Foodborne diseases in Switzerland: understanding the burden of illness pyramid to improve Swiss infectious disease surveillance

Schmutz, Claudia. Foodborne diseases in Switzerland: understanding the burden of illness pyramid to improve Swiss infectious disease surveillance. 2018, Doctoral Thesis, University of Basel, Faculty of Science.

Available under License CC BY-NC-ND (Attribution-NonCommercial-NoDerivatives).


Official URL: http://edoc.unibas.ch/diss/DissB_12894

Downloads: Statistics Overview


Background: Infectious diseases cause a considerable burden to population health worldwide. Different types of surveillance systems have been implemented to assess changes in disease frequency, to identify outbreaks, and to detect newly emerging diseases aiming at early detection of epidemics, disease control and prevention. Passive surveillance systems are mostly used, measuring the ‘incidence of notified cases’ rather than the incidence (frequency) of disease at population level. Foodborne pathogens, for example, do not always cause disease in infected individuals. Sick individuals – mostly presenting with acute gastroenteritis (AG) – do not always seek healthcare. Of those approaching a physician, aetiology of disease is investigated only in a fraction of patients. Finally, not all cases with a positive laboratory finding for a notifiable pathogen might be reported to the surveillance system. This “loss” of cases along the so-called burden of illness pyramid – from infection to actual notification in the surveillance system –, or the factor of underestimation, depends on the pathogen and the local health (care) system.
Two surveillance systems are implemented in Switzerland which are important for infectious disease surveillance and early detection: the National Notification System for Infectious Diseases (NNSID) and the Swiss Sentinel Surveillance Network (Sentinella). The NNSID is based on the Epidemics Act and is the only mandatory surveillance system in Switzerland covering the entire nation and involving all physicians and diagnostic laboratories. The Epidemics Act defines which observations have to be reported to the NNSID and stipulates the time frame for reporting. Sentinella is a network where information from a subset of
voluntarily participating physicians is collected to study diseases and health issues at the primary care level.
It was estimated that 31 foodborne hazards caused 33 million Disability Adjusted Life Years (DALYs) and 600 million cases of illness worldwide in 2010. In European countries, the incidence of AG was estimated at 0.3–1.5 disease episodes per person-year. Campylobacter spp. is the most frequent, notifiable, bacterial foodborne disease, both in the European Union (EU) and in Switzerland and showed increasing trends in the past decade. In contrast, Salmonella spp. notifications were decreasing in the EU and in Switzerland while enterohaemorrhagic Escherichia coli (EHEC) notifications were increasing in Switzerland but remained stable in the EU.
In Switzerland, information on foodborne diseases is mostly restricted to data obtained through the NNSID. Many factors contribute to underestimation and hence, it is unclear how well notification rates reflect disease incidence.
Objectives: This work aimed at contributing to a better understanding of the burden of illness pyramid for foodborne infections in Switzerland and, thus, contributing to improve infectious disease surveillance and control. It sought to investigate the frequency of cases of foodborne disease or AG at different levels of the burden of illness pyramid.
Further, it should describe trends and understand factors leading to case registration. Finally, a better understanding of disease epidemiology will lead to improvements in early disease detection and control.
Methods: This research work consisted of several projects characterising different levels of the burden of illness pyramid from its tip to the wide (population) base. In a first step, notification data of Campylobacter, Salmonella and hepatitis A were analysed to describe trends since 1988. Considering that the number of tests conducted can strongly)
influence the number of cases detected, we studied the trend in the proportion of positive tests out of all tests performed – the positivity rate – for Campylobacter, Salmonella and EHEC over a 10-year period. Personnel of diagnostic laboratories was consulted to assess current laboratory practices, focussing on the diagnosis of EHEC infections. Furthermore, we conducted a qualitative study among Swiss general practitioners (GPs) to understand physicians’ approaches towards anamnesis (including diagnosis) and treatment of AG in general and campylobacteriosis in particular. Subsequently, these findings were complemented by a study within Sentinella, where the number of patient consultations due to AG at primary care level was assessed. Physicians reported all first consultations due to AG including information on hospitalisation, stool diagnostics, treatment and inability to work.
Findings of the aforementioned studies, expert consultations and publicly available data were used to explore healthcare costs for AG and campylobacteriosis in Switzerland for the first time. Four distinct patient management models were defined for which frequency and individual case management costs were estimated. Extrapolations of these results were used to assess total direct healthcare costs for Switzerland.
Finally, bringing together all study results of the above-mentioned studies, we identified the need to understand the burden of AG at the basis – at the level of the general population. Therefore, a study protocol to investigate the lowest level of the burden of illness pyramid – the incidence and aetiology of AG at population-level – was developed.
Results: Campylobacter case notifications increased between 1988 and 2013 while Salmonella case notifications decreased. Highest case numbers for Campylobacter were recorded in 2012 with 8’480 cases. For Salmonella, peak levels were observed in 1992 with 7’806 cases. While showing inverse long-term trends, both pathogens follow a similar seasonality pattern with higher case numbers during summer months. In winter, a short but pronounced peak over Christmas and New Year was observed for Campylobacter. Positivity rates for Campylobacter increased from 2003 to 2012 while they decreased for Salmonella. At the same time, the number of tests conducted increased for both pathogens. Hepatitis A case notifications decreased between 1988 and 2016 in Switzerland, similar to Salmonella. The strongest decline was observed in the early 1990’s, starting even before active immunisation was introduced in 1992. At the same time, there was a shift in reported risk exposures for hepatitis A: Intravenous drug use was the most frequently mentioned risk exposure at the beginning of reporting while, more recently, contaminated food and beverages were mentioned predominantly as possible sources of infection. Notification forms and content were changed multiple times during this 29-year period.
Laboratory experts unanimously think that the increase in EHEC notifications which is observed in the NNSID can be explained by the introduction of multiplex gastrointestinal PCR panels. Those panels also test for EHEC while traditional culture-based stool testing mostly considered Campylobacter spp., Salmonella spp. and Shigella spp. only.
Nevertheless, there was also an increase in positivity rate observed for EHEC from 2007 to 2016 apart from an increase in testing frequency. Preliminary analysis of surveillance data on testing frequency, which was collected since the implementation of the new Epidemics Act in 2016, reveals several issues regarding data quality related to the complex and heterogeneous “laboratory landscape” in Switzerland.
AG case management of Swiss GPs is diverse. Nevertheless, four distinct strategies could be identified. The majority of patients is managed with a “wait & see” approach based on the knowledge that AG is usually self-limiting. Two of the four approaches include microbiological investigation (stool testing), with antibiotic treatment started either before
or after availability of stool test results. Swiss GPs perceive AG and campylobacteriosis as diseases of minor importance in their daily work but acknowledge that they can be disturbing and debilitating for the individual patient. Surveillance of AG in Sentinella revealed that 8.5% of AG patients received antibiotic therapy, for 12.3% stool testing was initiated and 86.3% of employees were not able to work. Extrapolation of case numbers suggested an incidence of AG at primary care level of 2’146 first consultations per 100’000 inhabitants in Switzerland in 2014. Direct healthcare costs of AG and campylobacteriosis in Switzerland were estimated at €29–45 million in 2012. Of these, €8.3 million were attributed to the 8’480 laboratory-confirmed campylobacteriosis patients registered in the NNSID. It was estimated that 233’000–629’000 patients consulted a physician without further stool testing resulting in healthcare costs of €9.0–24.2 million in 2012. Work-loss and other non-healthcare costs associated with AG and campylobacteriosis were not assessed in this study. However, this socio-economic burden will be explored in more detail in an upcoming study on the burden of gastroenteritis in Switzerland (“BUGS study”). The BUGS study was developed to explore the “true” incidence, burden of disease, aetiology and socio-economic impact of AG in Switzerland; to finally understand the entire burden of AG at population level and the level of underestimation of cases notified to the NNSID. BUGS is a prospective cohort study weekly following up individuals of the general population during a 52-week period. Furthermore, the presence of four pathogenic bacteria (Campylobacter, Salmonella, Shigella and EHEC) and of bacteria harbouring selected antibiotic resistances (fluoroquinolone, extended-spectrum beta-lactamase (ESBL), carbapenemase and mobilised colistin resistance-1 (mcr-1)) is assessed in cohort participants during an asymptomatic period.
Conclusions: The NNSID is a useful and stable surveillance system and health system component which is well accepted by stakeholders. Surveillance data from the NNSID suggest increasing trends for Campylobacter and EHEC and decreasing trends for Salmonella and hepatitis A. Our complementary research studies come to the same conclusion even though trends might appear more pronounced (EHEC) or attenuated (Salmonella) in the notification system than the true incidence due to changes in diagnostic procedures. Hence, from what we know we cannot fully explain the increase of Campylobacter and EHEC seen in the notification system. Therefore, an increase in disease incidence or an outbreak must be considered from an epidemiological perspective. Furthermore, underestimation is probably substantial. Cases seen in the NNSID are more likely to be severe, have co-morbidities or present with well-known risk factors. Assessing all factors contributing to underestimation on a regular basis is hardly possible. Instead, complementary research such as the proposed BUGS study are needed.
The information on disease trends and individual cases obtained through the NNSID should be restricted to the minimum (with high data quality) rather than expanded to keep the system as simple and responsive as possible, providing reliable information. This enables the system to stay alert to and be prepared for a rapid response in the event of changing case numbers. Maintaining systems like Sentinella and fostering strategic research partnerships for action is important to be able to react immediately once an outbreak or a change in disease epidemiology is suspected. Pathways to provide good evidence for public health policy and distribute information to stakeholders should be established.
Advisors:Utzinger, Jürg and Mäusezahl, Daniel and Chappuis, François
Faculties and Departments:09 Associated Institutions > Swiss Tropical and Public Health Institute (Swiss TPH) > Former Units within Swiss TPH > Health Impact Assessment (Utzinger)
UniBasel Contributors:Mäusezahl, Daniel
Item Type:Thesis
Thesis Subtype:Doctoral Thesis
Thesis no:12894
Thesis status:Complete
Number of Pages:1 Online-Ressource (xxxix, 359, XXXVIII Seiten)
Identification Number:
edoc DOI:
Last Modified:30 Jun 2020 01:30
Deposited On:22 Jan 2019 15:37

Repository Staff Only: item control page