REVIEW 5 major objections 6 minor 35 references
A Mathematical and Optimal Control Model for Rabies Transmission Dynamics Among Humans and Dogs with Environmental Effects
T0 review · 5 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read Vaccinating domestic dogs and giving bite victims post-exposure prophylaxis can eliminate rabies within five years, according to an optimal control model fitted to Tanzanian data.
desk verdict The optimal-control framework is standard and the model builds on their earlier work, but the five-year elimination claim is contradicted by their own fitted parameters and the reproduction number drops the environmental route entirely. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is a 12-compartment deterministic ODE system: susceptible, exposed, infected, and recovered classes for humans, free-ranging dogs, and domestic dogs, plus a state $M$ for rabies virus in the environment with saturating contamination $\lambda(M)=M/(M+C)$. Controls enter multiplicatively: $(1-u_1-u_3)$ scales human exposure, $(1-u_1-u_2)$ scales domestic-dog exposure, and PEP ($u_4$) moves exposed humans and dogs into recovered classes. Two tools carry the argument: the next-generation matrix gives the reproduction number $R_e$ that sets the stability threshold, and Pontryagin's maximum principle gives explicit characterizations of the four optimal controls in terms of adjoint variables, solved numerically by a forward-backward fourth-order Runge–Kutta scheme.
What would settle it
Compute the reproduction number of the free-ranging dog subsystem alone, $R_{21}$, from the fitted parameters; if $R_{21}>1$, run the optimal-control simulation with only domestic-dog vaccination and human PEP active. If infected free-ranging dogs and environmental virus do not decay to zero, the five-year elimination claim is contradicted.
Extended reading notes
Core claim
On the paper's own terms, the discovery is that domestic-dog vaccination and post-exposure prophylaxis are the two controls that carry the elimination, and that they work without a direct control on free-ranging dogs. The effective reproduction number takes the form $R_e = (R_{33}+R_{21}+\sqrt{R_{21}^2-2R_{33}R_{21}+4R_{31}R_{23}+R_{33}^2})/2$, where $R_{21}$ is free-ranging dog transmission, $R_{33}$ is domestic-dog transmission, and $R_{31}, R_{23}$ are cross-transmission terms; the dog terms are multiplied by $(1-u_1-u_2)$ and the human terms by $(1-u_1-u_3)$, so controls enter the threshold directly. The paper proves that the rabies-free equilibrium is globally stable when $R_e<1$ and that the endemic equilibrium is globally stable when $R_e>1$. In simulations with the fitted Tanzanian parameters, the optimal control problem keeps $u_2$ and $u_4$ at 100% for the first several to sixteen years and gradually reduces them, and all exposed and infected compartments, including free-ranging dogs and environmental virus, fall to near zero by year five.
Load-bearing premise
The elimination result assumes that free-ranging dogs can be cleared indirectly by controlling domestic dogs and humans, even though no control acts directly on the free-ranging dog transmission terms.
Editorial extensions
If this is right
- If $R_e<1$ is achieved by the optimal controls, the disease-free equilibrium is globally stable, so once rabies is cleared it does not rebound while the controls are sustained.
- Vaccination of domestic dogs ($u_2$) and post-exposure prophylaxis ($u_4$) are the controls held at 100% the longest, making them the main drivers of the five-year elimination.
- The integrated strategy also clears the free-ranging dog and environmental compartments, even though no term in the model directly reduces those infections.
- Higher dog-to-dog contact rates ($\psi_1$, $\psi_2$) raise $R_e$, so the same package must run longer or harder in high-contact settings.
- Deterrence factors ($\rho_1$, $\rho_2$, $\rho_3$) lower the exposed and infected domestic-dog counts, so discouraging contact complements vaccination.
Reading between the lines
- If the paper's fitted parameters leave the free-ranging dog subsystem with $R_{21}>1$, the five-year result rests on the cross-species and environmental links being interrupted indirectly; a closed free-ranging-population variant of the model is the natural stress test.
- Because the paper reports a real mass-vaccination campaign at $1.61 per dog, its quadratic cost weights could be replaced by unit costs to turn the elimination claim into a budget projection.
- The same control framework could be applied to imperfect PEP uptake, which the paper notes leaves 20–30% of bite victims untreated, to see how much the five-year horizon stretches.
- Setting the environmental shedding rates $\nu_1=\nu_2=\nu_3=0$ would isolate how much of the result depends on the virus reservoir $M$ rather than on direct bites.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proposes a deterministic 12-compartment ODE model for rabies transmission among humans, free-ranging dogs, domestic dogs, and an environmental virus compartment, with four time-dependent controls (health practice, domestic-dog vaccination, community education, and post-exposure prophylaxis). The authors derive a disease-free equilibrium, define an effective reproduction number R_e, state global stability results for the disease-free and endemic equilibria, perform a PRCC sensitivity analysis, solve an optimal control problem with Pontryagin's maximum principle, fit the model to reported human rabies incidence in Tanzania, and simulate four control strategies. The central claim is that an integrated strategy, especially vaccination of domestic dogs and PEP, can eliminate rabies within five years.
Significance. If the model and its quantitative predictions were correct, the paper would provide useful evidence for combining dog vaccination, education, and PEP in rabies control programs, and the inclusion of an environmental virus compartment addresses a genuinely underexplored transmission route. The authors also provide a standard optimal-control formulation and a data-fitting exercise. However, the central quantitative claims are undermined by internal inconsistencies: the reproduction number omits the environmental route that is a focus of the paper, the analytic R_e formulas contain incorrect pathway compositions, and the reported figures are inconsistent with the paper's own fitted parameters. The five-year elimination claim is not supported by the stated model equations.
major comments (5)
- [Sec. 3.2, Eq. (14)] The next-generation matrix F in Eq. (14) has zero entries in every column corresponding to the environmental compartment M, despite the fact that the incidence terms in system (1) contain τ3λ(M), κ3λ(M), and ψ3λ(M)/(1+ρ3). Since λ'(0)=1/C ≠ 0, the partial derivatives ∂F_i/∂M at the disease-free equilibrium are nonzero. The environmental transmission route is therefore absent from R_e, which invalidates the threshold condition R_e<1 as a statement about interruption of transmission through the environment emphasized in the title and conclusions.
- [Sec. 3.2, Eq. (16)] The formulas for R23, R31, and R33 in Eq. (16) are incorrectly composed. The path from infected domestic dogs to infected free-ranging dogs is ID→SF→EF→IF, so R23 should carry the factor γ/[(μ2+γ)(σ2+μ2)], not the domestic-dog factor a3=γ/[(μ3+γ1+γ2+u4)(σ3+μ3)]. Similarly, the paths from IF and ID to ID/H-domestic infection are IF→SD→ED→ID and ID→SD→ED→ID, so R31 and R33 should carry γ1/[(μ3+γ1+γ2+u4)(σ3+μ3)], not the free-range progression rate γ. As written, Eq. (16) is not the spectral radius of F V^{-1} even for the dog-only subsystem.
- [Sec. 4.1, Table 2 and Sec. 4.2, Figs. 5–7] Using the fitted values in Table 2 (κ1=2.0×10^-5, θ2=1004.12, γ=0.1664, μ2=0.0663, σ2=0.0896) gives R21≈1.39>1. Because Eq. (16) expresses R_e as R21 plus nonnegative terms, the model with Table 2 parameters implies R_e>1 at every control level, and u2 and u4 do not appear in R21 at all. Yet Figures 5–7 plot R_e between approximately 0.72 and 0.9. These figures therefore cannot have been generated from Eq. (16) with the parameters stated in Table 2; the paper does not disclose which parameters or formula were actually used.
- [Sec. 4.4, Strategy A, Figs. 9–11] The claim that the integrated strategy eliminates rabies within five years is inconsistent with system (1) and Table 2. Controls u1–u4 appear only in the human and domestic-dog equations; the equations for S_F, E_F, and I_F contain no control term. With R21≈1.39>1, the free-ranging dog subsystem is self-sustaining: even if domestic dogs, humans, and the environment are all set to zero, a small positive I_F(0) grows back to a positive endemic level and reseeds the other compartments through τ1, ψ1, and ν2. No control in the model acts on that subsystem, so the elimination trajectories in Figures 9–11 cannot be solutions of (1) with the fitted parameters unless the simulations used different parameters or a modified model.
- [Sec. 4.1, Eqs. (30)–(31) and Fig. 4] The parameter estimates that drive all later simulations are obtained by fitting the single equation (30) for I_H to data labelled 'Cumulative incidence of rabies' in Figure 4, but I_H in system (1) is an infected-state variable, not a cumulative incidence. No goodness-of-fit statistic, residual analysis, or uncertainty interval is reported, and the full model (1) is not fitted. The fitted parameters therefore provide a weak empirical basis for the quantitative predictions in Section 4.4.
minor comments (6)
- [Sec. 3.7.1, Theorem 6] The uniqueness proof is invalid: δu_j = \bar u_j - \tilde u_j may be negative, so the lower bound 0 ≤ 2A_j \tilde u_j δu_j in Eq. (27) is false; the expansion of J also has the wrong coefficients (2A_j and A_j instead of A_j and A_j/2). The uniqueness claim is therefore not established.
- [Sec. 3.1, Lemma 1] The proof of Lemma 1 establishes Lipschitz continuity of the vector field, which gives local existence and uniqueness, but it does not prove positivity of solutions; a standard invariant-region argument for the nonnegative orthant is missing.
- [Sec. 3.7.2, Eq. (28)] In the Hamiltonian, the cost term for u4 is written as (1/2)u4^2 A3; it should be (1/2)u4^2 A4. Also, the V matrix in Eq. (15) uses δ3 where the model and Table 2 use σ3.
- [Sec. 3.6, Eq. (22)] The objective functional contains the term -K6 S_D, and the text states that K6 is the weight for vaccinated domestic dogs, but S_D is the susceptible domestic-dog compartment. The modeling intention should be clarified.
- [Sec. 3.4, Eq. (18)] The expression for E_H^* in Eq. (18) is independent of the transmission rates λ_H, τ1, τ2, and τ3, which is not plausible for an endemic equilibrium; the formula also appears dimensionally inconsistent. This part of the endemic-equilibrium analysis needs correction.
- [Data Availability] The data availability statement says the data came from existing literature, but Figure 4 presents Tanzania incidence data from 1990–2018 without a specific source or citation; please identify the dataset.
Circularity Check
No significant circularity: the control and stability results are derived from the stated model and calibrated data, not defined into existence.
full rationale
The paper's main derived quantities—the effective reproduction number (16), the DFE stability Theorem 3, the Lyapunov-based endemic stability, and the Pontryagin characterization (29)—are computed from the model system (1) and its next-generation/adjoint structure, not imported from a fitted output. The Tanzania incidence data are used only to estimate parameters in Section 4.1; the subsequent optimal-control runs are forward simulations of the same system with controls, so the elimination scenario is a modeling consequence rather than a renamed fit. The base model is taken from the authors' own [3], but the present paper's novel optimal-control conclusions do not reduce to that citation: they are evaluated numerically here. There is a serious non-circularity concern that the fitted Table 2 values imply R21 ≈ 1.39 > 1 with no control acting on the free-ranging dog equations, which appears inconsistent with the Re < 1 surfaces and five-year elimination figures; however, that is a correctness/reproducibility issue, not a circularity of the kind where a conclusion is equivalent to its input by construction.
Assumptions & free parameters
free parameters (8)
- kappa1 =
0.000020
- kappa2 =
0.000081
- kappa3 =
0.000040
- tau3 =
0.000303
- nu1, nu2, nu3 =
0.001958, 0.008971, 0.005735
- C =
0.003011
- theta1, theta2, theta3 =
1993.38, 1004.12, 1203.84
- rho1, rho2, rho3 =
9.92, 8.12, 14.92
assumptions (5)
- domain assumption Homogeneous mixing within each population
- domain assumption Constant recruitment rates greater than natural mortality
- domain assumption No PEP or PrEP for free-ranging dogs
- standard math Validity of the next-generation matrix and Metzler matrix stability criteria
- standard math The Lyapunov construction used in Appendix A
invented entities (1)
-
Environmental rabies virus compartment M(t)
Cite this review
Pith. "Pith review of A Mathematical and Optimal Control Model for Rabies Transmission Dynamics Among Humans and Dogs with Environmental Effects." pith.science (2026). https://pith.science/paper/Q2H7VPI6
@misc{pith2026250708895,
author = {Pith},
title = {Pith review of: A Mathematical and Optimal Control Model for Rabies Transmission Dynamics Among Humans and Dogs with Environmental Effects},
year = {2026},
howpublished = {\url{https://pith.science/paper/Q2H7VPI6}},
note = {Machine review of arXiv:2507.08895}
}
abstract
This study presents a deterministic model to investigate rabies transmission dynamics, incorporating environmental effects and control strategies using optimal control theory. Qualitative and quantitative analyses reveal that the disease-free equilibrium is stable when the effective reproduction number $\mathcal{R}_e < 1$, and unstable when $\mathcal{R}_e > 1$. Mesh and contour plots illustrate an inverse relationship between $\mathcal{R}_e$ and control strategies, including dog vaccination, health promotion, and post-exposure treatment. Increased intervention reduces transmission, while higher contact rates among dogs raise $\mathcal{R}_e$. Numerical simulations with optimal control confirm the effectiveness of integrated strategies. Vaccination and treatment are identified as key interventions for achieving rabies elimination within five years.
Figures
Figures from the paper (17 more)
Reference graph
Works this paper leans on
-
[3]
M. Charles, V. G. Masanja, D. F. M. Torres, S. G. Mfinanga, G. A. Lyakurwa, Mathematical model to assess the impact of contact rate and environment factor on transmission dynamics of rabies in humans and dogs, Heliyon 10 (11) (2024) e32012. doi:10.1016/j.heliyon.2024.e32012. 31
- [1]
-
[2]
S. Ghai, T. Hemachudha, Continued failure of rabies elimination—consideration of challenges in applying the one health approach, Frontiers in Veterinary Science 9 (2022) 847659. doi:10.3389/fvets.2022. 847659
-
[4]
H. B. Baghi, R. Rasizadeh, J. S. Nahand, A. Shamekh, H. Fathi, A debate on elimination of dog- mediated human rabies in developing countries by 2030, Higieny i Medycyny Do´ swiadczalnej 78 (1) (2024) 5–12. doi:10.2478/ahem-2023-0026
-
[5]
A. R. Fooks, A. C. Jackson, Rabies: scientific basis of the disease and its management, Academic Press, 2020
work page 2020
-
[6]
S. Cleaveland, S. Thumbi, M. Sambo, A. Lugelo, K. Lushasi, K. Hampson, F. Lankester, Proof of concept of mass dog vaccination for the control and elimination of canine rabies, Revue scientifique et technique (International Office of Epizootics) 37 (2) (2018) 559
work page 2018
-
[7]
D. Minja, Tanzania: Ministry to raise awareness on rabies disease, Tanzania Daily News (Dar es Salaam)Accessed on June 26, 2025 (September 20 2023). URL https://allafrica.com/stories/202309200038.html?utm_source=chatgpt.com
-
[8]
C. E. Rupprecht, R. S. Mani, P. P. Mshelbwala, S. E. Recuenco, M. P. Ward, Rabies in the tropics, Current Tropical Medicine Reports 9 (1) (2022) 28–39. doi:10.1007/s40475-022-00257-6
Show all 35 references
-
[9]
O. C. Akinsulie, O. O. Adebowale, R. O. Adesola, O. G. Banwo, I. Idris, S. C. Ogunleye, O. Fasakin, A. Bakre, I. P. Oladapo, V. A. Aliyu, et al., Holistic application of the one health approach in the prevention and control of rabies: plausible steps towards achieving the 2030...
2024 doi
-
[10]
Velasco-Villa, L
A. Velasco-Villa, L. E. Escobar, A. Sanchez, M. Shi, D. G. Streicker, N. F. Gallardo-Romero, F. Vargas-Pino, V. Gutierrez-Cedillo, I. Damon, G. Emerson, Successful strategies implemented to- wards the elimination of canine rabies in the western hemisphere, Antiviral research 1...
2017 doi
-
[11]
C. O. Odero, P. M. Kitala, P. B. Gathura, Estimates of human cases of dog bites and rabies and economic burden of dog bites in Homa Bay County, Kenya, East African Medical Journal 99 (11) (2022)
2022
-
[12]
Hampson, L
K. Hampson, L. Coudeville, T. Lembo, M. Sambo, A. Kieffer, M. Attlan, J. Barrat, J. D. Blanton, D. J. Briggs, S. Cleaveland, P. Costa, C. M. Freuling, E. Hiby, L. Knopf, F. Leanes, F.-X. Meslin, A. Metlin, M. E. Miranda, T. M¨ uller, L. H. Nel, S. Recuenco, C. E. Rupprecht, C....
2015 doi
-
[13]
Changalucha, R
J. Changalucha, R. Steenson, E. Grieve, S. Cleaveland, T. Lembo, K. Lushasi, G. Mchau, Z. Mtema, M. Sambo, A. Nanai, N. J. Govella, A. Dilip, L. Sikana, F. Ventura, K. Hampson, The need to improve access to rabies post-exposure vaccines: Lessons from Tanzania, Vaccine 37 (2019...
2019 doi
-
[14]
C. S. Bornaa, B. Seidu, M. I. Daabo, Mathematical analysis of rabies infection, Journal of Applied Mathematics 2020 (1) (2020) 1804270. doi:doi.org/10.1155/2020/1804270
2020 doi
-
[15]
E. G. Hudson, V. J. Brookes, S. D¨ urr, M. P. Ward, Modelling targeted rabies vaccination strategies for a domestic dog population with heterogeneous roaming patterns, PLOS Neglected Tropical Diseases 13 (7) (2019) e0007582. doi:doi.org/10.1371/journal.pntd.0007582
2019 doi
-
[16]
Abdulmajid, A
S. Abdulmajid, A. S. Hassan, Analysis of time delayed rabies model in human and dog populations with controls, Afrika Matematika 32 (5) (2021) 1067–1085. doi:10.1007/s13370-021-00882-w
2021 doi
-
[17]
D. D. Hailemichael, G. K. Edessa, P. R. Koya, et al., Effect of vaccination and culling on the dynamics of rabies transmission from stray dogs to domestic dogs, Journal of Applied Mathematics 2022 (2022). doi:10.1155/2022/2769494
2022 doi
-
[18]
Castillo-Chavez, Z
C. Castillo-Chavez, Z. feng, W. huang, On the computation of ro and its role on, Mathematical Ap- proaches for Emerging and Reemerging Infectious Diseases: An Introduction 1 (2002) 229. 32
2002
-
[19]
Stoddart, Existence of optimal controls, Pacific Journal of Mathematics 20 (1) (1967) 167–177
A. Stoddart, Existence of optimal controls, Pacific Journal of Mathematics 20 (1) (1967) 167–177. doi:10.2140/pjm.1967.20.167
1967 doi
-
[20]
T. D. Sagamiko, N. Shaban, C. L. Nahonyo, O. D. Makinde, Optimal control of a threatened wildebeest- lion prey-predator system in the Serengeti ecosystem, Open Journal of Ecology 5 (4) (2015) 110–119. doi:10.4236/oje.2015.54010
2015
-
[21]
C. J. Silva, C. Cruz, D. F. M. Torres, A. P. Mu˜ nuzuri, A. Carballosa, I. Area, J. J. Nieto, R. Fonseca- Pinto, R. Passadouro, E. a. S. d. Santos, W. Abreu, J. Mira, Optimal control of the COVID-19 pandemic: controlled sanitary deconfinement in Portugal, Scientific Reports 11...
2021
-
[22]
Yamada, K
A. Yamada, K. Makita, H. Kadowaki, N. Ito, M. Sugiyama, N. C. Kwan, K. Sugiura, A comparative re- view of prevention of rabies incursion between Japan and other rabies-free countries or regions, Japanese journal of infectious diseases 72 (4) (2019) 203–210. doi:10.7883/yoken.J...
2019 doi
-
[23]
S. W. Teklu, Impacts of optimal control strategies on the HBV and COVID-19 co-epidemic spreading dynamics, Scientific Reports 14 (1) (2024) 5328. doi:10.1038/s41598-024-55111-8
2024 doi
-
[24]
L. S. Pontryagin, V. G. Boltyanskii, R. V. Gamkrelidze, E. F. Mishchenko, The mathematical theory of optimal processes, The Macmillan Company, New York, 1964
1964
-
[25]
Campos, C
C. Campos, C. J. Silva, D. F. M. Torres, Numerical optimal control of HIV transmission in Oc- tave/MATLAB, Math. Comput. Appl. 25 (1) (2020) Paper No. 1, 20. doi:10.3390/mca25010001
2020 doi
-
[26]
H. Tian, Y. Feng, B. Vrancken, B. Cazelles, H. Tan, M. S. Gill, Q. Yang, Y. Li, W. Yang, Y. Zhang, et al., Transmission dynamics of re-emerging rabies in domestic dogs of rural China, PLoS Pathogens 14 (12) (2018) e1007392. doi:10.1371/journal.ppat.1007392
2018 doi
-
[27]
Zhang, Z
J. Zhang, Z. Jin, G.-Q. Sun, T. Zhou, S. Ruan, Analysis of rabies in China: transmission dynamics and control, PLoS one 6 (7) (2011) e20891. doi:10.1371/journal.pone.0020891
2011 doi
-
[28]
W. H. Organization, et al., Working to overcome the global impact of neglected tropical diseases: first WHO report on neglected tropical diseases, World Health Organization, 2010
2010
-
[29]
W. H. Organization, WHO expert consultation on rabies: second report, Vol. 982, World Health Orga- nization, 2013
2013
-
[30]
K. M. Addo, An SEIR mathematical model for dog rabies; case study: Bongo district, Ghana, Ph.D. thesis, Kwame Nkrumah University of Science and Technology (2012)
2012
-
[31]
Hampson, F
K. Hampson, F. Ventura, R. Steenson, R. Mancy, C. Trotter, L. Cooper, B. Abela-Ridder, L. Knopf, M. Ringenier, T. Tenzin, et al., The potential effect of improved provision of rabies post-exposure prophylaxis in Gavi-eligible countries: a modelling study, The Lancet Infectious...
2019 doi
-
[32]
Ruan, Spatiotemporal epidemic models for rabies among animals, Infectious disease modelling 2 (3) (2017) 277–287
S. Ruan, Spatiotemporal epidemic models for rabies among animals, Infectious disease modelling 2 (3) (2017) 277–287. doi:10.1016/j.idm.2017.06.001
2017 doi
-
[33]
T. T. Yusuf, A. Abidemi, A. S. Afolabi, E. J. Dansu, Optimal control of the coronavirus pandemic with impacts of implemented control measures, Journal of the Nigerian Society of Physical Sciences (2022) 88–98 doi:10.46481/jnsps.2022.414
2022 doi
-
[34]
M. A. Stephano, J. I. Irunde, M. M. Mayengo, D. Kuznetsov, The significance of stochastic ctmc over deterministic model in understanding the dynamics of lymphatic filariasis with asymptomatic carriers, Journal of Applied Mathematics 2024 (1) (2024) 2130429. doi:10.1155/2024/2130429
2024 doi
-
[35]
G¨ um¨ u¸ s, S
M. G¨ um¨ u¸ s, S. W. Teklu, Cost-benefit and dynamical investigation of a fractional-order corruption popu- lation dynamical system, Fractal and Fractional 9 (4) (2025) 207.doi:10.3390/fractalfract9040207. 33
2025 doi
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.