Did the ban on serving raw beef liver in restaurants decrease Enterohemorrhagic Escherichia coli infection in Japan?: an interrupted time-series analysis

Abstract

Background: Enterohemorrhagic Escherichia coli (EHEC) is an important pathogen that causes diarrhea, hemorrhagic colitis, and hemolytic uremic syndrome (HUS). After an EHEC outbreak involving uncooked beef, serving raw beef liver dishes at restaurants was completely banned starting on July 1, 2012 in Japan. However, its long-term associations with the incidence rates of EHEC infections have never been assessed by formal interrupted time-series analysis (ITSA). Methods: A retrospective cohort study to assess the impact of banning raw beef liver provision at restaurants was conducted. The weekly incidence of asymptomatic and symptomatic EHEC infections, the incidence of HUS, and deaths were extracted from the national reportable diseases database from January 2008 to December 2017. ITSA was conducted to evaluate the impact of banning raw beef liver from July 2012. To account for a potential simultaneous external effect, the additional regulation on raw beef red meat handling (implemented in May 2011) and the seasonality were also incorporated into the model. Results: There were 32,179 asymptomatic and 21,250 symptomatic EHEC infections (including 717 HUS cases and 26 deaths) reported during the study period. During the pre-intervention period (before week 27, 2012), there were 0.45 asymptomatic EHEC infections per million-persons per week. The mean post-intervention asymptomatic EHEC infections were 0.51 per million-persons per week. ITSA revealed no baseline trend or change in the intercept and trend (0.002 infections per million-persons per week, 95% Confidence interval - 0.03-0.04, p = 0.93, 1.22, CI -1.96-4.39, p = 0.45, and - 0.006, CI -0.003-0.02, p = 0.68, respectively). For symptomatic EHEC infections, there were 0.30 cases per million per week during the pre-intervention period, and it became 0.33 cases per million per week after the intervention. Time series modeling again did not show a significant baseline trend or changes in the intercept and trend (0.0005, CI -0.02-0.02, p = 0.96, 0.69, CI -1.75-3.12, p = 0.58, and - 0.003, CI -0.02-0.01, p = 0.76, respectively). Conclusion: We did not find a statistically significant reduction in the overall incidence rates of both asymptomatic and symptomatic EHEC infections in Japan after implementing measures, including a ban on serving raw beef liver dishes in the restaurant industry.