Salmonellosis is an important public health problem causing substantial morbidity, and thus also having a significant economic impact. Although most infections cause mild to moderate self-limited disease, serious infections leading to death do occur. In the United States it is estimated that 1.4 million non-typhoidal Salmonella infections with 400 deaths occurs annually 1. Calculations from England and Wales regarding year 1995 resulted in an estimate of 102,227 indigenous cases, with 3,412 hospital admissions and 268 deaths 2.

The infective dose is usually high, but the bacteria grow well in most foodstuff. In food with a high fat content, e.g. chocolate and cheese, the infective dose is very low, and just a few bacteria may be sufficient to cause infection 34. The susceptibility to infection varies; in infants, elderly, or compromised hosts, the critical infective dose is lower 5. The onset of disease is often acute with diarrhoea, nausea and vomiting 6. The incubation period is 1–3 (range <1–10) days. The carrier state is normally 4–6 weeks, but a few percent may be asymptomatic carriers for months or even years 37. No vaccine is available against non-typhoidal salmonellosis. Using the Kauffmann-White scheme for antisera reaction to different bacterial O and H antigens, more than 2,500 Salmonella serovars have been identified, with prevalence patterns varying between different parts of the world 89. In Europe S. Enteritidis has been the most reported serovar since the middle of the 1980s 10. In the developed world salmonellosis due to S. Enteritidis is most often associated with consumption of poultry and eggs 1112.

The European Union (EU) consists of 25 member states along with 4 associated EEA/ EFTA (European Economic Area/ European Free Trade Area) countries and 4 candidate countries, with altogether 630 millions inhabitants. The recent establishment of the European Centre for Disease Prevention and Control (ECDC), shows that control and prevention of communicable diseases are high on the EU political agenda. The Salmonella species is one of the eight micro-organisms in the EU Zoonoses Monitoring Directive (2003/99/EG), which is mandatory to always monitor 13. The subject matter and scope of this directive is to ensure that zoonoses are properly monitored and that foodborne outbreaks receive proper epidemiological investigation, to enable the collection in the EU of the relevant information necessary to evaluate trends and sources. In the Zoonosis Control Regulation (2160/2003) which regulates the control of salmonella and other specified foodborne zoonotic agents, salmonella is the only agent specified 14. This directive and regulations shows that salmonellosis is a disease considered to have a high impact on the human health in the EU. However, public health actions need to be based on best possible surveillance data, but the ability to detect and report enteric zoonoses varies considerably between the European countries, hampering meaningful comparisons between countries 15.

In this study we are using Swedish notification data on all cases of salmonellosis, who acquired the infection during travel in Europe, and a unique database on travel patterns to give an estimate of the comparative burden of non-typhoidal salmonellosis in different European countries.

To measure the burden of non-typhoidal salmonellosis in different countries using the reported cases in each country will not produce comparable numbers, since the reporting systems in the different European countries are very disparate. In some countries all cases are reported, while in some other countries just the hospitalized cases. Furthermore, in many countries indigenous and imported cases are not considered separately.

To get comparable data, we used a different approach, basing our estimates on returning travellers to Sweden. Reliable denominator data on travel are essential when using travel-related disease data for calculations of risks 18. The TDB database gives information on each country/region visited with at least one overnight stay (with information obtained directly from the travellers), providing us with a unique opportunity to relate surveillance data to a better-suited travel denominator when calculating the burden of salmonellosis in European countries. This database has previously been used for estimating the risk of travel-related diseases in various parts of the world 91920212223, and the strengths and limitations of using this approach has been discussed in detail in these articles.

As the numerator we used the notifications in the national surveillance system. Since Sweden has a uniform national system for reporting infectious diseases and the notification database rests on both clinical and laboratory notifications, the sensitivity of the notification system is comparatively good. Using the capture-recapture technique to estimate the proportion of cases being reported by either clinicians or laboratories, more than 99 percent of all diagnosed Salmonella infections were found to be reported each year in 1998 to 2002 24. Even with a good notification system, the reported cases likely only represent the tip of the iceberg. Salmonellosis is an acute, mostly self-limited disease, and most cases never come under the attention of a physician, and even less are microbiologically investigated. This however, should not in any major way affect how the countries in this study compare to each other.

The data presented in this study are based on Swedish travellers and their behaviours and risks could never be fully representative for those of the native population in the respective countries. The risk estimates are also biased towards the parts of the countries with most tourists. However, since the data concerning the different countries are based on the same methods, we believe the risks estimates for the countries from this study are more suited for comparisons than the reported national incidence figures from the notification systems – with different structures and quality. Furthermore, a study from the UK estimated the incidence of salmonellosis presenting to general practitioners in 1995 to 157 per 100,000 and the incidence of laboratory confirmed cases to 80 per 100,000 25, figures in the same magnitude as our estimate of 119 per 100,000 in that country.

In this study certain countries in Europe stood out as having a comparatively very high burden of non-typhoidal salmonellosis. A north-south gradient could be clearly distinguished, with increasing incidences the more south in Europe the country is situated. Also a west-east gradient is shown with a higher burden in eastern countries/regions. Being the most common serovar in most regions of the world, S. Enteritidis is especially dominating in Europe, where almost 67% of all salmonellosis were due to this serovar. In Africa, Asia and America, this serovar has been less dominating and the variety of circulating serovars greater 9.

The duration of travel may affect the risk per 100,000 travellers since persons going to southern countries are mostly on a leisure trip with a longer duration compared to trips to neighbouring countries and thereby the risk for these southern countries could be overestimated. In this study the 4 countries with the highest burden of non-typhoidal salmonellosis were tourist destinations where Swedish travellers mostly stay 1–2 weeks. However, the country with the fifth highest burden, Poland, had travel durations similar to other neighbouring countries of Sweden. This indicates that our method is useful, but improvements can be introduced if duration of stay for the infected traveller is obtainable.

The sometimes very large differences between the estimated disease burden from this study, and the reported figures from the single countries are problematic from public health point of view, therefore the strategy on the future European-level surveillance presented by the ECDC in October emphasized the needs for more comparable surveillance data,.

Due to very rigorous control methods in the agricultural area, Sweden has an extremely low domestic incidence of salmonellosis in humans and well as in animals 2627; less than 0.1 % of food-producing animals are found to carry Salmonella 26. This favourable situation is shared with the EU member state Finland and the associated countries Iceland and Norway. These four countries have also managed to keep free from the pandemic of egg-associated S. Enteritidis that is contributing to a great extent to the burden of salmonellosis in other European countries. At least 75% of all notified episodes of salmonellosis in Sweden are travel-related. Even in neighbouring Denmark, the situation is the reversed, with only 25% of the most common serovar, S. Enteritidis being related to travel 28. When becoming members of the EU, Sweden together with Finland received specific Salmonella guarantees, which states that fresh beef, veal and pig meat together with poultry meat should be tested for the presence of Salmonella bacteria before importation. If any serovar of Salmonella is found, importation is not allowed. These guarantees have contributed to the maintenance of the excellent Salmonella situation in these member states and other states may apply for similar guarantees when they have approved control programmes in the agricultural area. Our study shows that these guarantees have a positive effect on public health, resulting in a very low level of indigenous cases of salmonellosis.

The two most common serovars, S. Enteritidis and S. Typhimurium, are well known to have different animals as reservoir for the disease in Europe. S. Enteritidis is shown to be the most frequent serovar isolated from poultry and poultry products including table eggs, while S. Typhimurium is the most frequent serovar in pigs, pork, beef and veal. In cattle S. Dublin is the most frequent isolated serovar but S. Typhimurium is the next frequent serovar. 29. However, S. Typhimurium is also rather common in poultry and S. Enteritidis is isolated from beef and pork.

Our data shows that focusing prevention of salmonellosis to prevention of S. Enteritidis infection will have a major impact from a public health perspective and will significantly lower the burden of salmonellosis in most European countries. S. Enteritidis is well known to originate from poultry products and products prepared with raw shell eggs. Aiming at reducing the presence of S. Enteritidis in poultry products and eggs will have a positive side-effect since not only S. Enteritidis will be reduced but also other serotypes of Salmonella present in these products resulting in an even lower burden of salmonellosis in Europe. However, since there was no direct correlation between the proportion of S. Enteritidis and the estimated overall incidence of salmonellosis, also other actions than those directed against infected poultry and eggs are needed to get down the burden of salmonellosis in those countries with the highest incidence. The new ECDC has identified zoonoses as one of its main priorities, and will together with the EU Member States, the European Commission and the European Food Safety Agency (EFSA) play a pivotal role in the future actions against salmonellosis in Europe.

The author(s) declare that they have no competing interests.

Both authors participated in the design of the study, performed the calculations and drafted the manuscript. Both authors has read and approved the final manuscript.