The design of a survey questionnaire to measure perceptions and behaviour during an influenza pandemic: the Flu TElephone Survey Template (FluTEST)

Abstract

BackgroundDuring the 2009–10 influenza (flu) pandemic, surveys to assess behaviour among the general public were designed quickly and suffered from methodological deficits as a result. To facilitate survey work in a future pandemic we (1) identified variables relating to behaviour, perceptions and presence of symptoms that were of relevance to policy-makers and other public health experts; (2) tested and refined the wording of questions to measure these variables; (3) assessed the reliability of responses to these questions; and (4) tested whether non-response bias due to attrition might prevent the use of a longitudinal design for future pandemic-related surveys.ObjectiveTo design, test and refine a set of questions to assess perceptions and behaviours in relation to a pandemic flu outbreak.MethodWe identified variables via existing systematic reviews and through consultation with pandemic flu planners from Public Health England, the English Department of Health, their advisory groups and academic colleagues. We adapted questions from existing scales or developed them afresh, and tested their clarity in three rounds of qualitative interviews with members of the public (totaln = 78). We used a random-digit dial telephone survey of adults from Great Britain (n = 1080) to assess the internal reliability of scales. We used a follow-up survey 1–2 weeks later to assess the test–retest reliability of responses and the differences between responders (n = 621) and non-responders (n = 459).ResultsWe identified seven core sets of outcome variables relating to the presence of flu-like illness and to various protective behaviours, as well as a set of likely predictor variables for the behaviours. Qualitative interviews identified multiple issues with our questions, most of which we resolved. Reliability of the items was largely satisfactory. Evidence of non-response bias was found, with non-responders being younger and less well educated than responders, and differing on several flu-related variables.ConclusionsIt would be ill-advised for public health bodies to enter the next pandemic without a plan for how to measure the public’s behaviours and perceptions. The extensive set of items that we compiled as part of this work has the benefit of being evidence based, policy relevant and readily understood. Although choosing how to gather data still requires consideration, these items can be used with confidence as soon as the next pandemic begins. Future work should consider the most appropriate method for conducting surveys using these items.Study registrationCurrent Controlled Trials ISRCTN40930724.FundingThis project was funded by The National Institute for Health Research (NIHR) Health Services and Delivery Research programme and will be published in full inHealth Services and Delivery Research; Vol. 2, No. 41. See the NIHR Journals Library website for further project information.