Surveillance of foodborne parasitic diseases in Europe in a One Health approach

In 2012, WHO/FAO ranked 24 foodborne parasites (FBP) using multicriteria decision analysis (MCDA) to provide risk assessors with a basis for prioritising control of highly ranked FBP on the global level. One conclusion was that ranking may differ substantially per region. In Europe, the same methodology was used to rank FBP of relevance for Europe. Of the 24 FBP, the top-five prioritised FBP were identified for Europe as Echinococcus multilocularis, Toxoplasma gondii, Trichinella spiralis, E. granulosus, and Cryptosporidium spp., all of which are zoonotic. The objective of the present study was to provide an overview of surveillance and reporting systems in Europe for these top five prioritised FBP in the human and animal populations, to identify gaps, and give recommendations for improvement. Information on the surveillance systems was collected from 35 European countries and analysed according to the five different regions. For most FBP, human surveillance is passive in most countries and regions in Europe and notification differs between countries and regions. Adequate surveillance programmes for these FBP are lacking, except for T. spiralis, which is notifiable in 34 countries with active surveillance in susceptible animals under EU directive. Although human and animal surveillance data are available for the five prioritised FBP, we identified a lack of consistency in surveillance and reporting requirements between national experts and European bodies. Recommendations for improved surveillance systems are discussed.

is commonly used in human surveillance and an example of passive surveillance. In active surveillance, analyses for a particular pathogen in human or animal populations are conducted following a specified sampling plan. Risk-based surveillance and surveillance systems to determine freedom of disease are increasingly being used in veterinary medicine (Stärk et al., 2016). Risk-based surveillance and control principles for FBP have recently been described by Alban et al. (2020), whereas the control of Trichinella spp. in Suidae in Europe, has been regulated at the EU level. The advantage of risk-based surveillance is increased cost effectiveness of the surveillance (European Commission, 2015).
The present study provides an overview of the different surveillance and reporting systems in Europe for the top-five prioritised FBP in human and animal populations, identifies gaps, and suggests recommendations for improvement.

Materials and methods
The survey was conducted in the same five European regions as reported for prioritising FBP in Europe: Northern, Western, Eastern, Southeastern, and Southwestern as previously described (Bouwknegt et al., 2018). Regional coordinators of the Euro-FBP network identified and contacted experts or agencies in each country of their region, sending them a questionnaire for completion and a glossary document with explanations. The questionnaire requested information about disease notification and surveillance systems for humans and animals during the previous five years (2014)(2015)(2016)(2017)(2018), the populations (for humans) under surveillance, reporting sources, and case definitions as described in EU decision 2018/945 (Commission Implementing Decision (EU), 2018). Data of the various countries per parasite were combined in a common Excel database and afterwards analysed per region. Results from the different regions were combined per parasite and discussed in a common meeting of all regional coordinators. An additional online questionnaire was subsequently used to address gaps and inconsistencies identified at the meeting (Supplementary file). We did not wish to include project-based surveillance, as this may be non-systematic and short-term, unless conducted according to governmental requirements (e.g., E. multilocularis in wildlife to document freedom from infection).
The information collected used the following pre-agreed definitions and criteria: • Notifiable: by law of that country.
• Passive surveillance: gathering data from clinical or laboratory diagnosed cases of infection.
• For humans, notifiable diseases are based on passive surveillance of the whole population; for non-notifiable diseases, this can be for a restricted population (e.g., pregnant women). • For animals, only for infections that cause clinical symptoms, such as E. multilocularis infections in dogs causing alveolar echinococcosis; T. gondii infections in pregnant sheep, Cryptosporidium infection in ruminants. • Active surveillance: analyses for a particular pathogen in human or animal populations follow a statistical or probability-based sampling plan. This is important in surveillance of diseases in which subclinical cases/carriers predominate. Active surveillance in humans may be population-based for, e.g., government records. Data from meat inspection of slaughtered animals for food production in the abattoir is defined as active surveillance in this study. Population-based surveillance can also include monitoring for justifying derogation from regulations, such as compulsory treatment of imported dogs into countries with proven absence of E. multilocularis in wildlife (Commission Delegated Regulation (EU), 2018; European Food Safety Authority and European Centre for Disease Prevention and Control, 2019).
The various active and passive surveillance systems in the different countries were not evaluated for their effectiveness, and no data on this aspect were requested in the questionnaire.

General overview
In total, 35 countries including EU Member States (MS) and non-MS in five European regions were included in the study. For many countries, information was obtained from public health and animal health institutes and food-safety agencies. The different surveillance systems in humans and animals are combined by region (Tables 1-5), and also described by parasite for all regions.
Active surveillance during meat inspection was described in EU Regulation 854/2004 and, since 2019, replaced by EU Regulation 627/2019 (European Regulation (EC), 2004; Commission Regulation (EU), 2019) for the MS. In the EU Regulation, practical arrangements for official control for cysticercosis during post-mortem inspection caused by Taenia saginata in domestic bovine animals and by T. solium for Suidae and for Trichinella spp. in susceptible slaughter animals are included. Risk-based surveillance has been in place for T. saginata in bovine animals since 2019, and for Trichinella in slaughter pigs since 2015 (European Commission, 2015). For Echinococcus spp. no additional arrangements are included. As Echinococcus spp. cysts are identifiable during meat inspection, both E. granulosus and E. multilocularis lesions are included as active surveillance in Tables 1-5 for slaughter animals. However, E. multilocularis cysts are seldom present in slaughter animals.
All countries reported that they follow case definitions of the parasitic diseases under survey, according to EU decision 945/ 2018 in all regions (Commission Implementing Decision (EU), 2018).   Prevention and Control, 2019). DK is developing its notification system. The disease is also not notifiable in BH, BG and ME.  Alveolar echinococcosis in dogs and cats will be reported during passive surveillance in BE, CH, FR, LI and NL but is very rare in N and S Europe and therefore unclear if it will be reported in many regions.
Speciation of Echinococcus in animals is reported from CZ, DK, EE, FI, FR, HU, ICE, NL, NO, SK, and SWE.

Toxoplasma gondii and toxoplasmosis
In humans, the case definition is that of the EU (Commission Implementing Decision (EU), 2018) and sometimes regulated by national legislation for acquired toxoplasmosis, as, for example, in HU and PT.
Congenital toxoplasmosis is notifiable in most European countries, except for AU, BE, CH, DK, NO, NL, and SWE, while both acquired and congenital toxoplasmosis are notifiable in AL, BH, BG, CY, EE, ES, FI, GR, HR, HU, ICE, IE, ME, MK, LV, PL, PT, RO, RS, SL and SK. In IT, toxoplasmosis is notifiable without specifying the form of infection (congenital or acquired), whereas acquired toxoplasmosis is notifiable in CZ. Screening pregnant women is mandatory in AU, BE, FR, SK, HR, IT, PL, and SL, and, since 2017, in RS, but is voluntary in BG, HU, and also in CZ and DE where screening is not covered by statutory health insurance (European Centre for Disease Prevention and Control, 2019). In ES, screening is managed at the Autonomous Communities level. In PT, screening it is not mandatory but highly recommended.
Passive surveillance of human clinical cases occurs in most countries, but whether these cases are reported is unclear. Clinical cases are not reported in BE and NL. In some countries with passive surveillance (e.g., CZ, EE, HU, SK, SR) both hospitalised and other patients with clinical signs are reported. In E Europe, active surveillance is carried out in PL in pregnant women, HIVpositive patients, and organ donors/transplant recipients. Other countries in this region report passive surveillance, but in different target groups, with mainly pregnant women and HIV-positive patients in CZ, EE, HU, LV, SK; preterm neonates and infants in HU and SK; organ donors and transplant recipients in SK; and patients with clinical signs and hospitalised patients in PL, RO, SK.
Population-based serosurveillance studies have been reported from AU, BE, DE, ES, FR, ICE, NL, NO, PT and SWE (Hofhuis et al., 2011;Findal et al., 2015;Evengård et al., 2001;Birgisdóttir et al., 2006). Underreporting in most countries seems likely due to the lack of clear rules for reporting.
In animals, toxoplasmosis is notifiable in BE, CH, DE, FI, FR, ICE, IE, LI, LV, NL, MK, PL, RS and SL. Passive surveillance based on clinical cases in N Europe occurs only in ICE and FI. In W Europe, infections (abortions in small ruminants) are notifiable in BE, IE, and NL. Toxoplasmosis is reportable in livestock, companion animals, and zoo animals in CH, DE, and LI, but whether these cases are reported is unclear. Passive surveillance is carried out in LV, PL, RO, and SK in all animal species with clinical signs and/or to determine cause of death. In several countries in SE Europe, clinical cases (abortions) in sheep and goats are notifiable, and passive surveillance based on differential diagnosis in aborted small ruminants occurs in some Autonomous Communities in ES. In some countries, active surveillance is carried out (serology and molecular methods) in slaughtered pigs, cattle, and sheep. However, these programmes are not intended as control measures (European Food Safety Agency Panel on Biological Hazards (BIOHAZ) et al., 2018).

Trichinella spiralis and trichinellosis
Notification of trichinellosis in humans is mandatory in all countries, except BE, FR, ME, TR and UK that have voluntary surveillance. DK also does not report surveillance.
In all regions and almost all countries, Trichinella infection in animals is notifiable. Active surveillance is carried out in all countries during meat inspection according to EC regulation 1375(European Commission, 2015Commission Regulation (EU), 2019). In IE, there is (active) risk-based surveillance in wildlife (red foxes); in PT there is an official surveillance plan for hunted wild boar, and in ES and IT there is serological follow up in positive pig farms. In CZ, four red foxes are examined per 100 km 2 annually, and active surveillance of wild boars and other wild animals (e.g. beavers, lynx, bear) used for human consumption is reported.

Echinococcus granulosus and cystic echinococcosis
CE is notifiable in most countries in Europe, except BE, CH, DK, LI, NL, and UK, but there is no active surveillance. The case definition of EU decision 2018/945 is used by those countries where CE is notifiable, but with no separate case definitions for AE and CE and speciation of human cases as previously described.
In most countries, passive surveillance is based on reporting clinical cases. In ES, passive surveillance is conducted through the "Red Nacional de Vigilancia Epidemiológica". In IT, a register of CE that contains information from clinical cases has been extended to the European Register of Cystic Echinococcosis (Rossi et al., 2016), to collect harmonised clinical data in the EU on a voluntary basis.
In animals, E. granulosus is notifiable in most countries except IT and ES. Active surveillance is mandatory in all countries by visual inspection of cysts in liver and lungs during meat inspection of ruminants, horses, and other susceptible animals, according to European Regulation 2019/627 (Commission Regulation (EU), 2019). PCR confirmation of suspected cysts is reported by FR and NL.

Cryptosporidium spp. and cryptosporidiosis
Human surveillance is passive in almost all countries and regions of Europe. There is wide variation both within and among regions regarding notification of cases, with cryptosporidiosis being notifiable in most countries in N Europe except DK, three countries in W Europe (DE, IE, UK), all countries in E Europe except PL, RO, and SK, all countries in SW Europe except IT, and five countries (BG, HR, MK, ME, SL) in SE Europe. Most countries report all diagnosed patients with clinical signs, but SK reports only hospitalised cases. Some countries (CZ, DK, FR, NL, NO, SWE) report that they identify and report specific Cryptosporidium species. However, there are differences regarding Cryptosporidium speciation: in SWE cases are speciated at the public health agency in the summer period and during outbreaks. In NO speciation is sometimes conducted, particularly during outbreak investigations.
Animal surveillance for Cryptosporidium infection is passive (animals presenting with clinical signs) in all countries, apart from active surveillance of cervids in DK and cattle in PL. Cryptosporidium infection in animals is not notifiable in any country except in CH. Some countries (CZ, PL, FR, DK, NL) report specific Cryptosporidium species to EFSA.

Discussion
Here we provide an overview of different surveillance systems in Europe for the top-five prioritised FBP in human and animal populations. These FBP, all of which are zoonotic, were prioritised using a similar approach as that used by WHO/FAO on the global scale, and based on different, mostly public health-based, criteria (Bouwknegt et al., 2018). Among these five FBP, infections caused by E. multilocularis, E. granulosus, and T. gondii tend not to cause acute clinical disease. For instance, most T. gondii infections are asymptomatic or only cause mild or unspecific symptoms; long-term sequelae, such as ocular toxoplasmosis, can occur years later. CE and AE have long incubation periods, ranging from five to fifteen years, making it extremely difficult to study outbreak situations or determine source attribution for sporadic cases. For cryptosporidiosis, with a relatively short incubation period (5-7 days), outbreaks can be more readily noticed and reported (European Food Safety Agency Panel on Biological Hazards (BIOHAZ) et al., 2018). However, even for Cryptosporidium, reporting is a problem in many countries due to lack of routine diagnostics and speciation of human and animal cases.
In general, although human and animal surveillance data are available for all five FBP, the surveillance and reporting requirements vary among and within regions and countries, and among national experts and European bodies.
For instance, AE in humans, which ranked as the most important FBP (Bouwknegt et al., 2018), is notifiable in N, E, and SE Europe, but in only four countries in W Europe. Moreover, echinococcosis speciation is not routinely conducted and, therefore, when human cases are diagnosed, many countries report only "echinococcosis", despite major differences in pathology, epidemiology, and disease progress between CE and AE. Underreporting seems to be a major issue for CE in SE Europe, according to a cross-sectional study (Tamarozzi et al., 2018), ultrasound screening of people in rural areas in Bulgaria, Romania and Turkey showed a prevalence of abdominal CE of 0.41% in Bulgaria and Romania and 0.59% in Turkey. Bulgaria accounted for 26% of all confirmed echinococcosis cases reported to ECDC in 2017 and for 53% of the E. granulosus cases, Romania only for 2%. The incidence data reported to ECDC differed from these values by a factor of 10 for Bulgaria and a factor 700 for Romania (Tamarozzi et al., 2018). Similar results have been shown in Italy and Greece comparing research data and ECDC reports (European Food Safety Authority and European Centre for Disease Prevention and Control, 2013;Brundu et al., 2014). In animals, Echinococcus cysts (mainly E. granulosus) are not mentioned in EU regulation 2019/627 (Commission Regulation (EU), 2019) for the control of animal products intended for human consumption, although the public health impact is huge in some regions in Europe. Consequently, E. granulosus in slaughter animals is underreported in many countries due to the low sensitivity of meat inspection, the lack of confirmation and PCR-based speciation of suspected lesions, and an absence of data registration systems. Due to the many neglected CE cases in SW and SE Europe (Tamarozzi et al., 2018;Brundu et al., 2014), improved veterinary public health control measures are particularly needed in these regions to prevent human exposure. Effectiveness of the improved control measures can be monitored by active surveillance of the human population.
Active surveillance for E. multilocularis in wildlife (red foxes) occurs in many countries, but often only on voluntary basis, despite EFSA recommending harmonised surveillance in wildlife (Boué et al., 2010). Thus, understanding the prevalence and spread of this FBP is difficult. Furthermore, clinical cases in humans and infections in animals are not notifiable in many countries, and, consequently, there is substantial underreporting of this emerging parasite in Europe (Conraths et al., 2017). In conclusion, surveillance of both AE and CE in humans and animals is variable and fragmentary, with potential underreporting. In addition, suspected human cases are not confirmed by PCR in many countries, and potential E. granulosus findings during meat inspection are often not confirmed by PCR nor reported. Whereas E. multilocularis is a concern in N Europe according to the ranking as reported by Bouwknegt et al. (2018), CE is considered to be a major neglected zoonosis in S Europe with significant economic losses in the public health sector (Piseddu et al., 2017). Here the life cycle of E. granulosus s.s. continues, the disease burden is substantial, and improved veterinary control is needed.
Mandatory and separate notification of E. granulosus and E. multilocularis infections in both humans and relevant animals is strongly recommended in all countries, such that clearer insights into the extent of the problem are obtained and trends can be analysed.
Congenital toxoplasmosis is notifiable in 29 of the 35 countries and pregnant women are screened in some countries; nevertheless, underreporting is still a major problem in many countries. Sero-surveillance studies have been conducted in DE, ICE, ES, NL, IT, NO, PT, and SW. Data from these studies in NL has shown that toxoplasmosis has one of the highest disease burdens among foodborne diseases (Havelaar et al., 2015). Although (sero)positivity is reported in slaughtered animals in most countries, there is no mandatory control in Europe to prevent human infections via consumption of meat from infected livestock. This reinforces the need for risk-based surveillance systems in livestock. A social cost-benefit analysis of two interventions in NL showed that freezing raw meat products would be beneficial (Suijkerbuijk et al., 2019).
As trichinellosis is notifiable in humans in most countries and there is mandatory control in animals, surveillance and reporting of T. spiralis infections in humans and animals are the best of all five prioritised FBP. However, although serious clinical cases are probably diagnosed, mild cases may be missed because active surveillance is lacking and outbreaks continue to occur, mainly associated with meat products from pigs reared under non-controlled housing conditions and hunted wild boar. It has recently been shown that pigs from controlled housing systems in Europe pose a negligible risk, and control should be focussed on pigs reared outdoors and on wildlife intended for human consumption (Franssen et al., 2018). The EU directive 2015/1375 includes the possibility of this risk-based surveillance (European Commission, 2015). Such risk-based surveillance will not focus on pigs in controlled housing with negligible risk, but on animal populations with a high risk for Trichinella, such as pigs reared outdoors and wildlife intended for human consumption, such as wild boar.
Although there is general concordance across Europe regarding passive surveillance of both humans and animals for Cryptosporidium, there are considerable differences in reporting, resulting in a skewed impression of the distribution of this parasite. Several countries claimed to have voluntary notification, but it is unclear what such notification entails, and how the data are recorded or to whom they are accessible. Information regarding speciation of Cryptosporidium was also variable among and within regions; in some cases, it appeared that responsible agencies did not have a clear overview of whether speciation was conducted when cases were diagnosed. This obviously affects the data quality. Although outbreaks of cryptosporidiosis are likely to be identified due to the acute disease onset, underreporting of cases during outbreaks occurs, while many sporadic cases are probably not diagnosed.
Based on this data gathering, recommendations regarding surveillance and reporting of the top-five FBP in Europe may be defined. Improved diagnosis and reporting of human CE and AE cases in Europe is needed with emphasis on the highly endemic countries in southern Europe. Special attention should be given to species identification of the different echinococcus species by molecular methods. In addition to human CE, also the reporting and the speciation of E. granulosus cysts mandatory during animal slaughterhouse surveillance should be improved, and early warning and surveillance systems to determine the prevalence need to be harmonised in wildlife for E. multilocularis throughout Europe. EFSA (2007) recommended a harmonised surveillance scheme in wildlife with estimated prevalence sizes of targeted wildlife species, harmonised detection methods and targeted regions for surveillance throughout Europe (Boué et al., 2010), but so far surveillance has often been project based and fragmented in the different countries, making comparisons among countries extremely difficult. T. gondii causes is an important foodborne infection that ranked high in Europe according to the MCDA methods (Bouwknegt et al., 2018) and DALY estimates for toxoplasmosis (Havelaar et al., 2015). However, reporting of congenital toxoplasmosis is absent in many countries in Europe. More efforts are needed to obtain better insight into congenital toxoplasmosis cases and the seroprevalence in Europe to improve estimation of the disease burden and thus adequately prioritize the control measures to be taken. In animals, surveillance is underdeveloped in livestock species and risk-based surveillance of livestock needs to be improved to reduce human meat-borne infections. Riskbased surveillance in meat producing animals can be based on risk assessment studies that are available for Italy and the Netherlands (European Food Safety Agency Panel on Biological Hazards (BIOHAZ) et al., 2018) and could be used to prevent human infections and thus reduce the disease burden in humans. Trichinellosis is the only FBP where reporting is mandatory in most countries in humans, despite a low global disease burden . Therefore, how much effort is needed to control this FBP in animal populations is debatable. However, half of the reported global human cases occur in Europe . Nowadays, a risk-based surveillance system can be implemented under EU reg. /1375(European Commission, 2015, which is also harmonised with guidelines for the Codex Alimentarius and the OIE (Alimentarius, 2015). Despite this, in almost all countries, all slaughtered pigs originating from controlled housing are tested. Therefore, riskbased surveillance in animals should be given more priority to be implemented in Europe to improve cost effectiveness. Improvement of human cryptosporidiosis reporting is needed, because only limited data are available about the presence and prevalence of cryptosporidiosis in human populations. Moreover, harmonised detection methods will improve comparison of reporting between countries. Determination of whether human infections are C. parvum, C. hominis, or another species is recommended.

Declaration of Competing Interest
Authors declare that they have no conflict of interest.