In this systematic review and meta-analysis, we undertook an electronic search in five databases (MEDLINE, Web of Science, Cumulative Index to Nursing and Allied Health Literature, Embase, and Wiley Cochrane Library) from Jan 1, 2016, to Aug 31, 2020, and a grey literature search. Included studies reported on the prevalence of ENDS or ENNDS use in nationally representative samples in populations younger than 20 years and collected data between the years 2016 and 2020. Studies were excluded if they were done in those aged 20 years or older, used data from specialist panels that did not apply appropriate weighting, or did not use methods that ensured recruitment of a nationally representative sample. We included the most recent data for each country. We combined multiple national estimates for a country if they were done in the same year. We undertook risk of bias assessment for all surveys included in the review using the Joanna Briggs Institute Critical Appraisal Checklist (by two reviewers in the author list). A random effects meta-analysis was used to pool overall prevalence estimates for ever, current, occasional, and daily use. This study was prospectively registered with PROSPERO, CRD42020199485.
The most recent prevalence data from 26 national surveys representing 69 countries and territories, with a median sample size of 3925 (IQR 1=2266, IQR 3=10 593) children and adolescents was included. In children and adolescents aged between 8 years and younger than 20 years, the pooled prevalence for ever (defined as any lifetime use) ENDS or ENNDS use was 17·2% (95% CI 15–20, I2=99·9%), whereas for current use (defined as use in past 30 days) the pooled prevalence estimate was 7·8% (6–9, I2=99·8%). The pooled estimate for occasional use was 0·8% (0·5–1·2, I2=99·4%) for daily use and 7·5% (6·1–9·1, I2=99·4%) for occasional use. Prevalence of ENDS or ENNDS use was highest in high-income geographical regions. In terms of study quality, all surveys scored had a low risk of bias for the sampling frame used, due to the nationally representative nature of the studies. The most poorly conducted methodological feature of the included studies was subjects and setting described in detail. Few surveys reported on the use of flavours or types of ENDS or ENNDS.
There is significant variability in the prevalence of ENDS and ENNDS use in children and adolescents globally by country income status. These findings are possibly due to differences in regulatory context, market availability, and differences in surveillance systems.
World Health Organization and the Bill & Melinda Gates Foundation.