Strategies to improve global influenza surveillance: A decision tool for policymakers

Background: Global pandemic influenza preparedness relies heavily on public health surveillance, but it is unclear that current surveillance fully meets pandemic preparedness needs.
Methods: We first developed a conceptual framework to help systematically identify strategies to
improve the detection of an early case or cluster of novel human influenza disease during the pre-pandemic
period. We then developed a process model (flow diagram) depicting nine major pathways through which
a case in the community could be detected and confirmed, and mapped the improvement strategies onto
this model. Finally, we developed an interactive decision tool by building quantitative measures of
probability and time into each step of the process model and programming it to calculate the net
probability and time required for case detection through each detection pathway. Input values for each
step can be varied by users to assess the effects of different improvement strategies, alone or in
combination. We illustrate application of the tool using hypothetical input data reflecting baseline and 12-
month follow-up scenarios, following concurrent implementation of multiple improvement strategies.
Results: We compared outputs from the tool across detection pathways and across time, at baseline and
12-month follow up. The process model and outputs from the tool suggest that traditional efforts to build
epidemiology and laboratory capacity are efficient strategies, as are more focused strategies within these,
such as targeted laboratory testing; expedited specimen transport; use of technologies to streamline data
flow; and improved reporting compliance. Other promising strategies stem from community detection –
better harnessing of electronic data mining and establishment of community-based monitoring.
Conclusion: Our practical tool allows policymakers to use their own realistic baseline values and
program projections to assess the relative impact of different interventions to improve the probability and
timeliness of detecting early human cases or clusters caused by a novel influenza virus, a possible harbinger
of a new pandemic. Policymakers can use results to target investments to improve their surveillance
infrastructure. Multi-national planners can also use the tool to help guide directions in surveillance system
improvements more globally. Fi