Smoking restrictions and hospitalization for acute coronary events in Germany

Abstract

Aims: To study the effects of smoking restrictions in Germany on coronary syndromes and their associated costs.

Methods and results: All German states implemented laws partially restricting smoking in the public and hospitality sectors between August 2007 and July 2008. We conducted a before-and-after study to examine trends for the hospitalization rate for angina pectoris and acute myocardial infarction (AMI) for an insurance cohort of 3,700,384 individuals 30 years and older. Outcome measures were hospitalization rates for coronary syndromes, and hospitalization costs. Mean age of the cohort was 56 years, and two-thirds were female. Some 2.2 and 1.1% persons were hospitalized for angina pectoris and AMI, respectively, during the study period from January 2004 through December 2008. Law implementation was associated with a 13.3% (95% confidence interval 8.2, 18.4) decline in angina pectoris and an 8.6% (5.0, 12.2) decline in AMI after 1 year. Hospitalization costs also decreased significantly for the two conditions-9.6% (2.5, 16.6) for angina pectoris and 20.1% (16.0, 24.2) for AMI at 1 year following law implementation. Assuming the law caused the observed declines, it prevented 1,880 hospitalizations and saved 7.7 million Euros in costs for this cohort during the year following law implementation.

Conclusions: Partial smoking restrictions in Germany were followed by reductions in hospitalization for angina pectoris and AMI, declines that continued through 1 year following these laws and resulted in substantial cost savings. Strengthening the laws could further reduce morbidity and costs from acute coronary syndromes in Germany.

Fig. 1

Trend for hospitalization that included an ICD10 code for angina. Time is shifted by state, such that each state’s hospitality ban begins at time 0. The gray line reflects day-to-day variation in the crude hospitalization rate; the blue line reflects month-to-month variation in the crude hospitalization rate. The black lines are the best fit linear regression lines for the hospitalization trend before and after implementation of the law. The inset shows slope estimates with 95% confidence intervals for both trend lines, with the estimate scaled to reflect change in the monthly hospitalization rate. Asterisk before the law, there was no significant time trend (slope = 0.06 [−0.55, 0.66]); after the law, there was a statistically significant downward time trend, −5.3 (95% CI −7.2, −3.5) admissions per month; p-value for the difference in slope<0.0001

Fig. 2

Trend for hospitalization that included an ICD10 code for acute myocardial infarction. Time is shifted by state, such that each state’s hospitality ban begins at time 0. The gray line reflects day-today variation in the crude hospitalization rate; the blue line reflects month-to-month variation in the crude hospitalization rate. The black lines are the best fit linear regression lines for the hospitalization trend before and after implementation of the law. The inset shows slope estimates with 95% confidence intervals for both trend lines, with the estimate scaled to reflect change in the monthly hospitalization rate. Asterisk before the law, there was a statisitcally significant upward time trend, with hospital admissions for angina increasing by 1.7 (95% CI 1.5, 1.9) per month; after the law, there was no significant time trend (slope = 0.05 [−0.57, 0.66]); p-value for the difference in slope <0.0001)