Examining the role of patients' experiences as a resource for choice and decision-making in health care: a creative, interdisciplinary mixed-method study in digital health

Conceptual and literature review

We used reviews and secondary analysis to clarify the theoretical mechanisms and the domains of health that we expected to be affected by exposure to online PEx. An approach drawing on ‘realist’ literature review methods was conducted to identify theories, mechanisms of action and the potential impact of PEx and to establish the conceptual and theoretical framework for the whole programme. The review identified seven domains through which access to other people’s experiences of a health condition could affect health; these comprised finding information, feeling supported, maintaining relationships, using health services, changing behaviours, learning to tell the story and visualising illness.

Secondary analysis of narrative interviews

Interviews from a unique Oxford archive of interviews, held by the Health Experiences Research Group (HERG), were used for a qualitative secondary analysis. At the time of the study, the archive included over 60 health conditions, which were sampled to (1) gather evidence through thematic analysis about how and why information based on real experience is sought and used, and (2) select quotations from interview transcripts which illustrate participants’ views on their use of the internet for health information, to inform the item pool for a questionnaire tool.

Development and piloting of the e-Health Impact Questionnaire

Trials to evaluate the impact of online PEx need to identify the most appropriate outcomes to be measured. The range of outcomes is potentially large and may be addressed, in part, by existing measures. However, as research on online patient experiential information was in its infancy at the time this WP commenced, there was no valid and reliable assessment tool to capture the health effects of using websites which contain PEx. We therefore developed the e-Health Impact Questionnaire (eHIQ) to assess health-related websites which contain either experiential or factual information. The tool was developed and tested for use across a range of health groups (e.g. people with long-term conditions, carers and those viewing websites aimed at changing health behaviour) and can also be used for websites that include different styles of online information (e.g. factual or experiential information and discussion forums). This WP was written up by LK as her doctorate (awarded in 2014, supervised by CJ and SZ).

Online ethnographic observations

In this WP, we explored internet behaviour using in-depth observational research methods to establish how different types and formats of online PEx are used and how online PEx material might be optimised. We studied how and why people share experiences of health and illness on the internet in the context of MS. Through a content analysis of online settings, ethnographic observation of online activity and telephone interviews, we studied how people living with MS and those who care for them seek out others’ experiences, how they share their own experiences, and how this affects their understanding of MS, their relationship with health-care practitioners, their health-care decision-making and their general well-being.

Observations and experimental studies

Acknowledging the diverse quality of online PEx, we conducted three studies using observational and internet laboratory café methods to discover how patients find and use PEx to inform health and lifestyle choices. The first study was expanded from that stipulated in the proposal and included, in addition to the asthma and MS carers elements, two further elements with people who wished to stop smoking. The second study, which focused on people with asthma and MS carers, involved the collection of eye-tracking data, to identify where people spend time, what holds their attention and how they organise their searches, and a comprehensive interview. From these studies, we developed a model of patients’ peer-to-peer engagement online, which was used to develop the prototype PEx websites that were used in the feasibility trials in the final WP.

The third study involved a sample of smokers. We sought to examine the effect of (1) message type (i.e. PEx vs. information only), (2) the palatability of material in terms of pre-existing beliefs (attitude towards quitting by using or not using aids and support) and (3) perceived similarity (based on gender) between the reader and the author of the PEx on a participant’s readiness to accept the message. The study adopted a longitudinal approach to examine the potential impact of these variables on behavioural intentions and actions 2 weeks and 3 months after exposure.

Development and testing of patient experience web interventions for exploratory trials

We developed six prototype multimedia websites featuring either PEx information (intervention) or non-PEx factual information (comparator) for each of three exemplar health issues (adults with asthma, carers of people with MS and people considering giving up smoking cigarettes). We used established qualitative research methods, led by experienced qualitative social scientists based in the Oxford team, to collect unstructured, narrative interviews about the experiences, information and support needs of people in each of the three exemplar groups. The design and presentation of PEx information on the websites were developed in accord with the guidelines developed in WP2. The comparator websites were based on non-PEx material from NHS Choices and were presented in a similar design to that of the intervention sites.

Exploratory trials

As originally proposed, we conducted a phase II pilot randomised single-blind trial, in which eligible participants were randomly allocated to a condition-specific PEx-based health information website or to a comparator website which contained no PEx information, stratified by one of three health conditions. The original application described three separate pilot randomised trials, one for each health condition. For operational efficiency, the three trials were run under the governance of a single trial protocol and analysed by condition (smoking cessation, asthma or caring for someone with MS). As this was an exploratory study, our main aim was to establish the feasibility of undertaking this research and to identify any emergent evidence of efficacy or harm. We measured recruitment and retention rates, usage of the intervention and comparator, and any adverse events. We measured the impact of all six websites on health status, and we assessed attitudes towards the websites using the new instrument, the eHIQ. We also included some condition-specific measures as secondary outcomes (such as self-efficacy measures). We also undertook qualitative interviews with 30 purposively selected trial participants to explore why and how they took part and used the websites (or did not).

Literature review

Our approach was informed by realist review, a method for synthesising research evidence regarding complex interventions. 5,6 It is based on the idea that research should pass on collective wisdom about the successes and failures of previous initiatives in particular policy domains ‘before the leap into policy and practice’. 5 Its main purpose is to identify explanations through which complex social programmes might operate so that policy-makers can learn how to introduce or adapt programmes based on a good understanding of why and how they might work.

This approach was suited to our task because the science of the role of information based on online PEx is epistemologically complex and methodologically diverse. 6 However, in our case, there was no set of theories or even defined social or health programmes to identify and evaluate: the field is too new. We therefore adapted the approach to allow us to identify and describe the domains of the territory rather than to develop programme theories of how it may work. Our approach was iterative and collaborative; SZ and SW worked intensively face to face, through e-mails and by telephone over 6 months in 2010. Box 1 summarises the review’s five overlapping steps.

Public panel of health website users

To check our interpretation of the literature, we convened a public panel which consisted of 30 participants primarily recruited through an invitation from Oxfordshire Primary Care Trust sent to a list of people who had previously agreed to help with research (see Chapter 11). We selected the respondents to ensure that they were representative of a cross-section of the community (gender, age, occupation and ethnic group) and included those who had used the internet for heath information in a variety of ways (websites, forums, blogs) and for a variety of health conditions (either for themselves or on behalf of family and friends).

We first showed examples of health information with and without the inclusion of ‘personal experiences of health and illness’. The panel was divided into four groups, each with a rapporteur, and asked to think about how people might be positively and negatively affected by experiential information. The rapporteurs delivered the feedback to the whole group. In a final plenary session, we summarised the results of the conceptual review and discussed how they compared with the results of the group discussions. A final questionnaire asked participants what, in their view, were the most important ways in which experiences might affect people, both positively and negatively, from the perspective of someone with a condition and that of a caregiver of someone with a condition. A report was circulated and feedback was sought.

Final review and revision

During our final search for relevant papers we presented our findings at a specialists’ workshop at the 25th British Computing Society conference on Human–Computer Interaction, July 2011, Northumbria University, Newcastle upon Tyne, as part of the workshop ‘Online patient experience (PEx) and its role in e-health’, organised by the authors. During discussion, we identified the importance of the visual presentations of online PEx. We had not identified this topic in our reading but on realising its importance we included it in our matrix. Finally, we agreed on the seven ‘domains’ of our conceptual map of the potential health effects of online PEx.

The potential health effects of seeing and sharing experiences online: seven domains

Our approach allowed us to identify seven domains which broadly map the role of online PEx:

1. finding information

2. feeling supported

3. maintaining relationships with others

4. experiencing health services

5. learning to tell the story

6. visualising disease

7. affecting behaviour.

For each of these we considered the potential for both positive and negative effects.

Stage 1: preliminary analysis

Given the range of conditions contained in the archive, it was not possible to undertake an exhaustive analysis of all available interviews. Key findings of HERG interview collections are published as online resources (www.healthtalk.org). These website sections were explored to gain an insight into the topics that the primary researchers had identified in each interview collection and to help select which interviews to use in the secondary analysis. We decided, in the first instance, to focus our analysis on those deemed to have closest relevance to the three exemplar conditions being investigated by the iPEx programme (smoking cessation, asthma and carers of people with MS). To maximise the chances of finding information about internet technologies, only interviews published from 2005 onwards were included. Based on this, seven complete collections and a further nine ‘pilot’ interviews with people affected by MS were included in the first round of analysis (Table 1). These were selected through discussions with key members of the project team, including those who were very familiar with the content of the HERG archive.

With the exception of the MS-related interviews, thematic summaries and interview extracts are available online for all of these conditions (see www.healthtalk.org). The summaries contain sections of direct relevance to the secondary analysis (e.g. ‘information and support’, ‘support and support groups’). The interviews were read (by FM or LK and in many cases by both) and all information relating to internet use and/or sharing or seeking experiences (online and off) was extracted. Scoping summary reports were compiled to identify the extent to which the collections had content relating to using the internet for health. A list of key emergent themes and associated statements (examples of the exact wording and language used by interviewees) was created for each condition. This list was subjected to an additional round of constant comparison coding, and key cross-cutting themes and areas for further investigation were summarised. This summary was not intended to be an exhaustive assessment of all the interviews. Instead, alongside the conceptual review (WP1a; see Chapter 2), it was used to highlight areas of interest for more in-depth secondary analysis and to index interviews and excerpts for the development of the assessment tool (WP1c; see Chapter 4).

The sample design and suitability of the data was further informed by consulting the original researchers of the selected interview collections. The researchers (most of whom are still employed in the HERG) provided input and advice based on their experiences of the collections they had created and their expertise in terms of online PEx more generally, which was incorporated into the analysis. Following discussions about the suitability of an interview collection for the research purpose, the original researchers provided coding reports [generated at the time of their own research using either the NUD*IST or NVivo (both QSR International, Warrington, UK) qualitative analysis software] that they believed would be particularly relevant to the topic. Coding reports took the form of electronic reports of sections of interview transcripts which related to a specific topic and were subsequently reviewed to inform the sample selection. Examples of coding reports were entitled ‘Web forums’, ‘Information’, ‘Support groups’ and ‘Information carer’.

The steps outlined above aided familiarisation with the data sets and helped us to assess the suitability of particular interview collections for further analysis. Sampling and selection of interview transcripts were not finalised at the outset, but, as is usual in qualitative analysis, remained continuous throughout the process.

Stage 2: further secondary analysis

Key findings from the preliminary secondary analysis (stage 1)

The conceptual review presented in Chapter 2 drew attention to how practices of seeking and sharing PEx can have numerous and often profound (both positive and negative) consequences for people affected by different conditions. 4 Our secondary analysis of patient interviews reinforced and supported the findings of the review.

The first five domains of how online PEx could affect health (‘information’, ‘feeling supported’, ‘relationships with others’, ‘experiencing health services’ and ‘affecting behaviour’) identified in the review were evident across all the modules analysed. The sixth (‘learning to tell your story’) emerged in interviews with people who had actively shared their experiences online or in other forums. As many more people seek out others’ experiences rather than share their own, this domain was less prevalent than the first five. Interestingly, the seventh domain (‘visualising disease’), although not common across the corpus of interviews, was evident in certain conditions with visible and highly variable symptoms (namely MND and Parkinson’s disease). Given their importance in both our secondary analysis and the extant literature, the first five domains were used as a key structuring device for the development of the assessment tool for measuring the impact of health-related websites. 66 Owing to the lack of research on the sixth and seventh domains, they were selected for further secondary analysis. 55,64,65

In our secondary analysis we found both variations and similarities between how people affected by different conditions used the internet and responded to PEx. This is significant, as a great deal of current research on health-related use of the internet is condition specific and hence unlikely to pick up on wider cross-condition trends and differences. Although it is not possible to provide an exhaustive list of all the factors that influence how people respond to online PEx, we list some particularly notable ones that emerged from our high-level analysis.

The first is the condition in question and, linked to this, the age of the people affected by it. This was especially striking in relation to carers of people with dementia, where, despite the fact that interviewees spoke about benefiting from sharing PEx in face-to-face environments, there was very little evidence of them using online PEx. This may be because interviewees tended to be older and used the internet less, but it may also be related to the nature of the condition and the types of PEx available about dementia. In contrast, young people are particularly active online, seeking and sharing PEx across a variety of technologies and platforms [e.g. patient forums, Facebook (Facebook Inc., Menlo Park, CA, USA), YouTube]. However, in the case of young people with serious chronic conditions, such as cystic fibrosis, interviewees spoke of being upset by and in some cases actively avoiding ‘negative’ or distressing information, including PEx.

It has long been recognised that PEx can be a highly ambivalent source of information with the potential to bring about feelings of both hope and fear, provide support and increase isolation, and be perceived as ‘true’ knowledge or discarded as irrelevant. 31,78 In our secondary analysis we found that this was most definitely the case, with interviewees speaking of having had both ‘positive’ (e.g. feeling ‘supported’, ‘empowered’, ‘inspired’) and ‘negative’ (e.g. feeling ‘depressed’, ‘upset’, ‘devastated’) responses to online PEx. In relation to the former, interviewees said that PEx was an extremely valuable source of information that was simply not available elsewhere. Key areas of information provided by PEx discussed across interviews were practical hints and tips about actually living with a condition, information about diagnosis, prognosis, treatment, side effects, family and personal relations, finances, jokes and humour. Furthermore, PEx affected the health and well-being of interviewees not only by virtue of its informational content (important as this was), but from the sense of support and mutual understanding that was gained from speaking to, sharing with and identifying with other people affected by the same condition. In terms of ‘negative’ responses to PEx, the two most common concerns were fears about the danger of misinformation and the emotional difficulties of dealing with PEx. Interestingly, we found that people with serious chronic or terminal conditions tended to respond especially strongly to PEx, seeing it as an extremely valuable source of information and support, but at the same time often finding others’ experiences very distressing.

It is important to note that what were perceived as ‘negative’ versus ‘positive’ experiences of PEx were not static but, as with other forms of health information, changed over time, sometimes radically. For many people diagnosis was a crucial time for seeking out PEx, while others preferred to wait and seek out PEx once they had adjusted to their new diagnosis. The fact that for many people the internet is available at any time and can be customised to filter and select different PEx was seen as an important feature of the technology. Indeed, interviewees were selective about the source of the PEx they found online (who and where was it coming from), as well as the format and medium through which it was communicated (e.g. text written on a patient forum or a video on YouTube). Thus, the value and, crucially, evaluation of PEx as either ‘good’ or ‘bad’ was highly relational and contingent not only across interviews, but over time, and in relation to different modes of communication. For example, in the case of MND and Parkinson’s disease, a number of interviewees expressed a strong preference for PEx communicated in writing rather than face to face to avoid seeing physical symptoms of the disease that they found upsetting.

Moreover, when interviewees spoke of how online PEx had affected their health and well-being they situated it in relation to their wider health-care environment and informational landscape (both online and offline). In some rare cases health-care practitioners had encouraged patients to look up PEx, occasionally suggesting websites and other sources; in others (again often in relation to more severe conditions), the use of the internet was actively discouraged by medical practitioners. Either way, across all of the collections we analysed, interviewees interpreted and used PEx in conjunction with information they received from numerous different sources, including the information they received from their health-care practitioner(s), NHS services, charitable organisations, the media, family and friends.

Key findings from the further secondary analysis (stage 2, strand 1)

Findings from further secondary analysis for the development of the e-Health Impact Questionnaire (stage 2, strand 2)

As discussed above, it was clear from the analysis that preferences for various types of online health information varied between participants. This was viewed as an individual’s ‘baseline’ acceptance or openness towards using the web for health information which could potentially have an influence on using a specific website. Therefore, a pool of items assessing general attitudes towards online health information was constructed (eHIQ part 1).

To inform the second pool of items relevant to using a specific health-related website (eHIQ part 2), themes derived from the secondary data analysis were mapped onto five of the seven domains identified in the review of literature in WP1a (see Chapter 2). These themes are outlined briefly here, but are discussed further in Kelly et al. 66 The first theme, ‘information’, incorporated learning about a health condition, becoming informed of what to expect and recognising decisions that may need to be made in the future. Health-related websites were also a valuable resource for finding practical tips and advice to help manage health.

The second theme, ‘feeling supported’, related to finding support through others going through similar health experiences. In some cases, accessing a particular health-related website was the first time participants learnt that other people had similar health concerns. Increased hope and confidence was also found through gaining and lending support.

The third theme, ‘relationships with others’, referred to a person feeling less isolated and that others understood what they were experiencing (e.g. in an online community). Using websites also had the potential to ease offline relationships, as these provided an area in which they could discuss their health concerns with people going through similar experiences and learn to articulate what they were feeling.

The fourth theme, ‘experiencing health services’, concerned the use of websites when deciding whether or not medical advice should be sought. The use of the internet could help participants to navigate the health system and helped to prepare people for upcoming decisions. Participants could corroborate advice, find out more information and reflect on decisions made after accessing health services.

The final theme, ‘affecting behaviour’, incorporated the motivational effect some website content may have on the individual. For example, learning of the consequences of ignoring best practice when managing a long-term condition encouraged people to improve their own health-care regime.

The Northumbrian focus group transcripts and the public panel comment forms were subsequently reviewed to ensure that each theme identified had been fully explored and that no additional themes were evident. No new overarching themes were evident; however, the confirmatory sources did provide greater insight into some themes identified. For example, the focus group transcripts provided further insight to the fourth theme, ‘experiencing health services’, as participants discussed the dangers of using misleading information found online instead of consulting a health professional. These concerns were also highlighted in the public panel group.

Overall, the five themes were consistent with the domains identified in the WP1a literature review, providing a strong basis for the conceptual underpinning of the eHIQ part 2 item pool. Although the two remaining themes (‘learning to tell your story’ and ‘visualising disease’) were less pervasive across the transcripts analysed, issues relevant to each domain were dispersed throughout the five confirmed themes. For example, ‘learning to tell your story’ was incorporated into ‘relationships with others’ when a participant discussed being able to communicate more succinctly with their partner after using an online forum. ‘Visualising disease’ was incorporated throughout the five themes identified when participants described images reinforcing people’s stories or messages. It was therefore felt that the issues addressed in domains 6 and 7 were adequately represented in the item pool.

In total, 376 participant statements were extracted from the transcripts to represent all the themes present. LK identified 149 generic statements that could be answered by people across health conditions. Statements were recast as questionnaire items and reduced to 67 items in an iterative process involving LK, SZ and CJ. LK, CJ and SZ added a further 15 items to ensure that all themes were adequately represented. On review, the expert group recommended changes to two items to make them acceptable to people with a low level of health literacy. All members agreed that the items reflected the impact of using health-related websites. Eighty-two items were, therefore, entered into patient interviews and psychometric testing (see Chapter 4 for a full description of the development of the eHIQ).

Participants

Participants were aged ≥ 18 years with access to the internet and living in the UK. To ensure that items were appropriate for inclusion in a generic questionnaire, we included participants with different types of experience (e.g. family carers, people with chronic conditions and people interested in changing a health behaviour).

Recruitment

Ethics

Recruitment methods for stages 1, 2 and 3b were approved by the University of Oxford’s Medical Sciences Division Research Ethics Committee (reference numbers MSD/IDREC/C1/2011/98, MSD/IDREC/C1/2011/77 and MSD/IDREC/C1/2013/063). Stage 3a was approved by the NHS Research Ethics Committee (reference number 12/SW/0209).

Instrument development and testing

In each stage, participants were asked to spend 10–15 minutes browsing a pre-selected health website; these included government websites (e.g. NHS Choices), charity websites (e.g. HealthTalk) and commercial websites (e.g. BootsWebMD). The websites were chosen to test the items with information based on facts, figures and personal experiences and other features such as discussion boards, video clips and ratings.

Stage 1: cognitive interviews

Participants (n = 21) were six men and 15 women with a mean age of 45 years [standard deviation (SD) 16.2 years]. Five were carers and 16 had one of a number of chronic health conditions. Three rounds of cognitive interviewing were carried out to ensure that (1) the instructions were easy to understand and the rubric clearly indicated how participants were supposed to answer items, (2) participants considered items relevant to the topic and acceptable to answer, (3) the response options were appropriate to the item stem and adequately covered the potential range of agreement, and (4) the electronic format was appropriate for use among participants with varying levels of confidence in using computers. Twenty-nine items were deleted and nine items were added, leaving 62 items to enter the next stage.

Two independent item pools were identified to enter psychometric testing. The first item pool contained 23 items on general attitudes towards health websites, while the second item pool contained 39 items asking about attitudes towards a specific health website. All items were given a 5-point response scale (strongly disagree to strongly agree).

Stage 2: pilot questionnaire

The two-part questionnaire was completed by 167 participants. The mean age of the sample was 38.8 years (SD 14.04 years); 125 participants were female and 41 were male (the sex of one participant was unknown) (see Table 4). Checks for floor and ceiling effects confirmed that no item had > 40% of respondents selecting one of the extreme response options and that no item had a large number of missing data (> 10% non-response). Therefore, no items were removed owing to floor or ceiling effects or missing responses.

Seven items (eHIQ part 1) and five items (eHIQ part 2) were iteratively removed owing to poor correlations (< 0.2) with a large number of items in the same item pool and decreasing the overall Cronbach’s alpha value. An EFA was carried out on the remaining eHIQ part 1 items. Five factors with eigenvalues of > 1 were initially extracted. These factors explained 63.85% of the variance. The first two factors were believed to constitute conceptually meaningful factors. These two factors explained 61.06% of the total variance.

In a further separate analysis, the eight items which had been removed following the EFA were examined to investigate the presence of any further conceptually relevant scales. This analysis resulted in a scale of six items which had Cronbach’s alpha value of 0.64. Therefore, 14 items in total for eHIQ part 1 entered stage 3.

An EFA was carried out on the remaining eHIQ part 2 items, resulting in six factors explaining 66.26% of the variance. All factors had a Cronbach’s alpha value of ≥ 0.76. The 34 items asking about using a specific health-related website were, therefore, identified as suitable to enter the next stage of development.

The translatability assessment suggested several minor changes to the instructions and item wording. Changes were suggested to improve consistency with wording, to ensure that all wording was culturally and linguistically relevant to the selected countries and to avoid ambiguities which could jeopardise future translations. Seven minor modifications were made to improve grammar and consistency with wording. Two items were amended to improve their cultural or linguistic suitability and six items were amended to avoid ambiguous wording. Three further items were identified as potentially problematic for translation. These items were:

1. I feel I have a sense of solidarity with other people using the website.

2. I can identify with other people using the website.

3. I feel I have a lot in common with other people using the website.

Although the above items had similar, yet subtly different meanings in the English language, translators were concerned that some languages may not have the capacity to express those differences (e.g. because of fewer adjectives or forms of expression in the target language). Following discussions, we decided to retain the items to capture the intended concept in English.

Stage 3: validation questionnaire

The reduced questionnaire was completed by 170 participants (see Table 4). Stage 3a recruited 47 participants, stage 3b recruited 96 participants and the separate website evaluation study recruited 27 participants. Of the 170 respondents, 59 were men and 107 were women (four unknown). Of those who consented to take part in either stage 3a or 3b, 71.1% (n = 143) completed the full questionnaire.

The items for each questionnaire part were entered into an EFA to examine the questionnaire structure and find the optimal factor solutions. An EFA for items in the eHIQ part 1 confirmed four factors. An item which did not load with other similar items and had poor distributions across response options (i.e. no participant selected the ‘strongly disagree’ response category) was removed. A further EFA suggested three factors; however, two items which formed the third factor were removed due to poor internal consistency (α = 0.59). The two remaining subscales explained 56.58% of the variance with the structure and pattern matrices (Table 5) demonstrating high loadings on their respective factors. The two remaining subscales were entitled ‘attitudes towards online health information’ and ‘attitudes towards sharing health experiences online’. Both subscales demonstrated good internal consistency (≥ 0.77) and were conceptually relevant to overall attitudes towards online health information (Table 6).

An EFA for items in the eHIQ part 2 identified six subscales explaining 66.34% of the variance. One of the six factors consisted of two items which exhibited poor internal consistency (Cronbach’s alpha = 0.50). Inspection of the structure and pattern matrices indicated that one of the two items loaded on another, conceptually relevant, factor, while the remaining item did not. The item which did not load on any other factor was, therefore, removed and all remaining items were entered into a further EFA producing five factors. Reflections on the fourth and fifth factors concluded with a consensus between the team (LK, CJ and SZ) that, grouped together, items did not constitute theoretically sound factors and explained very little variance. The seven items in factors four and five were, therefore, removed and a final EFA confirmed the three remaining factors explaining 61.68% of the variance. The structure and pattern matrices (Table 7) were used to interpret the final factor structure and confirmed high loadings of each item on their respective factor. The three subscales were entitled: ‘Confidence and identification’, ‘Information and presentation’ and ‘Understanding and motivation’. All scales had good internal reliability (≥ 0.80) and are listed in Table 8.

Stage 4: summary score

The factor correlation matrix for eHIQ part 1 was entered into a factor analysis. One factor with an eigenvalue of > 1 was extracted, explaining 76.2% of the variance. No rotation was needed, as all subscales loaded sufficiently on the factor extracted (Table 10, eHIQ part 1). The higher order factor, referred to here as ‘attitudes towards the use of the internet in relation to health’ was assessed for internal consistency achieving a Cronbach’s alpha of 0.78 (n = 169). This supported summing all eHIQ part 1 subscales to create a total index score. After the summed score for the higher-order factor had been standardised to 0–100 metric, the mean summary score was 67.0 [SD 15.04, 95% confidence interval (CI) 64.72 to 69.28; n = 169].

The factor correlation matrix for eHIQ part 2 was subsequently entered into a factor analysis. One factor with an eigenvalue of > 1 was extracted, explaining 63.1% of the variance. No rotation was needed, as all subscales loaded sufficiently on the factor extracted (see Table 10, eHIQ part 2). The higher-order factor, referred to as ‘overall impact of using a specific health-related website’, was assessed for internal consistency, achieving a Cronbach’s alpha of 0.87 (n = 163). The mean summary score was 64.22 (SD 14.20, 95% CI 62.02 to 66.41; n = 163).

Observation and analysis of patient forums and online information sharing

Between June 2011 and July 2012, one researcher (FM) spent extended periods of time (approximately 1–3 hours per day, 3–4 days per week) reading MS-related websites, following news items and analysing forums. She familiarised herself with different platforms and key debates, analysing current as well as historic data within and across different websites. Although this involved moving between websites and sources of information, three forums were chosen as key ‘sites’ to focus on: the MS Society forum (www.mssociety.org.uk/forum); Shift.MS (http://shift.ms); and ThisIsMS (referred to as TIMS by its members) (www.thisisms.com).

The forums were analysed in terms of topics and themes (what people were talking about); the ways in which PEx were articulated and responded to (discourse and imagery, the types of language used, colours, avatars and emoticons), and patterns of communication (e.g. what types of topics received most responses, how many people engaged in different types of conversations, and when, if ever, did moderators intervene). Notes were taken and, where relevant, discussions were copied and stored as text files for further analysis. Rather than limiting our analysis to the patient forums (as is often the case in the analysis of online health forums), links and references to other sources of information, such as biomedical publications or newspaper articles, YouTube videos or blogs, were followed up. In this way the content of the forums was contextualised in relation to broader information practices and topics being debated in the field of MS research and care at the time.

Analysis of YouTube videos

While conducting our research we became aware of the large number of MS-related videos online, many of which contain PEx. As there is relatively little research on YouTube as a source of PEx, and the visual dimensions of PEx were highlighted in both the literature review (see Chapter 2) and the secondary analysis (see Chapter 3), we decided to pursue further research on YouTube.

In January 2012, we searched for the term ‘CCSVI’ using the YouTube search facility. Four thousand videos were returned, and we selected the top 100 most viewed videos for a quantitative and qualitative analysis. A sample of the 15 most viewed videos were analysed by FM, Braden O’Neill and JP. Each researcher developed their own coding scheme and after discussion a final coding scheme was agreed. Next, FM and Braden O’Neill separately coded the full sample of 100 videos. Thirty videos were excluded because either they were not about CCSVI107 or they were not in English (videos with English subtitles were kept). 115 This left 70 videos, with number of views per video ranging from 7103 to 79,956.

The title, channel, number of views, date uploaded and country of origin were recorded for each of the videos. We also categorised videos into two broad groups: ‘patient’ and ‘non-patient’. Videos that focused on the experiences or thoughts of a particular person with MS were categorised as ‘patient’ videos; videos that discussed CCSVI in ways that did not draw on the experiences of individual patients were categorised as ‘non-patient’ videos. Where possible, the gender, type of MS and current medical treatment were additionally recorded for each ‘patient’ video. All videos were then coded in terms of the type of information they provided and whether they portrayed CCSVI positively, negatively, neutrally or ambiguously. Coding was consistent across both coders, with a basic percentage agreement intercoder reliability of 90%. 116 Any discrepancies were resolved through discussion.

In addition, we performed a qualitative thematic analysis on the full sample of ‘patient’ videos (n = 46). Some ‘patient’ videos belonged to a ‘channel’ that contained more videos uploaded by the same YouTube user, often posted over a period of time as a video diary. When this was the case, we analysed the entire channel to contextualise specific videos. We focused on what patients said and how they said it. For each video we noted key emergent themes, transcribed portions of the video and read the comments posted by people who had watched the videos. For further details on the methods used and results please refer to the article published on this work. 84

Interviews

The analysis of online material provides only a partial view of the dynamics of sharing PEx online. It cannot provide insight into people’s motivations for, or perceptions of, sharing their experiences. Moreover, only observing what is available online misses out on the experiences of ‘lurkers’: those who read PEx, but do not post. 117 We used telephone interviews as a method for exploring aspects of PEx that could not be studied through the analysis of online data. The telephone interviews captured the views of two groups of informants: first, people affected by MS (people with MS, and carers) and, second, people who worked on the production or ‘back-end’ of patient forums and online services (e.g. forum moderators, web managers, researchers and designers).

To recruit people affected by MS, the study was advertised on the MS Society, MS Trust and MS Resource Centre research pages, the MS pages of NHS Choices, Shift.MS and a MS National Therapy Centre blog. Regional MS Society groups were e-mailed and asked to circulate information about the study to their user groups. We recruited 17 people affected by MS (five carers and 12 patients) to the study (Table 11). The interviewees included people who had been diagnosed relatively recently and those who had had MS for many years. Some interviewees used the internet and shared PEx prolifically, while others were more circumspect in their use. Three interviewees had undergone venoplasty in response to a diagnosis with CCSVI and one was the mother of a young woman who had had this treatment. In all four of these cases, the interviewees had shared some of their experiences online.

To recruit people who worked on the production and management of patient forums and online services, we made direct contact with representatives from major websites and organisations involved in MS-related PEx. In all cases our informants agreed to take part in an interview (seven interviewees representing five organisations in total; see Table 11).

Both sets of interviews were semistructured and used a common interview protocol that was based on themes that emerged from the earlier analysis of the online material. Each interview lasted for approximately 1 hour. Informed consent was taken and interviews were audio recorded and transcribed verbatim. Constant comparison thematic coding was performed by FM on the interviews with the assistance of the qualitative analysis software tool NVivo. The findings were analysed in conjunction with the online material and observations and discussed at regular intervals with JP and other members of the project team.

Multiple sclerosis-related patient experience on the internet

The MS-related PEx circulating online is extremely large and diverse, ranging from YouTube video diaries and brief postings on Facebook pages through to detailed biographical information provided on blogs. Accounts of PEx are not limited to patient forums, blogs and social networking sites, but permeate the internet, appearing regularly on the websites of pharmaceutical companies, health service providers (including the NHS), charities and patient organisations (e.g. in the UK the MS Society, MS Trust, MS Resource Centre and Shift.MS). Despite this, only a relatively small number of people actually share PEx on the internet as producers of content. The majority are consumers (sometimes referred to as ‘lurkers’): people who draw on other people’s experiences, but do not share their own. As our study focused on online PEx, the majority of our participants had shared their experiences online at some stage, but a few said that they preferred not to do so. The most frequently cited reasons for this were concerns about privacy; fears of being subjected to negative or aggressive responses; the desire to maintain an emotional distance from the condition; and, for carers, concerns about breaking trust or breaching the privacy of their family member. Nonetheless, even participants who preferred not to share their experiences online or expressed concern about the reliability of PEx at times searched for other people’s experiences.

Three inter-related categories and associated practices of PEx emerged as particularly prominent in our analysis of both online and interview material. The first, to adopt Corbin and Strauss’ terminology,118 was PEx dealing with the everyday ‘work’ of living with MS. This includes seeking and sharing experiences about practical concerns and the challenges of living with MS that fall beyond the remit of medicine and health care as strictly defined (e.g. employment, financial concerns, home adaptations, mobility and transport). This type of PEx was frequently mentioned in interviews as a relatively uncontroversial source of very useful information that supplemented biomedical knowledge and expertise.

The second category was PEx directly linked to health-care practices and/or biomedical treatment and research. This takes many forms. It includes the sharing of PEx-based personal experiences and decision-making in relation to treatment and care (e.g. when someone was experiencing particularly troubling symptoms or side effects), discussions about interactions with health-care services and practitioners, and debates on biomedical research. As MS is a degenerative condition for which there are a number of different treatment options, but no cure, it is unsurprising that treatments and research are frequently discussed online. At times, these discussions include an activist component, with people advocating for different theories and treatments, including those that are controversial (such as goat serum, stem cell therapy or long-term antibiotic therapy).

Traditionally, a distinction has been made in social studies of medicine between ‘subjective’ experiences of illness and ‘objective’ biomedical knowledge about disease. 119 However, at least in Western Europe and North America, the majority of people experience illness in relation to some aspect of orthodox biomedicine (even if that relationship is an antithetical one). This is reflected in MS-related PEx on the internet where what might be more narrowly conceived of as ‘medical’ information is combined and supplemented with ‘experiential’ information, and vice versa, and personal accounts often include discussions on biomedical disease models, research and treatments. Despite this blurring, in both interviews and online comments, distinctions were frequently made between ‘facts’ and ‘opinions’, with people frequently caveating what they said with statements such as ‘This is just my subjective opinion’; ‘I know this is not scientific, but . . .’; and ‘I’m not a medical practitioner, so please check this with your doctor.’

Yet there was a strong sense (espoused online and in interviews) that individual, personal experiences of living with MS were a valuable source of knowledge and constituted a distinct way of knowing and source of knowledge that needed to be taken into consideration. In most cases this ‘experiential’79 or ‘patient knowledge’120 was seen as (by interviewees and in online comments) as complementary to rather than as a challenge to medical knowledge and traditional medical expertise. However, some patients did draw on their experiences to question or even directly challenge biomedical theories, treatments and practitioners (discussed in more depth in YouTube and the sharing of patient experience in relation to chronic cerebrospinal venous insufficiency).

The third category and associated practices of PEx shared online (very much inter-related and inseparable from both of the above) centred on self-expression, community building and mutual support. The focus here was less on imparting specific information (although this was included) than on emotional and creative expression. This included, for example, initial forums posts in which people introduced themselves to the community, and creative writing based on experiences of MS. These types of interaction differed between platforms and spaces, and in many cases were linked to forum cultures that encouraged particular types of engagement.

We have separated these three categories for analytical purposes only. In practice one online post will often combine elements of all three. Indeed, we contend that it is precisely through the interweaving of lived experience, biomedical knowledge, creative expression and emotional support in how it is constructed that makes PEx so valuable, and distinguishes it from other sources of health-related information and support. However, although the blurring of categories is a key characteristic of PEx, we found that these three categories and their various combinations were not equally distributed online. For example, with regard to the three forums we analysed, one slanted toward non-mainstream MS research and treatments, whereas another was geared towards providing a platform for sharing everyday experiences of living with MS and peer support, and the third specifically encouraged creative engagement (e.g. through shorts films and video clips). This highlights the role that the design and management of online spaces for sharing PEx play in how and what PEx is shared on the internet, a topic that has been relatively understudied in e-health research and which we discuss in more depth below in the specific context of MS.

Constructing patient experience: the role of the design and management of online spaces

In the early days of the internet, the most common sites for sharing PEx were blogs and text-based health forums. In a relatively short space of time, social media platforms such as Facebook and YouTube have emerged, and e-health and tele-health services have grown and diversified. Consequently, how and what PEx is shared online has changed radically. ‘Traditional’ text-based health forums still exist and continue to flourish, but they operate as part of a complex ‘polymedia’ landscape in which different mediums and platforms are situated in relation to each other. 121 Users can choose between platforms and technologies, while service providers learn from and adapt in relation to each other. In our interviews, participants (especially those who used the internet regularly in relation to MS) often expressed a strong preference for a specific site or service that they had become familiar with and found helpful to them. It would, however, be inaccurate to characterise this as a free or ‘open’ market where people can simply choose the service they prefer.

It is well recognised that search engines [predominantly Google^TM (Mountain View, CA, USA)] and large health-care incumbents (the NHS, pharmaceutical companies, established patient organisations) play a significant role in how and what health information people access online. 122 However, considerably less is known about how the design and management of online forums shape online experiential information sharing. Our data suggest that various, often subtle and pervasive, factors shape these different services, and that this affects who ends up using these services, what PEx they share online, how they share it and the consequences this use has for them. This was recognised and clearly articulated by interviewees working on sites and services for sharing PEx online, for example in the quotation below:

[T]he context in which you encourage people to submit data has a massive impact on the culture that’s developed there as well as the type of people who are attracted to your system. So if you want Daily Strength and you make a system where people can give each other hugs and smiley faces, you attract people who want hugs and smiley faces and put butterflies in their email. If you make a system that tries to mimic clinical trials and does scientific studies you’re going to attract highly educated, quantified-self type people who want to manage their illness better.

Interviewee 7, research director of health information organisation

Although the people we interviewed who worked on websites and forums were aware of and very sensitive to the fact that the technical, aesthetic and management decisions they made had an effect on their users and could even alienate some people, as one interviewee pointed out ‘nobody can be everything to everyone’, and, often, difficult decisions had to be made. In some cases the individual, group or organisation providing the platform, forum or online service (such as an organisational Facebook page) was able to make design and management decisions that reflected and supported their aims, ambitions and relative positioning in the field. Their ability to do so was, however, constrained by numerous factors; four key ones which emerged from our interviews were the nature and aims of the individual or organisation providing the service (e.g. funding, capacity, time); the condition in question and the service provider’s position in the wider field of information and support providers (e.g. competition, target users); technical dynamics, challenges and opportunities, often ones beyond their control (e.g. changes to underlying infrastructure and platforms such as Google and Facebook); and user dynamics, initiative and resistance.

Moreover, the rise of Facebook and YouTube as popular sites for sharing PEx has resulted in a decline in the number of people using more ‘traditional’ patient forums. However, Facebook and YouTube have not simply replaced these forums; instead, they are used for other purposes, and in some cases are directly linked to patient forums. For example, in the case of one charity that has both a Facebook page and a forum, Facebook was used for shorter messages, campaigns and less private information, while forums were perceived as more anonymous and intimate spaces:

I think that the overwhelming difference is that the forums allow people to be much more anonymous and we even go so far as not only can you create an online identity which conceals who you are, but you can actually ask a question completely anonymously on top of that as well. The older generation particularly don’t trust Facebook as a way of putting personal issues out there and even, to a certain extent, the kind of traffic we get on Facebook is quite different really.

Interviewee 1, social media officer, MS charity

However, platforms such as Facebook are not neutral or stable. They have their own commitments and politics,123 and regularly update their services in ways that can have unexpected effects on user interaction. For example, a change made by Facebook to their settings, something beyond the control of the MS charity using it, had a dramatic effect on the usage patterns of the page:

[I]t [the group] hasn’t actually been this active until the last year or so because Facebook have changed the settings or something. Once they changed their group settings so the groups were automated in their updates so suddenly people would see that something had happened on the group. People started looking much more at them and they became much more involved. So I would say up to about 2010 we didn’t have an awful lot of activity going on on there. There was, you know, now and again, but now it is . . . back then we might have had to send one question a week to information where now we have to send say five a day, so it is quite different.

Interviewee 6, web editor, MS charity

As these sites are often important sources of information and support for people affected by MS, changes, intentional or unintentional, can have dramatic effects, in some cases leaving users feeling powerless and bereft. This was brought forcibly to our attention in August 2011 when one of the forums we were observing underwent an extensive upgrade. The forum had been in operation for approximately 10 years and had a very dedicated group of users. When its layout, colour and functionality changed, many users responded very critically. The distress and upheaval brought about by the changes were evident in the large number of posts in which members expressed their unhappiness, in some cases stating that they would no longer use the forum. Some of the criticism was attributed to problems with functionality (e.g. users struggling to move their profiles over, customise their avatars and usernames, send private messages), but many focused on aesthetics, such as the change in the colours – from pastel to brighter shades – and the overall ‘feel’ of the forum. Numerous members stated that the forum no longer felt ‘like home’ and complained that it was now ‘stark and clinical’ as opposed to warm and friendly: for example, ‘I just feel emptiness now when I go in this site. I hardly come here now . . .’; ‘It’s cold in here. The sense of community and support has gone.’ The people working on the ‘back-end’ of this forum were obviously very aware of these responses, and were surprised by them.

This case illustrates how what might seem as ‘trivial’ aesthetic decisions, such as colour or where branding is placed, can influence what PEx is shared and how people respond to it. This was reiterated by research participants who worked on websites and forums, who spoke of how important aesthetics, including, colour, wording, visuals, branding, links and so on, were in not only attracting and retaining users, but in the creation of appropriate online spaces for sharing PEx that not only provided information, but a sense of care and support. Previous work has indicated that visual design elements unrelated to the actual health information content can influence trust and credibility;124 in a review of design features used in effective e-health interventions aesthetics, was raised as salient feature. 121 Similarly, in our research, what can be broadly classified as aesthetic factors played a key role in how people affected by MS perceived and used online spaces for sharing PEx. What counts as a ‘good’ design is not, however, easy to pin down. Some interviewees spoke of being instantly ‘attracted’ to a specific site or service, while others felt that the same service was unappealing or off-putting.

Familiarity with a given service also played an important role in how users perceived and used it. Different forum cultures and networks of exchange develop over time among specific groups of people in ways that can differ considerably from the original intentions of the service or platform provider. In all three of the forums we studied a core group of users played a significant role in shaping the overall tone and type of PEx shared on the forum. Sometimes they were volunteers and moderators with an official, if usually unpaid, role, but in the many cases they were simply prolific users.

These ‘super users’ spent considerable amounts of time online: welcoming people, responding to their queries, creating a friendly environment, and sometimes policing the forums. It is well recognised that health forums have specific cultures and that users play a crucial role in developing this. 125,126 However, the forms of emotional labour – broadly defined as labour involved in dealing with other people’s emotions and regulating your own – involved in the creation and maintenance of supportive and effective platforms for sharing PEx are rarely discussed in the e-health literature. 127 The importance of this form of ‘work’ was especially evident when there were breakdowns in relations, for example when a user or group of users ‘flamed’ – engaging in hostile and disruptive behaviour – a particular service or an individual. When this happened it not only disrupted the service (sometimes for a considerable time), but also prevented people from using it, or indeed any online platform for sharing experiences, in the future. Forums have different procedures for handling disruptive users and behaviours, from extremely light-touch moderation through to actively removing posts and banning some users. In the majority of cases when conflicts or disagreements arose on the forums we observed that they were resolved relatively swiftly with minimal intervention. However, during the period we were conducting our research the topic of CCSVI was a particular source of conflict and disagreement, and we observed many heated exchanges on it. It was also a topic in relation to which a great deal of PEx had been and was being shared. We, therefore, chose to focus on it as a specific area of interest.

YouTube and the sharing of patient experience in relation to chronic cerebrospinal venous insufficiency

In 2009 Zamboni proposed that abnormal venous drainage, what he called CCSVI, was a possible cause for MS. Furthermore, he suggested that venoplasty of the azygous and jugular veins – dubbed the ‘liberation procedure’ by its supporters – might improve symptoms and slow disease progression. 128 Although Zamboni’s research ran counter to much orthodox biomedical thinking about MS, between 2009 and 2012 it received a great deal of media and popular attention, with thousands of people across the world undergoing venoplasty in private clinics and/or as part of research studies. 29,129

Many people affected by MS shared their thoughts about CCSVI and venoplasty on blogs, forums, Facebook and YouTube, with some using their experiences to lobby for further research and treatment. 84,115,129,130 In particular, YouTube emerged as an important medium for sharing CCSVI-related PEx. To study this further we analysed the top 100 most viewed YouTube videos relating to CCSVI. The results of this analysis are published in detail elsewhere;84 therefore, we provide just a brief summary of the key findings here.

Although many neurologists and the representatives of international MS Societies expressed skepticism about Zamboni’s research, the YouTube videos we analysed adopted an overwhelmingly positive attitude towards it (67/70; 96%). Many of the videos were uploaded by patients (46/70; 66%). Videos in which people posted vivid pre- and post-treatment comparisons were particularly popular. Here, video posters did not simply share their experiences verbally, but tried to quantify and communicate the changes they experienced through performing various tests, such as walking up stairs, balancing on one leg or bending. Many of these tests drew on clinical and diagnostic practices as well as medical explanations and terminology, with patients frequently discussing MS treatments, medical professionals and the role of pharmaceutical companies. These videos, some of which form part of extensive video diaries that document people’s experiences over years, were often highly personal: videoed in people’s homes, showing their families, and embedding PEx not only in the clinical aspects of MS, but in everyday living with the condition.

Three key themes emerged from our analysis of these videos: (1) the visual medium enabled vivid depictions of pre- and post-treatment comparisons, often drawing on medical explanations, terminology and tests adapted from clinical practice; (2) patients not only displayed their own medical knowledge, but discussed current MS treatments, medical professionals and ‘big pharma’; and (3) videos were situated in relation to specific people’s experiences, conferring a sense of authenticity and personal immediacy. Thus, patients drew on medical knowledge to explain and reinforce their message, but, at the same time, their status as patients rather than practitioners conferred their thoughts, experiences, and, in some cases, advice, with a particular type of authority. The evidence generated through the YouTube videos we analysed was, therefore, predicated both on the language and practices of contemporary biomedicine and personal experiences of living with MS. This was most notably actualised in personal experience diaries, through which trust and legitimacy can be particularly developed, enhancing the strength of the ‘evidence’ portrayed.

Our aim in the study was not to comment on the scientific validity of or treatment efficacy for CCSVI. Nonetheless, an obvious question prompted by our findings is why were the YouTube videos (at least at the time of our study) so positive? There are of course numerous possible answers to this. Many of the most-viewed videos were uploaded by early adopters and strong advocates of CCSVI before later, more critical studies were published. 115 Thus, simply by virtue of being online for longer these videos had higher view counts and emerged more prominently in our sample. There is, moreover, a more general bias towards reporting positive experiences with medical treatments online. 131 Additionally, a few interviewees (including some who were supportive of CCSVI and Zamboni’s work) noted that those who might have been more critical of CCSVI often chose to stay silent because they were afraid of being criticised by their fellow patients or did not want to dampen other people’s hopes. This was reinforced by CCSVI advocates who worked actively to provide positive information about CCSVI on different online spaces, including forums and Facebook pages:

We got to the point on Facebook where there were people initiating conversations about it [CCSVI] and then somebody was replying to it and somebody would reply to that reply and all members of the ‘pro’, if you like, camp, and basically it was a very sort of stage managed dialogue and it got a lot of people’s backs up because they kind of felt like they were being patronised by a special interest group.

Interviewee 1, social media manager, MS charity

This resulted in a disconnect between the very positive information about CCSVI being posted online and the far less positive response of mainstream neurologists, policy-makers and funding bodies. Consequently, many people affected by MS who were interested in CCSVI or personally had a positive experience with venoplasty became disillusioned with the critical mainstream biomedical response to it. At the same time, other patients and practitioners expressed frustration that an unproven theory received so much attention and funding, a situation they saw as a hijacking of the MS research agenda by a partisan group, with detrimental effects on patients’ well-being. 132

Although CCSVI is a specific and controversial case, the way that CCSVI activists used internet technologies to generate new forms of knowledge is not an isolated event, but indicative of a wider phenomenon in which online ‘communities of practice’ can generate new forms of knowledge, becoming, albeit non-traditional, ‘epistemic communities’. 133 For example, in addition to the use of YouTube described above, users on the TIMS forum (which was and continues to be a hub for discussions on CCSVI) created a thread to collect experiences in a more systematic manner than is typical on patient forums (this is discussed in depth in a book on CCSVI written by one of the community members);130 and a group based in the Netherlands developed a tracking website (www.ccsvi-tracking.com) to aggregate and visualise outcomes post venoplasty. This results in the collection and interlinking of PEx that might not be ‘evidence-based’ as typically understood in medicine, but for many people serves as a form of ‘experiential evidence’, which they turn to in order to help them make decisions and, to a lesser extent, advocate for policy changes and particular research agendas.

Qualitative studies

Consistent with the focus of the wider iPEx programme, our participants were drawn from three different health populations: smokers wanting to quit, people with asthma and carers of people with MS. Participants were recruited using a number of sources, including newsletters and flyers and through various health support groups, and were compensated £20 for their participation. In total, we recruited 36 participants (20 male) aged between 20 and 65 years, with a mean age of 38.7 years. All were experienced internet users (9.92 average number of years using the internet). Sixteen of the participants took part in the follow-up interviews.

Breaking these down into health groups, we recruited:

• 10 smokers (nine male and one female) aged between 20 and 47 years, with a mean age of 29.5 years, who had been smoking an average of 12 years

• 15 people with mild to severe asthma (seven male and eight female) aged between 22 and 60 years, with a mean age of 34.07 years and an average length of asthma diagnosis of 20.58 years

• 11 MS carers (four male and seven female) aged between 39 and 65 years, with a mean age of 53 years, comprising 10 spouses and one mother of someone with MS. They had been carers for an average of 13.61 years.

The participants visited a laboratory styled as an internet café at Northumbria University. They were asked to search a subset of the health materials available on the internet for approximately 1 hour, seeking information and advice on issues relevant to their personal circumstances (smoking, asthma or MS caring). They were then asked to take part in a group discussion led by a facilitator and to complete log books detailing their impressions of the sites they visited. These sites were also automatically logged by the computer. Each participant was given access to seven predetermined sites during the search phase. Every site contained user-relevant material, but only four of the seven sites contained personal experiences. This selection was informed by content analyses of available e-health material reported in an earlier study. 150 Sites were chosen to be representative of those generally available to users, with the proviso that those requiring registration before access were excluded.

Following the initial search, when first impressions were recorded, participants were asked to choose two sites to explore in more detail. They then shared their experiences in group discussions that addressed the following main areas: (1) site selection and rejection; (2) site likes and dislikes; (3) aspects of personal experience; and (4) trusted elements. Participants were then directed to personal experiences sections on two further sites and took part in a final discussion around the content of these last two sites which addressed: (1) recall of experiential material; (2) level of interest and usefulness of personal experiences; (3) variety and type of experiences (i.e. individual accounts such as blogs or testimonials vs. interactive forum-style content containing multiple experiences); and (4) trust and behaviour.

Two weeks later, participants were contacted for a 1-hour semistructured telephone interview. The discussion covered the following main areas: (1) information searching behaviours since phase 1; (2) recall of the websites and their content; (3) information sharing activities; and (4) decision-making and current behavioural intentions.

We also conducted a follow-up qualitative study with two of our groups (people with asthma and MS carers). For this second study, we recruited 14 additional participants to the asthma group (five male and nine female) aged between 19 and 64 years, with a mean age of 28.21 years, and we recruited nine additional MS carers (three male and six female) with an age range of 40–73 years and a mean age of 58.9 years. These additional participants followed the same protocol as in the first study, but all were then asked to take part in a more extensive, structured interview that also involved the use of the repertory grid technique151 as a means of eliciting uniquely personal constructs about the ways in which they viewed the sites they found. This work has been published. 152

All focus group discussions and interviews were audio recorded and transcribed for analysis. The transcripts were read and reread separately by two of the research team and then were subject to thematic methods153 to identify both anticipated themes, for example design and reputation factors, and emergent issues such as comparison processes. The themes were then subject to discussion with the rest of the team to ensure validity in the interpretation and representation of data. For ethical reasons all participants were given a participant number and all quotations used in the results have been stripped of any information that could potentially identify the participant.

Resulting framework

The analysis revealed a three-phase process by which users engage with online personal experiences. Phase 1 is a ‘gating’ process, whereby people determine the suitability of the material they encounter; phase 2 comprises an ‘engagement loop’, whereby people look for support or advice from others with similar or relevant experience; and phase 3 is an ‘evaluation and outcomes’ process, whereby individuals make a judgement about the costs and benefits of engaging with the site in the longer term and in some cases experience outcomes related to that engagement. This three-phase process is captured in the framework (Figure 4), with the three phases described in more detail below.

FIGURE 4.

Schematic representation of the online health user experience (PEx) engagement framework. The stages are shown on the left-hand side of the framework (i.e. gating, engagement loop, and then evaluation and outcomes) and the dominating factors influencing engagement at each respective stage are shown on the right-hand side. Each box displays the questions and factors that users address during each stage of the engagement process. Reproduced with permission from Sillence E, Hardy C, Harris P, Briggs P. Modelling Patient Engagement in Peer-to-Peer Healthcare. 23rd International World Wide Web Conference, Korea, April 2014. 154

Case study 1: re-engaging with your asthma

Asthma patients were initially drawn to websites showing good design quality. In line with previous studies,8,155 we noted that participants preferred sites that were easy to navigate and looked professional. Participants were then subsequently drawn to those sites containing PEx, but only if the contributions seemed to come from individuals judged to be similar to themselves (e.g. in terms of age and length of time since diagnosis) and could offer ‘relevant’ stories. Before the study, participants were unsure what else they could or should know about asthma but, as they started to engage with the online personal experiences, many chose to explore certain stories in more detail by comparing their own experiences with those of others online. This comparison process allowed for recognition of their experiential asthma knowledge,157 with participants able to reflect on the ways in which they had experienced and handled certain situations with regard to their asthma. Despite many participants reporting a high level of asthma knowledge before the study, many were still able to acquire hints and tips from others regarding the management of their condition. For others it was a case of confirming or even fine-tuning their existing knowledge. However, one consequence of this comparison process was that participants sometimes dismissed potentially useful hints and tips for managing their asthma on the grounds that the online experience was written by someone different from them. Most participants thought that the online contributions came from people who were worse off or had more severe asthma than they did. This process of downward social comparison has been noted in other health settings in which people under threat evaluate themselves favourably in comparison with others. 158 A consequence of this process is that, for some, the personal experiences served as a reminder of the potentially serious nature of the condition and triggered some re-engagement with the self-management programme through the asthma nurse.

Case study 2: multiple sclerosis carers and coping in the here and now

Shared PEx online can be used to support both the problem-focused and the emotion-focused coping strategies of the MS carers. Problem-based coping was seen as the norm when caring for people in the early stages of MS, as carers needed to understand both the symptoms and behaviours they might expect of someone with MS, and may also need practical support to establish daily care routines or understand entitlements. This is consistent with other findings showing that coping strategies for MS carers are often focused on the current situation, the ‘here and now’ of coping. Carers felt that problem-coping strategies would give way to emotion-focused strategies as they became reconciled to the progression of the disease in the person with MS, accepting that knowing more about MS would not change anything. Consistent with this shift to emotional-coping, carers valued personal stories that fuelled empathy and a greater understanding of what the person with MS was experiencing. Participants recognised that, ultimately, individuals would have different needs and that each would come to employ their own distinctive coping strategies, better served by separate online resources. A sense that carers should be able to pick and choose from the stories available was strong here. Carers recognised that although they were not always likely to be receptive to some of the more difficult messages from other carers, they valued the fact that the PEx offered both practical and emotional support.

Case study 3: smokers and their mistrust of the commercial testimonial

The smokers’ engagement with PEx was characterised by strong preferences for certain kinds of experiential material. The interactive forums were seen as valuable in terms of providing support and motivation for those wishing to quit. Smokers identified with those authors who seemed well matched in terms of gender, age and location. Smokers noted unusual tips and hints for giving up smoking but, overall, preferred realistic yet positive support messages. Such messages acknowledged the difficulties associated with quitting but often included an inspirational message and a positive, successful outcome. The testimonials were viewed with more suspicion. For smokers, at least, there was a sense that this form of PEx might not always be genuine and in large part this reflected the strong commercial overtones associated with PEx around this topic area. Many of the websites featured commercial advertising and products associated with quitting. Celebrity and non-celebrity endorsements or testimonials were seen as difficult to trust and the smokers thought it highly unlikely that they would return to this kind of website. Overall, there was less evidence of longer-term engagement with the PEx for the smokers. The participants commented on the difficulty of navigating to useful sections of the forums or finding relevant experiences in the dense message threads.

Implications of the framework for further studies

Our framework highlights both the source and the message of PEx (the ‘who’ and ‘what’ of PEx) as critical factors in user engagement. We believe that people are quick to make judgements about the person offering information and advice, judgements that reflect the extent to which they are seen to be ‘like me’ but also judgements that take into account their motivation in shared storytelling that in turn raise issues around trust. Individuals are also concerned about the extent to which the content of a health message fits with their own experience and pre-existing beliefs and reflects their particular journey. For example, the MS carers we observed in the framework studies above tended to exhibit information-based coping strategies if they were caring for someone in the early stages of MS, but more emotion-based strategies when caring for someone in the later stages of MS, and their responses to available PEx varied accordingly. 159 In the two studies that follow, we explore these source and message issues in more detail by exploring smoking cessation.

Method

We used a similar paradigm to that reported earlier in this chapter (A framework for user engagement with online patient experience, Qualitative studies), asking 15 smokers to search for health-related smoking material online and then discuss their findings. Participants were recruited using a number of sources including council newsletters, leaflets and flyers. The smokers (seven male and eight female, with a mean age of 23 years) were all experienced internet users. They had been smoking on average for 7 years and smoked an average of 11 cigarettes per day.

Each participant was allocated to one of four focus groups and asked to attend a 2-hour session held in an internet café-styled laboratory in Newcastle upon Tyne, UK. During the first hour of the session, participants were invited to search the internet freely, looking for information and advice on smoking and health. We logged the sites they visited, but also asked them to record their perceptions of each site in a log book and to use this information during a subsequent group discussion with a facilitator. These group discussions covered four main themes: (1) selection and rejection factors, (2) liked and disliked websites, (3) PEx/account features and (4) trust and behaviour. Thematic analysis and coding practice followed that of the earlier qualitative studies. Coders were primed to highlight comments in relation to trust or mistrust and like or dislike of sites, and were also interested in the presence or absence of a ‘commercial voice’.

Results and discussion

Participants visited 8.5 websites on average. The most commonly viewed sites (in order of number of participants who visited the site) were the UK NHS Smokefree website (www.nhs.uk/smokefree), Netdoctor (www.netdoctor.co.uk), Wikipedia [www.en.wikipedia.org/wiki/Smoking_cessation (accessed 1 July 2011)] and the British Broadcasting Corporation’s smoking website [www.bbc.co.uk/health/physical_health/conditions/smoking_health_effects.shtml (accessed 1 July 2011)]. The consistency in sites viewed suggested that participants may have employed similar strategies and/or been influenced by similar factors when deciding which sites to examine. From the discussions, data logs and log books, all participants attempted to locate relevant websites using Google, although some participants also employed a more targeted search via the NHS.

As anticipated from our earlier work, we found that people said that they very much liked having access to other people’s experiences online and found the material engaging.

Well the patient UK one did have experiences of people who’ve smoked and who’ve quit and health issues that they’ve had whilst smoking and once giving up and things like that and that was quite good I think – reading about other people’s experiences.

Focus group 1, participant 3

We also found support for the claims that such experiences can help make the health issues seem more real and engaging as people felt that they could relate more to other people who are in similar situations.

[It’s more helpful] If they’re more realistic if they’re not just literally a doctor’s sat and written it like ‘Mrs C has done this’. If it’s an actual person saying ‘look it’s really hard. I’ve shouted a lot. I’ve done this, I’ve done that,’ then you can relate to them.

Focus group 1, participant 2

However, the presence of advertising was almost universally raised as something that was disliked about some of the sites.

There’s lots of adverts as well which annoyed me. It’s like they get some form of gain out of telling people to quit. There’s adverts and even some adverts weren’t even related to smoking.

Focus group 1, participant 2

As predicted, we found considerable evidence that users were suspicious about site content and that advertising had an important role to play in this. Sites that appeared to be selling a product or that were otherwise viewed as commercial created very negative first impressions in the sample, and participants often felt that information contained on such a site would not be impartial or trustworthy.

More tellingly, participants became particularly suspicious if it seemed that the PEx had been used cynically to advertise a particular product or manipulate beliefs.

There is one particular one that got like a few . . . a review of stories of how people are doing with the product, but I think those websites try to sell their products rather than they are really trying to be help.

Focus group 3, participant 10

Participants also seemed very aware of the potential for PEx to be manipulated and were concerned about the extent to which the experiences might be honest, questioning if they might have been scripted rather than represent an actual experience:

That kind of story from an actual person or something would have way more impact than some typed words on the internet that could have been put there by anyone. You don’t know who typed the website.

Focus group 1, participant 7

These discussion points are revealing as they show that people are drawn to PEx material, but can also mistrust PEx when it appears to be used for commercial gain. Indeed, users may come to question the genuineness of PEx that sits in an overly commercial context (at least in a context such as the UK, where health care is provided free at the point of delivery). In our subsequent study with smokers, we explore the extent to which matching the source and recipient characteristics of PEx can influence smokers’ perceptions of the acceptability of the advice presented to them and also directly influence health outcomes in terms of reducing the number of cigarettes smoked.

Method

We recruited smokers to an online study using poster advertising, e-mail and word of mouth. A total of 113 participants (mean age 24.7 years, n = 63, 55.7% female) completed all three phases of the study. Most (n = 74; 65.5%) were students.

Results

For the gender-match analyses, participants were matched on the basis of the correspondence between their gender and that of the ostensible author of the PEx. For the preference-match analyses, participants were matched on the basis of the correspondence between the information and their preferences for quitting supported (match) or unsupported (no match). Alpha was set at p ≤ 0.05 for all analyses.

Discussion

We explored the effect of PEx in conjunction with an exploration of the effect of matching PEx to the gender of the participant. We found little effect of gender matching and only observed a PEx effect in the men. Men found PEx preferable in terms of both mood and message believability, but it mattered little whether the experience seemed to come from a man or woman. PEx did not affect message knowledge or website evaluation as measured by the eHIQ in either sex.

Over time, other effects emerged. Men showed increases in quit attempts and decreases in nicotine dependency over the 3-month period. Women also showed decreases in nicotine dependency over this period. There were also some significant effects of condition on nicotine dependency. However, these effects may reflect the lingering impact of gender differences in dependency at baseline.

We also found some initial effects (at baseline) of matching for preference in terms of the way people might choose to give up smoking (supported or unsupported). Health information expressed in a way that is preference consistent led to higher levels of engagement with online content, which supports the general health literature on the role of homophily as a component of effective health behaviour change.

Primary objective

• To evaluate the feasibility issues in an online randomised study providing health information websites containing PEx information compared with matched health information websites that do not contain experiential information.

Secondary objectives

• To assess the efficacy of two types of online health information (PEx accounts compared with matched health information websites that do not contain experiential information) on a range of self-reported outcomes.

• To explore whether or not the interventions have differential effects on pre-specified subgroups of participants.

• To measure the impact of the intervention and comparator websites using the eHIQ.

Design

A phase II pilot randomised single-blind trial, in which eligible participants were randomly allocate to a PEx website or a comparator ‘facts and figures’ website stratified by health topic.

Interventions

As described in the previous chapter, for each condition (asthma, smoking cessation, MS carer) two websites were developed: one containing PEx information and the other containing information solely based on facts and figures with no experience-based content. Each were password protected, multimedia internet sites based on guidelines which were developed by the team working on WP2b. The comparator sites shared the design and multimedia features (such as video) of the intervention sites but the content excluded any accounts of personal experience. They used source material from the national health information portal NHS Choices with all experiential information removed.

Participants

Trial participants had asthma, were smokers with a willingness to quit or were carers of a person with MS. To be eligible for the asthma condition, participants were required to have clinically diagnosed asthma as coded in their primary care electronic record, and have been prescribed inhaled corticosteroids for at least 3 months in the previous year. To be eligible for the smoking cessation condition, participants were required to be current smokers, who had been smokers for at least 1 year, and who indicated some willingness to quit, including those referred to smoking cessation services. To be eligible as carers of people with MS, participants were required to identify themselves as an unpaid caregiver (i.e. not a professional carer) for another person who had a diagnosis of MS. Participants who matched more than one condition could be included only once, and only one participant per household could be included.

We included participants who were male or female, over the age of 18 years, resident in the UK, with access to the internet and able to use websites. We excluded people who could not understand English (owing to the nature of the intervention), who were terminally ill or who had a significant disease or disorder that may have put that person at risk because of participation in the research, or may have influenced the result of the research, or which affected that person’s ability to participate.

Recruitment

The recruitment of participants eligible for the asthma and smoking cessation conditions was predominantly through primary care practices in two regions of England, facilitated by the local primary care research networks. GPs who agreed to take part were asked to identify eligible individuals from their practices and to send them full information on the research study. In addition, for the smoking cessation recruitment, posters advertising the study were placed in public places, such as GP waiting rooms and libraries, and online adverts were placed on Facebook. The recruitment invitations and posters gave contact details for the study team so that potential participants could make initial contact to indicate their willingness to take part.

The recruitment of participants eligible as carers of people with MS took a different approach. Being a carer for someone with MS is not routinely recorded in primary care records. We took a broad and iterative approach to the recruitment of carers, helped by our patient and public advisory panel. Recruitment was not easy, and over time we used many routes. We placed posters and leaflets in GP practices, neurology clinics, MS therapy centres, local enablement and well-being services, and local leisure centres. We enlisted the support of several third-sector organisations that distributed recruitment adverts on their websites and social media feeds, including the MS Society, Shift.MS, Carers UK and the Carers Trust. We met with the local carers group, advertised in their newsletter and staffed a stand at a local carers’ conference. We placed paid-for adverts in Enable magazine, in local newspapers, in the London Metro paper and on Facebook. We had our own study recruitment webpage and promoted this through various channels, including the Oxford Daily Info website and tweets sent to key organisations and public figures. We were supported by members of the programme public panel and the MS patient and carer panel who helped us to identify recruitment avenues as well as potential participants. Further participants were recruited through word of mouth and personal contact.

Study procedures

For each condition, potential participants who indicated a willingness to take part completed screening questionnaires to assess their eligibility against our inclusion/exclusion criteria (see above). All eligible participants were sent a patient information sheet and consent form by post to be signed and returned to the research office in a reply-paid envelope. Once consent had been received participants’ details were entered into a secure trial management portal and they were sent a unique ‘welcome code’ with a link to the trial registration webpage. If after 1 week the participant had not used their welcome code, a reminder was sent by automatic e-mail inviting them to visit the trial registration page. If the participant no longer wished to take part, they could simply ignore this reminder. If the participant did wish to take part they could use the link in the e-mail to access the trial registration page where they were asked to complete baseline questionnaires on the online portal. Once these measures had been completed they were invited to create a unique user identification and password and then they were randomised to have access to either the intervention or the comparator website, for a period of 2 weeks. Randomisation used a computer-generated random number sequence in a 1 : 1 ratio. The sequence was generated by a trial statistician independent of the study. Participants did not know whether they were receiving the comparator or the intervention website, only that we were evaluating two approaches to giving health information. Nevertheless, the nature of the intervention meant that they could not be blind to whether they were looking at personal experiences or facts and figures information. The study investigators undertaking the analyses were blind to allocation, which was an automated process. Participants were able to log on and access the website interventions as much or as little as they wanted. At the end of the 2-week period, participants were invited by e-mail to complete final follow-up measures, using the same portal. Up to two e-mails and one telephone call were used to remind those who did not initially respond. Participants’ usage of their allocated website was also tracked.

Measures

The outcome measures for each condition are shown in Table 16. As a feasibility trial, our primary measures concerned the number of participants consented and recruited, the usage of the websites (in terms of numbers of logins, page views and time on site) and the numbers with completed outcome measures or lost to follow-up. For secondary outcomes, we used condition-specific self-report measures as well as the widely used and well-validated Short Form questionnaire-36 items (SF-36) health status measure (including both physical and mental dimensions and all subscales), and the eHIQ, a new self-completion measure designed earlier in this programme of work to measure the effects of accessing online health information. The eHIQ part 1 contains 11 items covering general attitudes towards using the internet to access health information. The eHIQ part 2, which contains 26 items, is administered at follow-up, as it captures the attitudes of the respondents towards the website they have recently viewed. Both parts of the questionnaire use five-point response categories, from ‘strongly agree’ to ‘strongly disagree’, for all items. 66,99 The eHIQ part 2 scores are combined to make three subscales: confidence and identification (reflecting confidence to discuss health with others and a person’s ability to identify with the website); information and presentation (reflecting trust and suitability of website content); and understanding and motivation (reflecting understanding and learning about relevant information and motivation to take action (see Chapter 4). Each subscale is scored between 0 and 100, with higher scores indicating a more positive rating.

The choice of condition-specific outcome measures was informed by our theoretical understanding of what the benefit of shared personal experiences might be. So, for the asthma condition, we were specifically interested in whether or not other people’s experiences could have effects on aspects of self-management including self-efficacy and asthma control. For asthma we therefore used the Chronic Disease Self-Efficacy Scale (CDSES), which is a six-item questionnaire assessing confidence in self-managing a chronic condition. 168 (The score is the mean of the six items. If more than two items were missing, the scale was not scored.) We also used the 12-item Partners in Health (PIH) scale, which assesses knowledge and behaviour in relation to self-management for a chronic condition. 169 The 12 items cover four domains of competency in relation to self-management (knowledge, coping, management of condition and adherence to treatment), and the total sum score is used (the higher score indicating greater competency). We also used a single-item asthma control question taken from the 5-item Asthma Control Test (which was used as a baseline descriptive measure), whereby participants rated their asthma control on a five-point scale from ‘not controlled’ to ‘completely controlled’. 170

For smoking cessation, we were interested in whether or not other people’s experiences could be harnessed to achieve health behaviour change, namely to quit smoking or to change the motivation to quit. We therefore used the following self-report measures: abstinence rates (single self-report question), quit attempts (single self-report question), the single-item Motivation To Stop Scale (MTSS) to measure whether motivation had changed, on a 7-point scale (where a rating of 1 reflected the absence of any belief, desire or intention to stop smoking, and a rating of 7 reflected a strong desire and short-term intention to stop),164 and the Smoking Abstinence Self-Efficacy Questionnaire (SASEQ), a six-item questionnaire on which smokers are required to indicate, using a five-point Likert scale, how confident they are in abstaining from smoking in six example situations. 171

For carers of someone with MS, we wanted to assess whether or not other people’s experiences of caring could help carers feel more confident and supported in their own caring roles. Most of the validated instruments in this area are specific to dementia care and this restricted our choice of questionnaires. We did not find any instruments specific to the caring role in MS. We used the 8-item Preparedness For Caregiving Scale (PFCS), which assesses how well prepared participants think they are. 172 Each question is scored 0–4, with 4 representing ‘very well prepared’, and the higher the score the more prepared the caregiver feels for caregiving. The score for the PFCS is calculated as the mean of the eight items.

Sample size and statistical analyses

We planned to recruit a total of 300 participants, 100 for each of the three conditions. The focus of the study was to assess feasibility, which could usually be addressed using a smaller sample but we anticipated that this sample size would provide enough power to estimate acceptability judged by engagement with the PEx components of the site with a precision of ± 14% in each condition/arm (50 participants) and 8% for the whole trial/per arm (150 participants). This was based on a worst-case scenario of acceptability being 50%. We assumed the acceptability rate to be in the region of 80%. We also wanted sufficient participants to be able to adequately assess issues of recruitment and retention, as attrition rates are often high in fully internet-based trials.

Feasibility outcome measures were summarised using descriptive statistics such as rates reported as percentages. We undertook descriptive statistics to characterise participants at baseline; constructed a Consolidated Standards of Reporting Trials (CONSORT) diagram to show the flow of participants through the trial, including the proportion who completed each stage to trial entry; and measured usage of the interventions in terms of total visits to site, total time on site and total number of website pages visited. Formal hypothesis testing of the self-report outcome measures was not performed, as this was a feasibility trial and was not intended (or powered) to assess evidence of significant treatment effect. However, the estimated difference in outcome measures of efficacy and corresponding 95% CIs were calculated using analysis of covariance (ANCOVA) adjusting for baseline values. In terms of effect sizes, based on a balanced randomisation of intervention–comparator groups in a ratio of 1 : 1 (i.e. 50 intervention and 50 comparator for each condition), the trial would detect potential large effects of the intervention for each condition; for dichotomous outcomes these were equivalent to relative risks of ≥ 2.1 for a baseline rate of ≤ 30% given an alpha of 0.05 and 90% power, and for continuous outcomes these detectable differences would be of the order of 0.4 SD based on the same power and significance.

Qualitative study

Following completion of final follow-up measures, a subset of study participants was purposively selected and invited to be interviewed by one of three qualitative researchers from the HERG at the University of Oxford. The participants were selected for maximum variation of their demographic characteristics and use of the intervention (determined by usage data), to give us a diverse sample. Information sheets and consent forms were sent out by e-mail following an expression of interest by participants (who were asked to indicate their willingness to take part when filling in their online measures). The consent forms were completed in person. The interviews focused on reasons for participating in the trial, engagement with the online resources, and attitudes towards online interventions and generic health resources and support. Participants from both intervention and comparator arms in each of the three conditions were included to explore differences which may have been attributable to the allocated website. The interviews used two semistructured interview guides: one for participants allocated to the intervention websites and one for those allocated to the comparator websites. These were based on reviews of existing literature and earlier, exploratory work. The guides differed to ensure that the participant remained blind to trial allocation and purpose. For example, those participants in the intervention arm of the trial were asked specifically about how they chose which PEx clips to view. The guides covered the following topics: motivation for participation, acceptability of trial procedure, experience of taking part in the online intervention and trial, use of the internet for health-related purposes and attitudes towards different types of health information. We used a set of laminated ‘home pages’ from six different types of websites to facilitate this discussion. Open-ended questions and prompts were used to further explore participants’ experiences of taking part in this online study and their attitudes towards the website content. The researchers did not ‘unblind’ the participants by telling them more about the study until the end of the interview.

The interviews were transcribed and analysed by thematic analysis using the method of constant comparison. A specialist software package (NVivo 10) was used to organise the data. Recurrent themes and subthemes were identified, as they emerged from the data, by the three interviewers, who initially coded their own interviews. This was followed by extensive discussion and comparison of the emergent themes between the researchers.

Changes from the original proposal

The original protocol stated that follow-up measurements would be collected at 1 month and 3 months, and that participants would be randomised in a 2 : 1 active–comparator ratio. After discussion with our trial statistician and the rest of the trial team, we reduced the follow-up to one measurement immediately post exposure to the website, and changed the randomisation ratio to 1 : 1. The change to a shorter follow-up (immediately post exposure to the website intervention) was for a number of reasons: that this would maximise blinding and minimise contamination; that our primary aim was to establish feasibility; that our main measures related to the immediate effect of the website exposure; and that we wanted to minimise attrition. This decision was discussed and agreed with the funder.

Ethics and governance

Ethics approval for this study was provided by the NHS Health Research Authority Research Ethics Committee, reference 13/NW/0162. The trials were overseen by a Trial Steering Committee chaired by Professor Elizabeth Murray of University College London. The trial registration number is ISRCTN29549695 and it was registered on 17 May 2013.

Smoking cessation

Recruitment through primary care was relatively slow, and we engaged more primary care practices (23 in total) than anticipated. The first participant was recruited in June 2013, and the last participant completed follow-up in August 2014. The trial ended as we had recruited > 80% of target sample size and the funding was ending. A total of 129 invitations were sent in response to expressions of interest in the study, and 89 individuals (69%) completed registration (i.e. they consented and completed baseline questionnaires). Two participants withdrew before randomisation, leaving 87 randomised participants (recruitment rate of 87/89 = 98%). Of these, 44 (51%) were allocated to the PEx website. Of the randomised participants, 22 out of 87 (25%) were lost to follow-up (please see the CONSORT flow diagram in Figure 9).

FIGURE 9.

Work package 3b: CONSORT flow diagram – smoking cessation.

The baseline demographic, internet use and smoking characteristics of randomised participants are presented in Table 17. These were balanced between groups. Most were regular users of the internet with good to excellent self-reported internet abilities. Of the 87 participants, 45 (52%) were women. The majority of participants reported smoking < 20 cigarettes per day, with most smoking between 11 and 20 cigarettes per day (53% of the sample). Most had intentions to quit smoking, with only three participants reporting either not wanting to quit (3%) or knowing that they should quit but not ‘really’ wanting to (18%, n = 16). Twenty-seven participants reported having made an attempt to quit during the 2 weeks before they started the study (31%).

The usage of the websites is shown in Table 18. This shows that the median number of logins during the 2-week period for both smoking cessation sites was 2, with a median number of page views of 11.5 for the intervention site and 7 for the comparator site. The median duration of use was 15 minutes for the intervention site and 5 minutes for the comparator site. One user never logged in to their allocated site, and two other users logged in but did not record a page view or spend any time on the site. Table 18 also presents the results of post hoc significance tests for differences in usage between groups. This was a post hoc analysis suggested by the chairperson of our Trial Steering Committee. The differences between groups for number of logins and number of page views were not statistically significant. The total duration on site was significantly longer (p = 0.013) for the PEx website.

Figures 10 and 11 are dot plots of the total number of logins and total number of pages, respectively, visited by the randomised group. The solid blue line indicates the median, and the lower and upper quartiles are shown by the dotted lines.

FIGURE 10.

Work package 3b: dot plot graphs of total number of logins by the randomised group for smoking cessation. IQR, interquartile range.

FIGURE 11.

Work package 3b: dot plot graph of total number of pages visited by the randomised group for smoking cessation. IQR, interquartile range.

Figure 12 shows dot plots of the total duration on the website in minutes by randomised group. The solid blue line indicates the median, and the lower and upper quartiles are shown by the dotted lines.

FIGURE 12.

Work package 3b: dot plot graphs of total duration (minutes) on website by the randomised group for smoking cessation. IQR, interquartile range.

The mean SASEQ scores were similar between the randomised groups. The baseline mean score (SD) for the intervention group was 1.94 (1.02) and for the comparator group was 1.84 (0.89); at 2-week follow-up these scores were 1.89 (0.99) and 1.71 (0.95), respectively. Both groups had low mean SASEQ scores, indicating low self-confidence in abstaining from smoking in a variety of situations. An ANCOVA found an adjusted mean difference for the change in SASEQ scores between the two randomised groups of 0.155 (95% CI –0.262 to 0.573), adjusted for baseline SASEQ scores.

The findings from the other exploratory comparisons of the smoking-specific outcome measures are shown in Table 19. This table shows that the baseline and follow-up scores and the change scores for the MTSS, abstinence rates and quit attempts were similar between groups for baseline and follow-up measures and for change scores.

The results for the SF-36 were also similar between groups and are shown in Table 20. Participants in both arms showed non-significant improvement in health state from baseline in the mental component and a very small, non-significant worsening in the physical component. There was no visible pattern between the two groups. The adjusted mean differences between the two groups (ANCOVA, n = 64 participants with full data) for the absolute change in physical component summary (PCS) score of the SF-36 was 1.393 (95% CI –0.702 to 3.489) and for the absolute change in mental component summary (MCS) score of the SF-36 was 0.785 (95% CI –3.09 to 4.648).

Asthma

Recruitment was rapid through nine primary care practices. The first participant was recruited in June 2013, and the last participant completed follow-up in November 2013. The trial ended as we had recruited in excess of the target sample size. Two hundred invitations were sent in response to expressions of interest in the study, and 150 individuals (75%) completed registration (i.e. they consented and completed baseline questionnaires). Two participants withdrew before randomisation, leaving 148 randomised participants (a recruitment rate of 148/150 = 99%). Seventy-three of these (49%) were allocated to the PEx website. Of the randomised participants, 27 (18%) were lost to follow-up (please see the CONSORT flow diagram in Figure 13).

FIGURE 13.

Work package 3b: CONSORT flow diagram – asthma.

The baseline demographic and internet use characteristics of randomised participants are presented in Table 21 and were balanced between groups. Most were regular users of the internet with good to excellent self-reported internet abilities. Of the 148 participants, 87 (58.8%) were women.

In general, the participants’ asthma symptoms were well controlled and did not interfere with daily living to a great extent (Table 22).

The usage of the websites is shown in Table 23. This shows that the median number of logins in the 2-week period for both sites was 2, with a median number of page views of 14 for the intervention site and 15 for the comparator site. The median duration of use of the intervention site was 15 minutes and for the comparator site was 19 minutes. Three users never logged in to their allocated site. Table 23 also presents the results of post hoc significance tests for differences in usage between groups. This was a post hoc analysis suggested by the chairperson of our Trial Steering Committee. The differences between groups for number of logins and number of page views and total duration on the sites were not statistically significant.

Figures 14 and 15 show dot plots of the total number of logins by randomised group with and without the extreme observations, respectively. The solid blue line indicates the median, and the lower and upper quartiles are shown by the dotted lines.

FIGURE 14.

Work package 3b: dot plot graphs of total number of logins by the randomised group for asthma (with extreme values). IQR, interquartile range.

FIGURE 15.

Work package 3b: dot plot graphs of total number of logins by the randomised group for asthma (without the extreme values). IQR, interquartile range.

Figure 16 shows dot plots of the total number of pages visited by randomised group. The solid blue line indicates the median, and the lower and upper quartiles are shown by the dotted lines.

FIGURE 16.

Dot plot graph of total number of pages visited by the randomised group for asthma. IQR, interquartile range.

Figures 17 and 18 show dot plots of the total duration (minutes) on website by randomised group with and without the extreme observations, respectively. The solid blue line indicates the median, and the lower and upper quartiles are shown by the dotted lines.

FIGURE 17.

Work package 3b: dot plot graphs of total duration (minutes) on website by the randomised group for asthma (with extreme values). IQR, interquartile range.

FIGURE 18.

Work package 3b: dot plot graphs of total duration (minutes) on website by the randomised group for asthma (without extreme values). IQR, interquartile range.

Table 24 shows summary statistics for baseline, 2-week follow-up and mean PIH scores. The mean (SD) change from baseline in the PEx (intervention) group was an increase of 0.068 (0.61), compared with an increase of 0.073 (0.64) in the comparator group. Treatment effect (mean differences with CI) was estimated using ANCOVA adjusting for baseline mean PIH scores, and showed that there were no significant differences between groups.

For the CDSES, two participants completed five of the six items at baseline (the mean of the five items was calculated) and four participants completed five items at follow-up (the mean of the five items was calculated). Higher scores indicate higher self-efficacy. The mean change from baseline to 2-week follow-up in the comparator group was an increase of 0.093, compared with a decrease of –0.068 in the PEx group. Treatment effect (mean differences with CI) was estimated using ANCOVA adjusting for baseline mean CDSES scores, which showed no significant difference between groups (Table 25).

The results for self-rated asthma control at the follow-up measurement are shown in Table 26.

The results for the SF-36 were also similar between groups and are shown in Table 27. Participants in both arms showed a small, non-significant improvement in health state from baseline in both the mental and physical components. The adjusted mean differences between the two groups (ANCOVA, n = 119 participants with full data) for the absolute change in PCS score of the SF-36 was 0.196 (95% CI –1.454 to 1.846) and for the absolute change in MCS score of the SF-36 was 0.843 (95% CI –1.066 to 2.750).

Multiple sclerosis carers

The recruitment of carers of people with MS proved challenging and was reviewed by the project team and the Trial Steering Committee on several occasions. As detailed in Methods, multiple routes to recruitment were used. Recruitment was slow, with usually between one and four carers recruited per month over the course of the study. Targeted advertising proved relatively unsuccessful and snowballing through personal contacts was more helpful. The first participant was recruited in June 2013 and the last participant completed follow-up in August 2014. The trial ended as the funding was ending and we had exhausted approaches to recruitment. Eventually, 47 invitations were sent in response to expressions of interest in the study, and 42 individuals (89%) completed registration (i.e. they consented and completed baseline questionnaires). No participants withdrew before randomisation and there were, therefore, 42 randomised participants (a recruitment rate of 100%). Twenty-two of these (52%) were allocated to the PEx website. Of the randomised participants, 6 out of 42 (14%) were lost to follow-up (please see the CONSORT flow diagram in Figure 19).

FIGURE 19.

Work package 3b: CONSORT flow diagram – carers of people with MS.

The baseline demographic and internet use characteristics of randomised participants are presented in Table 28 and were balanced between groups. Most were regular users of the internet with good to excellent self-reported internet abilities. Of the 42 participants, 18 (42.3%) were women.

The usage of the websites is shown in Table 29. This shows that the median number of logins in the 2-week period for both sites was 4, with a median number of page views of 27.5 for both the intervention and comparator sites. The median duration of use was 31 minutes for the intervention site and 31.5 minutes for the comparator site. Table 29 also presents the results of post-hoc significance tests for differences in usage between groups. This was a post hoc analysis suggested by the chairperson of our Trial Steering Committee. The differences between groups for number of logins and number of page views and total duration on site were not statistically significant.

Figure 20 shows a dot plot of the total number of logins by the randomised group. The solid blue line indicates the median, and the lower and upper quartiles are shown by the dotted lines.

FIGURE 20.

Work package 3b: dot plot graphs of total number of logins by randomised group for carers of people with MS. IQR, interquartile range.

Figure 21 shows dot plots of the total number of pages visited by the randomised group. The solid blue line indicates the median, and the lower and upper quartiles are shown by the dotted lines.

FIGURE 21.

Work package 3b: dot plot graph of total number of pages visited by randomised group for carers of people with MS. IQR, interquartile range.

Figure 22 shows a dot plot of the total duration (minutes) on website by the randomised group. The solid blue line indicates the median, and the lower and upper quartiles are shown by the dotted lines.

FIGURE 22.

Work package 3b: dot plot graphs of total duration (minutes) on website by randomised group for carers of people with MS. IQR, interquartile range.

The mean PFCS scores were similar between the randomised groups. The summary of PFCS scores are shown in Table 30. The change in PFCS score from baseline was very similar in both groups (intervention –0.16, comparator –0.15), with both groups showing a small decrease in PFCS score at the 2-week follow-up.

The results for the SF-36 were also similar between groups and are shown in Table 31. Participants in both arms showed non-significant worsening in health state from baseline in the physical component scores. The adjusted mean differences between the two groups (ANCOVA, n = 36 participants with full data) for the absolute change in PCS score of the SF-36 was 0.365 (95% CI –2.756 to 3.486) and for the absolute change in MCS score of the SF-36 was 2.655 (95% CI –1.788 to 7.099).

e-Health Impact Questionnaire findings

The eHIQ part 2 is administered at follow-up only, as it captures the attitudes of the respondents towards the website they have recently viewed. Tables 32 and 33 show the eHIQ part 2 results comparing the responses from participants who viewed the PEx website with those of participants who viewed the comparator websites, for all three conditions combined. Table 32 shows that there were only small, non-significant differences between the PEx and the non-PEx groups, both overall and on the three subscales (confidence and identification; information and presentation; and understanding and motivation).

Qualitative interviews

The qualitative interviews are discussed in detail in Chapter 9. Here we present only brief findings in relation to the trial design and process. The demographic characteristics of the interview participants are discussed in Chapter 9. In brief, the 30 interview participants were slightly older, and more likely to be retired, than those in the feasibility trials. The interviews demonstrated that the interventions and the online trial process were acceptable and usable. We did not identify major barriers to participation in the trial nor to use of the interventions: the interviews confirmed the pilot study findings that the technical aspects of the online trial had worked well, that the reminders were considered appropriate and that the questionnaires were not too burdensome.

People were motivated to take part in the trial for a fairly narrow range of reasons, including because the condition was relevant to them, the study sounded interesting, they were asked by their general practice and they liked to help with research.

There were few problems reported with accessing the websites (underlining the value of the pilot work and troubleshooting), although a couple of the participants had had difficulty on at least one occasion with opening the video clips. During the pilot study this had been associated with particular browsers. In the interviews people sometimes appeared unsure which ‘browser’ they had been using. The amount of time needed to complete the questionnaires at baseline and follow-up and the frequency of reminders were generally acceptable. Several of the participants said that they would have preferred longer access to the website as the limited period of 2 weeks meant that they had not had time to visit it as much as they would have liked. One man commented that, despite the e-mail reminders, he had been taken aback to discover that he no longer had access.

Participants were generally positive towards the website content, navigation and layout (comparator and intervention). Participants confirmed that certain aspects of the guidelines which had been used to develop the sites, such as the importance of reputable branding and the lack of commercial advertising, supported their impression of the trustworthiness of the website.

The interviews also shed light on the nature of usage and engagement with the websites, and how this differed from a pragmatic ‘real-world’ use of online resources. Some participants (including those allocated to the intervention and to the comparator) reported methodically working through their allocated websites in the order that the material appeared, as they felt that this was the task required of them, rather than selecting the information that interested them. Similarly, some reported logging on to the site only in response to automated reminder e-mails.

Approach to analysis

The three researchers who had conducted the interviews and coinvestigator SW read the same few interviews and agreed a coding frame based on the elements of the engagement framework (gating, engagement and outcomes) and the seven domains from the conceptual review (finding information, feeling supported, maintaining relationships with others, affecting behaviour, experiencing health services, learning to tell the story and visualising disease). We also coded ‘trust’ and the emerging issue of ‘timing/need’, which reflected an important set of comments around the point(s) in any given health trajectory at which experiential information might be regarded as more relevant or helpful. Each of the researchers coded her own interviews using the coding frame and NVivo. We used constant comparison to develop our coding, and considered the reasons for any differences between the three conditions in preparing the thematic analysis for this chapter.

The qualitative analysis presented in this chapter is based on all of the interviews because people in both arms of the study were asked to discuss and compare the six different types of websites. People in the comparator arm as well as the intervention arm had some very interesting things to say about the pros and cons of PEx and facts and figures, sometimes prompted by the laminated cards that we used in the interviews and sometimes due to their reflections on the value, and limitations, of the website they had access to in the study. Quotations from the interviews are presented along with brief sociodemographic information about the participant and whether they had been in the comparator or intervention arm.

Finding information

The research interviewers were careful not to assume that people were approaching the websites primarily as a source of ‘health information’ (we were, after all, aware that there were at least six other domains that might be relevant). All of the participants made it clear that the need for ‘information’ was what often motivated them to look at health websites, and that they would all want to be able to find reliable factual information about their condition or health issue (see also Timing and need). There were people in both the comparator and intervention groups who said this kind of factual information was either the only or the main type of information they wanted. For example, one person in the asthma comparator arm volunteered the following comment in relation to not wanting to look at videos, before any interview prompts had been raised about the sharing of personal experiences:

I prefer reading information and I would view a site like this, not for chatty, touchy-feely stuff, but for information because I’m not particularly interested in other people’s experiences of asthma, because I’ve had it so long that I feel that, you know, they’re not going to be particularly interesting or helpful.

Asthma comparator, 58-year-old woman

This is also an indication of the importance of timing and need, to which we return below.

An asthma intervention participant commented on the relative priority they would attach to experiential information:

Well I would think that the most important part of it is the symptoms and the available choices and medication.

So the medical kind of factual?

Yeah. The other thing about people’s experiences and what have you, although interesting, are not as important. I mean what everybody wants is to alleviate their symptoms and if the money’s there to break it up a bit with some personal experiences so much the better because it does make it easier to read if you get my meaning?

Asthma intervention, 66-year-old man

As anticipated, for the MS carers learning through other people’s experiences included getting practical advice about equipment and adaptations to the home. The value of hearing from other people who have ‘been there’ was frequently raised among those who had access to the intervention site.

So, and what was useful was, it’s all very well going to a site that tells you what’s available but it doesn’t tell you how people have coped with it, and that’s why the carers’ thing is so much more valuable. You know the physio will tell me but, it’s not quite the same as somebody who’s been in that situation and said, ‘This has worked for us,’ and you think, ‘Ah yeah, that might work for me’.

MS intervention, 59-year-old woman

Hearing other people’s experiences also helped some MS carers to make sense of what they had been told by various health professionals.

Because it highlighted . . . you go to, you go to see various people, you know take my partner and you go to see a consultant and people say different things and that was really useful because then you could see exactly what it was that they were, you know, meaning – for example benign MS.

MS intervention, 50-year-old woman

One person with asthma thought that people’s ‘detailed descriptions’ on the website were useful for those who had not talked about their illness to others.

To hear other people’s detailed descriptions of what they went through if you can put it that way, you know, because I’ve never talked to anybody about it. You know I know quite a few people who have asthma in various degrees but it’s not that we sort of talk about it, but it was interesting to hear other people’s views.

Asthma intervention, 78-year-old man

Concerns about the potential for harm included that information drawn from other people’s experiences could be either irrelevant or misleading. People with asthma and those caring for someone with MS pointed out that these conditions affect people very differently and therefore other people’s advice may not be relevant. One MS carer said that when it comes to adaptations to the home, even information from health professionals can be misleading ‘particularly in terms of finances and things’ as they might not understand that people struggling with limited funds may be encouraged to make expensive mistakes if they do not know what is going to work.

People who had many years’ experience of living with asthma or caring for someone with MS thought that it was unlikely that they would find anything new on a website; however, some who had looked at the PEx material in detail had been pleased and surprised to discover something new.

It was really interesting to see how other carers were coping with the idea of getting people in and things like that and how much problems there can be.

. . . We’ve had that a while so it’s not a major problem but you have to then buy new beds and things. And also you need two people. Well, really he prefers, and I do, that we can cope on our own as best we can. So there was a gentleman who mentioned the ceiling hoists which a friend of mine with MS also has, and that confirmed that that’s the only way to go you know.

So that was a really useful piece of information otherwise we’d have perhaps got hoists in and . . . I’ve been and tried them; I had a nice afternoon trying them out but, I did realise that they have limited potential.

MS intervention, 59-year-old woman

The ‘smoking cessation’ group felt, somewhat wearily, that they already knew everything there was to know about why it would be beneficial to give up. Some were evidently resistant to any more ‘information’ about the harms of smoking being thrust at them.

Feeling supported

Much of the literature about the role of health websites has focused on the potential for ‘peer-to-peer support’, which is arguably one of the most transformational aspects of the internet, especially for people who might be socially or geographically isolated, or who find it difficult to talk to others about their emotional responses. In our data this was particularly evident in the MS carers interviews; the following extract is from a woman who felt that she had benefited from access to the trial PEx website.

As someone who’s on my own, basically because relatives are far away, it’s quite useful; reassuring; encapsulates the most helpful comments of other carers; helps me realise aspects of [my husband’s] symptoms and behaviour which are to be expected, because it’s very difficult for him to know if it’s the MS if he has a tummy problem, or if it’s something else and then it’s easier for both of us to accept quite regularly. Reinforces the path we’re on – like I mentioned with carers – and you can see any alternatives might have bigger problems than they’d solve. And it helped about equipment; it helped more about emotions and every carer seemed to be unique and was finding the best way through which helped me think about the best way through.

MS intervention, 59-year-old woman

One of the participants with asthma commented that, among the benefits of seeing what other people have written about and dealt with, it is:

. . . also helpful to know (a) that you’re not alone and (b) how other people cope in those circumstances.

Asthma comparator, 45-year-old woman

The MS carer interviews included several examples of when hearing from another carer had helped (or could help) people to feel less alone and better able to cope. The MS carer quoted above had attended a face-to-face support group but felt that the PEx site was ‘a more useful thing’ for her:

I’ve been to carers’ meetings and they’re very helpful about what might be out there but it doesn’t get you into the emotion of being a carer in the same way. And I found out things, I can’t specify, but information about [husband’s] emotions which helps me to deal with them, and they don’t crop up at, you know other venues. So, it’s really useful when people talk about how they’re . . . I don’t know what you call them . . . the person you’re caring for feels and how their emotions are affected and it helps you to deal with it as well.

MS intervention, 59-year-old woman

Hearing about what other people are going through can sometimes help people to feel that they are comparatively well off. A woman in her sixties who was in her first few weeks of giving up smoking had not seen the PEx website, but volunteered early in the interview that she would have liked to see some other people’s experiences of trying to give up. She wanted to see not only the positive accounts but also what it was like for those who were struggling to give up, partly for ‘downward comparisons’, but also because she would ‘mistrust’ a site that was only positive. A person with asthma commented on the value of these ‘downward comparisons’ with people on the PEx website:

You also are thankful that you’re not as bad as some people you know because for some people it is really debilitating [. . .] I mean it doesn’t stop me doing what I want to do.

Asthma intervention, 78-year-old man

Carers of people with MS were aware that looking at other people’s experiences could be either uplifting or depressing. Online peer-to-peer contact could be a ‘mixed blessing’ or a ‘double-edged sword’, and it would not always be easy to predict how someone might respond to hearing someone else’s account:

It’s a balance the stories isn’t it . . . I mean the personal dimension where people have struggled and have, you know, made some progress or have come through well is very positive and inspiring but then equally without being, you know, depressing is also useful to know about the sort of difficulties where people have where, you know, the challenges that emerge as people become doubly incontinent and what you have to do about caring for that and all the rest of it because it’s something’s real and you don’t want people feeling, ‘Oh god I . . .’ you know either the person who’s suffering or the person . . . sounds awful when the person’s dealing with this, or the carer, ‘Oh god I’m inadequate, I can’t cope with this’.

MS comparator, 65-year-old man

Apologetic that he might be sounding unsympathetic, one person with asthma pointed out that people can tend to ‘witter on’ in forums and on blogs. Of course, a benefit of an online group is that one can remove oneself without causing the offence that might ensue in a face-to-face meeting. 4 Others wondered why it seemed to be that some people wanted to live their lives, and rehearse their health problems, in public and online. Those in the smoking cessation group differed from the asthma and MS carer participants in that several said that they really did not like the idea of being advised what to do – and they were doubtful that other people would want to hear about their experiences either:

. . . because, to me, it’s self-evident – if you want to give up smoking, you give up smoking.

Smoking intervention, 59-year-old man

Another smoker said that they would not want to ‘shove my opinions down other people’s throats’.

I don’t know . . . if I had breast cancer I wouldn’t want to read people’s last diaries maybe necessarily while I was undergoing my chemotherapy.

Yeah.

But if we’re talking about something like this I think it’s really helpful to have people who’ve had bad experiences or who have failed at various things because then it makes you just feel a lot better about, you know, the fact that you’re not coping well or you know you’ve given up smoking seven times in the last eight months you know and each time you’ve lapsed and you think, ‘Oh I’m hopeless,’ and you think, ‘Well so and so that happened to them and they persisted or they tried this other thing and it really worked,’ so I think that’s really good. I would always . . . I would mistrust a site I think that only had positive things.

Smoking intervention, 78-year-old woman

This perspective was echoed by some of the people with asthma or MS carers who, although a little concerned about unmoderated content on online forums, reasoned that it also would not be helpful to exclude negative stories:

Well there is a danger there isn’t there that if you’re feeling vulnerable and in need of support anyway and you don’t have any other support network and you know if there are forums they’re usually . . . I wouldn’t say uncensored but they do tell [. . .] they shouldn’t exclude people who tell a negative story otherwise there’s not much point in having a forum, it just becomes a controlled network, so there is that danger. Now as an individual you might come away from there feeling, ‘Well at least I’m not as bad as that,’ which isn’t necessarily a good idea but it might give some people some support to say look there’s people who are worse off, you could do better or you might identify them as a bit, you know, you’re being just as good, as bad as they are. So I don’t know about that one, could be a double edged sword. Overall I’d think it would come down as perhaps helping people who are seeking support overall.

Asthma intervention, 60-year-old man

The same participant drew attention to the value of the curated collection of clips on the intervention PEx site, which avoided being repetitive or presenting only one perspective.

Those were interesting as well and I thought that would also give support to people who perhaps are a little bit anxious particularly if they’ve been recently diagnosed or it’s grown over, you know, in recent times, so yeah, so that’s where I first went yeah.

OK. And did you look at any of the clips, the video clips?

Yes yes. I suppose also those are well put together. I think the . . . sometimes if they’re too repetitive or the people aren’t explaining themselves very well that might put you off, but if they’re put together in a supportive way I think they would provide people with encouragement yeah [. . .] I thought overall they were well put together; people . . . it’s always a danger if you’re using non-professionals I suppose or people . . . they just look at it from their own angle but then I suppose that’s the point and as long as they put it over clearly then yeah I think they were . . . they added to the website yeah.

Asthma intervention, 60-year-old man

The PEx site was designed to cover many different perspectives on each topic so that people using the site could find material that resonated with their own views and experiences. This had not been successful for one of the MS carers, who felt that her perspective was not included because she cared for a partner with MS but they did not live together:

In some ways it was quite disappointing to use partly because the examples . . . I think I was looking for something that wasn’t there, possibly because our situation is so different from most people in that situation. Because we don’t live in the same place; because we’re not man and wife; so I’m much more like a friend and I appreciate that this is designed for friends as well. But obviously a lot of the things that were said were, where somebody was in a very intimate relationship 24/7. So there were times when I thought, ‘Ah now that might give me a clue,’ and it didn’t quite hit the spot.

MS intervention, 59-year-old woman

Maintaining relationships with others

Carers of people with MS talked about the unmet need for support for people who are caring for a spouse with MS. One participant (in the non-PEx arm) explained how much he had gained from meeting someone else who was caring for a wife with MS. He reflected that he had not found anything similar on a website, although the need, evidenced by the high divorce rate (especially in cases when the woman has MS), seemed clear to him:

And we’d talked about things and we had exactly the same experiences of, when you go through that diagnosis it’s a big shock and then you worry a lot for a long time, and then things become normal. But, you redefine yourself as the partner of someone who has MS, and I found myself telling people that all the time. If they asked who I was, it was almost the first thing that came out all the time. And, it’s like you’ve . . . someone’s flicked a switch and you’re not who you were before and your partner isn’t who they were before. And, it’s really, really weird. I had never spoken to anyone else before who’d had that. And, it was just really incredibly helpful.

Yeah.

And that’s not . . . I haven’t found that on any website anywhere or anything like that. All you do see is that, you know statistics like; I think this is roughly right but, men who are partners, 75% of men who are partners of women with MS, divorce them. And, it’s about 25% of women whose male partners got MS, divorce them. And there’s a very big difference there but, I think there’s probably . . . and there is probably a need for some sort of support there that isn’t, that just isn’t there at the moment. Because, I think, MS is like one of those things . . . you know people try and hide Alzheimer’s because we’ve been through this with parents and god knows what, and partners of people with Alzheimer’s try and cover it up and everything. MS is one of those completely hidden things . . .

MS comparator, 59-year-old man

Sometimes the online environment can offer welcome anonymity. One of the smoking participants commented that it might be easier to share personal issues with strangers rather than with friends, family or health professionals:

. . . with a professional it’s a bit more like more like going to confession, do you know what I mean? There’s a sort of barrier . . . sometimes it’s easier to share personal things with someone you don’t know.

Smoking intervention, 78-year-old woman

One of the asthma participants recognised the potential value of the online forum because of his daughter’s experience of online communities for people with fibromyalgia:

What she does find is that it’s easier to cope with now there is a fair bit of it on the web and things like Facebook where they discuss how their days have been and that sort of thing and she’s not alone . . . this is . . . she’s found very much a great help, you know, that there’s people she can talk to over the web whereas she doesn’t physically know anybody in her local friendly circle that’s got it but there are people out there that she can contact and talk to and sort of say, ‘Oh I’ve had a rotten day,’ or, ‘I’ve felt better today,’ or whatever you know.

Asthma intervention, 78-year-old man

Affecting behaviour

Stories can inspire people to cope with their health problems. One man who had been diagnosed with asthma as a child was inspired and reassured to discover that a famous cricketer also had asthma:

I do remember there was an interview with I can’t remember who it was – it was a famous cricketer – who had asthma or who has asthma, and I do remember thinking, ‘Gosh well you know if this person can be a, you know, a test cricketer and he’s got asthma, you know, that’s a good thing I can run,’ because as a child, although Ventolin had been invented my GP had obviously never heard of it so I didn’t have any treatment; you know the only treatment for an asthma attack was to lie down and gasp on the floor and, you know, try and get fresh air. And so I was always scared of running and that sort of thing and so then in the second phase when I was older and then got the drugs I suppose at that point seeing success stories and thinking, ‘I can run if these other people can do things’.

Asthma comparator, 58-year-old woman

Interestingly, this was the same person quoted above in Finding Information who said, ‘I’m not particularly interested in other people’s experiences of asthma, because I’ve had it so long that I feel that, you know, they’re not going to be particularly interesting or helpful.’ Again, this illustrates how shared experience at the right time – success stories ‘at that point’, as she puts it – may be valuable.

One participant (in the non-PEx arm of the smoking study) felt that other people’s experiences of quitting (or trying to quit), if they were ‘smokers like me’, might be a ‘worthwhile addition’ to a website for smokers trying to give up:

But a worthwhile addition to it would be, people who have been smokers like me, for 47-odd years, and you know what’s your experience been actually finally quitting?

Right.

And how have you stuck with it? How have you done it? And there are lots of little things that help; there’s one thing in my view except my own inner decision.

Yes.

My decision.

So what other, you know what other types of experiences would you like to see? People who’ve . . . just people who have been successful or people who’ve tried and . . .?

People who’ve been . . . people who tried and failed even and are willing to try again.

Yeah.

And even people who say, you know, ‘Look this is the hardest thing you’ll probably ever do in your life’. It is very hard.

Smoking comparator, 59-year-old woman

Experiencing health services

There were a few accounts from people who had been encouraged to seek health care, although not necessarily because of the 2-week exposure to the trial websites. An asthma study participant said that taking part had reminded them that they were overdue to see their asthma nurse, which was probably simply a beneficial side effect of taking part in the research rather than a consequence of exposure to the intervention.

It has been suggested that health websites (in general) might support self-care by increasing people’s confidence in their ability to look after themselves. One of the participants from the asthma study described being reassured by the PEx website that their current approach to their health care was satisfactory, and a MS carer said that they felt a ‘bit more confident to go to the GP’ to ask for help:

Perhaps if anything it reassured me that I was doing just about most of what I could do to control the asthma and keep it in check, and it perhaps confirmed that there was nothing contradictory to what I’d been told or any course of action that, you know, any medication that I’m taking at the moment. There was nothing contradicting what I’d been told – different terminology but mounting to the same thing so I suppose if anything I came away feeling that this is . . . I had more confidence, it was reinforcing. If there was something glaringly different to what I’d been following then I might have gone back to my doctors and said, ‘Oi’. [laughs]

Yeah OK. Has anything that you looked at prompted you to go and talk to anyone else; go back to your GP; talk to support groups?

I think it’s made me a little bit more confident to go to the GP with problems.

[. . .] Yeah and I suppose websites could go either way couldn’t they, you know that online stuff might put you into downward spiral or, it might prompt you to think, ‘Actually there’s other people out there who’ve gone and got help, maybe I should do that’.

MS intervention, 59-year-old woman

It has made me think more and more that that’s what I should be doing. [Interviewer: Hm mm.] And it’s my mind-set and also habit as well and I think these e-cigarettes is to look . . . well say a lot of friends are doing those now and that seems . . . they seem quite happy with that. [Interviewer: Yeah.] And I think that might be sort of one of the aids to . . . and also I need to go and see the doctor and say, ‘Right, can I have some patches as well,’ and work out the best way for me to stop smoking.

Smoking comparator, 47-year-old man

Learning to tell the story

Our conceptual review indicated that one of the consequences of accessing online PEx might be that people would learn new ways to talk about their health, which might help them to describe their condition more accurately (or more persuasively). Any such change would probably be gradual, diffuse and difficult to assign to any one influence. However, one man who was caring for his partner with MS explained how hearing men talk about their emotions might help others to express themselves. This is, of course, related to ‘support’ but provides a particular stimulus to action:

I think there’s a gender issue here too because women have traditionally, for historical reasons, had a greater caring role and they still do because we still have a patriarchal society [. . .] but many women find it more easy to talk about their emotions and their feelings and to exchange. Men tend to not do that so easily and tend to internalise and block down, and I think the opportunity for, you know, a male person, for everyone, but maybe to actually see other male carers talking [. . .], ‘Oh so you know, actually I’m not alone; actually those are issues and actually I have these feelings of, you know anger and upset and whatever and it’s actually OK to feel those things because that’s an understandable human reaction, but then I’ve got to sort of deal with those and channel onto . . .,’ you know so all of that’s . . . and I guess this must be true . . . lots of support and therapy across all sorts of issues and conditions, not just MS but yes, I’ll take the view that the ability to share that is quite important.

MS comparator, 65-year-old man

Another man caring for his wife with MS expressed a similar idea about how hearing about other people’s problems might encourage a person to ‘talk to someone else about how they feel’:

I think, actually getting things off your chest or talking about things is always helpful. And, if it’s something that you find difficult to talk about with other people then it’s probably even more helpful to you to get it off your chest. And, if someone experiences someone else doing that it can possibly help them. Either just to read about it or, hear about it or, to encourage them to go and talk to somebody about how they’re feeling I think. For their own sort of mental health really.

MS comparator, 59-year-old man

In contrast, several of the participants in the asthma and smoking cessation groups were sceptical about the value of telling other people about health problems. Some expressed amazement that anyone would want to write, or read, a health blog. A person with asthma commented:

People are all different. It might be a generational thing as well in that it was the way I was brought up – basically, you know, you sort yourself out, you don’t go lumbering other people with your problems; or nod and say yes but by and large they’re not interested, they have their problems and so why should they want to hear yours?

[laughs]

I don’t know yeah.

If you thought they were going to help someone.

I mean the worst person to meet in an office environment is when you ask them how are you and you get the whole nine yards and you think oh . . . that’s unkind to some extent but because there are people who definitely need help and support and you would do – give them support – but I mean as a general rule, you know I’ve done it myself – you know you’re feeling like death but yeah you’re fine.

Asthma intervention, 60-year-old man

Visualising disease

The conceptual review drew attention to the inherently visual nature of health websites. Many sites include pictures of people with a health condition, and other illustrations. These images, intentionally or otherwise, position the condition in particular ways.

In the interviews we asked people if there were any images they would prefer not to see on a health website. Responses were clearly divided by condition. While a few of the people with asthma discussed design issues and talked in general about the benefits of breaking up text with images, they did raise concerns about negative or distressing images. In comparison, MS carers were aware that images of people with advanced disability might be distressing, particularly for the person with MS, although one remarked that ‘with MS there aren’t so many gory pictures because it’s not that sort of disease’:

They don’t want it in their face; they’ve got it, they want access to things to help them but they don’t want to really be reminded of it in a sad sort of way and that’s . . . websites can do that too seeing people severely disabled in a wheelchair, for my husband for example, probably wouldn’t do him any favours and it doesn’t actually interest me that much other than from the practical point of view that that particular piece of kit being suitable for my husband, then I would want to see it.

MS comparator, 64-year-old woman

The smoking cessation group expressed some quite strong feelings about the images that have been used in anti-smoking campaigns, which some suggested could be counterproductive:

That I do not want to see? Pictures of diseased lungs. [Interviewer: Hm mm.] I know that and I don’t want to see it because it makes me instantly want to go and smoke a packet of cigarettes. [Interviewer: Hm mm.] Because it makes me so nervous.

Are there any things that you’d really rather not see on a site that’s supposed to be helping and supporting people to give up smoking?

Well yes. I mean we’ve all seen the photographs, the medical photographs of lungs of someone with cancer. [Interviewer: Yeah.] And I don’t think we need to see them any more. [Interviewer: No.] You know I mean that kind of shock tactic it didn’t make me want to stop. [Interviewer: No.]

I just didn’t bother to look at the picture; I would open the packet in such a way that I didn’t look at the picture. And so I think most people see through that.

Yeah. You know you can’t . . . ultimately there is only one way and that is really through understanding and loving support.

Smoking intervention, 59-year-old woman

In contrast, the next quotation was from the one participant who suggested that images should be more graphic to shock people – on the grounds that everyone now knows that smoking is ‘not a good thing’ and people may need an extra push to stop:

I think people use the pictures which have a more hard-hitting effect.

OK so by graphic you mean sort of like . . .?

Images of lungs and that sort of thing yeah.

Lungs yeah. OK yeah.

And also the effect it has on others. I think that’s more of a positive for people to say, ‘Right . . .’ because I do think it’s . . . it’s not so much educational because I think most people have got a general, you know, they know it’s not a good thing. But I think if you can actually show it through and through and show that, you know, from start to finish, this is what will happen, not could happen, will happen.

Smoking comparator, 47-year-old man

One participant in her mid-sixties, who had just given up smoking, also commented on the image of a professor and expert on smoking cessation research, who was filmed for the website, talking in his office. She reacted to the way that he was positioned ‘against his files and his study papers and all the rest of it; his research documentation’. As someone who admitted that she had rather enjoyed her self-image as a bit of a rebel, she was not persuaded by the image of ‘authority figures’ telling her what to do:

But the photograph is totally, it’s just totally inappropriate you know [. . .] You see it interests me why I’m an addictive personality but the photo . . . he’s a great guy, no problem with that, but why put that, why? What is the reason for showing the professor’s photograph against his files and his study papers and all the rest of it; his research documentation – why, why do that? What purpose does it serve me who’s trying to stop smoking?

Authority? [laughs]

It’s an authority figure. Well you know [makes a ‘V sign’ with fingers and laughs], they’re closed, they’re closed. I used to put a cigarette between those [laughs].

Smoking comparator, 66-year-old woman

Facts, figures, experiences and blogs

People were asked to compare and comment on the different types of health websites, using a series of laminated cards with screengrabs from different types of sites accompanied by an explanation from their interviewer. Among all three groups and both arms of the feasibility trial, there was general agreement that everyone needs the facts and figures from reliable, non-commercial sources such as the NHS, well-known universities and voluntary organisations. People were very alert to the dangers of advertising and commercial influence on the site content:

I quite like the idea of factual information; again it’s just telling it straight and being able to work your way through something logically and forming your own thought process over it and that way you’ve got something to take away and discuss with your health professional or friends and family, whatever, to try and ascertain the meaning. I suppose that’s the way I tend to work so I’d say that was quite useful.

Asthma comparator, 45-year-old woman

When it came to patients’ experiences, there was considerably more variation in response. As we have indicated previously, some of the people we interviewed were at a loss to understand why anyone would ever want to talk about, or hear, anyone’s experiences of their health issues. People who had experience of caring for someone with MS had often realised the value of hearing about other people’s experiences, but they and the people with asthma emphasised that the conditions affect people very differently, making it hard to learn from a single blog, or an individual posting on a forum. There were also concerns about unmoderated content which might attract highly damaging comment and activity from ‘trolls’ on forums.

Those who had visited the PEx repeatedly commented on the benefit of being able to see many different people’s experiences gathered together under topics so that they did not become too repetitive (see Feeling supported).

It was also clear from the interviews that although there are some functions that can be very well performed by a website, the material is seen as a supplement, not an alternative, to care from a professional. Similarly, those who valued the PEx material also wanted to see facts and figures:

And that’s where a website is very handy because you can back-up what you’ve been listening to and also give you that little bit . . . well no GP’s going to know exactly how much you can take in or how much you want to as some people don’t want to know do they?

Mm yes or don’t want to know right now but they might want to know later.

No they might want to know later yes quite. If you’d just had a really serious diagnosis you might not want to know immediately anything more but you might . . .

MS intervention, 59-year-old woman

I think things like this, well the factual ones should encourage people to use the services less but when they do go it’s for more important issue which can’t be resolved you know outside the health service.

OK and is that something that would make you think, ‘Ooh quite like to know more about that, I might ask the GP,’ or . . . or look it up further on Google or something or other?

I would probably try and read more about it I suppose yeah. I suppose if I wanted more information after I’d tried to do that then I’d probably would talk to the GP or the nurse who deals with it.

Asthma comparator, 57-year-old man

Yeah. I’d want the facts first, only then would I perhaps want to look at individual stories and that’s only if I could search them to follow somebody who’d perhaps got as near enough the same condition as myself and I’d then follow them through but if I had a condition that was concerning me I’d want facts and information about it and then I’d go to my doctor and take a conventional approach to it.

Asthma intervention, 60-year-old man

Timing and need

Timing and need was a strongly emergent theme from the interviews. We have already touched on this in the example of the 58-year-old woman in the asthma comparator group who said that she was now ‘not particularly interested in other people’s experiences of asthma, because I’ve had it so long that I feel that, you know, they’re not going to be particularly interesting or helpful’. The same participant also told us that at one stage in her illness she had found the example of a famous cricketer with asthma helpful in encouraging her to think she could still go running.

Many of the people recruited for the asthma feasibility trial had had the condition for some time; they generally felt well informed and their condition was well controlled. It is not surprising, therefore, that they often felt that they had little to gain from either the facts and figures or the personal experiences sites. However, they could see that they or others might have benefited from information at a different point in their illness, such as when they were newly diagnosed or experiencing changing symptoms. The same woman in the asthma comparator group as mentioned above explained further:

But of course somebody who’s newly diagnosed . . . or particularly if they had a child who was newly diagnosed, would be looking for totally different things . . . and probably with the other conditions you’ve been looking at as well, somebody who is newly diagnosed is going to be in a sort of, you know, shock/horror ‘What do I actually do? What does it mean? Am I the only one?’ sort of thing. It’s almost a completely different thing . . . I suppose it almost feels like making a bit too much of a fuss about it – talking to other people about it. But then you see my asthma’s pretty well controlled most of the time . . . I mean going back to a parent of a child who’d been newly diagnosed, that’s such a completely different sort of experience. I think if you’re afraid of the condition then I think having discussions with other people is going to be perhaps more effective. So I think again it comes back to this difference between, you know, if you’ve had asthma for, you know, over half a century as I have, versus if it’s all completely new and you don’t know anything about it – you’re going to want completely different sorts of support and information.

Asthma comparator, 58-year-old woman

This example is typical of how people talked about personal variation in the time at which they might need different sorts of information.

One woman in the intervention arm, who was caring for her husband with MS, explained that the trial had come at a particularly helpful point for some pieces of information and emotional support, as she had been experiencing some depression because of her sense of isolation and sadness at her husband’s disease progression. The following extracts show how she selected material for her own current needs. In addition, searching the site online was, for her, a more efficient way of gleaning others’ experiences than a support group would have been:

I find it difficult to find the time to talk to people a lot and go to MS meetings and things like that, and very often you’re talking about people’s relatives and things, whereas I find that the website condensed the sort of information that I needed. So I did find out things which could have taken quite a long time to find out otherwise . . . I had to by-pass a lot of things because my husband is now chair bound. So, I didn’t need to look at anything about mobility cars, scooters, you know anything like that because we’ve already been through that. So, it was particularly things for the future. He’s not too worried but I like to be prepared.

OK so quite selective, quite targeted?

Very targeted, yes, probably missed half of it because it’s no longer appropriate, which doesn’t mean it wasn’t of real value to people who are in a different position . . .

What kinds of things were useful?

Well I did look at some of the things that we’re sort of currently in, like [paid] carers and things, and it was really interesting to see how other carers were coping with the idea of getting people in and things like that, and how much problems there can be . . . I was particularly interested in hoists because he doesn’t need one yet, but we need to know what sort of hoists might be available [. . .] I think initially when you’re diagnosed you want to look at a whole lot of information. [partner] and I were very different; he didn’t want to know, he just wanted to get on. I wanted to be prepared . . . so I wanted to get hold of literature and find out as much as I could. So it’s very useful in those early stages I think . . . And what was useful was, it’s all very well going to a site that tells you what’s available but it doesn’t tell you how people have coped with it, and that’s why the carers’ thing is so much more valuable. You know, the physio will tell me but, it’s not quite the same as somebody who’s been in that situation and said, ‘This has worked for us,’ and you think, ‘Ah yeah, that might work for me’.

MS intervention, 59-year-old woman

At the same time, she was selective about experiences that she knew she did not want to see, perhaps because they were upsetting:

When people were in a similar situation to myself then I would, you know watch the whole clip. And I think I tried to look at every clip where it might be relevant. As opposed to clips where people were in the very early stages and talking about holidays – dream on [laughs] – you know, they were irrelevant and probably painful.

Work package 1a and overview

Ziebland S, Wyke S. Health and Illness in a Wired World: How Might Access to Patients Experiences on the Internet Affect People’s Health? Oral presentation given at the 25th British Computing Society conference on Human-Computer Interaction as part of an iPEx organised workshop, ‘Online patient experience (PEx) and its role in e-health’, Northumbria University, Newcastle upon Tyne, UK, July 2011.

Wyke S, Ziebland S. Health in (Y)our Hands. Oral presentation, Radboud University, Nijmegen, the Netherlands, October 2012.

Wyke S, Ziebland S. Health and Illness in a Connected World: How Might Patients’ Experiences on the Internet Affect People’s Health? Oral presentation, University of British Colombia, Vancouver, BC, Canada, November 2011.

Ziebland S. Health and Illness in a Connected World. Keynote. Qualitative Inquiry Conference, Ben Gurion University, Negev, Israel, February 2012.

Ziebland S. Health and Illness in a Connected World. Invited seminar, DIPEx International and Social Change, University of Monash, Melbourne, VIC, Australia, February 2012.

Ziebland S. The Rise of ‘Patients’ Experiences’: Evidence, Distraction or Final Arbiter. Cochrane Lecture, London School of Hygiene and Tropical Medicine, Society for Social Medicine, 2012.

Ziebland S, Wyke S. Health and Illness in a Connected World: How Might Sharing Experiences on the Internet Affect People’s Health? Abstract and oral presentation, Science and Technology Studies Unit Conference, University of York, York, UK, July 2012.

Ziebland S, Wyke S. Health and Illness in a Connected World: How Might Sharing Experiences on the Internet Affect People’s Health? Oral presentation, British Sociological Association Medical Sociology Annual Conference, University of Leicester, Leicester, UK, September 2012.

Ziebland S, Powell J, Wyke S, Locock L, Farmer A, Perera R, et al. Examining the Role of Patients’ Experiences as a Resource for Choice and Decision-Making in Health Care: An NIHR Programme for Applied Research. South West Society for Academic Primary Care, January 2013.

Work package 1b

Mazanderani F, Powell J, Locock L. Being Differently the Same in the Sharing of Experiences in Healthcare. Oral presentation given at the 25th British Computing Society conference on Human-Computer Interaction as part of an iPEx organised workshop, ‘Online patient experience (PEx) and its role in ehealth’, Northumbria University, Newcastle upon Tyne, UK, July 2011.

Mazanderani F, Powell J, Locock L. Biographical Value: Exploring the Economics, Epistemics and Ethics of Patients’ Experiential Narratives. Oral presentation given at the British Sociological Association, Medical Sociology Annual Conference, University of Chester, Chester, UK, September 2011.

Work package 1c

Hardy C, Hinton L. Mazanderani F, Kelly L. Communicating with Other Patients Online: Exploring the Effects of Sharing Health Experiences on the Internet. Oral presentation given at the European Association of Communication in Healthcare Annual Conference, as part of an iPEx organised symposium, ‘Communicating with other patients online: exploring the effects of sharing health experiences on the internet,’ University of St Andrews, St Andrews, UK, September 2012.

Kelly L, Jenkinson C, Ziebland S. Measuring the Effects of Exposure to Health Websites. Oral presentation given at the European Association of Communication in Healthcare Annual Conference as part of an iPEx organised symposium, ‘Communicating with other patients online: exploring the effects of sharing health experiences on the internet,’ University of St Andrews, St Andrews, UK, September 2012.

Kelly L, Ziebland S, Jenkinson C. Developing and Pre-Testing an Item Pool Relating to the Effects of Exposure to Health Websites. Abstract and poster presentation at International Society for Quality of Life Research, Budapest, Hungary, October 2012.

Kelly L, Simpson H, Verjee-Lorenz A, Clayson D, Churchman D, Jenkinson C, Ziebland S. Translatability Assessment of the e-Health Impact Questionnaire (eHIQ). Poster presentation at the 18th International Society of Pharmacoeconomics and Outcomes Research, New Orleans, LA, May 2013.

Kelly L, Simpson H, Two R, Ziebland S, Jenkinson C. Measuring the Effects of Using Health-Related Websites: Translatability Assessment of Questionnaire Items in the E-Health Impact Questionnaire (eHIQ). Poster presentation at Med 2.0, London, UK, September 2013.

Kelly L, Jenkinson C, Ziebland S. Developing a Tool to Measure the Effects of using Health-Related Websites (the e-Health Impact Questionnaire). Oral presentation, International Society for Quality of Life Research, Miami, FL, October 2013.

Kelly L, Ziebland S, Jenkinson C. The e-Health Impact Questionnaire: Developing a Tool to Measure the Effects of using Health-Related Websites. Value Health 2013;16:A605. Poster presented at International Society of Pharmacoeconomics and Outcomes Research 16th Annual European Congress, Dublin, Ireland, November 2013.

Work package 2a

Mazanderani F, Powell J. Bodies of Knowledge: Reconfiguring Multiple Sclerosis as a Vascular Disease. Making (In)Appropriate Bodies Conference, hosted by the Science Studies Department, University of Vienna, Vienna, Austria, December 2011.

Mazanderani F, Powell J. Anecdote and Evidence: Social Media and the Enactment of ‘Proof’ in a Controversial Theory about Multiple Sclerosis. Oral presentation, The Co-Production of Knowledge: Social Media, STS and . . ., hosted by the University of York, York, UK, July 2012.

Mazanderani F, Powell J. Platforms, Patients and the Production of Experience. Oral presentation given at the European Association of Communication in Healthcare Annual Conference as part of an iPEx organised symposium, ‘Communicating with other patients online: exploring the effects of sharing health experiences on the internet’, University of St Andrews, St Andrews, UK, September 2012.

Mazanderani F, Powell J. An Aesthetics of Care? Attraction, Attachment and Alignment in Online Multiple Sclerosis Communities. Abstract and oral presentation, Society for Social Studies of Science and European Association for the Study of Science and Technology Joint Conference, Copenhagen, Denmark, October 2012.

O’Neill BG, Mazanderani F, Powell J. ‘People Power’ or ‘Pester Power’? YouTube as a Forum for the Generation of Evidence and Patient Advocacy. Oral presentation, Royal Society of Medicine, London, UK, November 2012.

Mazanderani F. Illness Narratives, Experiential Knowledge and Contemporary Regimes of ‘Living With’ Disease. Oral presentation, Medical Sociology Annual Conference, University of York, York, UK, September 2013.

Mazanderani F, O’Neill B, Powell, J. YouTube and Patient Activism: Online Video and the Generation of ‘Experiential’ Evidence. Oral presentation, Med 2.0, London, UK, September 2013.

Mazanderani F and Powell J. When Bodies Fall Apart. Internet Use, Patient Activism and the Mediation of Medical Knowledge. Oral presentation, Medical Sociology Annual Conference, University of York, York, UK, September 2013.

O’Neill B, Ziebland S, Powell J. Health Literacy and the Internet: EXperiences Evaluated in a Randomised Trial (EXPERT). Oral presentation, Connected Life, Oxford, UK, 2014.

Work package 2b

Mo P, Sillence E, Briggs P, Harris P. What Kind of Information on Smoking Cessation is Available Online? A Content Analysis of Common Websites. Oral presentation, 25th British Computing Society conference on Human–Computer Interaction as part of an iPEx organised workshop ‘Online patient experience (PEx) and its role in ehealth’, Northumbria University, Newcastle upon Tyne, UK, July 2011.

Hardy C, Sillence E, Briggs P, Harris PR. A Toolkit Approach for eHealth research: Using Focus Groups, Interviews, Data Logging, and Eye Tracking to Examine Individuals across a Range of Health Conditions. Oral presentation, European Association of Communication in Healthcare Annual Conference as part of an iPEx organised symposium ‘Communicating with other patients online: exploring the effects of sharing health experiences on the internet’, University of St Andrews, St Andrews, UK, September 2012.

Hardy C, Sillence E, Briggs P, Harris P. Engaging With Online Patient Experiences: Exploring Differences Between Health Groups. Abstract and oral presentation, Global Health Conference, Venice, Austria, October 2012.

Briggs P, Hardy C, Sillence E, Harris P. An Engagement Framework for Understanding the Communication Needs of Different Health Groups. Poster and extended abstract, CHI Conference on Human Factors in Computing Systems, Paris, France, 27 April–2 May 2013.

Sillence E, Hardy C, Briggs P. Why Don’t We Trust Health Websites that Help Us Help Each Other? ACM Web Science 2013, Paris, May 2013.

Hardy C, Sillence E, Briggs P, Harris P. Online Peer-to-Peer Healthcare: Exploring Push-Pull Factors of Engagement. Oral presentation, Med 2.0, London, UK, September 2013.

Briggs P, Hardy C, Harris PR, Sillence E. Patient-led Perspectives on eHealth: How Might Hyperpersonal Data inform Design? Proceedings of HCIK, ACM Press, Korea, 2014.

Sillence E, Hardy C, Harris P, Briggs P. Modelling Patient Engagement in Peer-to-Peer Healthcare. 23rd International World Wide Web Conference, Korea, April 2014.

Sillence E, Harris P, Briggs P. Assessing Patient Experience and Patient Preference when Designing Web Support for Smoking Cessation. Extended abstract and oral presentation, Association for Computing Machinery conference on Digital Health, Florence, Italy, May 2015.

Work package 3a

Hughes N. Personal Identity and Caring: Stories Told by Relatives and Friends of People with Multiple Sclerosis. Oral presentation, Contemporary Narratives of Care, Birkbeck, London, October 2013.

Kirkpatrick S, Locock L, Salisbury H, Ziebland S. ‘I’d Always Seen Myself as Invincible but Suddenly I’d Got this Condition that Wasn’t Going to Go Away’: The Emotional Impact of a Diagnosis of Asthma in Adulthood. British Sociological Association Medical Sociology Annual Conference, Aston University, Birmingham, UK, September 2014.

Hughes N. Personal identity and role of ‘carer’ among relatives and friends of people with MS. Oral presentation, 6th National Conference for Psychologists Working in MS, Genoa, Italy, October 2014.

Work package 3b

Newhouse N, Ziebland S, Locock L, Perera R, Farmer A, Powell J. Sharing Experiences Online: Design of a Randomised Trial to Determine the Feasibility and Efficacy of Three Novel Internet-Based Interventions Based on Patient Experience. Oral presentation, Med 2.0, London, UK, September 2013.

Newhouse N. The EXPERT study: Patient Experience as Information Resource: What Does the Right ‘Dose’ Look Like? Oral presentation, UK Society for Behavioural Medicine, Oxford, UK, December 2013.

Powell J, Ziebland S, Newhouse N. The EXPERT Study: Randomised Controlled Trial of an Internet-Based Intervention Harnessing Patient Experiences to Support Self-Management of Asthma. Oral presentation, Med 2.0, Malaga, Spain, October 2014.

Newhouse N. The EXPERT Study. Patient Experience as Information Resource: What Does the Right ‘Dose’ Look Like? Oral presentation, Centre for Behaviour Change Conference, London, UK, February 2015.

Powell P, Newhouse N, Martin A, Jawad S, Yu L-M, Davoudianfar M, et al. Randomised Feasibility Study of a Novel Experience-based Internet Intervention for Smoking Cessation. Poster presentation, Association for Computing Machinery Conference on Digital Health, Florence, Italy, May 2015.

Powell P, Newhouse N, Martin A, Jawad S, Yu L-M, Davoudianfar M, et al. The EXPERT Study: Feasibility Randomised Controlled Trial of an Internet-Based Intervention using Patient Experience to Support Self-Management of Asthma. Oral presentation, Society for Academic Primary Care Conference, Oxford, UK, July 2015.

Overarching programme

Ziebland S. Health and Illness in a Connected World: Understanding the Role and Value of ‘Patients’ Experiences’ Online. British Sociological Association Medical Sociology South Coast Conference, Brighton, UK, July 2015.

Other outputs

Series of presentations for undergraduate students from Radboud University, Nijmegen, the Netherlands, October 2013:

• O’Neill B. ‘People Power’ or ‘Pester Power’? YouTube as a Forum for the Generation of Evidence and Patient Advocacy.

• Newhouse N. EXPERT Trial.

• O’Neill B. Technology of Stories: Health Literacy and EXPERT.

A secondary analysis of the asthma qualitative interviews in WP3a contributed to the evidence considered by the National Institute for Health and Care Excellence Quality Standard on asthma.

A new website on experiences of asthma: www.healthtalk.org/asthma.

A new website on experiences of giving up smoking: www.healthtalk.org/giving-up-smoking.

A new resource for the family and friends of people with MS: www.healthtalk.org/ms-friends-family.

Examining the Role of Patients’ Experiences as a Resource for Choice and Decision-Making in Health Care. iPEx programme dissemination event for senior policy-makers, London, June 2014.

Examining the Role of Patients’ Experiences as a Resource for Choice and Decision-Making in Health Care. iPEx programme final dissemination event, London, June 2015.

Preamble

You have been taking part in a trial of different types of health website. We’d like to hear your views about taking part in the study, and understand more about how people may use such sites and why.

Reassure that this is not a test.

• Could you start by telling me why you agreed to take part?

• What were you expecting would be involved?

• Did it meet your expectations?

• What browser did you use?

Part A: ‘process questions’

Part B: principles of PEx and other content – a key area

We’d like to talk a bit now about the different types of health-related content that people may visit online.

Broadly, we might distinguish between web site content that is:

• content provided by professionals or medical charities – facts and figures about conditions

• collections of people’s experiences of particular health issues or conditions

• individual people’s blogs about their health

• patients’ forums/chat rooms

• feedback and reputation sites about people’s experience of particular services.

PART C: sharing your own experiences

We are interested to know what you think about sharing your own experiences with other people (apart from health professionals).

• Have you ever told people who you do not know, or do not know well, about your experiences of asthma/being a carer/trying to give up smoking? This might be in a face to face support group, an interview, an article, in an online forum or blog, or in a book.

• Do you feel you ever would?

• Are there some types of experiences you think people might be more or less likely to tell others about (expand if possible).

• Do you feel any more likely to tell YOUR story since seeing these other people’s experiences – if so, why do you think this might be? If not, why not – has it put you off in any way? Do you feel that your experiences are already covered and there is no need to add?

• Other reasons – would be good to explore this quite thoroughly since not much known about why people do and don’t.

Final questions

‘Trade off‘ questions.

• If voluntary organisations and the NHS are going to fund websites what type of content do you think it is most important for them to provide (could ask respondents to rank the laminated cards and tell us why)?

• Would you like a summary of the findings in due course?

Preamble

You have been taking part in a trial of different types of health website. We’d like to hear your views about taking part in the study, and understand more about how people may use such sites and why.

Reassure that this is not a test.

• Could you start by telling me why you agreed to take part?

• What were you expecting would be involved?

• Did it meet your expectations?

• What browser did you use?

Part A: ‘process questions’

Part B: principles of patient experience and other content – a key area

We’d like to talk a bit now about the different types of health-related content that people may visit online.

Broadly, we might distinguish between website content that is:

• content provided by professionals or medical charities – facts and figures about conditions

• collections of people’s experiences of particular health issues or conditions

• individual people’s blogs about their health

• patients’ forums/chat rooms

• feedback and reputation sites about people’s experience of particular services.

PART C: sharing your own experiences

• We are interested to know what you think about sharing your own experiences with other people (apart from health professionals).

• Have you ever told people who you do not know, or do not know well, about your experiences of asthma/being a carer/trying to give up smoking? This might be in a face to face support group, an interview, an article, in an online forum or blog, or in a book. Do you feel you ever would?

• Are there some types of experiences you think people might be more or less likely to tell others about (expand if possible).

• Are you likely to tell YOUR story on a website? If not, why not?

• Other reasons – would be good to explore this quite thoroughly since not much known about why people do and don’t.

Final questions

‘Trade-off‘ questions:

• If voluntary organisations and the NHS are going to fund websites what type of content do you think it is most important for them to provide (could ask respondents to rank the laminated cards and tell us why)?

• Would you like a summary of the findings in due course?