Managing infectious diseases over connected populations: a non-convex optimal control

The paper develops an optimal control model to analyse various management options for infectious diseases that occur in metapopulations, under both Nash and cooperative behaviour. As pathogens are renewable resources with negative value, the problem may be non-convex. Since the disease can be transmitted across various connected populations, externalities are involved. Both aspects deserve attention as two issues arise: a) is eradication of the disease in finite time preferable to indefinite treatment? b) are cooperative solutions well-behaved? The problem is solved numerically and the results indicate that while eradication is likely to be an optimal strategy when initial levels of infections are relatively low, the internalisation of between-population externalities (as indicated by the first order necessary conditions of the cooperative optimal control problem) might not always be possible. Also, ignoring these two aspects can lead to inadequate policy design.