4. Discussion
To the best of our knowledge, this is the first study, both in human and veterinary medicine, that investigates the dynamics of vaccination compliance and the factors which influence its deterioration after an epidemic. The results of this study show that vaccination compliance diminishes as time elapses from an epidemic. The main factor influencing compliance deterioration is the reduction of the farmers’ perceived social pressures to vaccinate. Together with availability of manpower for vaccination and price considerations, these are the major causes of reduction in vaccination compliance. These findings open an opportunity for interventions to mitigate the compliance deterioration of vaccination for epidemic diseases.
Vaccination compliance is a well-studied issue in human medicine (Betsch et al., 2018 & Brewer et al., 2017) and recently is being studied also in veterinary medicine (Elbres et al., 2010; Eschle et al., 2020 & Gehrig et al., 2019). Like other studies in which the TPB was used to explain and predict behaviour, we found that the three main constructs - attitude, subjective norms and PBC- all predicted the intention to vaccinate (Agarwal, 2014 & Schmid et al., 2017).
Both in human medicine studies (with a particular focus on Influenza vaccination) and veterinary medicine studies (e.g. vaccination against Bluetongue disease in cattle), the intention and/or behaviour to vaccinate are significantly associated with the attitude of the decision maker towards vaccination (Schmid et al., 2017 & Sok et al., 2016). Similar to the current study, these studies show a significant association of social pressures with intention. The most important normative referent is the physician/veterinarian. At the level of attitudinal beliefs, there are also similarities. In the study of Sok et al. (2015) who surveyed Dutch farmers, it was indicated that the most influential attitudinal beliefs relate to being insured both economically and psychologically from the disease consequences. In our study, the most influential beliefs related to the vaccine economical benefit, reducing the severity of clinical signs, receiving money from the insurance company, having peace of mind and reducing the number of disease cases. However, as opposed to Sok et el. we haven’t found a significant association of intention to vaccinate with the relative risk attitude, or the yearly milk quota (another measure of the size of the herd) (Sok et al., 2018). This might stem from the low variance among the Israeli dairy farmers and herds which did not enable enough power for such a comparison.
The main novel contribution of the current study is to the understanding of vaccination compliance deterioration as time elapses from an epidemic. Given the consequences of LSD epidemics, one would expect that previous experiences with the disease should have been associated with higher compliance. Surprisingly, we did not find such direct association. However we did find lower vaccination compliance in the Centre of Israel, which was the least affected area during previous LSD epidemics. This may indicate again on the influence of the existing norms in the area on intention to vaccinate. When comparing of farmers’ direct and indirect measurements of attitude, subjective norms and PBC before and after the epidemic of 2019 we found both higher perceived social pressure to vaccinate and higher intention to vaccinate, after the 2019 epidemic compared to before. This result should however interpreted cautiously, as the situation right after the epidemic in 2019 is not the exact situation in 2016. In 2016, vaccination became voluntary three years after the last epidemic. We also lack the exact picture of the actual reduction in perceived social pressures to vaccinate after the epidemic.
According to the TPB, the PBC construct moderates the interaction between the intention and the behaviour (Armitage et al., 2010). This is supported by our results, as we found that among farmers who intended to vaccinate, PBC was the main differentiating factor between farmers who vaccinated and those who eventually didn’t vaccinate. The 50% reduction that we found from intention to behaviour is indeed within the range described before for different behaviours (Sheeran, 2002) but high compared to human vaccination studies (Brewer et al., 2011).
Another important finding in our study is that past vaccination is a reliable and strong predictor of future vaccination. The multivariable analysis on the data of the 566 dairy farms insured by the ‘Hachaklait’ organization indicates that farmers who vaccinated their herd both in 2017 and 2018 were about 120 times more likely to vaccinate also in 2019 than a farmer with no history of vaccination. This finding was independent of any other demographical characteristics. However, this past behaviour itself can be the result of certain attitudinal, normative and control beliefs (Brewer et al., 2017). This finding implies that most of the effort to increase vaccine compliance should be focused on farmers who withdrew from vaccination or never vaccinated their herd against the disease.
The current study suffers from several limitations. One limitation is the small sample size of farmers who answered the TPB questionnaire. These questionnaires were part of a larger study in which the behaviour of farmers was studied to four other livestock diseases. Ideally, each farmer would have been surveyed on all five diseases. However, since that would have made the questionnaire exceptionally long, we preferred to ask each farmer randomly about one disease. Another possible limitation is that the study does not follow the same persons over time and thus does not show directly the personal change of farmers’ beliefs. Rather we interviewed different farmers at different times and compared the answers. It should be noted, however, that interviewing the same persons at different times may create response biases due to the gained experience by the farmers who repeatedly answer the same questionnaires (Wetzel et al., 2016). It would be, however, interesting to perform similar studies using a repeated questioning methodology as well and to compare the results with the results of the current study.