, paired with pharmacist counseling, improved smoking abstinence rates among young adults.²⁷ An educational intervention via Telegram, based on the Health Belief Model, promoted smoking prevention behaviors,²⁹ while text messaging increased abstinence and reduced cigarette use among vocational students.²⁸ These findings highlight mHealth’s potential for addressing substance use in students.

Risk of bias in studies

The quality assessment ratings for the included studies are summarized in Tables 2 and 3 (refer to extended data). Out of the five studies, four were rated as having a moderate risk of bias by Cochrane EPOC tool.^26–29 The main factors contributing to these moderate ratings included concerns about the knowledge of allocated interventions potentially influencing the outcomes, selective outcome reporting, and other bias risks. However, one study was assessed as having a low risk of bias by a checklist from JBI.³⁰ These findings emphasize the need to carefully consider potential biases when evaluating the effectiveness of mHealth interventions for substance use.

Risk of bias assessments showed four studies had moderate risk due to concerns like intervention knowledge influencing outcomes and selective reporting,^26–29 while one study had low risk.³⁰ Careful consideration of biases is crucial when evaluating mHealth interventions.

The analysis of the effectiveness results

This review emphasizes the effectiveness of mHealth interventions, contributing to the growing body of evidence supporting digital platforms for the prevention and treatment of substance use. The findings reveal significant reductions in substance use, highlighting the potential of these interventions. However, the review also underscores the need for long-term outcome assessments and more personalized, comprehensive strategies to enhance their effectiveness across diverse populations.

By comparing these results with existing literature on mHealth interventions for substance use, a broader context is provided. A summary table of key statistical findings from each study offers a clearer perspective on how their performance compares to the broader research landscape.

Outcomes

The review found that IM apps were highly acceptable and engaging for alcohol reduction. Mobile chat-based messaging, when combined with Alcohol Brief Interventions (ABI), effectively reduced alcohol consumption. The “Quit with US” app, when paired with pharmacist counseling, led to improved smoking cessation rates. An educational intervention delivered via Telegram, based on the Health Belief Model, successfully promoted smoking prevention behaviors. Additionally, text messaging was found to increase abstinence and decrease cigarette consumption.

The review employed standardized scales and tools to assess alcohol consumption, smoking cessation, and health literacy, though differences in these tools may affect result comparability. Self-reported data on substance use, prone to social desirability or recall bias, could compromise reliability, particularly in stigmatized populations. Variability in follow-up periods—ranging from 4–8 weeks to 3 months that poses challenges in interpreting long-term versus short-term intervention effects. A comparative table clarifies each study’s approach, highlighting operational definitions, data collection timelines, and discrepancies in outcome measures that may influence effectiveness interpretation.

Diverse study designs and methodological approaches

The review highlights a diverse range of study designs, including qualitative research, randomized controlled trials (RCTs), and quasi-experimental approaches, each featuring varying sample sizes and follow-up durations. These methodologies contribute uniquely to understanding phenomena, and their combined use, known as methodological triangulation, strengthens the robustness and applicability of findings. This diversity underscores the importance of evaluating intervention effectiveness and understanding participant experiences across different contexts. However, the reliance on self-reported data introduces potential biases, raising concerns about the accuracy of the findings. Several methodological limitations were identified, such as the exclusion of non-smartphone users, which limited the inclusivity of the sample, particularly among university students. Recruitment was also confined to vocational schools, potentially reducing the representativeness of the broader youth demographic. Additionally, the lack of member checking in qualitative research and the infeasibility of blinding in the study protocol may have introduced bias. Despite a high retention rate of 85%, incomplete follow-up data and the short 12-week study duration hindered the assessment of long-term smoking cessation outcomes. The small sample size further limited the study’s power to detect significant differences. Behavioral factors, such as participants’ drinking habits and educational gaps, may have influenced the intervention’s efficacy, while practical constraints, like low phone penetration, affected scalability. The exclusive reliance on self-reported data, without biochemical validation, also weakened the evidence quality. Future research should adopt hybrid designs combining qualitative and quantitative methods to enhance understanding and address these limitations. Overall, the findings call for cautious interpretation and suggest improvements in recruitment diversity, longer follow-up periods, and greater methodological rigor.

Regional and demographic gaps in Mhealth interventions for substance use

The review reveals a regional focus on mHealth interventions for substance use, with all included studies conducted in Asian countries, particularly addressing cigarette smoking.³⁵ This regional emphasis likely reflects the urgent need for low-cost, scalable, and accessible solutions to combat the growing substance use epidemic in Asia. Notably, no interventions were implemented in Africa, where substance use is projected to rise rapidly, especially given the varying smoking prevalence across sub-Saharan Africa.³⁶ This gap highlights the necessity of testing mHealth interventions in high-burden regions with limited substance cessation infrastructure. Furthermore, all reviewed interventions were conducted in upper-middle-income countries, with none in low- or lower-middle-income nations, underscoring the need for stakeholder engagement and alignment with policy environments to scale digital health solutions effectively in LMICs.³⁷ For example, the “Quit with US” app in Thailand, supported by Payap University and organizations like the Thai Health Promotion Foundation, demonstrates the importance of early stakeholder involvement and policy alignment for scalability. Most interventions targeted university settings, focusing on young adults and addressing alcohol reduction, smoking cessation, and prevention. Future research should explore the applicability of these interventions in diverse populations and settings to enhance their impact. Overall, the review emphasizes tailored mHealth solutions for young adults in higher education, while highlighting the need to extend their reach beyond this demographic to maximize effectiveness.

Characteristics and content

The review highlights the diversity of mHealth interventions aimed at reducing substance use, showcasing variations in duration, frequency, platforms, and characteristics. Intervention durations typically spanned three months, with delivery frequencies ranging from three times weekly in the first month to once weekly in the third month. Platforms included instant messaging apps, mobile chat-based messaging, smartphone applications like “Quit with US,” Telegram, and traditional SMS, reflecting the evolving technological landscape for substance use interventions. Given that reducing substance use is a continuous process, two-way interactions have proven more effective than one-way messaging.³⁴

The review also explores diverse intervention characteristics, such as university students’ perceptions of IM apps for alcohol reduction, the effectiveness of combining brief alcohol interventions with mobile chat support, and the efficacy of educational interventions for smoking prevention. These approaches demonstrate the growing interest in mHealth solutions across various platforms and populations. Research indicates that mHealth interventions hold promise for the prevention, treatment, and aftercare of substance use.^38–40 Real-time, real-life implementations, particularly text messaging, have shown effectiveness in reducing substance use among vulnerable groups, including youth.³² Smartphone apps have also been effective in managing cravings and facilitating behavior change for individuals with substance use disorders.⁴¹

Text messaging (SMS) has emerged as an affordable, acceptable, and effective method for reducing alcohol consumption among young people, demonstrating significant potential for fostering positive behavioral changes.⁴² Additionally, many Vietnamese smokers interested in quitting expressed a willingness to use and pay for text messaging cessation services, highlighting the potential of mHealth strategies to deliver effective smoking cessation programs in Vietnam.⁴³ Overall, the review underscores the versatility and promise of mHealth interventions in addressing substance use across diverse contexts.

Limitations

The systematic review highlights the potential and limitations of mHealth interventions for substance use. While these interventions can reach broad audiences and promote self-regulation, further research is needed to enhance the validity of findings. Future studies should prioritize larger sample sizes for better generalizability and adopt longitudinal designs to assess long-term effects. The reliance on self-reported data in reviewed studies raises concerns about bias due to socially desirable responses. Incorporating collateral informants or drug screenings could improve outcome accuracy. The review itself has limitations, such as excluding non-English studies, which may limit its global applicability. However, its rigorous process including dual-reviewer screening, data extraction, and quality assessment, strengthened reliability. Efforts to include unpublished evaluations also broadened its scope. The review’s strength lies in its comprehensive analysis of psychometric properties, outcome assessments, behavioral change techniques, methodological challenges, and sampling methods across included studies.