Another look at silent circulation of poliovirus in small populations

Infect Dis Model. 2018 Jun 9;3:107-117. doi: 10.1016/j.idm.2018.06.001. eCollection 2018.


Background: Silent circulation of polioviruses complicates the polio endgame and motivates analyses that explore the probability of undetected circulation for different scenarios. A recent analysis suggested a relatively high probability of unusually long silent circulation of polioviruses in small populations (defined as 10,000 people or smaller).

Methods: We independently replicated the simple, hypothetical model by Vallejo et al. (2017) and repeated their analyses to explore the model behavior, interpretation of the results, and implications of simplifying assumptions.

Results: We found a similar trend of increasing times between detected cases with increasing basic reproduction number (R0) and population size. However, we found substantially lower estimates of the probability of at least 3 years between successive polio cases than they reported, which appear more consistent with the prior literature. While small and isolated populations may sustain prolonged silent circulation, our reanalysis suggests that the existing rule of thumb of less than a 5% chance of 3 or more years of undetected circulation with perfect surveillance holds for most conditions of the model used by Vallejo et al. and most realistic conditions.

Conclusions: Avoiding gaps in surveillance remains critical to declaring wild poliovirus elimination with high confidence as soon as possible after the last detected poliovirus, but concern about transmission in small populations with adequate surveillance should not significantly change the criteria for the certification of wild polioviruses.

Keywords: AFP, acute flaccid paralysis; CFP, case-free period; CNC, confidence about no circulation; CNCx%, time when the confidence about no circulation exceeds x%; DEFP, detected-event-free period; OPV, oral poliovirus vaccine; POE, Probability of eradication; Polio; Silent circulation; Small populations; Stochastic modeling; TBC, time between detected cases; TUC, time of undetected circulation after the last detected-event; TUCx%, xth percentile of the TUC; WPV, wild poliovirus.