Lange S, et al. - Authors gauged the prevalence of smoking during pregnancy by country, WHO region, and globally and the proportion of pregnant women who smoked during pregnancy, by frequency and quantity, on a global level. In many countries, smoking during pregnancy is still a prevalent behaviour. Smoking prevention programmes and health promotion strategies should be informed by these findings. An attention should be drawn towards the need for improved access to smoking cessation programmes for pregnant women.

Methods

• Experts, for this systematic review and meta-analysis did a comprehensive systematic literature search for studies reporting the prevalence of smoking during pregnancy in the general population, published between Jan 1, 1985 and Feb 1, 2016, using several electronic bibliographic databases (CINAHL, Embase, ERIC, Medline, Medline in process, PsychINFO, Scopus, and Web of Science), without language or geographical restrictions.
• They included original research studies published in a peer-reviewed journal and evaluated study quality using a tool specifically developed for use in systematic reviews addressing questions of prevalence.
• They excluded the studies if they did not include lifetime non-smokers in their sample or estimate, used a sample not generalisable to the general population of the respective country, or did not provide primary data.
• In order to evaluate the prevalence by country, country-specific random-effects meta-analyses was done for countries with two or more available empirical studies, and the prevalence was predicted using a multilevel fractional response regression model with country-specific indicators for countries with one or no study.
• They evaluated the proportion of female daily smokers who do not quit once pregnant by calculating the regional and global averages of the prevalence of daily smoking during pregnancy and of the prevalence of daily smoking in women.
• They did random-effects meta-analyses using available data from all countries and applied the respective proportions to the global prevalence estimate to estimate the global prevalence, by frequency and quantity.
• A time-trend analysis using a univariate multilevel fractional response model was done.

Results

• Results demonstrated that out of 21,329 studies identified, 295 were retained for data extraction.
• The estimates were calculated via meta-analysis for 43 countries and via statistical modelling for 131 countries.
• Findings suggested that the 3 countries with the highest estimated prevalence of smoking during pregnancy were Ireland (38·4%, 95% CI 25·4–52·4), Uruguay (29·7%, 16·6–44·8), and Bulgaria (29·4%, 26·6–32·2).
• During pregnancy the global prevalence of smoking was estimated to be 1·7% (95% CI 0·0–4·5).
• Results demonstrated that prevalence of smoking during pregnancy was 8·1% (95% CI 4·0–12·2) in the European Region, 5·9% (3·2–8·6) in the Region of the Americas, 1·2% (0·7–1·7) in the Southeast Asian Region, 1·2% (0·0–3·7) in the Western Pacific Region, 0·9% (0·0–1·9) in the Eastern Mediterranean Region, and 0·8% (0·0–2·2) in the African Region.
• Researchers noted that globally, out of the pregnant women who smoked 72·5% (95% CI 70·4–75·0) were daily smokers, and 27·5% (25·4–29·6) of them were occasional smokers; 51·8% (95% CI 50·0–53·5) women who smoked were light smokers, 34·8% (33·1–36·4) were moderate smokers, and 13·5% (12·3–14·7) were heavy smokers.
• Moreover, 52·9% (95% CI 45·6–60·3) was the proportion of women who smoked daily and continued to smoke daily during pregnancy, ranging from 30·6% (95% CI 25·6–36·4) in the European Region to 79·6% (44·2–100·0) in the Western Pacific Region.