Data sources and search strategy
The electronic databases, such as Medline, Embase, PsychInfo, Cochrane Central Register of Controlled Trials (CENTRAL), PubMed, Google Scholar, Academia, and ResearchGate, were explored. Reference lists of the selected articles were also screened. All relevant studies available on the topic were included irrespective of time duration. The systematic search was restricted to studies published in the English language. The keywords were "smoker or smokers OR smoking," "tobacco" OR cigarette OR nicotine OR smoking cessation OR "tobacco consumption OR cessation, OR abstain* OR quit* OR stop* OR computer OR computer-aided design, OR internet, OR computer, OR networks, OR media, OR cellular phone OR mobile, OR text OR message* OR SMS, OR web, OR electronic mail OR Chat, OR video recording.
Eligibility criteria
PICO Framework
Participants
Inclusion: Adults aged more than 18 years use the internet or face-to-face interventions to reduce or quit tobacco use. No ethnicity restrictions were applied.
Exclusion: Cochrane studies that compare the internet to face-to-face interventions with other interventions.
Intervention
Internet interventions (Phone, mobile, WhatsApp, Facebook, Online network group, Online Support group, text messaging, other internet media)
Comparator
Face-to-face interventions or no intervention in the comparator group.
Face-to-face interventions include counseling, cognitive behavior therapy, or health education forms with control or routine care.
Outcome
Post-intervention tobacco quitting – number of participants quitting tobacco after the intervention (internet use)
Study design
Randomized controlled trials
Time frame
No restriction to the time frame was applied.
Screening of eligible studies
A systematic search was done by two reviewers independently. After searching, studies were screened with titles and abstracts of respective studies. All selected studies were imported to Rayyan (https://rayyan.qcri.org), a free web-based software. [60] Two reviewers screened the full text of articles based on eligibility criteria determined as per review protocol. Any relevant discrepancy has been resolved by consensus with the help of a third reviewer. We adhered to the guidelines of Preferred Reporting Items for Systematic Review and Meta‑Analysis (PRISMA) 2009. [28] (Additional File. 1)
Data extraction
Two reviewers extracted the data from the full text of eligible studies. Corresponding authors of included studies were contacted for the relevant data. Data excel sheet was prepared to note the characteristics of selected studies. It includes the author’s name with publication year, country, sample size, participants' mean age, male to female ratio, baseline tobacco consumption, and follow-up period after the intervention (Table 1).
Eligible studies were exported to RevMan software for data analysis. Forest plots have been created to present the results with Odds ratio (OR), confidence interval (CI), and effect size.
Risk of bias assessment
Two reviewers independently assessed the quality of included studies. Risk of bias graph and summary has been created in Review Manager software 5.4 version under the heads of Random sequence generation (selection bias), Blinding of participants and personnel (performance bias), Allocation concealment (selection bias), Blinding of outcome assessment, (detection bias), Incomplete outcome data (attrition bias), Selective reporting (reporting bias), and Other bias (Figure 2). [29]
The GRADE approach was also followed to explore the quality of evidence on high, moderate, and low levels. [30] RevMan files were exported to the GRADE Profiler to assess the quality of studies and create a "Summary of Findings" table (Table 2).
Data analysis
Review Manager software 5.4 version was used for meta‑analysis. [29] The fixed‑effects model and effect measures were calculated as the OR with p-value < 0.05 considered statistically significant. I2 statistics with 25, 50, and 75 % were measured to compute statistical heterogeneity in low, moderate, and high grades. The tabulated data presented in a forest plot (Figure 3).
Funnel plots have also been created to assess the publication bias across studies. It measures an effect estimate against its standard error for an outcome (Figure 4).
Tobacco quitting among participants has been analyzed at one, three, six, and twelve months of follow-up and presented in a forest plot (Figure 3).