How do Risk Perceptions Drive Smokers to Completely Switch to a Smoke-Free Tobacco Product (IQOS™)? A Four-Country Cohort Study

Besides the reasons for using IQOS^™ that referred to RF or RH, depending on the country, participants could also select multiple other reasons out of lists for (i) using IQOS^™ (29 [Japan], 17 [Italy], 21 [Germany], or 25 [Russia] other reasons) or (ii) not using IQOS^™ (29 [Japan], 26 [Italy], 27 [Germany], or 28 [Russia] other reasons). Besides RF and RH, the two other main predictors of exclusive IQOS^™ use across all 4 countries were (i) IQOS^™ is different (i.e., more hygienic: no fire, no smoke, less smell, no odor and stains on hands, furniture) to smoking cigarettes and (ii) participants liked the (real tobacco) taste of IQOS^™.

Abbreviations: RF: perceived reduced formation of harmful chemicals; RH: perceived reduced risk of harm.

1 The statements on RF and RH were not consistent in the four countries because the information/wording of the statement (“claims”) had to be phrased in accordance with the local regulations and law in the individual countries.

Participants of PMI’s open online IQOS^™ user consumer cohorts in Japan (≥ 21 y), Italy (≥ 19 y), Germany (≥ 19 y), and Russia (≥ 19 y) were randomly selected from PMI’s country-specific IQOS^™ user databases and were followed-up between 2016 and 2020 during their first 48 weeks in the cohort. Chi-square tests and ANOVA were used to test for statistical differences in categorical and continuous parameters among countries, and the Bonferroni correction for multiple comparisons was applied.

Abbreviations: LAS: legal age for smoking in country +1 year (i.e., Japan ≥ 21 y; Italy, Germany, and Russia ≥ 19 y); N/A: not available; PMI: Philip Morris International; RF: perceived reduced formation of harmful chemicals; rH: perceived reduced risk of harm; SD: standard deviation.

1 Education categories (different category grouping than in Supplemental Table 1 due to harmonization of country-specific categories):

Japan: (1) Elementary School/Junior High School; (2) High School/Old Junior High School; (3) Junior College/Higher Professional School; (4) College/University Graduate School; (5) None of these

Italy: (1) Elementary School/Junior High School; (2) High school/Old Junior High School; (3) Junior College/Higher Professional School; (4) College/University/Graduate School; (5) None of these

Germany: (1) Primary Education/Secondary Education not completed/Secondary Education completed; (2) High School/University entrance; (3) Technical School with statements/Vocational School/Technical School/Supervisor; (4) University degree/Technical degree; (5) Another type of professional training/None of these

Russia: (1) Elementary Education; (2) Secondary Education (8–11 forms); (3) Vocational School or College; (4) Higher Education (University/Institute)/Academic Degree/Second Higher Education; (5) None of these

2 Different employment status category grouping than in Supplemental Table 1 due to harmonization of country-specific categories.

3 Income categories:

Japan: [Yen]: (1) ≤ 199,999; (2) 200,000–299,999; (3) 300,000–399,999; (4) 400,000–499,999; (5) 500,000–599,999; (6) 600,000–799,999; (7) ≥ 800,000

Italy: N/A

Germany: [Euro]: (1) < 1000; (2) 1000–2999; (3) 3000–3999; (4) 4000–4999; (5) ≥ 5000

Russia: [Ruble]: (1) ≤ 50,000; (2) 50,001–70,000; (3) 70,001–90,000; (4) 90,001–110,000; (5) 110,001–130,000; (6) 130,001–150,000; (7) > 150,000

4 No P value available due to different number of income categories across countries that could not be meaningfully harmonized and statistically compared

5 In Italy, for some of the participants RF was not available to be selected as a reason for using IQOS^™ from the beginning of the follow-up.

Data are the OR (95% CI) for the relationship (P < 0.0001 if not indicated otherwise by a superscript) between the number of RF and/or RH indications by participants within the first 48 weeks of follow-up and the likelihood of exclusive HTP and/or IQOS^™ use at week 48 in Japan (≥ 21 y), Italy (≥ 19 y), Germany (≥ 19 y), and Russia (≥ 19 y). Four groups of IQOS^™ users - categorized by the number of times they indicated RF and/or RH as a reason for using IQOS^™ over time - were compared to a reference group of IQOS^™ users who did not make any indication of RF and/or RH. Indicating any combination of the two RF and/or RH related reasons for using IQOS^™ in any week of the follow-up was regarded as a single indication.

Abbreviations: CI: confidence interval; HTP: heated tobacco product; OR: odds ratio; RF: perceived reduced formation of harmful chemicals; RH: perceived reduced risk of harm.

1 P = 0.096;

2 P = 0.011

Data are ORs (95% CI) for the relationship (P < 0.0001 if not indicated otherwise by a superscript) between the number of RF or number of RH indications (indications were not mutually exclusive, therefore, models were mutually adjusted for RF and RH) by the participants (≥ 21 y) within the first 48 weeks of follow-up and the likelihood of exclusive IQOS^™ use at week 48 in Japan. Four groups of IQOS^™ users categorized by the number of times they indicated either RF or RH as a reason for using IQOS^™ over time were compared to a reference group of IQOS^™ users who did not make any indication of RF or RH. Indicating any combination of the RF-related reasons for using IQOS^™ in any week of the follow-up was regarded as a single indication, and likewise for RH indications.

Abbreviations: CI: confidence interval; OR: odds ratio; RF: perceived reduced formation of harmful chemicals; RH: perceived reduced risk of harm.

1 P = 0.026

Data are (i) unadjusted and adjusted HR (95% CI) based on Cox regression analysis for the relationship (P < 0.0001) between indicating RF and/or RH (yes/no) as a reason for using IQOS^™ within 48 weeks of follow-up and the risk of becoming a stable non-exclusive IQOS^™ user (i.e., stable situational IQOS^™ user or stable abandoner IQOS^™ user); (ii) unadjusted percentages of IQOS^™ users becoming stable exclusive IQOS^™ users; and (iii) unadjusted average numbers of weeks to stable exclusive IQOS^™ use in Japan (≥ 21 y), Italy (≥ 19 y), Germany (≥ 19 y), and Russia (≥ 19 y). Adjusted models were adjusted for indicating (yes/no) taste and advantages in comparison with cigarettes (e.g., no smoke, no ash, less smell, less staining of curtains/furniture) as a reason for using IQOS^™, as well as for age, sex, marital status, living situation, employment, and occupational level if the covariates reached a significance level of at least P < 0.10 in the model. IQOS^™ users who were missing one of the covariates were excluded from the analysis.

Abbreviations: CI: confidence interval; HR: hazard ratio; RF: perceived reduced formation of harmful chemicals; RH: perceived reduced risk of harm; stable IQOS^™ user: participant who was at least the last 4 weeks of the 48 weeks of follow-up an (i) stable exclusive (100% IQOS^™), (ii) stable situational (> 0 to < 100% IQOS^™) or (iii) stable abandoner (0% IQOS^™) IQOS^™ user.

1 Here the term “risk” refers to the HR of Cox regression analysis.

2 In all four countries, > 96% of stable exclusive IQOS™ users were already stable exclusive IQOS™ users before the last 4 weeks of follow-up.

3 Sample sizes for the covariate-adjusted analysis were as follows: Japan (N = 3710, ≥ 21 y), Italy (N = 6232, ≥ 19 y), Germany (N = 4721, ≥ 19 y), and Russia (N = 4396, ≥ 19 y).