Analysing WAHIS Animal Health Immediate Notifications to Understand Global Reporting Trends and Measure Early Warning Capacities (2005–2021)

,


Introduction
Te increased mobility of animals, their products, and people poses a risk for the transmission of pathogens [1].Early warning systems aim to detect the incursion of a disease at a relatively early point in time in an attempt to reduce the consequences of the outbreak [2].Tis early detection of diseases is essential to successfully and timely manage disease spread at both national and international levels.
Te World Organization for Animal Health (WOAH, founded as OIE) was established in 1924 as the reference organization to coordinate and lead the animal health in the world [3] after a huge epidemic of rinderpest that afected Europe and the Americas.One of the WOAH's main objectives is the dissemination of ofcial information related to the presence of animal diseases in countries.It has played a crucial role in timely and transparent data sharing on the global animal health situation, allowing countries at risk to take appropriate action to prevent the spread of transboundary animal diseases (TADs) [4].Members are committed to submitting notifcation reports to WOAH related to the diseases listed by the organization and emerging diseases [5].Other international organizations, such as the Food and Agriculture Organization of the United Nations (FAO), through its Global Animal Disease Information System (EMPRES-i), or regional networks such as the European Commission through the Animal Disease Notifcation System (ADIS), also publish animal health alerts.
Before 2005, WOAH published early warning reports on a dedicated web page.In 2005, it launched its World Animal Health Information System (WAHIS), which had an early warning component [6].In 2021, a modernized WAHIS platform [7] was launched, containing all the disease information since 2005, with a technologically advanced yet user-friendly interface, plus new visualization, mapping, and data mining functions.
Since 2005, the disease notifcation system has consisted of two parts as defned in Chapter 1.1.of WOAH Terrestrial Animal Health Code [5]: (1) the early warning system, which includes immediate notifcations to notify alert messages of exceptional epidemiological events, as well as follow-up reports; and (2) the monitoring system comprising sixmonthly reports describing the situation of each listed disease in each country and annual reports providing contextual information on national animal populations; zoonotic cases in humans; and resources of animal health national authorities in terms of staf, diagnoses, and vaccine production.WOAH's list of diseases is approved each year by WOAH's General Assembly, based on the criteria identifed in the Aquatic and Terrestrial Animal Health Codes, including their international spread, impact on animals and humans, and reliable means of diagnosis [5].Te vast majority of reports are timely and transparently submitted by members to WOAH spontaneously.In complement, WOAH conducts an epidemic intelligence activity to track non-ofcial information from a variety of sources, ask WOAH members for confrmation of additional relevant events, and request the corresponding ofcial reporting [8][9][10].FAO, WOAH, and the World Health Organization (WHO), through the Global Early Warning System for Animal Diseases including major Zoonoses (GLEWS+), share information and use their organizational systems to detect threats and verify information via their respective networks [11].Trough these established platforms, transparency is promoted to prevent the spread of cross-border animal diseases, strengthen countries' networks, and provide continuous interaction and cooperation, which can contribute to efectively controlling disease outbreaks.
Based on the notifcation procedure, the reporting requirement of immediate notifcations should follow Article 1.1.3. of the WOAH Terrestrial Animal Health Code [5].Te exceptional epidemiological events related to listed diseases that must be notifed through the WAHIS system within 24 hours after confrmation are the following: (a) the frst occurrence of a listed disease in a country, zone, or compartment; (b) the recurrence of an eradicated listed disease in a country, zone, or compartment following the fnal report that declared the event ended; (c) the frst occurrence of a new strain of a pathogenic agent of a listed disease in a country, zone, or compartment; (d) the recurrence of an eradicated strain of a pathogenic agent of a listed disease in a country, zone, or compartment following the fnal report that declared the event ended; (e) a sudden and unexpected change in the distribution or increase in incidence or virulence of, or morbidity or mortality caused by, the pathogenic agent of a listed disease present within a country, zone, or compartment; and (f ) the occurrence of a listed disease in an unusual host species.Emerging diseases, as defned by WOAH, should also be reported by WOAH members through the early warning system.Te information is mainly submitted electronically through WAHIS where data are organized according to the concepts of events, reports, and outbreaks.
To enable smooth reporting through the online system, national focal points for disease notifcation receive specifc training on the use of this interface [12,13].After WOAH receives a report, the information is verifed, processed, translated, verifed, and publicly displayed in the WAHIS public interface.All WOAH members are also informed in real time of immediate notifcations published by the Organization, via emails.As ofcial international warnings, these immediate notifcations will allow other countries, trade partners, and other relevant stakeholders to put preventive and control measures in place (e.g., trade restrictions, disease surveillance, and others to protect human and animal health).
An analysis of the alerts for terrestrial animals submitted by countries and published by WOAH over the past few years can provide information on their patterns and help to assess the early warning capacity at both national and global levels.Te aim of this study was therefore to analyse the immediate notifcation reported to WOAH for terrestrial animal diseases between January 2005 and February 2021 to describe the following during the whole period and over time: the diseases reported, the reporting countries and regions, the most common sources, and two early warning indicators-the period between the start of the event and confrmation and the period between confrmation and reporting.According to the Terrestrial Code [5], countries and territories are required to send an immediate notifcation report, followed by weekly follow-up reports for a single outbreak or a group of epidemiologically related outbreaks of the diseases that are in the subject of a notifcation.Reports of an event include information on outbreaks defned as "the occurrence of a disease in an epidemiological unit," i.e., in an apiary, a backyard, a farm, a forest, a livestock market, a natural park, a slaughterhouse, a village, or a zoo.

Materials and Methods
Te information provided by the immediate notifcation reports included the susceptible species, the number of affected animals, and spatial and temporal information frst described, followed by analysis of information on the source 2 Transboundary and Emerging Diseases of infection.For a given event, multiple sources of infection can be selected by the reporting country/territory from a drop-down list provided in WAHIS.Reporting countries/ territories may also indicate that the source is unknown.
Ten, based on the information provided by WOAH members in the immediate notifcations, it was possible to identify two main periods of time crucial for the early warning of diseases: (1) the confrmation period (CT), defned as the time interval between the start of the disease event and the date when the disease was confrmed (usually by the national reference laboratory); and (2) the notifcation period (NT), represented by the time interval between the date of confrmation of the disease and the date when the report was submitted to WOAH.Te CT refects the country's capacity to detect and diagnose a disease event.It may serve as an indicator of early detection and early warning at the national level, and its length will strongly impact the ability to timely control a disease outbreak.On the other hand, NT reveals the timeliness of notifcation of animal disease events at international level.It may be used as an early warning indicator at the global scale.Any delay in the notifcation process will afect the efectiveness of regional and global prevention and response.For this analysis, nine events were excluded due to poor data quality on event times.
We assigned the economic status of the countries in four income groups (low, lower-middle, upper-middle, and high income) according to the World Bank classifcation of countries using the gross national income (GNI) per capita from 2020 [14], and the categorisation of countries within world regions was based on UN classifcation [15].

Statistical Analysis.
We used the Kruskal-Wallis oneway analysis of variance to test if there were signifcant diferences in confrmation/notifcation times (CT and NT), between years, world regions, or income groups.We evaluated the relationship between CT and NT with Pearson's product-moment correlation coefcient performing all statistical analyses on R 3.6.1 version [16].Furthermore, we produced the map showing reports' global distribution using QGIS 3.10.3-A[17].1).

Distribution of Events by
Te most frequently reported diseases are shown in Table 1.Troughout the analysis, infection with high pathogenicity avian infuenza virus (poultry) accounted for the most signifcant number of notifcations (560, 17% of the total) with peaks with more than 60 notifcations in 2006, 2016, and 2020; African swine fever (ASF) and foot-andmouth disease (FMD) each represented 11% of the immediate notifcations.Te annual notifcation frequency of the diseases shows that each disease had a diferent trend.
During the period of analysis, the vast majority of events (81%) were reported for domestic animals only, 12% of events were reported for wild animals only, and 7% of events were reported for both domestic and wild animals.
Over the period of analysis, 72% of immediate notifcations were submitted for disease recurrence, 15% for disease frst occurrence in a zone, 6% for the detection of new strains, 4% for disease frst occurrence in a country, 3% for emerging diseases, and less than 1% for a change in disease epidemiology or detection of disease in unusual host.

Geographical Distribution of Events.
Most events were notifed by countries and territories in Europe (43%), followed by Asia (31%), Africa (15%), the Americas (11%), and Oceania (1%).Te three European countries notifying most events were Ukraine (140 events), Germany (126 events), and the Russian Federation (126 events).In Asia, China (People's Rep. of China) notifed the highest number of events (160), followed by Israel (105) and Chinese Taipei (Taiwan) (67 events).In Africa and the Americas, only South Africa and the United States of America notifed more than Transboundary and Emerging Diseases In bold is the year with more reports from each disease.† Dates in parentheses are used to identify situations where a disease was not included in the list of reportable diseases until after 2005 (date before the dash) or was removed from the list of reportable diseases before 2021 (date after the dash).‡ As of 8 February. 4 Transboundary and Emerging Diseases 50 events (62 and 67, respectively).Only 37 events were notifed by countries and territories of Oceania, 22 of which were from Australia.Te geographical distribution of the 103,615 outbreaks reported through early warning system during the period of analysis is presented in Figure 2. Out of the outbreaks reported in wildlife, 83% were reported from European countries and territories and 16% were from Asian countries and territories.Other world regions accounted for 1% only.
Notifcations came from all but 18 WOAH members (mostly very small countries); in addition, 12 non-WOAH members countries or territories have also reported some immediate notifcation.

Infectious Sources of Primary Diseases.
For 71% of the events considered in this analysis, the reporting country/ territory did not select any known source of infection.For the remaining events, the most frequent source selected was "contact with wildlife" (9%); "illegal movement of animals," "introduction of new live animals," and "vectors" (7% each); "contact with infected animals at grazing/watering points" (4%); "fomites (humans, vehicles, feed, etc.)" and "legal movement of animals" (3% each); and "airborne spread," "animals in transit," and "swill feeding" (1% each).Te sources reported for the most frequently reported diseases are shown in Table 2.

Comparison of Diferences in Notifcation Time.
Table 3 shows the confrmation period (CT, days between onset of the disease event and laboratory confrmation) and the notifcation period (NT, days between disease confrmation and reporting to WOAH) along the study period.Te median distribution of CT was 5 days, while NT was 4 days.However, there were a high number of events with longer CT and NT periods (the 95th percentiles were 52 and 56 days, respectively).Te sum of both periods had median of 11 days.Values of CT reduced over the study period (p < 0.0001); in the frst 5 years, the median time of CT was 7-8 days, dropping to 3-5 days in the last 5 years.Te value of NT was reduced from 4 to 3 days (p � 0.014).Regarding the most frequently reported diseases, CTand NTperiods for ASF became 2 days shorter (for both periods) and, for infuenza, 4 and 3 days shorter, respectively.Te six most reported diseases have shorter periods than the rest of the diseases (p < 0.0001).Moreover, CT and NT showed a positive correlation through a linear regression model, except for FMD, Newcastle disease, and anthrax (Pearson's correlation for all diseases: 0.14, p value: <0.0001; values for the most frequent diseases are given in Table S.2).
In terms of world regions (Table 4), European countries showed the shortest CT and NT.For CT, the value above the median was smaller in Europe and Asia than in the other regions.Te diferences between regions were signifcant (p < 0.0001) except for CT between Africa and the Americas    Sources of infection most frequently reported are highlighted in bold.Some reports indicated more than one source of infection; therefore, the sum of them can be higher than 100%.† Dates in parentheses are used to identify situations where a disease was not included in the list of reportable diseases until after 2005 (date before the dash) or was removed from the list of reportable diseases before 2021 (date after the dash).

Transboundary and Emerging Diseases
6 Transboundary and Emerging Diseases and NT between Europe and Oceania.Regarding the time for laboratory confrmation of the disease (CT), 50% of the European reports were completed within 3 days.In the Americas and Oceania, more than 10% of reports took more than 2 months to get the laboratory confrmation (Table S.3.1).Table S.3.2 shows that 31% of all reports have an NT of one day.Furthermore, within a week, the notifcation rate in Europe was as high as 77%, while Africa and the Americas presented the lowest values (52% and 61%, respectively).On the other hand, there is a clear relationship between the economic level of these countries and both their CT and NT (p < 0.0001, both) (Table 5).High-income countries presented shorter periods, showing faster diagnostic procedures and notifcation time.

Discussion
One of the WOAH's missions is to improve the early detection and dissemination of information on the occurrence of animal diseases for the global community.It is important to provide countries with timely warnings of new outbreaks of diseases in the world.Tey allow for improvement of awareness, surveillance, and prevention measures to reduce the global spread of diseases and their consequences for the world economy, livelihoods, animal health, and public health.
Over the years, the number of notifcations submitted to WOAH has increased considerably [4], as confrmed also with the results of this study.Tis increase may be related to a higher awareness of the need to report (i.e., whether through pressure by the international community or by WOAH's eforts to follow up and confrm rumours through the epidemic intelligence activities) that leads to increased transparency.Besides, WOAH has also devoted eforts in training veterinary services in surveillance and notifcation and establishing a strong network with focal points for disease notifcation.It may also be partly attributed to improved capacities of veterinary services and laboratories to detect and diagnose, better preparedness of veterinary services, and improved online reporting systems at the national level.Tis increase in the number of reports could also partly be attributed to the global situation of animal disease epidemics.For example, highly pathogenic avian infuenza (HPAI) spread from Asia to Europe between 2004 and 2006, when at least 74 immediate cases were reported.In 2014, a new epidemic wave started with two peaks recorded in 2016 and 2020 with 75 and 140 immediate notifcations, Results include the number of reports, the quartiles, and 95th percentile (the 5th percentile is always zero).† As of 8 February.Results include the number of reports, the quartiles, and 95th percentile (the 5th percentile is always zero).
Transboundary and Emerging Diseases respectively.Furthermore, the ASF spread across Asia affecting domestic pigs and wild boar in 2018 [18] that led to 82 immediate reports sent to WAHIS only on that year.Due to these major pandemics, HPAI and ASF were the most frequently notifed animal diseases during the study period, as previously described by Mur et al. [6].
Te same factors may explain the discrepancies between the number of events reported in diferent regions of the world.Europe submitted the highest number of reports followed by countries from Asia.Oceania, probably due to the low number of countries and their insularity, had the least number of notifcations.
Tis analysis revealed that most events (81%) were reported only for domestic animals and that almost all outbreaks in wildlife were reported by countries in Europe and Asia.Tis clearly shows the disparities in surveillance and reporting between animal health sectors as veterinary services are primarily interested in sharing information for the international movement of livestock and their products.Information on livestock and wildlife is collected and managed by diferent ministries, and in many countries, limited resources are allocated to wildlife surveillance.
Nevertheless, these results have to be interpreted only as global trends.At country and even regional level, the number of reports should not be considered as a measure of the overall animal health status, but rather the result of the country's detection capacity and transparency.For the same reason, a low number (or absence) of immediate notifcations for specifc countries does not necessarily imply absence of disease.For example, many countries declare the diseases stable in their territories, and the WOAH standards enable them to continue to report these diseases through sixmonthly reports [4].On the other hand, factors infuencing the detection capacity might include the surveillance systems in place, the size of the susceptible populations, the ratio between veterinarians and the livestock population, the prevalent production systems (e.g., intensive vs. extensive), the clinical expression of the disease, the awareness level of the diferent stakeholders and their trust in the authorities (which may also vary by disease), whether there is a compensation policy in place, or the laboratory capacity and profciency.In addition, transparency may vary between diseases, i.e., countries may choose not to report diseases with higher trade consequences or those not notifable at national level.
Te median overall CT was 5 days and tended to decrease over time.Tis decrease shows an improvement in countries' early detection capabilities.Te median global NT was 4 days and similarly tended to decrease over time.Tis decrease shows an improvement in countries' transparency eforts.However, WOAH Standards (Aquatic and Terrestrial Animal Health Codes) require members to submit an immediate notifcation of an exceptional event within 24 hours of confrming the event; the results of this analysis show that this standard of excellence has not yet been achieved.Timely notifcation of disease outbreaks through WAHIS enables countries and stakeholders to take action to prevent further disease spread.It also includes the facilitation of safe trade in animals and their products.In this study, NT was within one day for only 31% of reports (38% in Oceania, 36% in Europe, and 32% in Asia).
High-income countries had relatively short CT and NT times.In comparison, low-income countries had a delay in the notifcation of disease outbreaks.Tis result could be related to a lack of good management, resourcing, and technical support, which infuence disease control strategies [19], and transparency is also likely correlated with the countries' level of income.Te low correlation observed between CT and NT indicates that, despite being related, the variables that afect both periods are substantially diferent.
It is uncertain why some individual events presented a long CT or/and NT period, even more than 60 days.It could be attributed to human fling errors or unawareness or a lack of attention to diseases not considered exceptional.It could also be related to data collection challenges in the feld if they are far from the medical/inspection centre [20].Te difculty to achieve a diagnostic in thedue time can be explained in some case by the shortage of resourcesdevoted by the countries to animal health.In Oceania, the very long CTs in several cases are related to diseases reported after accidental detection of new serotypes or diseases in the frame of some research studies (some of them analysed several months after samples were taken) and the lower number of notifcations from the region.CTs can also be very long for new emerging diseases, as surveillance protocols and diagnostic tools are often lacking.
Knowing the potential source of infection and exposure route can help prevent the spread of TADs [21].Tis analysis showed that reporting countries were only able to provide source information for 29% of events.Te source of an event is frequently identifed during trace-back activities.Tese activities are often not performed in the early stages of event detection but later.Tis may partially explain why little information is reported to WOAH through immediate notifcations and follow-up reports, which are primarily aimed at early warning.In addition, obtaining this Results include the number of reports, the quartiles, and 95th percentile (the 5th percentile is always zero). 8 Transboundary and Emerging Diseases information depends on country capacity, and reporting is infuenced by heterogeneities in transparency among reporting members.Because of the globalization of the market, international trade has presented a huge increase during the years of study (live pigs, poultry, and meat have increased by more than 80% and live cattle by 40%) [22].Terefore, the introduction of legal or illegal animals becomes a strong potential source of infection for each disease, and it is likely that the information ofcially provided by WOAH members on the role of animal movements in the spread of disease through WAHIS early warning system is underestimated.For instance, the source of FMD in Africa was reported as the illegal movement of animals [23].Wild animals accounted for a large proportion of the source of infuenza and rabies, since they act as an important reservoir of the virus.Only diseases from terrestrial animals have been analysed, and similar studies concerning diseases of aquatic animals would be useful to better understand the importance and efciency of the immediate notifcation system.Another limitation of the study is that it does not analyse the extension, the duration, and importance of the diferent events, nor does it look into specifc diseases.
In conclusion, there is a change in the diseases most reported over the period, with an increase in the cases of HPAI and ASF in the last four years.Data collected in this study showed a clear increase over time in terms of the number of disease notifcations and a concomitant decrease in the submitting time (both CT and NT).Most of the notifed reports were from Europe and Asia, and in general, shorter notifcation times have been observed in countries with higher incomes.Tere is still room for improvement by further reducing these periods, applying more resources to the veterinary services, especially in low-income countries.Moreover, these results show an improvement of the transparency indicators at the global level, and they could be useful to review the reporting periods included in Chapter 1.1.of WOAH Terrestrial Animal Health Code.

Figure 1 :
Figure 1: Number of immediate notifcation reports by submission year (between 1 January 2005 and 8 February 2021).
Disease, Year, Animal Category, and Reason for Notifcation.All the diseases are summarized in supplementary information (Table S.1).Te annual number of notifcation reports increased between 2005 (89 notifcations submitted) and 2009 (180 notifcations) and then gradually fell to 138 in 2013, before gradually rising to 313 in 2018.Tis number decreased slightly in 2019 (290 submitted notifcations) and reached a new peak in 2020 with 358 notifcations.As of 8 February, the number of notifcations in 2021 was already 77 (Figure

Table 2 :
Sources of infection for the most reported disease events between 1 January 2005 and 8 February 2021.

Table 3 :
Annual records of the confrmation period (CT), notifcation period (NT), and total time (CT + NT) in days.

Table 4 :
Five regions' records of CT and NT in days.

Table 5 :
CT and NT records in diferent economies.