Basis for a One Health Approach—Inventory of Routine Data Collections on Zoonotic Diseases in Lower Saxony, Germany

Simple Summary: To the authors’ knowledge, there is currently no standardised inventory on information on regularly collected routine data on zoonotic infections, meaning that they can spread from animals to humans and the other way around. We used a template with specific questions on the level of the data base and the different pathogens documented in each of these. To support further development in this sector, we gathered information on 37 different data bases, mainly from the human and veterinary sector, and showed opportunities and challenges in joint analysis of data from these data collections. Abstract: Two-thirds of human infectious diseases are zoonotic diseases and routine data collections exist for each sector (human, veterinary, environmental). However, these operate separately and the collected data are not integrated across sectors. Publicly available information on these routine data collections in terms of metadata and the information collected is sparse. The aim was to create an inventory of routine data collections in the Federal State of Lower Saxony, Germany. A systematic screening of existing routine data collections from the human and veterinary sectors on zoonotic infectious diseases was carried out on the basis of expert interviews. A standardised template was used to collect relevant metadata on data collections and pathogens they contain. The template was transferred to Research Electronic Data Capture tools. We recorded metadata for 19 veterinary, 16 human and 2 other data collections, and for 69 different zoonotic pathogens. The frequencies of a selection of metadata were analysed descriptively. The data collections, which served different purposes, differed, e.g., in underlying population and sampling strategy, export format and access to the original data. We identified challenges for integrated analyses of data from different collections, which need to be addressed to develop a One Health monitoring and surveillance system.


Introduction
Although approximately two-thirds of all human infectious diseases are zoonotic infections [1], data from public health and veterinary authorities are usually not routinely linked and analysed together.Zoonotic pathogens can cause both acute (e.g., gastrointestinal) and chronic diseases (e.g., endocarditis following a Q fever infection [2]) and are therefore of high public health relevance.In Germany, of the about 60 notifiable pathogens (according to the German infection protection law) about 40 can be classified as zoonotic.One of the most frequent notifiable zoonotic, thus One Health-relevant, infection of which food is the main source of infection, is Campylobacter spp. with 60,000-70,000 cases per year (80-90 cases per 100,000 inhabitants) in Germany [3].The control of zoonoses is a Zoonotic Dis.2024, 4 58 major One Health challenge.Globalisation, trade and climate change, among others, cause changes in living conditions and affect the transmission and occurrence of zoonoses [4][5][6].This can be observed at global, national and local levels.Multisectoral monitoring, meaning primarily the simultaneous consideration of the human, animal and environmental sectors, is therefore of particular importance.
In Germany, existing routine data collections as well as monitoring and surveillance systems (MOSS) for the documentation of zoonotic agents in the human, animal and environmental sectors are currently operated separately [7][8][9].Data on the national level are collected according to different regulations, e.g., the Act on the Prevention and Control of Infectious Diseases in Humans of 20 [10] or the General administrative regulation on the collection, analyses and publication of data about the occurrence of zoonoses and zoonotic agents along the food chain from 10 February 2012, last changed on 13 September 2023 (BAnz AT 13 September 2023 B1).Additional data collections are applied only in specific federal states.For example, in Lower Saxony, the (human medicine) Surveillance for influenza and other acute respiratory illnesses (ARE) or particular pathogen-based laboratory projects in veterinary medicine.
The linkage of data from human and veterinary medicine in the sense of a One Health approach has been repeatedly discussed and recommended [8,11,12].In Germany, the existing documentation systems serve different purposes and are operated under the legal responsibility of different authorities.So far, information exchange between human and veterinary health authorities in disease outbreak situations is only partially standardised and takes place in a rather informal way.However, a One Health MOSS, or framework, has only been established for basic data.
This study was carried out as part of the Connect One Health Data (Connect OHD) project.The overall aim of the Connect OHD project was to develop a concept, using underlying metadata structures of routine data collections, for a targeted integration and analysis of existing data collections to improve One Health surveillance.The aim of the present investigation is to provide an inventory of the existing routine data collections for zoonotic pathogens in Lower Saxony and corresponding metadata to set up the basis for secondary data use in the Connect OHD project.

Concept of the Study and Collaborating Partners
For this investigation, the Federal State of Lower Saxony, Germany, was chosen as a model region, because of the high livestock density and the existing cooperation between the sectors.The number of inhabitants in Lower Saxony is over 8,000,000 (31 December 2021 [13]).The number of food-producing animals is almost 100,000,000 (over 25,000 farms), including approx.2,360,000 dairy and beef cattle (over 15,000 farms), 8,600,000 pigs (over 6000 farms), 55,500,000 broiler (over 900 farms) and 20,200,000 laying hens (over 4000 farms) and 5,000,000 turkeys (over 300 farms) (1 March 2020 [14][15][16]).
In Lower Saxony, data on zoonotic pathogens are mainly collected, compiled and managed by the Public Health Agency of Lower Saxony (NLGA) for human data and the Lower Saxony State Office for Consumer Protection and Food Safety (LAVES) for data from animal, food and feed origin.As the overall aim is to link data on zoonotic pathogens from routine data collections, these important stakeholders have been involved as project partners.The Department of Biometry, Epidemiology and Information Processing at the University of Veterinary Medicine Hannover (TiHo) was responsible for project coordination, data integration, management and analysis.

Identification and Selection of Data Collections and Pathogens
A systematic screening of existing human and veterinary data collections in Lower Saxony, focusing on routine data collections to which we have access at LAVES and NLGA, was performed.We also included data collections on zoonotic pathogens held by other institutions, e.g., the Friedrich-Loeffler-Institute, Federal Research Institute for Animal Health (FLI).Criteria for the inclusion of data collections was the regular recording of data on at least one zoonosis or zoonotic pathogen, independent of the population studied, the objective of surveillance and it had to refer to Lower Saxony and Germany.In addition, data collections on other One Health-relevant information, e.g., weather data, were included.In order to ensure that we have a complete collection of routine data in Lower Saxony, various experts were consulted during the enrolment process and missing data collections were included.The search for data collections, documentation of metadata, as well as continuous updating was carried out from April 2020 to February 2022.

Collection of Metadata on Data Collections and Pathogens
In order to record information (metadata) of data collections, we developed a standardised template based on Wendt [17].The template was optimised after consultation with respective thematic experts and tested in a pilot phase, where it was applied to two data collections in each sector.It was used to collect standardised metadata for the hierarchical levels of data collections, pathogens and variables.The same template was used for each sector (human, veterinary and environmental).
The template included 71 items on the data collection levels.The information recorded included the responsible institution, the purpose of the data collection, the populations covered (e.g., animals, humans, environment), the lowest level of aggregation of the original data, type of data collection (e.g., sampling), as well as possibilities to access the data.
For each data collection, we collected information at the pathogen level on the sample material examined, the depth of differentiation of the pathogens, the legal basis of the sample collection and on the availability of geographical or temporal information.At the pathogen level, the metadata included 50 items.
For each variable recorded in the data collection, 13 items, e.g., on variable content, data type and format, were collected (the original version of the template used can be found in Supplement S1, the translated version in Supplement S4).
The translation of the original template used was done with the help of deepl.com(accessed on 31 January 2023) and additional screening and correction by the authors.Metadata were gathered by the study team based on publicly available information and with the help of the experts for the respective data collection.For the management of metadata, the template itself was transferred to the Research Electronic Data Capture (REDCap) tools [18,19] hosted at TiHo.Access to the REDCap database can be requested from the authors after formal approval.

Metadata of Data Collections
The metadata of 37 data collections were recorded.Nineteen data collections were assigned to the veterinary sector, sixteen to the human sector and two could not be assigned to either sector (see Supplement S2).For five of the human data collections, the NLGA is the competent authority, and for five of the veterinary data collections the LAVES.The other data collections are under the responsibility of different institutions.
Of the examined data collections, fourteen refer to the "human" population, four to the "animal" population, four to "food or feed", two to "environment or water", four to "other" targets, like demographics and nine data collections refer to a combination of different populations like "animal" and "food or feed" (Figure 1).Of the 37 data collections, 13 are not pathogen-based, i.e., do not contain specific pathogen information (but some health-related information, e.g., climate data and trade).For further six data collections, pathogen-related metadata were not collected because zoonotic pathogens are in practice rarely documented in these data collections (e.g., haemovigilance report, diagnoses in hospitals) or there was an overlap between data collections and pathogen metadata were already collected for another data collection, e.g., the laboratory information management system (LIMS), which data are used for different purposes (Table 1).The LIMS of the LAVES collects information on all processed samples, associated tests and results.Therefore, it contains information on samples from different investigation programs, e.g., the federal surveillance plan, the MOSS according § § 50-52 German Food and Feed Code, the National Residue Control Plan, the Monitoring of zoonoses and zoonotic agents, the Zoonoses Trend Reports and the Salmonella control programme.Thus, metadata of the data collections of the LAVES and the contained data collections were collected for each one, but, in order to avoid overlapping and double information, the pathogen part of the template was not used for these data collections.Furthermore, connections between data collections were documented.
The level (statistical unit) at which data is recorded in the various data collections varies.Thirty-one data collections contain individual data with an individual human case, an animal, a farm or a sample being the statistical unit.Six data collections contain aggregated data.Of the 37 data collections, 13 are not pathogen-based, i.e., do not contain specific pathogen information (but some health-related information, e.g., climate data and trade).For further six data collections, pathogen-related metadata were not collected because zoonotic pathogens are in practice rarely documented in these data collections (e.g., haemovigilance report, diagnoses in hospitals) or there was an overlap between data collections and pathogen metadata were already collected for another data collection, e.g., the laboratory information management system (LIMS), which data are used for different purposes (Table 1).The LIMS of the LAVES collects information on all processed samples, associated tests and results.Therefore, it contains information on samples from different investigation programs, e.g., the federal surveillance plan, the MOSS according § § 50-52 German Food and Feed Code, the National Residue Control Plan, the Monitoring of zoonoses and zoonotic agents, the Zoonoses Trend Reports and the Salmonella control programme.Thus, metadata of the data collections of the LAVES and the contained data collections were collected for each one, but, in order to avoid overlapping and double information, the pathogen part of the template was not used for these data collections.Furthermore, connections between data collections were documented.The level (statistical unit) at which data is recorded in the various data collections varies.Thirty-one data collections contain individual data with an individual human case, an animal, a farm or a sample being the statistical unit.Six data collections contain aggregated data.
The purposes of the data collections recorded were diverse (Figure 2).For human data collections, the most common purposes are identification of events, monitoring and prediction (n = 14; 38.9%), evaluation of activities (n = 10; 27.8%) and tracing or tracking (n = 4; 11.1%).The most common purposes for data collection in the veterinary sector are food safety (n = 11; 23.4%), prediction, identification, monitoring (n = 10; 21.3%) and identification of risk factors (n = 8; 17.0%).Concerning the sampling strategy, 5 out of 37 data collections (13.5%) are based on passive surveillance, 20 (54.1%) are based on active surveillance and 12 (32.4%)are surveys.Most of the veterinary data collections (52.6%, n = 10) gather their data via active surveillance, five (26.3%) via a survey and four (21.1%) via passive surveillance.For the human sector, eight (50%) out of sixteen data collections collect the data via active surveillance, seven (43.8%) via a survey and only one (6.2%)collects data via passive surveillance.All data collections that belong to other sectors collect their data in active surveillance.
The number of records in the data collections per year varied from low numbers, e.g., the nationwide system for collecting data on food involved in foodborne outbreaks with less than 50 records per year, to data sources with more than 1,000,000 records per year (LIMS LAVES, Trade Control and Expert System New Technology-TRACES NT and the diagnoses in hospitals).
Data can be extracted from the documented original data collections in a variety of export formats, including, Excel (n = 22), CSV (n = 19) and other formats (n = 12, often XML).However, six data collections only offer an export in PDF-format.Exporting data in two or more different data formats is possible for 25 data collections.For two data collections, information on the export format was not available.
Access to and export of original data is regulated in very different ways.Data directly available through open access are sparse.At the State Office for Statistics in Lower Saxony, For the human sector, eight (50%) out of sixteen data collections collect the data via active surveillance, seven (43.8%) via a survey and only one (6.2%)collects data via passive surveillance.All data collections that belong to other sectors collect their data in active surveillance.
The number of records in the data collections per year varied from low numbers, e.g., the nationwide system for collecting data on food involved in foodborne outbreaks with less than 50 records per year, to data sources with more than 1,000,000 records per year (LIMS LAVES, Trade Control and Expert System New Technology-TRACES NT and the diagnoses in hospitals).
Data can be extracted from the documented original data collections in a variety of export formats, including, Excel (n = 22), CSV (n = 19) and other formats (n = 12, often XML).However, six data collections only offer an export in PDF-format.Exporting data in two or more different data formats is possible for 25 data collections.For two data collections, information on the export format was not available.
Access to and export of original data is regulated in very different ways.Data directly available through open access are sparse.At the State Office for Statistics in Lower Saxony, access to data is generally possible online but sometimes with restrictions due to confidentiality.Data from the Climate Data Center from the Germany's National Meteorological Service are publicly available.
Results (aggregated data) from data collections in the human sector are usually publicly available (n = 14, 87.5%), while anonymised original data are available on request (n = 2; 12.5%).In the veterinary sector, access to data is more limited.Only three out of nineteen (15.8%) data collections provide publicly accessible data.For the remaining data collections (n = 16, 84.2%) access is denied.Here, data use is possible only after written request and its formal approval by the competent authorities.In addition, data export is restricted due to technical reasons.Overall, the results show that access to data in the veterinary sector, both for researchers and the public, is much more limited than in the human sector.
The results of 86.5% (32/37) of the data collections from all sectors are published in regular reports.

Metadata of Pathogens
Metadata were collected for about 69 different zoonotic agents.The number of pathogens per data collection varied considerably and depended, e.g., on the sector, the reference population and the legal basis or purpose of the data collection.However, Escherichia coli was covered by most data collections from both sectors compared to other pathogens (see Figure 3).access to data is generally possible online but sometimes with restrictions due to confidentiality.Data from the Climate Data Center from the Germany's National Meteorological Service are publicly available.
Results (aggregated data) from data collections in the human sector are usually publicly available (n = 14, 87.5%), while anonymised original data are available on request (n = 2; 12.5%).In the veterinary sector, access to data is more limited.Only three out of nineteen (15.8%) data collections provide publicly accessible data.For the remaining data collections (n = 16, 84.2%) access is denied.Here, data use is possible only after written request and its formal approval by the competent authorities.In addition, data export is restricted due to technical reasons.Overall, the results show that access to data in the veterinary sector, both for researchers and the public, is much more limited than in the human sector.
The results of 86.5% (32/37) of the data collections from all sectors are published in regular reports.

Metadata of Pathogens
Metadata were collected for about 69 different zoonotic agents.The number of pathogens per data collection varied considerably and depended, e.g., on the sector, the reference population and the legal basis or purpose of the data collection.However, Escherichia coli was covered by most data collections from both sectors compared to other pathogens (see Figure 3).The metadata collected on the level of pathogens for each data collection can be used to identify overlaps in the available information in order to plan an integrated analysis.A subset of the metadata collected is shown in Table 2 for Campylobacter spp., Listera spp.and Francisella tularensis as an example.Some information is consistently recorded in different data collections per pathogen while other information differs.Overall, the average number of records per pathogen and year varied between the different data collections (Table 2).The number of records depends on the sector and the sampling strategy.
An overview of the documented pathogens can be found in Supplement S3.

Discussion
The huge impact of zoonotic diseases requires an integrated One Health MOSS to map the entire infection process [20].Within Germany, the federal states are often responsible for the implementation and maintaining of human or animal health monitoring programmes.In more or less every monitoring program in Germany, the data are collected for a specific purpose, whereby an actual One Health integration of the data is currently not taken into account.However, the integrated use of data from different data sources, beyond the intended purpose, can add significant value to the prevention and management of zoonotic diseases [12].The added value is multiple and aims to identify associations between pathogen records in different sources, improve awareness of health problems, detect events early and ultimately save resources.

Strategies for a One Health MOSS and Inventory of the Existing Routine Data Collections
In general, two strategies to set up a One Health MOSS can be discussed: A newly constructed monitoring system based on a representative sampling frame (active surveillance), or a passive and secondary system that uses information from different existing collections.To our knowledge, an MOSS, whose primary task is the monitoring and surveillance in the One Health context does not currently exist.Existing One Health surveillance activities are not based on sharing original or secondary data but rely on more or less standardised, regular information exchange (e.g., meetings) between sectors [21].
Secondary data use is the use of data that were originally collected for another purpose, usually based on a specific (legal) foundation, and are used secondarily for another analysis.As a result, depending on the purpose of secondary data use, not all information that would be useful for the particular secondary analysis is usually available, and the quality of the data may be limited.On the other hand, the use of secondary data has the great advantage of reducing the additional effort for data collection.Other advantages are moderate costs, the usually large reference population and the often long time series of data available [22].
But it has to be taken into account that using secondary data can cause additional effort due to data protection and during data management, and the advantages and disadvantages must be considered.
Until now, an operating and integrated One Health system is far from being realised in Germany and the use of secondary data from existing data collections therefore currently seems to be the only option.However, in order to develop a One Health surveillance approach and strategies for integrated analyses, the existing data sources and their contents must first be known, comprehensively described, and finally be assessed for their applicability to a specific One Health use case.
In order to investigate the possibilities of a continuous integrated analysis strategy of data from human and veterinary authorities in Germany, the data collections available in Lower Saxony, as a pilot region within Germany, were described comprehensively for the first time, and the corresponding metadata were systematically documented.This metadata inventory can be recognized as a "One Health Data Hub", i.e., a systematic documentation of the available data (Tables 1 and 2, Supplement S2 and S3).Our metadata inventory generally is in line with recent "data-sharing activities" in Germany, like the National Research Data Infrastructure for Personal Health Data (NFDI4Health) for the documentation of general human health data [23,24], the "Datenintegrationszentren (DIZ)" of the "Medizininformatik-Initiative" for the usage of data from different systems and different sites or the "Forschungsdatenportal (FDPG)" of the "Medizininformatik-Initiative", which describes available research data from the sector of German health care [25].
Secondary usage of routine data is advisable and reasonable [26,27].The added value of our metadata inventory is that it can be used as basis for further One Health activities.Based on the metadata description, overlaps as well as limitations (see below) can be identified and characterised.The extensive collection of metadata down to variable levels facilitates the generation of ideas and the planning of integrated data analyses.With this transparent description of the metadata of existing data collections, it is possible to support One Health initiatives or developments [7].As it is often difficult to identify existing data collections, and detailed descriptions of metadata are lacking [17], the inventory presented here can be a basis for further work.As the knowledge of data collections in other sectors is sparse [17], the inventory can also help to identify relevant data collections in other sectors.
As our One Health inventory covers a variety of populations, the potential for linking data is quite diverse.Depending on the population or sector of "human", "animal/food/feed" and "environment", the most detailed statistical units are a "human", an "animal", a "farm", a "food sample" or an "environmental sample site".These different statistical units alone complicate a joint data analysis.In addition, for reasons of data privacy, the data are often processed for use as secondary data, e.g., through aggregation and pseudonymization.Therefore, linking the data is often only possible on an aggregated spatial level such as municipalities, counties or similar.
To decide which data collections are useful in a One Health approach, the application in specific use cases is essential.A use case can be defined as "a specific scenario that sets requirements for a concept and thus serves as the basis for its development" [28].Therefore, the pathogen of interest is an essential aspect to choose relevant data collections.For this reason, our inventory provides the possibility to search for pathogens and to find out in which data collections entries on these pathogens exist.Overlaps and differences between data collections can be assessed and a general framework for integrated data analysis based on a specific use case can be developed.Restrictions for conducting a use case can be the (low) number of detections or technical and data privacy restrictions.For example, Campylobacter spp. is the most frequently reported bacterium causing gastroenteritis in humans and is therefore an important pathogen to investigate.However, it does not cause infection in animals, like broilers.Thus, the number of diagnostic samples taken in the veterinary monitoring programs is low compared to the number of reported human cases.This is one of the reasons for the different number of records.Other reasons are the different underlying legal bases and sampling strategies.Also due to the processing and distribution of animals and food, pathogen findings in food production that are spatially correlated to human cases must be interpreted with caution.
The metadata inventory generated also documents information, e.g., on access to the original data, export format and processing of the original data.Thus, it provides necessary information for the development of secondary data analyses.Moreover, this framework may initiate further developments on existing data structures like frequency of data updates, harmonising of data fields and catalogues and others (Table 2).Based on the metadata collected in the inventory, use cases can be developed but also requirements for changes to the data collections can be identified in order to improve the usability for cross-sector analyses.
The many different legal bases of data collections as well as the different responsible data holders (see Table 1, Supplement S2 for a more detailed insight into the metadata), can hinder joint analyses across data collections, even when a specific use case has been developed.To carry out a legal data assessment for secondary data use, considering the data protection law, the exact purpose of data use must be defined each time original data are requested.This procedure is at odds with a rapid response to an outbreak, as it requires time and human resources.To develop a One Health MOSS in the future, solutions need to be created by the legislative body [29].
Diverse institutions and persons, such as competent authorities and administrators are involved in the cross-sectoral One Health concept.So far, the exchange of information between the sectors in Lower Saxony mainly takes place in the event of an outbreak in accordance with certain guidelines and administrative directives, but to some extent informally.The intersectoral exchange of information in general and of data can increase data quality and trust, improve communication and build collaborations [7,21,28].The "post-hoc-standardisation" of different data sources is not a technical problem.However, interfaces must be defined and inter-operability needs expertise, on both the good knowledge of the data and their origin and the technical implementation [30].These issues, and especially the personal capacities, need to be addressed in the development of a One Health MOSS as well.Nevertheless, to take advantage of the added value of integrated analysis (One Health), more standardised information exchange is needed.

International One Health MOSS
In general, developments in the area of One Health MOSS are difficult to track as they are rarely published or shared in detail.However, the challenges described are not only a German problem.Comparing the documented data collections with available data collections from other countries shows similar limitations [21,31,32].In addition, information on officially established One Health surveillance systems is rarely published [33].Furthermore, the uniqueness of the systems makes it difficult to compare them, even if they were developed for a similar purpose [7,20].
Ellis-Iversen et al. [21] present a selection of smaller One Health surveillance initiatives from different countries such as France, Finland, Switzerland, Denmark, the Netherlands and Sweden.The authors collected the information through questionnaires and interviews.They also describe that the responsible data holders often have separate databases on separate servers.This can lead to an increased effort to implement One Health surveillance.In addition, One Health surveillance initiatives face the problem that new developments in data privacy have to be adopted, which can have negative consequences for existing standard operating and information exchange procedures.
In the Dutch Signalling Forum Zoonoses, only signals of zoonotic events that pose a potential threat to the public are communicated, either at its monthly meeting or in urgent case at an ad hoc meeting.This initiative brings together stakeholders from the relevant sectors (human, veterinary, food and wildlife) [34].
International One Health initiatives deal with specific issues but do not have a comparable method.Nevertheless, they contain important information, for example on carriers of diseases like food or parasites.

Considerations for Cross-Sector One Health Data Curation
To enable data from different collections to be used more easily in the future for integrated data use beyond their original purpose, this data integration should already be taken into account when collecting the data.Based on our metadata inventory, we identified a list of issues that need to be considered for cross-sector One Health data curation and to achieve an integrated MOSS in the future, namely: 1.
Data-related a.
Data structure and level of aggregation: i.e., the level of spatial and temporal aggregation has to be harmonised; b.
Nomenclature and formats: i.e., catalogues and definitions have to be harmonised with easy to follow rules; c.
Completeness and data quality: i.e., according to the use cases addressed, variables under study have to be selected, the number of missing values has to be calculated and data quality rules have to be addressed; d.
Level of differentiation of pathogens: i.e., the available/needed level, e.g., genus, species, subspecies, genes, needs to be considered/harmonised.
Purpose of data collection: e.g., to improve knowledge and monitoring purposes, food safety; b.Type of data collection sampling: e.g., monitoring, surveillance, active or passive systems.

3.
Data privacy-related a.
Re-Identification: i.e., individual plants or people must not be identifiable.Therefore, access to the highest possible spatial resolution may be denied.b.
Purpose of data use: i.e., for legal data protection assessments, the purpose of data use must be precisely defined.

4.
Technical-related a. Implementation of data exchange: e.g., via individual exports, data-interfaces or a data warehouse.Most data collections are stand-alone solutions, which were developed for a specific purpose and have grown historically.This makes it difficult to use these systems for other purposes.b.
Development of analysis procedures: e.g., control charts/Shewart charts, time series analysis and expected values derived therefrom.
Experts for the original data: i.e., due to the variety of data and types of documentation, experts in each data collection are needed to explain and interpret the data.b.
Statistical-, data management-, and IT-experts: i.e., to develop analysis procedures and to implement data transfer and management.c.
Personal contacts between the sectors and other stakeholders: i.e., intersectoral exchange is essential to build a One Heath MOSS, e.g., via regular meetings or joint workshops.

Conclusions and Outlook
Common zoonoses pose a risk to human and animal health and are therefore the focus of monitoring and surveillance of public authorities.However, as zoonoses occur with different frequencies in human and animal populations, and also with different severity (from ubiquitous and asymptomatic to fatal), the actions of the competent authority will differ depending on the reference population.Original data from different sectors cannot be linked at the individual level because they relate to different populations and levels of production.In addition, legal and data privacy issues hinder immediate linkage between data.As a result, data from the various sectors have not yet been combined directly.Such data combination is made even more difficult by the lack of harmonisation between MOSS [21,35].A general solution is hard to create, as the different purposes of the data collections, the different underlying legal acts and the different competent authorities hinder a common approach.In addition, the primary purpose of data collection must remain assured.Therefore, connecting different data sources and joint analyses based on defined use cases is currently the only way to answer specific One Health questions.
At the moment, spatial and temporal analyses of aggregated data can be realised.Another way to use the existing routine data is to develop analysis procedures that generate signals to detect unexpected events (higher number of findings than usual).These signals could then be shared and discussed across sectors.Routine data have to be available in sufficient detail for authorised institutions.For One Health analyses additional information like food production, consumption or supply chains could be considered [36].
Overall, communication, trust and exchange between sectors need to be improved in the future to enable collaboration and joint analyses in the One Health context [17,33].These goals can be achieved through further and advanced training as well as collaborative projects [21,35].
We found that, several aspects have to be considered for cross-sectoral data analyses: First, harmonisation of existing data and integration of additional information seems to be crucial.Here, especially data from the environment sector should be complemented by data which are non-routine data, such as from temporary projects or off-schedule investigations.
July 2000 (BGBl.I p.1045) as amended on 17 July 2023 (BGBl.2023 I No. 190), the Directive 2003/99/EC of the European Parliament and of the Council of 17 November 2003 on the monitoring of zoonoses and zoonotic agents, amending Council Decision 90/424/EEC and repealing Council Directive 92/117/EEC

Figure 1 .
Figure 1.Overview of the number of recorded data collections as well as the corresponding sector and population of interest in Lower Saxony.

Figure 1 .
Figure 1.Overview of the number of recorded data collections as well as the corresponding sector and population of interest in Lower Saxony.

Figure 2 .
Figure 2. Purposes of data collection in the respective data collections of the different sectors in Lower Saxony (multiple selections were possible).

Figure 2 .
Figure 2. Purposes of data collection in the respective data collections of the different sectors in Lower Saxony (multiple selections were possible).Concerning the sampling strategy, 5 out of 37 data collections (13.5%) are based on passive surveillance, 20 (54.1%) are based on active surveillance and 12 (32.4%)are surveys.Most of the veterinary data collections (52.6%, n = 10) gather their data via active surveillance, five (26.3%) via a survey and four (21.1%) via passive surveillance.For the human sector, eight (50%) out of sixteen data collections collect the data via active surveillance, seven (43.8%) via a survey and only one (6.2%)collects data via passive surveillance.All data collections that belong to other sectors collect their data in active surveillance.The number of records in the data collections per year varied from low numbers, e.g., the nationwide system for collecting data on food involved in foodborne outbreaks with less than 50 records per year, to data sources with more than 1,000,000 records per year (LIMS LAVES, Trade Control and Expert System New Technology-TRACES NT and the diagnoses in hospitals).Data can be extracted from the documented original data collections in a variety of export formats, including, Excel (n = 22), CSV (n = 19) and other formats (n = 12, often XML).However, six data collections only offer an export in PDF-format.Exporting data in two or more different data formats is possible for 25 data collections.For two data collections, information on the export format was not available.Access to and export of original data is regulated in very different ways.Data directly available through open access are sparse.At the State Office for Statistics in Lower Saxony, access to data is generally possible online but sometimes with restrictions due to confidentiality.Data from the Climate Data Center from the Germany's National Meteorological Service are publicly available.Results (aggregated data) from data collections in the human sector are usually publicly available (n = 14, 87.5%), while anonymised original data are available on request (n = 2; 12.5%).In the veterinary sector, access to data is more limited.Only three out of nineteen

Figure 3 .
Figure 3.The top ten pathogens recorded in the gathered data collections.Figure 3. The top ten pathogens recorded in the gathered data collections.

Figure 3 .
Figure 3.The top ten pathogens recorded in the gathered data collections.Figure 3. The top ten pathogens recorded in the gathered data collections.

Table 1 .
Overview of One Health data collections on zoonotic pathogens examined for Lower Saxony.

Table 1 .
Cont. more detailed version of Table1can be found in Supplement S2.Selected variables have been made available to the public as an R-Shiny app.The Shiny app can be accessed at https://shiny.tiho-hannover.de/connectohd/(accessed on 14 December 2023). A

Table 2 .
Selected information for Campylobacter spp., Listeria spp.and Francisella tularensis.Veterinary information is based on data provided for Connect OHD.
* Average number from 2017 to 2019.