07/03/2017 - Press release
The efficacy of animal disease surveillance systems depends primarily on the quality of the data gathered in the field, yet the factors that influence their quality are seldom assessed. Two studies in Vietnam and Germany by researchers from CIRAD and their partners have revealed that it is vital to take account of socioeconomic and cultural parameters to limit the bias linked to under-declaration of diseases. This innovative economic evaluation work could serve as a basis for the authorities to improve surveillance systems, and the integrated approach taken could rapidly be transposed and adapted to current crises (avian influenza) or potential threats (African swine fever, lumpy skin disease).
The earlier an animal disease is detected, the more effective the health authorities' response and the more limited economic and human losses will be. Public authorities are therefore keen to introduce effective epidemiological analysis systems. However, at present, the surveillance strategies that supply data for such analyses are rarely assessed, much less on a scientific basis. Assessing the efficacy of a strategy, and its cost-effectiveness, based on scientific evidence is a novel, vitally important approach, at a time when financial constraints are increasing in both North and South.
"The question of the reliability of surveillance data supplied by field players (livestock farmers, veterinary surgeons, etc) is crucial" , explains Marisa Peyre, a researcher at CIRAD. "The trouble is that surveillance strategies generally do not take account of local constraints and socio-cultural factors." Yet they are often crucial to the success of those strategies. To what extent? To find out, researchers adopted an innovative economic assessment approach, using social science and experimental economics tools.
In Vietnam, researchers from CIRAD, CIRRD* and FARAH** used the assessment method to identify the socioeconomic factors that push pig farmers to report (or not to report) diseases to the authorities (1) . "Policy-makers generally consider that a fair level of compensation is enough to ensure farmers report cases of sick animals, but in practice, this is not so: they often prefer to sell their animals at a loss, through parallel channels, rather than going through official channels", Marisa Peyre explains. This attitude can partly be put down to uncertainty about whether compensation will really be paid, but that is not the only reason. In the event of a report, the relevant authorities, or the army in the event of an acute crisis, generally cull entire herds. "And such culls have a cost that is not covered by the promised compensation: loss of breeding animals obtained through a long selection process, stress for the farmer, confusion about apparently healthy animals being culled, etc." This is compounded by other problems: the vets in charge of reporting are part of the community and come under social pressure, livestock farmers no longer have confidence in the public authorities, and so on. In the end, these socioeconomic factors mean that what looks like the most effective policy for containing epidemics turns out to be counter-productive.
This situation is not restricted to southern countries, as shown by a second study in Germany in which Marisa Peyre was also involved (2) . This time, the researchers assessed the acceptability of different surveillance strategies for classical swine fever (CSF) in wild boar populations. They used an innovative participatory method developed by CIRAD (3) . CSF, which is a huge threat to European pig farms, can pass from wild pig populations to domestic populations. Each year, hunters therefore voluntarily supply the health authorities with a certain number of dead animals for testing for the disease. The scientists identified the best three reporting strategies in terms of cost effectiveness, out of a possible 69. However, after asking hunters about their behaviour, it transpired that two of them were impossible to apply in practice. They were too restrictive, and would not have been applied, resulting in under-reporting of CSF. This work demonstrated the importance of considering acceptability when building surveillance systems. "No matter how hard you try to improve your surveillance strategy, by changing the number and type of samples taken and when, if your measures are not accepted, they will be ineffective", Marisa Peyre concludes. "That seems obvious, but it is still difficult to get the message across to the authorities, or even to the scientific and technical community! E pidemiologists and surveillance network leaders still too often focus on biological and medical aspects, whereas it would be a good idea to make greater use of the human and social sciences in this field, in both North and South (4) ."
This integrated approach will be of use for surveillance of other health situations in the South. It could also rapidly be transposed to disease surveillance in the North, whether for current crises, such as the avian influenza epidemic, or potential threats, such as African swine fever or lumpy skin disease. These highly contagious animal diseases, which originated in sub-Saharan Africa, have reached Eastern Europe in recent years and are major risks for the livestock farming sector.
* Center for Interdisciplinary Research for Rural Development - Vietnam National University of Agriculture
** Fundamental and Applied Research for Animal & Health, Faculty of Veterinary Médicine, University of Liège
(1) Pham H.T.T et al
(2017), Preventive Veterinary Medicine
DOI : 10.1016/j.prevetmed.2017.01.002
(2) Schulz K. et al.
(2017), Nature Scientific Reports
DOI : 10.1038/srep43871
(3) Calba C. et al.
(2015), Preventive Veterinary Medicine
DOI : 10.1016/j.prevetmed.2015.10.001
(4) Goutard, F.L. et al.
(2015), Preventive veterinary medicine
DOI : 10.1016/j.prevetmed.2015.02.014