Punctuated interventions, such as SIAs, necessarily result in
fluctuating dynamics – e.g. the proportion susceptible, disease burden, or risk
of outbreak fluctuates in time – but this fluctuation can, in principle, be
predicted if the underlying rates (birth, vaccination, transmission) are known
and remain constant. Recognizing that local variation in demographic rates may
require locally tailored intervention frequency that is difficult to fully
asses a priori, or that population
and transmission dynamics may change in unpredictable ways in between scheduled
SIAs, Lessler et al Lessler 2016 suggested that campaigns may be triggered by
observable metrics, such as serological surveys, that are indicative of high
outbreak risk. This analysis suggests
that triggered campaigns can be more effective at reducing measles burden that
relying on a pre-determined SIA schedule.
Given the costs in both the necessary surveillance system with which to
trigger such campaign and the reactive implementation of campaigns, dynamic
models are a useful tool for a priori
assessment of the potential impact of such programs. Here, the interventions themselves are tied
to the dynamics of the system. Thus,
understanding of the expected dynamics, i.e. trends or cycles, and the
variation in those dynamics, i.e. stochastic fluctuations, is critical to
develop reactive systems that meet operational needs for sensitivity and
specificity. By comparison, prior to the
introduction of a conjugate vaccine, reactive vaccination campaigns for
outbreaks of meningococcal meningitis in the African meningitis belt were
triggered when reported cases increased above an operational outbreak
threshold; high thresholds result in slow response with low potential impact to
limit the outbreak, low thresholds trigger frequent, costly interventions in
locations that may not necessarily experience large outbreaks, which draws
resources away from other areas in need. Defining operational thresholds
required an understanding of the temporal variation in outbreak risk,
meningitis outbreaks are highly seasonal and are correlated with environmental
conditions, and the stochastic nature of outbreak spread Lewis 2001.