The impact of the Danish smoking ban from 2007 on smoking-related morbidity. A register-based natural experiment

Abstract

Introduction The number of countries introducing national tobacco control policies has increased. In terms of both countries and population covered, smoking bans in public settings that protect nonsmokers from second hand tobacco smoke, create supportive environments that encourage smoking cessation, and prevent smoking initiation, represents the one tobacco control policy with the greatest progress since 2007. Extensive research has documented the impact of smoke-free policies on preventing adverse health outcomes, especially cardiovascular diseases, reducing exposure to second hand tobacco smoke and decreasing smoking-related mortality. In 2007, Denmark introduced a national smoking ban, prohibiting all indoor smoking in workplaces and public places. However, little is known about the impact of this smoking ban over time on smoking-related morbidity in the Danish population. This study investigates the effect of the Danish smoking ban on hospital admissions for acute myocardial infarction (AMI), chronic obstructive pulmonary disorder (COPD) and cancer in oral cavity and pharynx. We hypothesized that admission rates decreased after the introduction of the Danish smoking ban in 2007 compared to pre-legislation. Methods This study was designed as a natural experiment in which the whole Danish population form the study population (> 30 years). Interrupted time series analysis was performed to quantify changes in-hospital admission rates for AMI, COPD and cancer in oral cavity and pharynx, pre- and post the introduction of the national smoking ban. Monthly age-standardised admission rates per 100,000 inhabitants were estimated for each diagnosis and two subperiods defined: pre-legislation (January 2000 to August 2007) and post-legislation (September 2007 to December 2015). As an immediate change in admission rates after the introduction of the smoking ban was not expected, increases or decreases in the slope of the trend after the introduction of the ban were compared with pre-legislative trends. Diagnostic information was obtained from the Danish National Patient Register and information on demographics was obtained from the Danish Civil Registration System. The exposure was defined as the introduction of the smoking ban in August 2007 and the outcome as hospital admissions during the study period 2000–2015 due to a primary diagnosis of AMI (I21), COPD (J44) and cancer in oral cavity and pharynx (C06-C06.9, C14). Results We found no significant changes in trends for hospital admissions for any of the included smoking-related diseases after the introduction of the smoking ban compared to post-legislation trends. To exemplify, a steady decrease in-hospital admissions for AMI was observed during the study period suggesting that the decrease in AMI admissions at the time the smoking ban was introduced represented a continuation of a pre-existing trend. Conclusion The findings may be explained by the fact that the smoking ban included several exceptions resulting in a less comprehensive ban compared to other countries.