Background

While cigarette smoking has declined globally, waterpipe smoking is rising, especially among youth. The impact of this rise is amplified by mounting evidence of its addictive and harmful nature. Waterpipe smoking is influenced by multiple factors, including appealing flavors, marketing, use in social settings, and misperceptions that waterpipe is less harmful or addictive than cigarettes. People who use waterpipes are interested in quitting, but are often unsuccessful at doing so on their own. Therefore, developing and testing waterpipe cessation interventions to help people quit was identified as a priority for global tobacco control efforts. 

Objectives

To evaluate the effectiveness of tobacco cessation interventions for people who smoke waterpipes.

Search methods

We searched the Cochrane Tobacco Addiction Review Group Specialized Register from database inception to 29 July 2022, using variant terms and spellings ('waterpipe' or 'narghile' or 'arghile' or 'shisha' or 'goza' or 'narkeela' or 'hookah' or 'hubble bubble'). We searched for trials, published or unpublished, in any language.

Selection criteria

We sought randomized controlled trials (RCTs), quasi‐RCTs, or cluster‐RCTs of any smoking cessation interventions for people who use waterpipes, of any age or gender. In order to be included, studies had to measure waterpipe abstinence at a three‐month follow‐up or longer.

Data collection and analysis

We used standard Cochrane methods. Our primary outcome was abstinence from waterpipe use at least three months after baseline. We also collected data on adverse events. Individual study effects and pooled effects were summarized as risk ratios (RR) and 95% confidence intervals (95% CI), using Mantel‐Haenszel random‐effects models to combine studies, where appropriate. We assessed statistical heterogeneity with the I² statistic. We summarized secondary outcomes narratively. We used the five GRADE considerations (risk of bias, inconsistency of effect, imprecision, indirectness, and publication bias) to assess the certainty of the body of evidence for our primary outcome in four categories high, moderate, low, or very low.

Main results

This review included nine studies, involving 2841 participants. All studies were conducted in adults, and were carried out in Iran, Vietnam, Syria, Lebanon, Egypt, Pakistan, and the USA. Studies were conducted in several settings, including colleges/universities, community healthcare centers, tuberculosis hospitals, and cancer treatment centers, while two studies tested e‐health interventions (online web‐based educational intervention, text message intervention). Overall, we judged three studies to be at low risk of bias, and six studies at high risk of bias.

We pooled data from five studies (1030 participants) that tested intensive face‐to‐face behavioral interventions compared with brief behavioral intervention (e.g. one behavioral counseling session), usual care (e.g. self‐help materials), or no intervention. In our meta‐analysis, we included people who used waterpipe exclusively, or with another form of tobacco. Overall, we found low‐certainty evidence of a benefit of behavioral support for waterpipe abstinence (RR 3.19 95% CI 2.17 to 4.69; I² = 41%; 5 studies, N = 1030). We downgraded the evidence because of imprecision and risk of bias.

We pooled data from two studies (N = 662 participants) that tested varenicline combined with behavioral intervention compared with placebo combined with behavioral intervention. Although the point estimate favored varenicline, 95% CIs were imprecise, and incorporated the potential for no difference and lower quit rates in the varenicline groups, as well as a benefit as large as that found in cigarette smoking cessation (RR 1.24, 95% CI 0.69 to 2.24; I² = 0%; 2 studies, N = 662; low‐certainty evidence). We downgraded the evidence because of imprecision. We found no clear evidence of a difference in the number of participants experiencing adverse events (RR 0.98, 95% CI 0.67 to 1.44; I² = 31%; 2 studies, N = 662). The studies did not report serious adverse events.  

One study tested the efficacy of seven weeks of bupropion therapy combined with behavioral intervention. There was no clear evidence of benefit for waterpipe cessation when compared with behavioral support alone (RR 0.77, 95% CI 0.42 to 1.41; 1 study, N = 121; very low‐certainty evidence), or with self‐help (RR 1.94, 95% CI 0.94 to 4.00; 1 study, N = 86; very low‐certainty evidence). 

Two studies tested e‐health interventions. One study reported higher waterpipe quit rates among participants randomized to either a tailored mobile phone or untailored mobile phone intervention compared with those randomized to no intervention (RR 1.48, 95% CI 1.07 to 2.05; 2 studies, N = 319; very low‐certainty evidence). Another study reported higher waterpipe abstinence rates following an intensive online educational intervention compared with a brief online educational intervention (RR 1.86, 95% CI 1.08 to 3.21; 1 study, N = 70; very low‐certainty evidence). 

Authors' conclusions

We found low‐certainty evidence that behavioral waterpipe cessation interventions can increase waterpipe quit rates among waterpipe smokers. We found insufficient evidence to assess whether varenicline or bupropion increased waterpipe abstinence; available evidence is compatible with effect sizes similar to those seen for cigarette smoking cessation. 

Given e‐health interventions' potential reach and effectiveness for waterpipe cessation, trials with large samples and long follow‐up periods are needed. Future studies should use biochemical validation of abstinence to prevent the risk of detection bias. Finally, there has been limited attention given to high‐risk groups for waterpipe smoking, such as youth, young adults, pregnant women, and dual or poly tobacco users. These groups would benefit from targeted studies.