Abstract
Many infectious agents transmitting through a contaminated environment are able to persist in the environment depending on the temperature and sanitation determined rates of their replication and clearance, respectively. There is a need to elucidate the effect of these factors on the infection transmission dynamics in terms of infection outbreaks and extinction while accounting for the random nature of the process. Also, it is important to distinguish between the true and apparent extinction, where the former means pathogen extinction in both the host and the environment while the latter means extinction only in the host population. This study proposes a stochastic-differential equation model as an approximation to a Markov jump process model, using Escherichia coli O157:H7 in cattle as a model system. In the model, the host population infection dynamics are described using the standard susceptible-infected-susceptible framework, and the E. coli O157:H7 population in the environment is represented by an additional variable. The backward Kolmogorov equations that determine the probability distribution and the expectation of the first passage time are provided in a general setting. The outbreak and apparent extinction of infection are investigated by numerically solving the Kolmogorov equations for the probability density function of the associated process and the expectation of the associated stopping time. The results provide insight into E. coli O157:H7 transmission and apparent extinction, and suggest ways for controlling the spread of infection in a cattle herd. Specifically, this study highlights the importance of ambient temperature and sanitation, especially during summer.
Similar content being viewed by others
References
Alejandro RC (2004) Flushing and scraping freestalls and drylot pens. In: News D (ed) University of California cooperative extension, CA
Allen L (2007) An introduction to mathematical biology. Prentice-Hall, Upper Saddle River
Allen LJS, Allen EJ, Jonsson CB (2006) The impact of environmental variation on hantavirus infection in rodents. In: Gumel AB, Castillo-Chavez C, Mickens RE, Clemence DP (eds) Contemporary mathematics series. Proceedings of the joint summer research conference on modeling the dynamics of human diseases: emerging paradigms and challenges, vol 410. AMS, Providence, pp 1–15
Allen LJS, Lahodny GE Jr (2012) Extinction thresholds in deterministic and stochastic epidemic models. J Biol Dyn 6:590–611
Ayscue P, Lanzas C, Ivanek R, Grohn YT (2009) Modeling on-farm Escherichia coli O157:H7 population dynamics. Foodborne Pathog Dis 6:461–470
Bani-Yaghoub M, Gautam R, Döpfer D, Kaspar CW, Ivanek R (2012a) Effectiveness of environmental decontamination in control of infectious diseases. Epidemiol Infect 140(3):542–553
Bani-Yaghoub M, Gautam R, Shuai Z, van den Driessche P, Ivanek R (2012b) Reproduction numbers for infections with free-living pathogens growing in the environment. J Biol Dyn 6(2):923–940
Barkocy-Gallagher GA, Arthur TM, Rivera-Betancourt M, Nou X, Shackelford SD, Wheeler TL, Koohmaraie M (2003) Seasonal prevalence of Shiga toxin-producing Escherichia coli, including O157:H7 and non-O157 serotypes, and Salmonella in commercial beef processing plants. J Food Prot 66:1978–1986
Berry ED, Miller DN (2005) Cattle feedlot soil moisture and manure content: II. Impact on Escherichia coli O157. J Environ Qual 34:656–663
Chase-Topping M, Gally D, Low C, Matthews L, Woolhouse M (2008) Super-shedding and the link between human infection and livestock carriage of Escherichia coli O157. Nat Rev Microbiol 6:904–912
Clancy D (2005) A stochastic SIS infection model incorporating indirect transmission. J Appl Probab 42:725–737
Cray WC et al (1998) Effect of dietary stress on fecal shedding of Escherichia coli O157:H7 in calves. Appl Environ Microbiol 64:1975–1979
Cushing JM, Costantino RF, Dennis B, Desharnais RA, Henson SM (2003) Chaos in ecology. Academic Press, San Diego
Davis MA, Cloud-Hansen KA, Carpenter J, Hovde CJ (2005) Escherichia coli O157:H7 in environments of culture-positive cattle. Appl Environ Microbiol 71:6816–6822
Diekmann O, Heesterbeek JAP, Metz JAJ (1990) On the definition and the computation of the basic reproduction ratio \(R_0\) in models for infectious diseases in heterogeneous population. J Math Biol 28:365–382
Döpfer D, Geue L, Schares S, Mintel B, Hoffmann B, Fischer EAJ (2012) Dynamics of shiga-toxin producing Escherichia coli (STEC) and their virulence factors in cattle. Prev Vet Med 103(1):22–30
Durrett R (1996) Probability: theory and examples. Duxbury Press, Belmont
Ewald PW (2004) Evolution of virulence. Infect Dis Clin North Am 18:1–15
Ferrari MJ, Grais RF, Bharti N, Conlan AJK, Bjornstad ON, Wolfson LJ, Guerin PJ, Djibo A, Grenfell BT (2008) The dynamics of measles in sub-Saharan Africa. Nature 451:679–684
Friedman A (1964) Partial differential equations of parabolic type. Prentice-Hall, Englewood Cliffs
Gardiner CW (2009) Handbook of Stochastic Methods for physics, chemistry, and the natural sciences, 4th edn. Springer, Berlin
Gautam R, Bani-Yaghoub M, Neill WH, Döpfer D, Kaspar CW, Ivanek R (2011) Modeling the effect of seasonal variation in ambient temperature on the transmission dynamics of a pathogen with a free-living stage: example of Escherichia coli O157:H7 in a dairy herd. Prev Vet Med 102(1):10–21
Gillespie DT (2002) The chemical Langevin equation and Fokker–Planck equation for the reverisble isomerization reaction. J Phys Chem A 106:5063–5071
Heesterbeek JAP, Roberts MG (2007) The type-reproduction number T in models for infectious disease control. Math Biosci 206:3–10
Hussein HS, Sakuma T (2005) Prevalence of shiga toxin-producing Escherichia coli in dairy cattle and their products. J Dairy Sci 88:450–465
It\(\hat{\text{ o }}\) K (1944) Stochastic Integral. Proc Imp Acda Tokyo 20:519–524
Jiang X, Morgan J, Doyle MP (2002) Fate of Escherichia coli O157:H7 in manure-amended soil. Appl Environ Microbiol 68:2605–2609
Karatzas I, Shreve SE (1991) Brownian motion and stochastic calculus. Springer, New York
Karmali MA, Petric M, Lim C, Fleming PC, Arbus GS, Lior H (1985) The association between idiopathic hemolytic uremic syndrome and infection by verotoxin-producing Escherichia coli. Inf Dis 151:775–782
Kudva IT, Blanch K, Hovde CJ (1998) Analysis of Escherichia coli O157:H7 survival in ovine or bovine manure and manure slurry. Appl Environ Microbiol 64(9):3166–3174
Kurtz TG (1971) Limit theorems for sequences of jump Markov processes approximating ordinary differential processes. J Appl Probab 8:344–356
Kurtz TG (1981) Approximation of population processes. In: CBMS-NSF regional conference series in applied mathematics, vol 36. SIAM, Philadelphia
Lande R, Engen S, Saether B-E (2003) Stochastic population dynamics in ecology and conservation. Oxford University Press, Oxford
LeJeune JT, Wetzel AN (2007) Preharvest control of Escherichia coli O157 in cattle. J Anim Sci 85:E73–E80
Liu WC, Jenkins C, Shaw DJ et al (2005) Modelling the epidemiology of verocytotoxin-producing Escherichia coli serogroups in young calves. Epidemiol Infect 133:449–458
Liu WC, Shaw DJ, Matthews L et al (2007) Modelling the epidemiology and transmission of verocytotoxin-producing Escherichia coli serogroups O26 and O103 in two different calf cohorts. Epidemiol Infect 135:1316–1323
MacDiarmid BN, Watkin BR (1972) The cattle dung patch. 3. Distribution and rate of decay of dung patches and their influence on grazing behaviour. J Br Grossl Soc 27:48–54
Maule A (2000) Survival of verocytotoxigenic Escherichia coli O157 in soil, water and on surfaces. Symp Ser (Soc Appl Microbiol) 29:71–78
Matthews L, Low JC, Gally DL et al (2006a) Heterogeneous shedding of Escherichia coli O157 in cattle and its implications for control. Proc Natl Acad Sci 103:547–552
Matthews L, McKendrick IJ, Ternent HE, Gunn GJ, Synge BA, Woolhouse MEJ (2006b) Super-shedding cattle and the transmission dynamics of Escherichia coli O157. Epidemiol Infect 134:131–142
McGee P, Bolton DJ, Sheridan JJ, Earley B, Leonard L (2001) The survival of Escherichia coli O157:H7 in slurry from cattle fed different diets. Lett Appl Microbiol 32:152–155
Mead PS, Griffin PM (1998) Escherichia coli O157-H7. Lancet 352:1207–1212
Mechie SC, Chapman PA, Siddons CA (1997) A fifteen month study of Escherichia coli O157:H7 in a dairy herd. Epidemiol Infect 118:17–25
Mitchell AR, Griffiths DF (1980) The finite difference method in partial differential equations. Wiley, New York
Murray JD (2002) Mathematical biology, vol I. Springer, Berlin
Nataro JP, Kaper JB (1998) Diarrheagenic Escherichia coli. Clin Microbiol Rev 11:142–201
Ripa J (2012) Stochasticity, environmental. In: Hastings A, Gross LJ (eds) Encyclopedia of theoretical ecology. University of California Press, Berkeley, pp 712–718
Roberts MG, Heesterbeek JAP (2003) A new method for estimating the effort required to control an infectious disease. Proc R Soc Lond B 270:1359–1364
Shane EM, Endres MI, Janni KA (2010) Alternative bedding materials for compost bedded pack barns in Minnesota: a descriptive study. Appl Eng Agric 26:465–473
Schuss Z (2009) Theory and applications of stochastic processes. In: Springer series on applied mathematical sciences, vol 170
Smith HL, Waltman P (1995) The theory of the chemostat. Cambridge University Press, Cambridge
Turner J, Begon M, Bowers RG, French NP (2003) A model appropriate to the transmission of a human food-borne pathogen in a multi-group managed herd. Prev Vet Med 57:175–198
Turner J, Bowers RG, Begon M, Robinson SE, French NP (2006) A semi-stochastic model of the transmission of Escherichia coli O157 in a typical UK dairy herd: dynamics, sensitivity analysis and intervention/prevention strategies. J Theor Biol 241:806–822
Turner J, Bowers RG, Clancy D, Behnke MC, Christley RM (2008) A network model of E.coli O157 transmission within a typical UK dairy herd: the effect of heterogeneity and clustering on the prevalence of infection. J Theor Biol 254:45–54
van den Driessche P, Watmough J (2002) Reproduction numbers and sub-threshold endemic equilibria for compartmental models of disease transmission. Math Biosci 180:29–48
Verhulst PF (1838) Notice sur la loi que la population poursuit dans son accroissement. Corresp Math Phys 10:113–121
Vital M, Hammes F, Egli T (2008) Escherichia coli O157 can grow in natural freshwater at low carbon concentrations. Environ Microbiol 10:2387–2396
Vosough Ahmadi B, Frankena K, Turner J, Velthuis AGJ, Hogeveen H, Huirne RBM (2007) Effectiveness of simulated interventions in reducing the estimated prevalence of Escherichia coli O157:H7 in lactating cows in dairy herds. Vet Res 38:755–771
Wang G, Zhao T, Doyle MP (1996) Fate of enterohemorrhagic Escherichia coli O157:H7 in bovine feces. Appl Environ Microbiol 62:2567–2570
Wood JC, McKendrick IJ, Gettinby G (2007) A simulation model to assess herd-level intervention strategies against Escherichia coli O157. Epidemiol Infect 135:749–764
Zhang XS, Chase-Topping ME, McKendrick IJ, Savill NJ, Woolhouse MEJ (2010) Spread of Escherichia coli O157:H7 infection among Scottish cattle farms: stochastic models and model selection. Epidemics 2:11–20
Acknowledgments
We thank three anonymous referees and the editor for their suggestions that improved this paper. This work was supported by the National Science Foundation grant NSF-EF-0913367 to RI funded under the American Recovery and Reinvestment Act of 2009. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. This publication is based in part on work supported by Award No. KUS-C1-016-04, made by King Abdullah University of Science and Technology (KAUST).
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
Below is the link to the electronic supplementary material.
Appendices
Appendix A: Proof of Theorem 3.1
System (2.2) can be written as \(d{{\varvec{x}}}/dt=f({{\varvec{x}}})\), where \({{\varvec{x}}}=(I,E)^T\) and \({\varvec{f}}=(f_1,f_2)^T\) is continuously differentiable in region \(R\). Every positive semi-orbit of this system lies in \(R\), a closed and bounded subset of \(\mathbb R ^2\). In addition, there are a finite number of equilibria in \(R\). We apply the powerful Poincaré–Bendixson theory for two-dimensional autonomous systems of differential equations, in particular, a result known as the Poincaré–Bendixson trichotomy (Smith and Waltman 1995). This result states that for system \(d{{\varvec{x}}}/dt=f({{\varvec{x}}})\) with the preceding properties that the omega-limit set of a positive semi-orbit beginning in \(R\) must approach one of the following limiting sets: an equilibrium, a periodic orbit, or a cycle graph (equilibria connected by solution trajectories). We show that the omega-limit set for every solution of (2.2) beginning in \(R\) is the DFE in the case \(T\le 1\) or the endemic equilibrium in the case \(T>1\).
Proof of Theorem 3.1
For the case \(T\le 1\), there is only one equilibrium in \(R\), namely, the DFE. For \(T<1\), the DFE is locally asymptotically stable and for \(T=1\), the DFE is neutrally stable (one zero eigenvalue and one negative eigenvalue). Since every periodic orbit must contain at least one equilibrium in the interior of its orbit, there cannot exist any periodic orbits for \(T\le 1\) in \(R\). Moreover, there does not exist any cycle graphs (homoclinic orbit) when \(T\le 1\) because the origin is not a saddle point. Thus, by the Poincaré–Bendixson trichotomy, solutions of (2.2) must approach the DFE. The DFE is globally asymptotically stable when \(T\le 1\).
For the case \(T>1\), system (2.2) has two equilibria in \(R\), namely, the DFE and the endemic equilibrium. First, we rule out periodic orbits encircling the endemic equilibrium by applying the Dulac threshold (Smith and Waltman 1995). No periodic orbit can touch the boundary \(E=0\) or \(I=0\) because it is repelling and the origin is an equilibrium. Given the Dulac function \(B=1/(IE)\), the expression \(\partial (Bf_1)/\partial I+\partial (Bf_2)/\partial E<0\) is of one sign in the interior of \(R\) which by Dulac’s threshold implies there does not exist any periodic orbits in \(R\) (Smith and Waltman 1995). Second, we rule out cycle graphs. When \(T>1\), the DFE is a saddle point with stable manifold outside of \(R\) and unstable manifold in \(R\). In addition, the endemic equilibrium is locally asymptotically stable. Hence, there cannot be any homoclinic orbits or heteroclinic orbits and consequently no cycle graphs. Solutions cannot approach the DFE because it is repelling in \(R\). Hence, by the Poincaré–Bendixson trichotomy, solutions of (2.2) must approach the endemic equilibrium. The endemic equilibrium is globally asymptotically stable in the interior of \(R\) when \(T>1\).
Appendix B: Estimation of the maximal carrying capacity \(K_{E}\) of free-living pathogen in the environment
According to Jiang et al. (2002), the carrying capacity of the free-living pathogen in fecal debris is \(10^{5}\) CFU per gram of feces. Notice that a host will on average produce 30 kg of feces per day. So, if there are \(N\) hosts, the maximal carrying capacity of the free-living pathogen in fecal debris will be about \(30,000\times N\times 10^5=3\times 10^{9}\times N\) CFU. In reality, the maximum carrying capacity for the system is not known because the total amount of fecal debris in the cattle environment is unknown. Our assumption underlying the maximum carrying capacity implies that at most there could be \(30,000\times N\) grams of feces in the cattle environment on any given day and since each gram of feces has the carrying capacity of \(10^5\) CFU/g the approximated maximum carrying capacity is \(3\times 10^{9}\times N\) CFU. We tested the robustness of the model results and conclusion to this assumed maximal carrying capacity by running numerical analysis with different values of the maximal capacity. While the proportion of pathogens in the environment over time relative to the maximal carrying capacity obviously changed for different assumed values of this parameter, the qualitative behavior was identical, thus confirming the robustness of the model to the assumption.
Appendix C: Estimation of the pathogen growth rate \(g_{E}\)
Gautam et al. (2011) show that the relationship between the growth rate \(Z\) and the temperature \(T\) can be expressed as
Taking the expectation of both sides of (C.1), we have
Plugging the average temperature into (C.2) gives us the estimates of the average growth rate in (5.5).
Appendix D: Estimation of the pathogen cleaning rate \(r_{E}\)
Four scenarios of cleaning, including cleaning every day, twice per week, once per week and no cleaning, are considered to reflect this wide variation in cleaning frequencies across the cattle industry. The cleaning rate for each of these scenarios was calculated using the formula described in Vosough Ahmadi et al. (2007) as
where \(C\) represents the number of CFU of E. coli O157:H7 per milliliter of bedding materials at time \(t\) (in days), \(I\) is the intercept or initial number of bacteria, and \(q\) is the rate of bacterial depletion. Vosough Ahmadi et al. (2007) obtained a death rate of \(0.46\) CFU per day for cleaning practice of two times per week by fitting an exponential distribution to the data reported by Davis et al. (2005). To account for cleaning of water, \(0.17\) CFU per day is added to the estimated death rate. Using this approach we estimate the overall death rate for different cleaning frequencies as given in Table 1.
Rights and permissions
About this article
Cite this article
Wang, X., Gautam, R., Pinedo, P.J. et al. A stochastic model for transmission, extinction and outbreak of Escherichia coli O157:H7 in cattle as affected by ambient temperature and cleaning practices. J. Math. Biol. 69, 501–532 (2014). https://doi.org/10.1007/s00285-013-0707-1
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00285-013-0707-1
Keywords
- Escherichia coli O157:H7
- Stochastic \({SIS}_\mathrm{E} {{ model}}\)
- Kolmogorov equations
- Extinction
- outbreak