Exploring e-cigarette policy recommendations and the role of evidence in international public health guidelines: a citation network analysis

Abstract

BackgroundPublic health guidelines internationally aspire to be evidence based. However, different public health bodies have recommended adopting different approaches to e-cigarettes. While some have actively encouraged e-cigarette use, others have cautioned that e-cigarettes might introduce new health risks and renormalise tobacco smoking. We investigated the sources of evidence used by public health bodies when making recommendations about e-cigarette policy. MethodsWe did a citation network analysis of international public health guidelines across four purposively selected diverse policy contexts. We systematically identified relevant public health guideline documents focused on e-cigarette policy. Information about all cited references was extracted from 18 public health guidelines, eight from the UK, four from Australia, four from the USA, and two from WHO (worldwide data). The primary outcome was the source of evidence drawn upon by public health bodies. We did a network analysis of the citations (using R Studio, version 1.2.1335) to understand if and how evidence use varied across guidelines. Findings2802 references (2332 unique references) were included across the 18 public health guidelines, with none to 958 references cited per guideline document. Results from the citation network show that the four different contexts report from similar evidence. 74 references were cited across three or more guidelines. Of these, the most common study design was toxicology (30 references), followed by survey-based studies (12 references); with non-systematic literature reviews being the least popular (one reference). Among the most cited 74 studies, 18 (24%) reported conflicts of interest with pharmaceutical companies and 13 (18%) reported conflicts of interest with e-cigarette companies. InterpretationWe showed that public health guidelines across different contexts and countries draw on similar studies, despite articulating different policy approaches. Conflicts of interest are common among the studies that have been most influential in decision making. Limitations of our study include the reliance on citations rather than broader forms of evidence use. Greater transparency is needed in how recommendations are being developed, with particular consideration of how to manage conflicts of interest. FundingMedical Research Council and Chief Science Office.