- Email: [email protected]
A longitudinal study on transmission of Staphylococcus aureus genotype B in Swiss communal dairy herds
Preventive Veterinary Medicine 136 (2017) 65–68
Contents lists available at ScienceDirect
Preventive Veterinary Medicine journal homepage: www.elsevier.com/locate/prevetmed
A longitudinal study on transmission of Staphylococcus aureus genotype B in Swiss communal dairy herds Bart H.P. van den Borne a,∗ , Hans U. Graber b , Verena Voelk c , Carlotta Sartori b , Adrian Steiner c , M. Christina Haerdi-Landerer d , Michèle Bodmer c a
Veterinary Public Health Institute, Vetsuisse Faculty, University of Bern, Schwarzenburgstrasse 151, 3097 Liebefeld, Switzerland Agroscope, Institute for Food Sciences, Schwarzenburgstrasse 161, 3003 Bern, Switzerland c Clinic for Ruminants, Department of Clinical Veterinary Medicine, Vetsuisse Faculty, University of Bern, Bremgartenstrasse 109a, 3012 Bern, Switzerland d Institute of Agricultural Sciences (IAS), ETH Zurich, Universitaetsstrasse 2, 8092 Zurich, Switzerland b
a r t i c l e
i n f o
Article history: Received 15 August 2016 Received in revised form 4 November 2016 Accepted 15 November 2016 Keywords: Dairy cattle Mastitis Intramammary infection Staphylococcus aureus Transmission parameter Reproduction ratio
a b s t r a c t Staphylococcus aureus is a common mastitis causing pathogen of dairy cattle. Several S. aureus genotypes exist, of which genotype B (GTB) is highly prevalent in Swiss dairy herds. Dairy farming in mountainous regions of Switzerland is characterised by the movement of dairy cattle to communal pasture-based operations at higher altitudes. Cows from different herds of origin share pastures and milking equipment for a period of 2 to 3 months during summer. The aim of this longitudinal observational study was to quantify transmission of S. aureus GTB in communal dairy operations. Cows (n = 551) belonging to 7 communal operations were sampled at the beginning and end of the communal period. Transmission parameter ˇ was estimated using a Susceptible-Infectious-Susceptible (SIS) model. The basic reproduction ratio R0 was subsequently derived using previously published information about the duration of infection. Mean transmission parameter ˇ was estimated to be 0.0232 (95% CI: 0.0197–0.0274). R0 was 2.6 (95% CI: 2.2–3.0), indicating that S. aureus GTB is capable of causing major outbreaks in Swiss communal dairy operations. This study emphasized the contagious behaviour of S. aureus GTB. Mastitis management in communal dairy operations should be optimized to reduce S. aureus GTB transmission between cows and back to their herds of origin. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Staphylococcus aureus is a mastitis causing pathogen that remains prevalent in dairy herds worldwide despite decades of research and control programs. A wide spectrum of S. aureus strains has been reported to cause intramammary infections (IMI). In Switzerland, 17 genotypes were identified in a convenience sample of milk samples (Fournier et al., 2008). S. aureus genotype B (GTB) and C were most commonly identified (Fournier et al., 2008). In a follow up study, S. aureus GTB was associated with a high within-herd prevalence whereas S. aureus genotype C was associated with a low within-herd prevalence (Graber et al., 2009). S. aureus GTB is the cattle-adapted form of clonal complex 8 (Boss et al., 2016) which is a well-known clonal complex causing infections in multiple species (van den Borne et al., 2010a), including humans (Deurenberg and Stobberingh, 2008). S. aureus GTB made
∗ Corresponding author. E-mail address: [email protected] (B.H.P. van den Borne). http://dx.doi.org/10.1016/j.prevetmed.2016.11.008 0167-5877/© 2016 Elsevier B.V. All rights reserved.
the human-to-bovine jump approximately 38 years ago (Boss et al., 2016) and is now mainly detected in dairy herds located in central Europe (Cosandey et al., 2016). During the summer months dairy farming in mountainous regions of Switzerland is characterised by the commingling of cows from several different herds of origin in large communal pasture-based dairy operations. Cows are brought together on dairy operations at higher altitudes where they share pastures, milking equipment and housing facilities for a period of 2 to 3 months. At the end of the communal period, cows are transported back to their herds of origin at lower altitudes. Hereafter, these high altitude summer dairy operations are referred to as communal operations. Most cows are transported to a communal operation together on a single date. However, some cows may arrive later or leave earlier, with the latter occurring more frequently. Seasonal calving is common for dairy herds located in mountainous regions of Switzerland and most cows on communal operations are approaching the dry-off period when they return to their herds of origin. Farmers transport cows that need to be dried off or that are severely diseased cows back to their herds of origin earlier. This
66
B.H.P. van den Borne et al. / Preventive Veterinary Medicine 136 (2017) 65–68
commingling of cows from different herds of origin contributes to the spread of contagious pathogens (Presi et al., 2011; Spergser et al., 2013) but quantitative data about S. aureus GTB transmission in communal operations is lacking while this would add to the development of intervention strategies. The aim of this study was to quantify transmission of S. aureus GTB among Swiss dairy cows in communal pasture-based operations. 2. Material and methods 2.1. Data collection and description of communal operations Data were obtained during a previously described longitudinal study in which 9 communal dairy operations (labelled A to I) were selected by convenience. For a detailed description of the herd selection and sampling procedure see Voelk et al. (2014). Briefly, all dairy cows in these operations were sampled at the beginning and the end of the communal period during the summer of 2012. Clean composite milk samples were taken from all cows by trained study personnel at the first or second milking after they arrived at their communal operation. Milk sampling was performed again 3 to 7 days before transporting cows back to their herds of origin at the end of the communal period. After receiving a training, farmers were instructed to collect clean composite milk samples from cows that arrived late or left the communal operation early to ensure that all cows were sampled. Cows were milked twice daily and premilking udder preparation consisted of the cleaning of udders with paper towels or wood shavings and fore-stripping. Milking gloves were not used. All milking machines were serviced before the communal period. Cows were not vaccinated against any mastitis pathogens. Post-milking teat disinfection (PMTD) was applied in all 9 communal operations. Cows with clinical mastitis were treated with intramammary antimicrobials and remained on the communal operation but were not sampled for S. aureus GTB determination. Treated cows were milked separately using a different milking cluster. 2.2. Laboratory analysis Identification of S. aureus GTB in collected composite milk samples was performed as described previously (Voelk et al., 2014). In short, milk samples were diluted 1:10 with Chapman medium and incubated at 37 ◦ C for 18 h. After this enrichment step, the bacterial culture was processed using the mericon DNA Bacteria Plus Kit (Qiagen Instruments AG, Hombrechtikon, Switzerland) for the isolation of DNA from bacteria. Real-time PCR was performed to detect the presence of three S. aureus GTB specific genes. LukEB, sea and sed code for the GTB-typical polymorphism within the lukE gene and for the two staphylococcal enterotoxins A and D, respectively. Diagnostic sensitivity and specificity of this test are reported to be close to 100% (Boss et al., 2011; Syring et al., 2012). 2.3. Statistical analysis A Susceptible-Infectious-Susceptible (SIS) model was used to quantify cow level transmission of S. aureus GTB within communal dairy operations (Lam et al., 1996; Zadoks et al., 2002; Barlow et al., 2013; Schukken et al., 2014). Cows that tested positive for S. aureus GTB classified as being infectious (I), and S. aureus GTBnegative cows were classified as susceptible (S). Cows could not become immune and were assumed to mix randomly. Segregation of cows with deprived udder health (i.e., milking them last or with a separate milking cluster) was not practiced on communal operations (Voelk et al., 2014). Susceptible cows were all assumed to be equally susceptible and infectious cows were assumed to be equally
infectious throughout the infectious period. New S. aureus GTB IMI were assumed to have occurred halfway through the study period. Cows that left the communal operation earlier and were not sampled at the end of the summer period were not included in the analysis. In some cases, cows left the communal operation earlier (or arrived there later) and were sampled by the herdsmen, but did not have the date of sampling recorded. These cows were assumed to stay, on average, for two-third of the time the majority of cows remained on the communal operation. Dynamics of S. aureus GTB IMI can be quantified by transmission parameter ˇ which represents the number of secondary IMI resulting from one S. aureus GTB infectious cow per unit of time (i.e., day in this study). In a population of size N (where N = S + I), the number of secondary cases (C) per time interval (t) depends on the number of susceptible (S) animals at the start of the interval, the average number of infectious (I) animals per interval, and transmission parameter ˇ. Transmission parameter ˇ can then be estimated as a function of I, S, C, N, and t. Data were analysed using a generalized linear model (PROC GENMOD) in SAS 9.4 (SAS Institute Inc., Cary, NC, USA) with a complementary log–log link function, C as the number of new cases and S as the number of trials in the binomial process. The term log ((I/N) x t) was used as an offset, with I denoting the average number of infectious cows, N the total number of cows at the start of the interval, and t the mean duration of the sampling interval for each communal operation. Because the number of new cases C at the end of each interval can be estimated using 1–e−ˇ(I/N)t , the linear relationship for the statistical model equals (Velthuis et al., 2003; Schouten et al., 2009): c log log
C S
= log ˇ + log
I N
· t
Exponentiation of logˇ (i.e., the intercept of the statistical model) yields the transmission parameter ˇ. The deviance dispersion parameter was forced to one because some underdispersion was present (i.e., the deviance dispersion parameter was 0.78). Evaluation of model fit gave no reasons for concern; both the Pearson and deviance 2 test were non-significant. The reproduction ratio (R0 ) is the average number of secondary infections arising from one infected individual during its entire infectious period in a fully susceptible population (Keeling and Rohani, 2008; Vynnycky and White, 2010). R0 was estimated by multiplying ˇ with 110, the arithmetic mean duration of S. aureus IMI (in days) reported in literature (Barlow et al., 2009). Significance of log(I), when added to the model with log(t/N) being the offset (Lam et al., 1996; Zadoks et al., 2001; Schukken et al., 2014), was calculated in order to test the hypothesis that S. aureus GTB depends on the number of S. aureus GTB infected cows (and thus can be considered a contagious genotype). 3. Results Communal operations C and G were S. aureus GTB-negative at first sampling (Voelk et al., 2014) and thus excluded from transmission parameter estimation. A further 82 cows were excluded because they were not sampled when they arrived later (n = 8) or left the communal operation earlier (n = 74). The final total study population consisted of 551 dairy cows originating from 83 individual herds of origin, and transported to 7 communal operations for the summer of 2012. Of those 551 cows, 10 late arriving cows did not have their arrival date recorded and 77 cows that left early did not have their leaving date recorded. The number of susceptible (S), infectious (I) and new cases of S. aureus GTB (C) and the average interval duration per communal pasture are presented in Table 1. A total of 144 new S. aureus GTB cases were observed. Transmission parameter ˇ was estimated to be 0.0232 (95% CI: 0.0200–0.0269) given a model intercept estimate
B.H.P. van den Borne et al. / Preventive Veterinary Medicine 136 (2017) 65–68 Table 1 Number of cows according to their Staphylococcus aureus genotype B status in 7 Swiss communal operations. Data served as input for the transmission model. Communal operation
S
I
C
t
A B D E F H I
82 62 69 52 66 40 94
20 24 3 12.5 29.5 37.5 22
22 26 3 11 27 26 29
67.9 65.9 70.5 71.5 65.8 61.7 71.0
t = average number of days between samplings; S = number of susceptible cows at start of interval; I = average number of infectious cows during the interval; C = number of new cases during interval.
of −3.7632 (SE = 0.0746). This corresponded to a R0 of 2.6 (95% CI: 2.2–3.0) assuming a mean duration of IMI of 110 days (Barlow et al., 2009). Log(I) was highly significant (P < 0.0001) when added as a predictor to the model. This confirmed the hypothesis that transmission of S. aureus GTB is dependent on the number of infected cows, suggesting cow-to-cow transmission. 4. Discussion The aim of this study was to quantify transmission of S. aureus GTB in a longitudinal field study. It was estimated that one S. aureus GTB infectious cow was able to infect 2.6 other cows, in a fully susceptible population, during its entire infectious period. Since R0 was significantly above 1, we expect that S. aureus GTB would be able to cause outbreaks in communal dairy operations where mastitis management is generally poor (Keeling and Rohani, 2008; Vynnycky and White, 2010). This was the case in this study since there was a strong increase in the cow level and between-herd of origin prevalence of S. aureus GTB during the communal period as reported previously (Voelk et al., 2014). The current and previous study (Voelk et al., 2014) therefore reemphasized the contribution of communal operations to the spread of contagious pathogens among dairy herds (Presi et al., 2011; Spergser et al., 2013). Transmission of S. aureus has been previously quantified in a few studies. Barlow et al. (2009) summarized data of Lam et al. (1996) and Zadoks et al. (2002) quantifying transmission of major gram-positive bacteria (i.e., pathogens other than S. aureus were also included) in 10 herds. Mean reported transmission parameter ˇ was 0.0087 for quarters with PMTD and 0.0362 for quarters not receiving PMTD. In an intervention study, the effect of antimicrobial treatment on transmission dynamics of S. aureus in 2 herds applying PMTD was evaluated (Barlow et al., 2013). Transmission parameter ˇ was 0.0080 in one herd and 0.0045 in the other but was not affected by antimicrobial treatment. Transmission parameters differed between S. aureus strains in that study but the study did not include strains belonging to clonal complex 8, the clonal complex to which S. aureus GTB belongs (Boss et al., 2016). A recent study evaluated efficacy of a new polyvalent mastitis vaccine in 2 herds with PMTD by quantifying the reduction in S. aureus transmission as a result of the vaccine (Schukken et al., 2014). Mean transmission parameter ˇ of S. aureus among control cows not receiving the vaccine was 0.0080 in the first herd and 0.0098 in the second. Transmission of S. aureus was lower among vaccinated cows. In our study, transmission of S. aureus GTB was almost as high as the estimate of Barlow et al. (2009) for cows not receiving PMTD (0.0362), despite the fact that cows in our study were receiving PMTD. The quality of PMTD (coverage, product efficacy, etc.), the implementation of other preventive measures, the virulence diversity of various S. aureus strains, and possible differences in the susceptibility of hosts hamper a true comparison between herds
67
and studies. Finally, it should be noted that aforementioned studies determined S. aureus transmission at quarter level whereas we measured it at cow level. This is a result of a difference in sampling strategy. Our estimate should therefore be considered an underestimation because transmission between quarters within the same cow could not be estimated in our study. There were some limitations in our study. First, samples were only collected in 7 communal operations at 2 occasions. This limited the dataset to 7 intervals from which transmission parameter ˇ could be estimated. The final population size of 551 dairy cows made it possible to produce estimates with rather narrow confidence intervals nonetheless. The high precision may also have been a result of there being a similar force of infection in the 7 communal operations. It was not possible to statistically compare transmission parameters between the 7 communal operations because only 1 interval could be evaluated per communal operation. Moreover, the relatively short study period limited our ability to calculate cure rates. Cure rates could have been calculated based on the available data (some cows did indeed clear their infection; data not shown) but this would have resulted in a strong underestimation. Cows were only followed for a period of 2 to 3 months during lactation. Their dry off period, in which they have a high probability of being cured, could not be included in the study. Duration of IMI used to calculate R0 was therefore based on input from literature, which may not have been completely generalizable to the communal operations followed in the current study. The herds in the current study had an open structure, as cows did not remain on pasture for the entire study period. There were a few late arrival cows and a considerably higher number of cows that left early. Farmers were instructed to take milk samples from these cows but compliance was moderate, resulting in loss of some cows to follow up. Seventy-four of the cows sampled at the first occasion did not have a second sample collected. An additional 77 cows were sampled but had missing sampling dates, and the sampling interval for these cows had to be inferred. This inference may have biased our estimation of the mean transmission parameter in an unknown direction and with an unknown magnitude. Cows in the study that left their communal operation early had a higher prevalence of S. aureus GTB (Voelk et al., 2014). This suggest that the transmission parameter ˇ estimate (and thus R0 ) would have been higher if those cows had been included in the study. Finally, cows with clinical mastitis were sampled neither. Those cows may have experienced a new IMI and a subsequent cure during the communal period; a result of an antimicrobial treatment. We could thus have missed such short IMI resulting in an underestimation of transmission parameter ˇ. This study confirmed the hypothesis that genotype B is a contagious S. aureus genotype. Including log(I) as a predictor into the statistical model revealed a strong association with the number of new S. aureus GTB cases. Moreover, R0 was significantly greater than 1 and S. aureus GTB is therefore able to cause outbreaks as observed in the communal dairy herds participating in this study. Quantification of transmission parameter ˇ aids in designing cow level intervention strategies to reduce transmission of S. aureus GTB within communal operations and between cows’ herds of origin. Interventions to limit transmission of S. aureus GTB within communal operations include PMTD (Lam et al., 1996), vaccination (Schukken et al., 2014), and segregation of infectious cows by milking them last or with a separate milking cluster. Another possibility to reduce spread of S. aureus GTB within communal operations is to shorten duration of IMI. Given that transmission parameter ˇ was estimated to be 0.0232 (95% CI: 0.0200–0.0269), duration of IMI should be below 50 days to achieve a R0 significantly below 1. Within communal operations, this can be achieved by antimicrobial treatment (Barkema et al., 2006; van den Borne et al., 2010b) or transporting cows back to their herds of origin. Communal dairy
68
B.H.P. van den Borne et al. / Preventive Veterinary Medicine 136 (2017) 65–68
operations could additionally restrict the movement of infectious animals in order to limit between-herd of origin transmission of S. aureus GTB. Conflict of interest The authors declare no conflict of interest. Acknowledgements The authors would like to thank all farmers and personnel of the communal operations for their participation in the study and John Berezowski (Veterinary Public Health Institute, Liebefeld, Switzerland) for commenting on a previous version of the manuscript. Funding for the original study (Voelk et al., 2014) was provided by the Canton of Graubünden (Switzerland). References Barkema, H.W., Schukken, Y.H., Zadoks, R.N., 2006. Invited Review: the role of cow, pathogen, and treatment regimen in the therapeutic success of bovine Staphylococcus aureus mastitis. J. Dairy Sci. 89, 1877–1895, http://dx.doi.org/ 10.3168/jds.S0022-0302(06)72256-1. Barlow, J.W., White, L.J., Zadoks, R.N., Schukken, Y.H., 2009. A mathematical model demonstrating indirect and overall effects of lactation therapy targeting subclinical mastitis in dairy herds. Prev. Vet. Med. 90, 31–42, http://dx.doi.org/ 10.1016/j.prevetmed.2009.03.016. Barlow, J.W., Zadoks, R.N., Schukken, Y.H., 2013. Effect of lactation therapy on Staphylococcus aureus transmission dynamics in two commercial dairy herds. BMC Vet. Res. 9, 28, http://dx.doi.org/10.1186/1746-6148-9-28. Boss, R., Naskova, J., Steiner, A., Graber, H.U., 2011. Mastitis diagnostics: quantitative PCR for Staphylococcus aureus genotype B in bulk tank milk. J. Dairy Sci. 94, 128–137, http://dx.doi.org/10.3168/jds.2010-3251. Boss, R., Cosandey, A., Luini, M., Artursson, K., Bardiau, M., Breitenwieser, F., Hehenberger, E., Lam, T., Mansfeld, M., Michel, A., Mösslacher, G., Naskova, J., Nelson, S., Podpeˇcan, O., Raemy, A., Ryan, E., Salat, O., Zangerl, P., Steiner, A., Graber, H.U., 2016. Bovine Staphylococcus aureus: subtyping, evolution, and zoonotic transfer. J. Dairy Sci. 99, 515–528, http://dx.doi.org/10.3168/jds.20159589. Cosandey, A., Boss, R., Luini, M., Artursson, K., Bardiau, M., Breitenwieser, F., Hehenberger, E., Lam, T., Mansfeld, M., Michel, A., Mösslacher, G., Naskova, J., Nelson, S., Podpeˇcan, O., Raemy, A., Ryan, E., Salat, O., Zangerl, P., Steiner, A., Graber, H.U., 2016. Staphylococcus aureus genotype B and other genotypes isolated from cow milk in European countries. J. Dairy Sci. 99, 529–540, http:// dx.doi.org/10.3168/jds.2015-9587. Deurenberg, R.H., Stobberingh, E.E., 2008. The evolution of Staphylococcus aureus. Infect. Genet. Evol. 8, 747–763, http://dx.doi.org/10.1016/j.meegid.2008.07. 007. Fournier, C., Kuhnert, P., Frey, J., Miserez, R., Kirchhofer, M., Kaufmann, T., Steiner, A., Graber, H.U., 2008. Bovine Staphylococcus aureus: Association of virulence genes, genotypes and clinical outcome. Res. Vet. Sci. 85, 439–448, http://dx.doi. org/10.1016/j.rvsc.2008.01.010.
Graber, H.U., Naskova, J., Studer, E., Kaufmann, T., Kirchhofer, M., Brechbühl, M., Schaeren, W., Steiner, A., Fournier, C., 2009. Mastitis-related subtypes of bovine Staphylococcus aureus are characterized by different clinical properties. J. Dairy Sci. 92, 1442–1451, http://dx.doi.org/10.3168/jds.2008-1430. Keeling, M.J., Rohani, P., 2008. Modeling Infectious Diseases, first ed. Princeton University Press, Princeton, New Jersey, USA. Lam, T.J.G.M., De Jong, M.C.M., Schukken, Y.H., Brand, A., 1996. Mathematical modelling to estimate efficacy of postmilking teat disinfection in split-udder trials of dairy cows. J. Dairy Sci. 79, 62–70. Presi, P., Struchen, R., Knight-Jones, T., Scholl, S., Heim, D., 2011. Bovine viral diarrhea (BVD) eradication in Switzerland-experiences of the first two years. Prev. Vet. Med. 99, 112–121, http://dx.doi.org/10.1016/j.prevetmed.2011.01. 012. Schouten, J.M., Graat, E.A.M., Frankena, K., Van Zijderveld, F., De Jong, M.C.M., 2009. Transmission and quantification of verocytotoxin-producing Escherichia coli O157 in dairy cattle and calves. Epidemiol. Infect. 137, 114–123, http://dx.doi. org/10.1017/s0950268808000320. Schukken, Y.H., Bronzo, V., Locatelli, C., Pollera, C., Rota, N., Casula, A., Testa, F., Scaccabarozzi, L., March, R., Zalduendo, D., Guix, R., Moroni, P., 2014. Efficacy of vaccination on Staphylococcus aureus and coagulase-negative staphylococci intramammary infection dynamics in 2 dairy herds. J. Dairy Sci. 97, 5250–5264, http://dx.doi.org/10.3168/jds.2014-8008. Spergser, J., Macher, K., Kargl, M., Lysnyansky, I., Rosengarten, R., 2013. Emergence, re-emergence, spread and host species crossing of Mycoplasma bovis in the Austrian Alps caused by a single endemic strain. Vet. Microbiol. 164, 299–306, http://dx.doi.org/10.1016/j.vetmic.2013.02.007. Syring, C., Boss, R., Reist, M., Bodmer, M., Hummerjohann, J., Gehrig, P., Graber, H.U., 2012. Bovine mastitis: the diagnostic properties of a PCR-based assay to monitor the Staphylococcus aureus genotype B status of a herd, using bulk tank milk. J. Dairy Sci. 95, 3674–3682, http://dx.doi.org/10.3168/jds.2011-4968. van den Borne, B.H.P., Nielen, M., van Schaik, G., Melchior, M.B., Lam, T.J.G.M., Zadoks, R.N., 2010a. Host adaptation of bovine Staphylococcus aureus seems associated with bacteriological cure after lactational antimicrobial treatment. J. Dairy Sci. 93, 2550–2558, http://dx.doi.org/10.3168/jds.2009-2971. van den Borne, B.H.P., van Schaik, G., Lam, T.J.G.M., Nielen, M., 2010b. Therapeutic effects of antimicrobial treatment during lactation of recently acquired bovine subclinical mastitis: two linked randomized field trials. J. Dairy Sci. 93, 218–233, http://dx.doi.org/10.3168/jds.2009-2567. Velthuis, A.G.J., De Jong, M.C.M., Kamp, E.M., Stockhofe, N., Verheijden, J.H.M., 2003. Design and analysis of an Actinobacillus pleuropneumoniae transmission experiment. Prev. Vet. Med. 60, 53–68, http://dx.doi.org/10.1016/S01675877(03)00082-5. Voelk, V., Graber, H.U., van den Borne, B.H.P., Sartori, C., Steiner, A., Bodmer, M., Haerdi-Landerer, M.C., 2014. A longitudinal study investigating the prevalence of Staphylococcus aureus genotype B in seasonally communal dairy herds. J. Dairy Sci. 97, 4184–4192, http://dx.doi.org/10.3168/jds.2013-7291. Vynnycky, E., White, R.G., 2010. An Introduction to Infectious Disease Modelling, first ed. Oxford University Press, Oxford, New York. Zadoks, R.N., Allore, H.G., Barkema, H.W., Sampimon, O.C., Gröhn, Y.T., Schukken, Y.H., 2001. Analysis of an outbreak of Streptococcus uberis mastitis. J. Dairy Sci. 84, 590–599, http://dx.doi.org/10.3168/jds.S0022-0302(01)74512-2. Zadoks, R.N., Allore, H.G., Hagenaars, T.J., Barkema, H.W., Schukken, Y.H., 2002. A mathematical model of Staphylococcus aureus control in dairy herds. Epidemiol. Infect. 129, 397–416.
Contents lists available at ScienceDirect
Preventive Veterinary Medicine journal homepage: www.elsevier.com/locate/prevetmed
A longitudinal study on transmission of Staphylococcus aureus genotype B in Swiss communal dairy herds Bart H.P. van den Borne a,∗ , Hans U. Graber b , Verena Voelk c , Carlotta Sartori b , Adrian Steiner c , M. Christina Haerdi-Landerer d , Michèle Bodmer c a
Veterinary Public Health Institute, Vetsuisse Faculty, University of Bern, Schwarzenburgstrasse 151, 3097 Liebefeld, Switzerland Agroscope, Institute for Food Sciences, Schwarzenburgstrasse 161, 3003 Bern, Switzerland c Clinic for Ruminants, Department of Clinical Veterinary Medicine, Vetsuisse Faculty, University of Bern, Bremgartenstrasse 109a, 3012 Bern, Switzerland d Institute of Agricultural Sciences (IAS), ETH Zurich, Universitaetsstrasse 2, 8092 Zurich, Switzerland b
a r t i c l e
i n f o
Article history: Received 15 August 2016 Received in revised form 4 November 2016 Accepted 15 November 2016 Keywords: Dairy cattle Mastitis Intramammary infection Staphylococcus aureus Transmission parameter Reproduction ratio
a b s t r a c t Staphylococcus aureus is a common mastitis causing pathogen of dairy cattle. Several S. aureus genotypes exist, of which genotype B (GTB) is highly prevalent in Swiss dairy herds. Dairy farming in mountainous regions of Switzerland is characterised by the movement of dairy cattle to communal pasture-based operations at higher altitudes. Cows from different herds of origin share pastures and milking equipment for a period of 2 to 3 months during summer. The aim of this longitudinal observational study was to quantify transmission of S. aureus GTB in communal dairy operations. Cows (n = 551) belonging to 7 communal operations were sampled at the beginning and end of the communal period. Transmission parameter ˇ was estimated using a Susceptible-Infectious-Susceptible (SIS) model. The basic reproduction ratio R0 was subsequently derived using previously published information about the duration of infection. Mean transmission parameter ˇ was estimated to be 0.0232 (95% CI: 0.0197–0.0274). R0 was 2.6 (95% CI: 2.2–3.0), indicating that S. aureus GTB is capable of causing major outbreaks in Swiss communal dairy operations. This study emphasized the contagious behaviour of S. aureus GTB. Mastitis management in communal dairy operations should be optimized to reduce S. aureus GTB transmission between cows and back to their herds of origin. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Staphylococcus aureus is a mastitis causing pathogen that remains prevalent in dairy herds worldwide despite decades of research and control programs. A wide spectrum of S. aureus strains has been reported to cause intramammary infections (IMI). In Switzerland, 17 genotypes were identified in a convenience sample of milk samples (Fournier et al., 2008). S. aureus genotype B (GTB) and C were most commonly identified (Fournier et al., 2008). In a follow up study, S. aureus GTB was associated with a high within-herd prevalence whereas S. aureus genotype C was associated with a low within-herd prevalence (Graber et al., 2009). S. aureus GTB is the cattle-adapted form of clonal complex 8 (Boss et al., 2016) which is a well-known clonal complex causing infections in multiple species (van den Borne et al., 2010a), including humans (Deurenberg and Stobberingh, 2008). S. aureus GTB made
∗ Corresponding author. E-mail address: [email protected] (B.H.P. van den Borne). http://dx.doi.org/10.1016/j.prevetmed.2016.11.008 0167-5877/© 2016 Elsevier B.V. All rights reserved.
the human-to-bovine jump approximately 38 years ago (Boss et al., 2016) and is now mainly detected in dairy herds located in central Europe (Cosandey et al., 2016). During the summer months dairy farming in mountainous regions of Switzerland is characterised by the commingling of cows from several different herds of origin in large communal pasture-based dairy operations. Cows are brought together on dairy operations at higher altitudes where they share pastures, milking equipment and housing facilities for a period of 2 to 3 months. At the end of the communal period, cows are transported back to their herds of origin at lower altitudes. Hereafter, these high altitude summer dairy operations are referred to as communal operations. Most cows are transported to a communal operation together on a single date. However, some cows may arrive later or leave earlier, with the latter occurring more frequently. Seasonal calving is common for dairy herds located in mountainous regions of Switzerland and most cows on communal operations are approaching the dry-off period when they return to their herds of origin. Farmers transport cows that need to be dried off or that are severely diseased cows back to their herds of origin earlier. This
66
B.H.P. van den Borne et al. / Preventive Veterinary Medicine 136 (2017) 65–68
commingling of cows from different herds of origin contributes to the spread of contagious pathogens (Presi et al., 2011; Spergser et al., 2013) but quantitative data about S. aureus GTB transmission in communal operations is lacking while this would add to the development of intervention strategies. The aim of this study was to quantify transmission of S. aureus GTB among Swiss dairy cows in communal pasture-based operations. 2. Material and methods 2.1. Data collection and description of communal operations Data were obtained during a previously described longitudinal study in which 9 communal dairy operations (labelled A to I) were selected by convenience. For a detailed description of the herd selection and sampling procedure see Voelk et al. (2014). Briefly, all dairy cows in these operations were sampled at the beginning and the end of the communal period during the summer of 2012. Clean composite milk samples were taken from all cows by trained study personnel at the first or second milking after they arrived at their communal operation. Milk sampling was performed again 3 to 7 days before transporting cows back to their herds of origin at the end of the communal period. After receiving a training, farmers were instructed to collect clean composite milk samples from cows that arrived late or left the communal operation early to ensure that all cows were sampled. Cows were milked twice daily and premilking udder preparation consisted of the cleaning of udders with paper towels or wood shavings and fore-stripping. Milking gloves were not used. All milking machines were serviced before the communal period. Cows were not vaccinated against any mastitis pathogens. Post-milking teat disinfection (PMTD) was applied in all 9 communal operations. Cows with clinical mastitis were treated with intramammary antimicrobials and remained on the communal operation but were not sampled for S. aureus GTB determination. Treated cows were milked separately using a different milking cluster. 2.2. Laboratory analysis Identification of S. aureus GTB in collected composite milk samples was performed as described previously (Voelk et al., 2014). In short, milk samples were diluted 1:10 with Chapman medium and incubated at 37 ◦ C for 18 h. After this enrichment step, the bacterial culture was processed using the mericon DNA Bacteria Plus Kit (Qiagen Instruments AG, Hombrechtikon, Switzerland) for the isolation of DNA from bacteria. Real-time PCR was performed to detect the presence of three S. aureus GTB specific genes. LukEB, sea and sed code for the GTB-typical polymorphism within the lukE gene and for the two staphylococcal enterotoxins A and D, respectively. Diagnostic sensitivity and specificity of this test are reported to be close to 100% (Boss et al., 2011; Syring et al., 2012). 2.3. Statistical analysis A Susceptible-Infectious-Susceptible (SIS) model was used to quantify cow level transmission of S. aureus GTB within communal dairy operations (Lam et al., 1996; Zadoks et al., 2002; Barlow et al., 2013; Schukken et al., 2014). Cows that tested positive for S. aureus GTB classified as being infectious (I), and S. aureus GTBnegative cows were classified as susceptible (S). Cows could not become immune and were assumed to mix randomly. Segregation of cows with deprived udder health (i.e., milking them last or with a separate milking cluster) was not practiced on communal operations (Voelk et al., 2014). Susceptible cows were all assumed to be equally susceptible and infectious cows were assumed to be equally
infectious throughout the infectious period. New S. aureus GTB IMI were assumed to have occurred halfway through the study period. Cows that left the communal operation earlier and were not sampled at the end of the summer period were not included in the analysis. In some cases, cows left the communal operation earlier (or arrived there later) and were sampled by the herdsmen, but did not have the date of sampling recorded. These cows were assumed to stay, on average, for two-third of the time the majority of cows remained on the communal operation. Dynamics of S. aureus GTB IMI can be quantified by transmission parameter ˇ which represents the number of secondary IMI resulting from one S. aureus GTB infectious cow per unit of time (i.e., day in this study). In a population of size N (where N = S + I), the number of secondary cases (C) per time interval (t) depends on the number of susceptible (S) animals at the start of the interval, the average number of infectious (I) animals per interval, and transmission parameter ˇ. Transmission parameter ˇ can then be estimated as a function of I, S, C, N, and t. Data were analysed using a generalized linear model (PROC GENMOD) in SAS 9.4 (SAS Institute Inc., Cary, NC, USA) with a complementary log–log link function, C as the number of new cases and S as the number of trials in the binomial process. The term log ((I/N) x t) was used as an offset, with I denoting the average number of infectious cows, N the total number of cows at the start of the interval, and t the mean duration of the sampling interval for each communal operation. Because the number of new cases C at the end of each interval can be estimated using 1–e−ˇ(I/N)t , the linear relationship for the statistical model equals (Velthuis et al., 2003; Schouten et al., 2009): c log log
C S
= log ˇ + log
I N
· t
Exponentiation of logˇ (i.e., the intercept of the statistical model) yields the transmission parameter ˇ. The deviance dispersion parameter was forced to one because some underdispersion was present (i.e., the deviance dispersion parameter was 0.78). Evaluation of model fit gave no reasons for concern; both the Pearson and deviance 2 test were non-significant. The reproduction ratio (R0 ) is the average number of secondary infections arising from one infected individual during its entire infectious period in a fully susceptible population (Keeling and Rohani, 2008; Vynnycky and White, 2010). R0 was estimated by multiplying ˇ with 110, the arithmetic mean duration of S. aureus IMI (in days) reported in literature (Barlow et al., 2009). Significance of log(I), when added to the model with log(t/N) being the offset (Lam et al., 1996; Zadoks et al., 2001; Schukken et al., 2014), was calculated in order to test the hypothesis that S. aureus GTB depends on the number of S. aureus GTB infected cows (and thus can be considered a contagious genotype). 3. Results Communal operations C and G were S. aureus GTB-negative at first sampling (Voelk et al., 2014) and thus excluded from transmission parameter estimation. A further 82 cows were excluded because they were not sampled when they arrived later (n = 8) or left the communal operation earlier (n = 74). The final total study population consisted of 551 dairy cows originating from 83 individual herds of origin, and transported to 7 communal operations for the summer of 2012. Of those 551 cows, 10 late arriving cows did not have their arrival date recorded and 77 cows that left early did not have their leaving date recorded. The number of susceptible (S), infectious (I) and new cases of S. aureus GTB (C) and the average interval duration per communal pasture are presented in Table 1. A total of 144 new S. aureus GTB cases were observed. Transmission parameter ˇ was estimated to be 0.0232 (95% CI: 0.0200–0.0269) given a model intercept estimate
B.H.P. van den Borne et al. / Preventive Veterinary Medicine 136 (2017) 65–68 Table 1 Number of cows according to their Staphylococcus aureus genotype B status in 7 Swiss communal operations. Data served as input for the transmission model. Communal operation
S
I
C
t
A B D E F H I
82 62 69 52 66 40 94
20 24 3 12.5 29.5 37.5 22
22 26 3 11 27 26 29
67.9 65.9 70.5 71.5 65.8 61.7 71.0
t = average number of days between samplings; S = number of susceptible cows at start of interval; I = average number of infectious cows during the interval; C = number of new cases during interval.
of −3.7632 (SE = 0.0746). This corresponded to a R0 of 2.6 (95% CI: 2.2–3.0) assuming a mean duration of IMI of 110 days (Barlow et al., 2009). Log(I) was highly significant (P < 0.0001) when added as a predictor to the model. This confirmed the hypothesis that transmission of S. aureus GTB is dependent on the number of infected cows, suggesting cow-to-cow transmission. 4. Discussion The aim of this study was to quantify transmission of S. aureus GTB in a longitudinal field study. It was estimated that one S. aureus GTB infectious cow was able to infect 2.6 other cows, in a fully susceptible population, during its entire infectious period. Since R0 was significantly above 1, we expect that S. aureus GTB would be able to cause outbreaks in communal dairy operations where mastitis management is generally poor (Keeling and Rohani, 2008; Vynnycky and White, 2010). This was the case in this study since there was a strong increase in the cow level and between-herd of origin prevalence of S. aureus GTB during the communal period as reported previously (Voelk et al., 2014). The current and previous study (Voelk et al., 2014) therefore reemphasized the contribution of communal operations to the spread of contagious pathogens among dairy herds (Presi et al., 2011; Spergser et al., 2013). Transmission of S. aureus has been previously quantified in a few studies. Barlow et al. (2009) summarized data of Lam et al. (1996) and Zadoks et al. (2002) quantifying transmission of major gram-positive bacteria (i.e., pathogens other than S. aureus were also included) in 10 herds. Mean reported transmission parameter ˇ was 0.0087 for quarters with PMTD and 0.0362 for quarters not receiving PMTD. In an intervention study, the effect of antimicrobial treatment on transmission dynamics of S. aureus in 2 herds applying PMTD was evaluated (Barlow et al., 2013). Transmission parameter ˇ was 0.0080 in one herd and 0.0045 in the other but was not affected by antimicrobial treatment. Transmission parameters differed between S. aureus strains in that study but the study did not include strains belonging to clonal complex 8, the clonal complex to which S. aureus GTB belongs (Boss et al., 2016). A recent study evaluated efficacy of a new polyvalent mastitis vaccine in 2 herds with PMTD by quantifying the reduction in S. aureus transmission as a result of the vaccine (Schukken et al., 2014). Mean transmission parameter ˇ of S. aureus among control cows not receiving the vaccine was 0.0080 in the first herd and 0.0098 in the second. Transmission of S. aureus was lower among vaccinated cows. In our study, transmission of S. aureus GTB was almost as high as the estimate of Barlow et al. (2009) for cows not receiving PMTD (0.0362), despite the fact that cows in our study were receiving PMTD. The quality of PMTD (coverage, product efficacy, etc.), the implementation of other preventive measures, the virulence diversity of various S. aureus strains, and possible differences in the susceptibility of hosts hamper a true comparison between herds
67
and studies. Finally, it should be noted that aforementioned studies determined S. aureus transmission at quarter level whereas we measured it at cow level. This is a result of a difference in sampling strategy. Our estimate should therefore be considered an underestimation because transmission between quarters within the same cow could not be estimated in our study. There were some limitations in our study. First, samples were only collected in 7 communal operations at 2 occasions. This limited the dataset to 7 intervals from which transmission parameter ˇ could be estimated. The final population size of 551 dairy cows made it possible to produce estimates with rather narrow confidence intervals nonetheless. The high precision may also have been a result of there being a similar force of infection in the 7 communal operations. It was not possible to statistically compare transmission parameters between the 7 communal operations because only 1 interval could be evaluated per communal operation. Moreover, the relatively short study period limited our ability to calculate cure rates. Cure rates could have been calculated based on the available data (some cows did indeed clear their infection; data not shown) but this would have resulted in a strong underestimation. Cows were only followed for a period of 2 to 3 months during lactation. Their dry off period, in which they have a high probability of being cured, could not be included in the study. Duration of IMI used to calculate R0 was therefore based on input from literature, which may not have been completely generalizable to the communal operations followed in the current study. The herds in the current study had an open structure, as cows did not remain on pasture for the entire study period. There were a few late arrival cows and a considerably higher number of cows that left early. Farmers were instructed to take milk samples from these cows but compliance was moderate, resulting in loss of some cows to follow up. Seventy-four of the cows sampled at the first occasion did not have a second sample collected. An additional 77 cows were sampled but had missing sampling dates, and the sampling interval for these cows had to be inferred. This inference may have biased our estimation of the mean transmission parameter in an unknown direction and with an unknown magnitude. Cows in the study that left their communal operation early had a higher prevalence of S. aureus GTB (Voelk et al., 2014). This suggest that the transmission parameter ˇ estimate (and thus R0 ) would have been higher if those cows had been included in the study. Finally, cows with clinical mastitis were sampled neither. Those cows may have experienced a new IMI and a subsequent cure during the communal period; a result of an antimicrobial treatment. We could thus have missed such short IMI resulting in an underestimation of transmission parameter ˇ. This study confirmed the hypothesis that genotype B is a contagious S. aureus genotype. Including log(I) as a predictor into the statistical model revealed a strong association with the number of new S. aureus GTB cases. Moreover, R0 was significantly greater than 1 and S. aureus GTB is therefore able to cause outbreaks as observed in the communal dairy herds participating in this study. Quantification of transmission parameter ˇ aids in designing cow level intervention strategies to reduce transmission of S. aureus GTB within communal operations and between cows’ herds of origin. Interventions to limit transmission of S. aureus GTB within communal operations include PMTD (Lam et al., 1996), vaccination (Schukken et al., 2014), and segregation of infectious cows by milking them last or with a separate milking cluster. Another possibility to reduce spread of S. aureus GTB within communal operations is to shorten duration of IMI. Given that transmission parameter ˇ was estimated to be 0.0232 (95% CI: 0.0200–0.0269), duration of IMI should be below 50 days to achieve a R0 significantly below 1. Within communal operations, this can be achieved by antimicrobial treatment (Barkema et al., 2006; van den Borne et al., 2010b) or transporting cows back to their herds of origin. Communal dairy
68
B.H.P. van den Borne et al. / Preventive Veterinary Medicine 136 (2017) 65–68
operations could additionally restrict the movement of infectious animals in order to limit between-herd of origin transmission of S. aureus GTB. Conflict of interest The authors declare no conflict of interest. Acknowledgements The authors would like to thank all farmers and personnel of the communal operations for their participation in the study and John Berezowski (Veterinary Public Health Institute, Liebefeld, Switzerland) for commenting on a previous version of the manuscript. Funding for the original study (Voelk et al., 2014) was provided by the Canton of Graubünden (Switzerland). References Barkema, H.W., Schukken, Y.H., Zadoks, R.N., 2006. Invited Review: the role of cow, pathogen, and treatment regimen in the therapeutic success of bovine Staphylococcus aureus mastitis. J. Dairy Sci. 89, 1877–1895, http://dx.doi.org/ 10.3168/jds.S0022-0302(06)72256-1. Barlow, J.W., White, L.J., Zadoks, R.N., Schukken, Y.H., 2009. A mathematical model demonstrating indirect and overall effects of lactation therapy targeting subclinical mastitis in dairy herds. Prev. Vet. Med. 90, 31–42, http://dx.doi.org/ 10.1016/j.prevetmed.2009.03.016. Barlow, J.W., Zadoks, R.N., Schukken, Y.H., 2013. Effect of lactation therapy on Staphylococcus aureus transmission dynamics in two commercial dairy herds. BMC Vet. Res. 9, 28, http://dx.doi.org/10.1186/1746-6148-9-28. Boss, R., Naskova, J., Steiner, A., Graber, H.U., 2011. Mastitis diagnostics: quantitative PCR for Staphylococcus aureus genotype B in bulk tank milk. J. Dairy Sci. 94, 128–137, http://dx.doi.org/10.3168/jds.2010-3251. Boss, R., Cosandey, A., Luini, M., Artursson, K., Bardiau, M., Breitenwieser, F., Hehenberger, E., Lam, T., Mansfeld, M., Michel, A., Mösslacher, G., Naskova, J., Nelson, S., Podpeˇcan, O., Raemy, A., Ryan, E., Salat, O., Zangerl, P., Steiner, A., Graber, H.U., 2016. Bovine Staphylococcus aureus: subtyping, evolution, and zoonotic transfer. J. Dairy Sci. 99, 515–528, http://dx.doi.org/10.3168/jds.20159589. Cosandey, A., Boss, R., Luini, M., Artursson, K., Bardiau, M., Breitenwieser, F., Hehenberger, E., Lam, T., Mansfeld, M., Michel, A., Mösslacher, G., Naskova, J., Nelson, S., Podpeˇcan, O., Raemy, A., Ryan, E., Salat, O., Zangerl, P., Steiner, A., Graber, H.U., 2016. Staphylococcus aureus genotype B and other genotypes isolated from cow milk in European countries. J. Dairy Sci. 99, 529–540, http:// dx.doi.org/10.3168/jds.2015-9587. Deurenberg, R.H., Stobberingh, E.E., 2008. The evolution of Staphylococcus aureus. Infect. Genet. Evol. 8, 747–763, http://dx.doi.org/10.1016/j.meegid.2008.07. 007. Fournier, C., Kuhnert, P., Frey, J., Miserez, R., Kirchhofer, M., Kaufmann, T., Steiner, A., Graber, H.U., 2008. Bovine Staphylococcus aureus: Association of virulence genes, genotypes and clinical outcome. Res. Vet. Sci. 85, 439–448, http://dx.doi. org/10.1016/j.rvsc.2008.01.010.
Graber, H.U., Naskova, J., Studer, E., Kaufmann, T., Kirchhofer, M., Brechbühl, M., Schaeren, W., Steiner, A., Fournier, C., 2009. Mastitis-related subtypes of bovine Staphylococcus aureus are characterized by different clinical properties. J. Dairy Sci. 92, 1442–1451, http://dx.doi.org/10.3168/jds.2008-1430. Keeling, M.J., Rohani, P., 2008. Modeling Infectious Diseases, first ed. Princeton University Press, Princeton, New Jersey, USA. Lam, T.J.G.M., De Jong, M.C.M., Schukken, Y.H., Brand, A., 1996. Mathematical modelling to estimate efficacy of postmilking teat disinfection in split-udder trials of dairy cows. J. Dairy Sci. 79, 62–70. Presi, P., Struchen, R., Knight-Jones, T., Scholl, S., Heim, D., 2011. Bovine viral diarrhea (BVD) eradication in Switzerland-experiences of the first two years. Prev. Vet. Med. 99, 112–121, http://dx.doi.org/10.1016/j.prevetmed.2011.01. 012. Schouten, J.M., Graat, E.A.M., Frankena, K., Van Zijderveld, F., De Jong, M.C.M., 2009. Transmission and quantification of verocytotoxin-producing Escherichia coli O157 in dairy cattle and calves. Epidemiol. Infect. 137, 114–123, http://dx.doi. org/10.1017/s0950268808000320. Schukken, Y.H., Bronzo, V., Locatelli, C., Pollera, C., Rota, N., Casula, A., Testa, F., Scaccabarozzi, L., March, R., Zalduendo, D., Guix, R., Moroni, P., 2014. Efficacy of vaccination on Staphylococcus aureus and coagulase-negative staphylococci intramammary infection dynamics in 2 dairy herds. J. Dairy Sci. 97, 5250–5264, http://dx.doi.org/10.3168/jds.2014-8008. Spergser, J., Macher, K., Kargl, M., Lysnyansky, I., Rosengarten, R., 2013. Emergence, re-emergence, spread and host species crossing of Mycoplasma bovis in the Austrian Alps caused by a single endemic strain. Vet. Microbiol. 164, 299–306, http://dx.doi.org/10.1016/j.vetmic.2013.02.007. Syring, C., Boss, R., Reist, M., Bodmer, M., Hummerjohann, J., Gehrig, P., Graber, H.U., 2012. Bovine mastitis: the diagnostic properties of a PCR-based assay to monitor the Staphylococcus aureus genotype B status of a herd, using bulk tank milk. J. Dairy Sci. 95, 3674–3682, http://dx.doi.org/10.3168/jds.2011-4968. van den Borne, B.H.P., Nielen, M., van Schaik, G., Melchior, M.B., Lam, T.J.G.M., Zadoks, R.N., 2010a. Host adaptation of bovine Staphylococcus aureus seems associated with bacteriological cure after lactational antimicrobial treatment. J. Dairy Sci. 93, 2550–2558, http://dx.doi.org/10.3168/jds.2009-2971. van den Borne, B.H.P., van Schaik, G., Lam, T.J.G.M., Nielen, M., 2010b. Therapeutic effects of antimicrobial treatment during lactation of recently acquired bovine subclinical mastitis: two linked randomized field trials. J. Dairy Sci. 93, 218–233, http://dx.doi.org/10.3168/jds.2009-2567. Velthuis, A.G.J., De Jong, M.C.M., Kamp, E.M., Stockhofe, N., Verheijden, J.H.M., 2003. Design and analysis of an Actinobacillus pleuropneumoniae transmission experiment. Prev. Vet. Med. 60, 53–68, http://dx.doi.org/10.1016/S01675877(03)00082-5. Voelk, V., Graber, H.U., van den Borne, B.H.P., Sartori, C., Steiner, A., Bodmer, M., Haerdi-Landerer, M.C., 2014. A longitudinal study investigating the prevalence of Staphylococcus aureus genotype B in seasonally communal dairy herds. J. Dairy Sci. 97, 4184–4192, http://dx.doi.org/10.3168/jds.2013-7291. Vynnycky, E., White, R.G., 2010. An Introduction to Infectious Disease Modelling, first ed. Oxford University Press, Oxford, New York. Zadoks, R.N., Allore, H.G., Barkema, H.W., Sampimon, O.C., Gröhn, Y.T., Schukken, Y.H., 2001. Analysis of an outbreak of Streptococcus uberis mastitis. J. Dairy Sci. 84, 590–599, http://dx.doi.org/10.3168/jds.S0022-0302(01)74512-2. Zadoks, R.N., Allore, H.G., Hagenaars, T.J., Barkema, H.W., Schukken, Y.H., 2002. A mathematical model of Staphylococcus aureus control in dairy herds. Epidemiol. Infect. 129, 397–416.