Truace
Subscribe
70/100

Moderate evidence

Cite
More
Export CSVExport JSONEmailLinkedInX

Chatbots helped more smokers quit at 6 months, review of five trials finds

Study title: Conversational artificial intelligence interventions to support smoking cessation: A systematic review and meta-analysis

Evidence type
Systematic review or meta-analysis
Sample
n = 58,796
Sources
1 sourcePeer-reviewed
Published
2 min read
Truvace record
TRV-2026-2759cite or check it

Study title: Conversational artificial intelligence interventions to support smoking cessation: A systematic review and meta-analysis

A systematic review searched six databases in June 2022 and found 819 studies, of which five randomised controlled trials qualified, with a combined 58,796 participants. A pooled analysis of three of them found smokers given conversational AI-enhanced support were 29% more likely to be abstinent at 6 months than controls (RR 1.29, 95% CI 1.13 to 1.46).

Read the study abstract

Background Conversational artificial intelligence (chatbots and dialogue systems) is an emerging tool for tobacco cessation that has the potential to emulate personalised human support and increase engagement. We aimed to determine the effect of conversational artificial intelligence interventions with or without standard tobacco cessation interventions on tobacco cessation outcomes among adults who smoke, compared to no intervention, placebo intervention or an active comparator. Methods A comprehensive search o…

The gain

A meta-analysis of three randomised trials found smokers using conversational AI-enhanced cessation interventions were 29% more likely to be abstinent at 6 months than control participants (RR 1.29, 95% CI 1.13 to 1.46).

In brief

The five trials tested chatbots inside smartphone apps (three trials), a social media chatbot (one) and an internet-based avatar (one). The authors say all five studies reported benefits from the conversational AI interventions.

The evidence is still thin. The studies differed in setting, methods, participants and end-points, loss to follow-up was generally high and overall risk of bias was high. The authors call the evidence limited but promising and want better-designed trials that follow standardised reporting guidelines.

Main points

  1. Six databases were searched in June 2022, yielding 819 studies; five randomised controlled trials published since 2005 met the inclusion criteria.
  2. The five trials had a combined sample of 58,796 adults who smoke.
  3. Three trials used chatbots embedded in smartphone apps, one used a social media chatbot and one used an internet-based avatar.
  4. The primary outcome was sustained abstinence, self-reported or biochemically validated, for at least 6 months.
  5. The pooled risk ratio for quitting at 6 months across three studies was 1.29 (95% CI 1.13 to 1.46, p < 0.001).
  6. Two authors independently extracted data and assessed risk of bias; risk of bias was high overall and loss to follow-up was generally high.

What the review found when it combined the studies

These are the review’s own pooled results, copied from its abstract. Truvace did not recalculate them.

  • Smoking abstinence at 6-month follow-up among adults using conversational AI cessation interventions versus controls

    Risk ratio 1.29 (95% CI 1.13 to 1.46)

    search to June 2022

    The review’s own words
    The random effects meta-analysis ( Figure 2 ) of abstinence outcomes at 6-month follow-up ( n = 3) found that participants in the conversational AI intervention were significantly more likely to quit smoking compared to participants in control conditions (RR = 1.29, 95% CI (1.13, 1.46), p < 0.001).

Code checks every number above against the review’s text before it is shown. The short labels are an AI model’s reading of that text, so the review’s own sentence sits under each one. A pooled figure is only as strong as the studies behind it. How Truvace uses pooled results

The full story

The review compared conversational AI interventions, alone or alongside standard tobacco cessation support, with no intervention, a placebo or an active comparator. Only randomised controlled trials were eligible, and the primary outcome was abstinence sustained for at least 6 months. Secondary outcomes were point-prevalence abstinence and sustained abstinence of less than 6 months.

Of the five included trials, three were pooled in a random effects meta-analysis of 6-month quitting. Participants in the AI-enhanced arms were significantly more likely to have quit than controls (RR 1.29, p < 0.001). The authors report that every included study found benefits from the AI interventions.

The studies varied widely in setting, methodology, intervention, participants and end-points, producing high heterogeneity. High loss to follow-up and high overall risk of bias mean the results should be read with caution, according to the authors, who call for further randomised trials that follow standardised reporting guidelines.

Sources

1 on file
  1. DIGITAL HEALTHdoi:10.1177/20552076231211634
    Peer-reviewed2023-01-01

Talk it over

Four questions to start onWho gets this benefit, and who was left out of the study or story?

Start with a question

Chatbots helped more smokers quit at 6 months, review of five trials findshttps://truvace.com/item/conversational-artificial-intelligence-interventions-to-support-smoking-cessatio

  1. Who gets this benefit, and who was left out of the study or story?

    The gain A meta-analysis of three randomised trials found smokers using conversational AI-enhanced cessation interventions were 29% more likely to be abstinent at 6 months than control participants (RR 1.29, 95% CI 1.13 to 1.46).

    Answer in the thread

  2. How do we know? What can this kind of evidence show, and what can’t it?

    The evidence This rests on a systematic review or meta-analysis with a sample of 58,796, from DIGITAL HEALTH.

    Answer in the thread

  3. What does it leave open? What else would you want to know before acting on it?

    The record’s own caveat The five trials differed in setting, methods, interventions, participants and end-points, loss to follow-up was generally high and risk of bias was high overall, so the authors urge caution.

    Answer in the thread

  4. Have you seen AI used this way where you study, work or live? Did it go the same way?

    Answer in the thread

The questions are the same on every record. The details under them are taken from this record as written.

What readers say

More from Health