Posted: June 15th, 2012 - 12:11pm
Source: EFSA Journal 2012;10(6):2726 [25 pp.]

Abstract
The European Union (EU) Summary Reports on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2009 and 2010 – specifically for the data on Salmonella, Campylobacter, verotoxigenic Escherichia coli, Listeria monocytogenes and foodborne outbreaks was reviewed. The main conclusions and recommendations are reported. Comparison between EU Member States (MSs) was found to be difficult due to the differences of the methods used, sampling schemes and reporting systems. Methods, sampling schemes and reporting systems among MSs should therefore be harmonised. When comparing MS-specific trends, the impact of sample sizes, weight of samples and methodologies should be considered, as these variables could otherwise lead to misinterpretation of the data. Incidence data alone do not provide a full picture of the public health burden of zoonotic diseases. Fatalities provide another important insight. Ultimately, summary measures of public health such as disability adjusted life years (DALYs) and cost-of-illness estimates should be presented. Travel information was found to be still incomplete in many MSs. For many pathogens this hampers source attribution. To better understand the public health problems related to food and animal sources in the EU, it is desirable to differentiate between travel within and outside the EU. This would also be useful to better evaluate the public health impact of EU-wide food safety measures. Whenever possible the data/results should be analysed using proper statistical tools. When data do not allow for this, the text should be kept to presenting the data without implying any patterns or trends.
Summary
The European Food Safety Authority (EFSA) asked the Panel on Biological Hazards (BIOHAZ) to deliver a Scientific Opinion on a review on the European Union (EU) Summary Reports on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2009 and 2010 – specifically for the data onSalmonella, Campylobacter, verotoxigenic Escherichia coli, Listeria monocytogenes and foodborne outbreaks.
Comparison between EU Member States (MSs) is difficult due to the differences of the methods used, sampling schemes and reporting systems. Methods, sampling schemes and reporting systems among MSs should be harmonized.
Data on human cases were reported via The European Surveillance System (TESSy) to ECDC by the 27 MSs and three EEA/EFTA countries (Iceland, Lichtenstein and Norway). Switzerland reported human cases directly to EFSA. The completeness of reporting varied between both MSs and pathogens.
The data reported to ECDC did not provide for an accurate picture of the epidemiological situation in the EU as they did not account for underreporting and under-ascertainment. The ratio between true cases and reported cases is known to vary widely between MSs and between pathogens, making comparisons extremely difficult. These limitations also affect the accuracy of risk assessment and attribution studies and furthermore may lead to inappropriate estimates of the cost-benefit ratio of interventions.
Incidence data alone do not provide a full picture of the public health burden of zoonotic diseases. Fatalities provide another important insight. Ultimately, summary measures of public health such as disability adjusted life years (DALYs) and cost-of-illness estimates should be presented.
Trends in reported cases are assumed to provide insight in the changes in disease incidence, and thus for the evaluation of control programs. Supporting evidence for the assumption that the sensitivity of surveillance systems does not change over the years has not been provided. Evaluation of the public health impact of food safety programmes is further hampered by a lack of risk factor information.
Travel information is still incomplete in many MSs. For many pathogens this hampers source attribution. To better understand the public health problems related to food and animal sources in the EU, it is desirable to differentiate between travel within and outside the EU. This would also be useful to better evaluate the public health impact of EU-wide food safety measures.
Whenever possible the data/results should be analysed using proper statistical tools. When data do not allow for this, the text should be kept to presenting the data without implying any patterns or trends. This also relates to use of words or phrases like “only” and “very low” contamination in e.g. RTE turkey and bovine meat, respectively. Such words imply an acceptance of the contamination level, which is outside the scope of the report. Phrases like “lower or higher than” are only considered appropriate to use when comparing trends from e.g. year to year, if supported by proper statistical analyses.
When comparing MS-specific trends, the impact of sample sizes, weight of samples and methodologies should be considered, as these could otherwise lead to misinterpretation of the data. In addition data from MSs, where sample unit and sample size were unspecified could be presented in an annex.