Method
Design and Recruitment
Semi-structured focus groups and interviews were conducted with a diverse sample of the general public. To ensure that participants were from a range of social and ethnic backgrounds we recruited from a variety of organizations in London and Southampton including children's centres, AgeUK lunch clubs, community centres, students from a university, voluntary organisations and support groups for those with underlying conditions such as diabetes, COPD (Chronic Obstructive Pulmonary Disease) and PSC (Primary Sclerosing Cholingitis). Advertisements were placed in these centres explaining the purpose of the study, who was eligible, how to participate and offering a small monetary compensation for participation. The managers of the centres where interviews were held advertised the study and made rooms available for the focus groups to take place.
Ethical approval for the study was granted by University College London (Reference: 5081/001) and the University of Southampton (Reference: 7387) ethics committees.
Sample
Sampling was purposeful and individuals who varied in their risk status were recruited. Of the 71 participants, 23 were men and 48 were women; Details of the demographic profile are shown in Table 1. Thirty-five were from designated at-risk groups of whom 10 had an underlying condition, and six were pregnant. Of the 36 participants not designated as being at-risk, nine were specifically recruited because they were mothers with young children. Thirty-eight of the participants were vaccinated for seasonal influenza regularly (of whom 20 were from clinical at-risk groups) and two had been vaccinated for seasonal influenza for the first time this year. Eighteen people who did not consider themselves to be at risk had been vaccinated at least once before for seasonal influenza. Reasons for being vaccinated among those who were designated as not being at high risk included recommendation by a GP, and being offered the vaccine at work. 12 participants had received monovalent H1N1 vaccine and three had antiviral medicines during the 2009–2010 pandemic.
It should be noted that groups were not always mutually exclusive. For example, some individuals who had been recruited as 'elderly' (over 65 years of age) also reported that they had other underlying conditions that would put them in another at-risk category.
Materials
Two scenarios were developed: an uncertain and a severe scenario. The severe scenario was based on that used by Teasdale and Yardley (2011) which described a severe level of risk, severe health consequences and the national impact of the pandemic. The uncertain scenario was developed to reflect the early conditions that occurred during the 2009/10 pandemic. This described an uncertain situation, uncertain health consequences and uncertain public impact of the pandemic (see Table 2).
Short messages promoting the uptake of vaccinations and antiviral medicines for pandemic influenza were developed to reflect evidence from prior research that identified barriers to uptake but also to reflect the key drivers of behaviour as defined in the COM-B framework. These were presented as advice from official sources (see Table 3).
Procedure
Data collection took place in London and Southampton from November 2013 to March 2014. Nine focus groups, three paired interviews and six individual interviews were conducted by the first two authors at the centres from which participants were recruited. Written informed consent was obtained from all participants who received a small monetary compensation for their involvement. Interviews lasted between 20 and 65 minutes and were audio recorded with the participants' consent.
An interview schedule structured into two sections was used to guide the discussion. The first section was to establish what participants knew about pandemic influenza, vaccinations, and antiviral medicines for pandemic influenza, and personal experiences of pandemic influenza. The second section focused on responses to two scenarios and advice concerning vaccinations and antiviral medicines. Participants were asked to imagine that they were in a given situation and to consider what they would think, feel and do if this were to occur. The Uncertain scenario (Table 2) was always shown first, followed by the advice about antiviral medicines (Table 3). The Severe scenario (Table 2) was shown second followed by the advice on vaccinations (Table 3), and then antiviral medicines. All participants were debriefed in full at the end of the interview and reassured that these were fictional scenarios.
Data Analysis
Audio recordings were transcribed verbatim and NVivo 10 was used to code and to maintain a trail of memo and theme development. Analysis was iterative and each transcript was read and re-read numerous times by the first two authors independently. Transcripts were coded line by line and analysed comparatively to identify similarities and differences. A data audit was conducted by the first two authors to clarify meanings, remove duplicated codes and identify data that did not match the coding scheme.
Inductive analysis was used to identify responses to the uncertain and severe scenarios. Deductive analysis was used to identify facilitators and barriers to following recommended advice to be vaccinated and take antiviral medicines. In addition, code names were assigned to the six COM-B components: physical and psychological capabilities; automatic and reflective motivations, and social and physical opportunities (see Additional files 1 and 2 – code frames http://www.biomedcentral.com/1471-2458/15/190/additional). For the purposes of analysis, the Theoretical Domains Framework was used. This is a variant of the COM-B which subdivides the themes into 14 detailed components that map directly onto COM-B. These are: 'knowledge'; 'skills'; 'memory, attention and decision processes'; 'behavioural regulation'; 'social/professional role and identity'; 'beliefs about capabilities'; 'optimism'; 'beliefs about consequences'; 'intentions'; 'goals'; 'reinforcement 'emotion'; 'environmental context and resources'; and 'social influences'.
The facilitators and barriers to being vaccinated and take antiviral medicines were reviewed separately. Responses to accepting advice were also investigated according to two broad categories – those designated as being in a priority group (35 people – men and women over 65 years, pregnant, underlying illnesses) and those not designated as being in a priority group (36 people – men and women under 65 years, mothers with young children).