Nilpotent singularities and dynamics in an SIR type of compartmental model with hospital resources (Q2634268)

The authors study the SIR model \[ \begin{aligned} \frac{dS}{dt} &= A-dS-\frac{\beta SI}{S+I+R},\\ \frac{dI}{dt} &= -\left( d+\nu\right) I-\mu\left( b,I\right) +\frac{\beta SI}{S+I+R},\\ \frac{dR}{dt} &= \mu\left( b,I\right) -dR, \end{aligned} \] where \(S\) is the number of susceptible individuals, \(I\) is the number of infected individuals, \(R\) is the number of recovered individuals, \(A\) is the recruitment rate, \(\beta\) is the contact transmission rate, \(\nu\) is the per capita disease-induced death rate and \(d\) is the per capita natural death rate. Contrary to classic epidemic models where the per capita recovery rate \(\mu\) is constant, it is assumed in the paper that \(\mu\) is a function of available medical resource \(b\) and the number of infected individuals \(I.\) More precisely, the per capita recovery rate is selected to satisfy five conditions corresponding to limited availability of medical resource and several possible scenarios with the consequent decrease or increase of the number of infected individuals. Starting with a rational function \[ \mu\left( b,I\right) =\mu_{0}+\left( \mu_{1}-\mu_{0}\right) \frac {DI+E}{I^{2}+BI+C}, \] where constants \(B,\) \(C,\) \(D,\) and \(E\) depend on the available medical resource \(b\) and a certain parameter \(p\in\left( 0,1\right) ,\) after some simplifying assumptions, the authors finally choose \[ \mu\left( b,I\right) =\mu_{0}+\left( \mu_{1}-\mu_{0}\right) \frac{3b^{2} }{I^{2}+3b^{2}}. \] Due to the presence of the nonlinear recovery rate, the dynamics of the system becomes much more complex. It is noted that small variations of the available medical resource \(b\) result in structural changes in the system's dynamics. The authors conclude that for sufficiently large values of \(b,\) there exist no periodic solutions with a small amplitude whereas large amplitude oscillations provide a more reasonable explanation for disease recurrence. An elaborated bifurcation analysis combining analytic tools and numeric simulations reveals presence of multi-steady states and multi-periodic regimes, as well as transcritical, sadle-node, cusp, and Hopf bifurcations and Bogdanov-Takens bifurcations of codimensions 2, 3, and 4. It is argued that if the basic reproduction number \[ \mathcal{R=}\frac{\beta}{d+\nu+\mu_{1}}<1, \] where \(\mu_{1}=\lim_{I\rightarrow0+}\mu\left( b,I\right) ,\) the disease is eradicated for the values of \(b\) that exceed a certain threshold value \(b^{+}\left( \mu_{1}\right) \) determined in the paper. However, if \(\mathcal{R>}1,\) even substantial increase in available medical resource \(b\) cannot eradicate the disease resulting only in the reduction of the number of infected individuals.