The role of flavored electronic nicotine delivery systems in smoking cessation: A systematic review

Alex C Liber,Marie Knoll,Christopher J Cadham,Mona Issabakhsh,Hayoung Oh,Steve Cook,Kenneth E Warner,Ritesh Mistry,David T Levy
DOI: https://doi.org/10.1016/j.dadr.2023.100143
2023-03-16
Abstract:Background: Electronic nicotine delivery systems (ENDS) come in numerous flavors and may aid smoking cessation. This systematic review examines evidence on the role of ENDS flavors in smoking cessation. Methods: We searched EMBASE OVID, PsychInfo, and Medline databases for studies that: 1) examined cigarette cessation outcomes for persons using ENDS (intent, attempts, and success) and 2) reported results separated by respondent's ENDS flavor used. We extracted crude and adjusted odds ratios for associations between cessation outcomes and types of ENDS flavors used (nontobacco vs. tobacco/unflavored; nontobacco and nonmenthol vs. tobacco/unflavored and menthol). We did not consider cessation outcomes among people not using ENDS. We evaluated the evidence using the GRADE approach, focusing on consistency and reliability of findings across studies. Results: 29 studies met inclusion criteria, producing 36 odds ratios (ORs) comparing cessation outcomes across ENDS flavor groups. Three ORs examined quit intent, five examined quit attempts, and 28 examined quit success. Using GRADE, we reached Low levels of certainty that there was not an association between ENDS flavor use and intention to quit smoking or making a quit attempt. There were Very Low levels of certainty that nontobacco flavored versus tobacco/unflavored ENDS use was not associated with smoking cessation success, with similar findings for nonmenthol and nontobacco compared to tobacco and menthol flavored ENDS. Conclusion: The evidence about the role of different flavored ENDS use and smoking cessation outcomes is inconclusive, reflecting highly heterogeneous study definitions and methodological limitations. More high-quality evidence, ideally from randomized controlled trials, is required.
What problem does this paper attempt to address?