Study design
The expert elicitation workshop integrated current knowledge, structured
the thinking and enabled a coherent set of hypotheses to be defined
which could feasibly be investigated. Eight of these hypotheses were
subsequently identified for investigation by this study. These
hypotheses effectively clustered into two areas: population reservoirs
of AIV (commercial and backyard chickens and ducks) and system linkages
(environmental sampling in LBMs and collector yards; traders as links
between enterprise types; movements and contacts between poultry
populations). Due to the complexity of designing one study which could
accommodate all these aspects, in combination with the challenge of
investigating a disease agent which is maintained at low levels in the
source poultry populations, the design was complex, with multiple
strands of investigation being incorporated. The Phase 1 surveys led to
a better description and understanding of the structure of the poultry
production system as well as the value chain. Social Network Analysis
provided further evidence on contact structures which may influence
transmission, although it is unfortunate that the nomadic duck contact
structure could not be determined. Subsequently, the Phase 2 active
disease sampling aimed to determine if, when and where AIV could be
detected in an endemic context. An approach combining these
methodologies in a mixed methods framework allowed these strands of
investigation to be drawn together, enabled several of the study
hypotheses to be refined, and could inform further discussion.
Triangulation strengthened the external validity of the combined results
and enabled a synthesis of these disparate outputs to form a coherent
narrative.