Perceptions of the authority of health information ... - Wiley Online Library

Perceptions of the authority of health information. Case study: digital interactive television and the Internet Blackwell Publishing Ltd.

David Nicholas, Paul Huntington, Peter Williams & Barrie Gunter, Ciber, Department of Information Science, City University, London, UK

Abstract As part of a Department of Health funded project nearly 2000 people were surveyed as to their use of two digital health information services, one on the Web and the other on digital interactive television (DiTV). The website was of a commercial company—Surgerydoor—and the DiTV service NHS based. This paper concentrates on the issue of trust in digital health information. Two of the main findings were that advertising was found to have an effect on trust, though the quality and type of advertising will impact in different ways on trustworthiness. DiTV subscribers who had either used the Living Health channel which carried NHS branded health information or had heard of the service, were more likely to say that the NHS was a symbol of trust for them compared with DiTV subscribers who had not used the service.

Introduction If the number of website quality rating systems published on the topic of health is anything to go by, health information professionals are deeply concerned with the quality and authority of health information.1 Digital health information offered directly to the consumer is of special concern. However, does the general public share the concerns of the health and information professionals? The digital environment is a complicated one, and with its own unique problems. The first of these is that it is a relatively new and fast-changing environment with new sites appearing all the time. This situation has led one commentator to observe that authority is ‘up for grabs’.2 Secondly, many people—especially the young—feel that if something is on a digital platform, then it must be authoritative. This may be particularly true with Correspondence: Ciber, Department of Information Science, City University, London EC1V 0HB, UK. E-mail: [email protected] http://www-digitalhealth.soi.city.ac.uk/isrg/doh.htm

regard to DiTV. Lastly, it may be difficult to judge authority because there are so many parties associated with the production of a digital information service. This paper represents an early attempt to obtain an understanding of how the public perceives the quality and authority of digitally delivered information. It is one of a number of studies being undertaken as part of a 2-year research project examining the impact of digital health information provision on the consumer, conducted on behalf of The Department of Health (The Web, the kiosk, digital TV and the changing face of consumer health information provision: a national impact study. April 2000–January 2003). Its aim is to determine the views of the general public in respect of digital health information, specifically with regard to issues of ownership, trust and the perceived authority of the information delivered via the Internet and digital interactive television (DiTV). The views of the users of two digital services were canvassed and evaluated, those of the commercial health website Surgerydoor (www.surgerydoor.co.uk),

© Health Libraries Group 2003 Health Information and Libraries Journal, 20, pp.215 –224

215

216

Authority of health information, David Nicholas et al.

and those of Birmingham subscribers to the cable TV broadcaster Telewest. Previous research Research into issues of the authority of Internet sites has almost exclusively centred around examinations of the innate quality of information itself, and how this can be assured, rather than the perceptions and attitudes of the information users. Whilst this might be natural, given the importance of authoritative and accurate information in the health field, the apparent lack of interest in public perception is a serious omission. This is particularly true when one considers the importance of targeting different groups with different messages, e.g. with information about influenza vaccinations, giving up smoking, etc., and the array of platforms available to information providers. The reason for this interest is that the tremendous growth of the Internet has ensured that there are thousands of unregulated, unsourced and, possibly, unscrupulous sites and documents accessible at ones fingertips through the World Wide Web. The authority of these sites has come under much scrutiny. Impicciatore,3 for example, assessed the quality of Internet sites, which focused on one particular medical condition. The researchers undertook Internet searches for ‘parent-orientated web pages relating to home management of feverish children’. The information given on the 41 sites retrieved was checked by comparison with published guidelines. Only four sites ‘adhered closely’ to official recommendations, the largest deviations being in sponging procedures and how to take a child’s temperature. Their most worrying finding was that two sites recommended practices that may actually induce coma. Complete and accurate information for the condition was ‘almost universally lacking’. Similarly, Griffiths and Christensen4 surveyed 21 websites that provided information about depression and assessed the quality of information against a number of criteria. They classified sites according to their stated purpose, ownership, involvement with major drug companies, and whether they showed evidence of a professional editorial board. They also scored site information against US federal best practice guidelines embod-

ied in the code published by the Agency for Health Care Policy and Research. They also assessed the identification, affiliations and credentials of authors associated with the sites. Findings indicated that the quality of content varied and was often poor in terms of these criteria. Furthermore, accountability criteria as indicated by the reported credentials of content authors might be poor quality guarantees. Instead, evidence of ownership and the existence of an independent editorial board were more useful quality indicators. Ways of guaranteeing the quality—and therefore, the authority—of Internet information are badly needed, and not just to enhance the public’s view of the authority of the provider. Wrongly diagnosed ailments or other manifestations of poor information provision could have fatal consequences. It is no surprise, therefore, that a number of health bodies and information providers have attempted to formulate policy statements, guidelines and principles regarding web-based health information.5,6 Kim et al.1 have surveyed these in order to identify concerns, recommendations and areas of consensus. Their ‘top ten’ aspects of sites were: • content (quality, reliability, accuracy, scope, depth); • design and aesthetics (layout, interactivity, presentation, graphics, multimedia); • disclosure of authors, sponsors, etc.; • currency of information and frequency of update; • authority/reputability of source; • usability, navigability; • accessibility and availability; • links, and quality of links; • attribution and documentation (references, balanced evidence); • intended audience. Concern has been expressed, somewhat ironically, about the quality of quality rating bodies and systems themselves. For example, HernándezBorges et al.7 in a study looking at the rating criteria of a number of systems, found that only three gave information about their own editorial boards—despite attribution, authority and openness generally being stated criteria for evaluating medical sites. Jadad and Gagliardi8 analysed sources that reviewed and rated health information sites,

© Health Libraries Group 2003 Health Information and Libraries Journal, 20, pp.215 – 224


and concluded that the evaluation instruments were not comprehensive and many did not actually measure what they claimed. Their presence therefore was not necessarily as informative as desired. Although no work appears to have been undertaken on the public perception of the quality of the NHS, with regard to information in terms purely of information provision, the present authors have looked at authority, albeit peripherally, in a usability study of the Surgerydoor website (on which the questionnaire reported later was posted).9,10 Participants were asked for their opinions on navigation, information content, aesthetic appeal, etc. and interviews revealed that advertisements on the site considerably lowered its level of authority in the eyes of the users. On Surgerydoor, these are displayed as banners across the top of each page, where they occupy a larger area than the title/logo of the company itself. Further, they contain flashing, eye-catching and distracting images. This gave the impression to many users that the key function of the site was for profit rather than to promote healthy living or provide health information services: ‘health shouldn’t be like that—an excuse to make money’. The extent of advertising led some to wonder how the organization was funded, and what the links were with the Department of Health and/or National Health Service. One respondent excused the advertisements on the grounds that the company had to obtain revenue ‘otherwise, how are they going to maintain the site? It’s better to have the information—even if surrounded by adverts—than to not have it at all’. Others said, noticing that content was provided by health-related bodies, that the company should be subsidized by these, to lower the extent of advertising. This sensitivity to the commercial aspects of the site may be due to respondents feeling that health care (and, by implication, health information) should be free at the point of delivery. For many people, the idea that health information can be regarded as a means of moneymaking is still anathema.10 It may be, however, that other factors related to the nature of the online environment come into play here. The issue of branding, advertising and brand recognition has taken on a new dimension on the Web. Simon11 writes that

the surfeit of choice online produces ‘a concomitant change in consumer attitudes’, moving them from what he describes as ‘receptive space’ to ‘sceptical space’. It may be that, with such a glut of information—including that concerned with health— users feel they do not have to tolerate the interference of advertising. They can move effortlessly from one brand to another. Travis12 suggests from a number of usability studies undertaken by Forrester Research, that fewer than 20% of website visitors look for a favourite brand—in keeping with the finding that the attribution of information and therefore authority was not important for the users studied. Issues of the perception of health information and of whether advertising and sponsorship lead to actual or perceived bias of content are clearly important. The NHS are keen to explore this issue as their current digital TV initiative13 includes health information appearing under the auspices of commercial service providers. The confusion about how Surgerydoor was funded showed a lack of easily available information about this on the website, and also the confusion caused by having links to external websites within the Surgerydoor frame, including pages produced by the NHS. Background to service content The Living Health service consisted of an onscreen database (mostly text) on a wide range of health topics largely adapted from NHS Direct Online, supplemented by content from other suppliers. The main menu offered seven information topics: • Today’s Health News (consisting of several stories each day, summarizing and explaining current items appearing in the popular press and /or the medical journals and linking, where appropriate, to additional, relevant information elsewhere on the service). • Healthy Living (general information pages on diet, smoking, exercise, etc.). • Men’s Health. • Women’s Health. • Children’s Health. • Illness and Treatment (pages on specific diseases). • Local Health Services (an information directory).


217

218


There were two ‘transactional’ services, both of which helped reinforce the linkage between Living Health and the NHS: (i) NHS Direct In-Vision and (ii) an On-Line Surgery Appointment Booking Service. In-Vision provides a one-way video link between a nurse in a NHS Direct Call Centre and the user at home; the video link is supplemented by a telephone link to provide oral communication between the two parties. The On-Line Appointments Booking Service allows users to book an appointment with their GP. Three GP surgeries in Birmingham are partners in this venture. The general service and GP booking element was launched on 28 June 2001. The In-Vision element was rolled out across Birmingham cable hubs starting 30 July 2001, and was available in all cable homes. Moving on to Surgerydoor, this has described itself as a website offering ‘the UK’s first electronic versions of official NHS information and the country’s biggest online health multistore’.14 It comprises over 5000 pages of content with access to an additional 40 000 pages of local health service maps and listings via web links to approved (and approving) external organizations. Apart from information provision, the site sells products and services ranging from pharmacy items to health foods and services. Revenue is generated from sponsorship, advertising and content syndication. Site content is divided into the following sections, each of which appear as main content entries on a side bar and contain a series of subheadings: • Health Daily: this contains a news and health alert, weather, and tips (such as ‘what to look for when considering joining a gym’). • Medical: comprises ‘Emergencies’, a Symptoms Index, ‘Diseases in Depth’, a Medical Dictionary and a Prescription Drug Guide. • Healthy Living: discusses Preventing Accidents, Dental Health Drugs and advice on alcohol consumption. • NHS and Benefits: provides listings on ‘Health in your area’ (searchable by postcode), including surgeries, dentists, hospitals, etc. • Complementary Medicine: includes an ‘A to Z’ of complementary medicine. • Travel Health: includes a section on vaccinations, giving a list of what vaccinations are required for each country, and a ‘Traveller’s

Health Kit’—a list of items that may be useful during a holiday. • Community and Fun: including a ‘Surgerydoor Health Magazine’, feedback and suggestions, health surveys, quizzes and competitions. • Shopping: an area where users can buy online. • Today’s Selection of Features and Topics: includes ‘This week’s Radio Times article from Dr Mark Porter’, who is one of the website team. Methods Data for this study were based on two large questionnaire surveys: the first was an online questionnaire of web users of a commercial Internet health site (Surgerydoor); the second was a postal questionnaire delivered to digital interactive television subscribers to the Telewest cable network in Birmingham. The online questionnaire was hosted on the Surgerydoor website for the month of November 2000, which the 2700 subscribers to the site’s newsletter were invited to complete. (3W Marketing Ltd were responsible for gathering the information, though not for its analysis.) However, only one-third of respondents said they were subscribers to the newsletter, implying that two-thirds were casual users and responded to the questionnaire as a result of their use of the site. People were made aware of the survey as a result of a pop-up box on various pages of the site. In total, 1068 users answered the questionnaire, 5% of the 21 118 unique IP addresses that were recorded as visiting the site in November 2000. This was not a random sample, however: respondents were a self-selected sample, and hence results can only offer an indication of likely patterns of uses of health information Internet sites. For digital television (Telewest) subscribers, data were obtained from a postal questionnaire sent by City University with literature promoting the ‘Living Health’ channel, a publicly funded commercial online health information channel. This was sent to all potential Telewest Birmingham subscribers, approximately 45 000 households. Seven hundred and twenty-three were returned. The two studies were very different. The Web study was an online study of a commercial website

© Health Libraries Group 2003 Health Information and Libraries Journal, 20, pp.215– 224


while the DiTV study was a postal questionnaire sent out to all DiTV subscribing households who could access the Living Health service. The age profiles between the studies were similar. Both recorded few respondents under the age of 15—less than half a per cent fell into this age group for both studies. Furthermore, both studies recorded that most respondents fell into the 35–54 age group; 49% for the website and 41% for the DiTV study. The DiTV did record more respondents in the over 55 age group; 30% fell in this age group compared with 18% for respondents of the website survey. The website recorded relatively more respondents as falling in the 15 –34 age group. Differences between the studies were apparent for gender. A large proportion (79%) of respondents to the Web questionnaire were women, whilst for the DiTV study only 53% were women. This is a large difference. However, it is not felt that this impacted on the trust variables that are at the heart of the investigation, as statistical analysis showed that gender was not found to be a significant factor in any of these variables.

Figure 1 We are particularly interested in your views about the Surgerydoor site. Please rate the following: trustworthiness

Results Website Users of the Surgerydoor website were asked how they rated the trustworthiness of the site. Perhaps unsurprisingly, only two respondents rated Surgerydoor’s trustworthiness as poor (Fig. 1). There were four possible responses: poor, OK, good and excellent. Three-quarters of respondents rated the site as either good or excellent. Respondents were further asked how many sites they actually visited when they last used the Internet to search for health information (see Fig. 2). Just under a third said that they visited just the one site, suggesting that these users are core users of the Surgerydoor site and trust the site sufficiently not to visit other sites for their health enquiries. Seventy per cent of users said that they had visited more than one site. For these information seekers visiting a number of sites may be a way of checking the authority of the sites. Figure 3 reports the relationship between how the respondent heard about Surgerydoor and the site’s rated trustworthiness. As can be seen, those

Figure 2 How many health Internet sites did you visit when you last used the Internet to search for health information?

Figure 3 How the respondent heard about Surgerydoor, by site trustworthiness


219

220


A person’s trust in the authenticity of a site was also found to be a significant factor on health outcome, with those users demonstrating the greatest trust being more likely to claim a positive health outcome.15 It was found that those rating the site’s trustworthiness as either good or excellent were more likely to say that they had been helped a lot and were less likely to say that the site was of no help. In all cases, users who did not rate the site as good or excellent were twice as likely to report ‘No help’ compared with those users who rated the trustworthiness of the site. Figure 4 User attitude towards the advertisement, by site trustworthiness

respondents who were recommended the site were least likely to say that site trustworthiness was poor or OK. Users who arrived at the site via an advert or a search engine were most likely to rate trustworthiness as either poor or OK (see Fig. 4). Those who read about the site were more likely to rate the site as excellent. Advertisements on the site also impacted on the site’s trustworthiness, as with the qualitative usability study the authors carried out earlier.9,10 Users were asked if the amount of advertising on the site was excellent, good, OK or poor. Users finding the number of advertisements as either poor or OK were more likely to report that the site’s trustworthiness was also only poor or OK. One-third of users who found the advert content poor also rated the trustworthiness as poor or OK.

Digital interactive television Most of the health information content for the Living Health channel delivered by Telewest is provided by the NHS and all pages are branded with the NHS symbol. A number of questions on the questionnaire sought respondents’ views of the NHS and are given in Fig. 5. Less than two-thirds (63%) of all respondents knew that the NHS was involved in providing content for the Living Health channel. Over three-quarters (77%) felt that the NHS was a symbol of trust and an even higher proportion (83%) agreed that they trusted the information because the NHS was involved. However, only half (51%) said that they would carry on using the service if the NHS were not involved. The number of times the respondent visited their doctor in the last 12 months and the user’s interest in health information were significant in

Figure 5 Views on NHS on Living Health © Health Libraries Group 2003 Health Information and Libraries Journal, 20, pp.215 – 224


which, as mentioned before, have on occasion not been very good. Factors that might influence views on trust

Figure 6 Respondent’s trust in digital health information knowing the NHS produced it, by the number of doctors’ visits made by them in the last 12 months

determining the respondent’s perception of the NHS. People visiting the doctor less frequently and those less interested in health information were less likely to accept the NHS as a symbol of trust, were less likely to recognize the NHS symbol, and were less likely to say that the NHSbranded information could be trusted. Thus, in the case of those people who said they had not seen a doctor in the last 12 months, over a quarter (27%) said that they did not trust the information knowing the NHS was involved (Fig. 6). This compares with, respectively, 13 and 11% of those people who had visited the doctor once or twice, or three or more times, respectively, in the last 12 months, who said that they did not trust the information knowing that the NHS was involved. A similar pattern emerges with regard to the users’ interest in health information. Eighty-nine per cent of respondents who said they were very or quite interested in health information said they trusted NHS information. This falls to 71% of those who said that they were not very or not at all interested. Plainly, a significant proportion of the population does not trust the NHS. This suggests that people who do not come into contact with the NHS, or who do not have an interest in health information are less likely to perceive the NHS as a symbol of trust. Those who use the NHS tend to trust the NHS symbol. Those who don’t, appear to trust the symbol less. Perhaps these users are picking up their opinion of the NHS from the press and media reviews of the organisation,

An alternative way of modelling trust is to look at the probability of impact of each variable in determining the users’ trust of the NHS symbol. Multiple logistic regression was used to identify factors likely to be important in explaining why the NHS was a symbol of trust for the user. The independent variables included questions on the health information source used (including doctors, nurses, books, magazines, Internet, television, and friends and family) and the health topics in which they were interested (including user interest in prescription drugs, new treatments, healthy living, complementary medicine, diet, medical news, a particular condition, research, general health, alternative health, and exercise). Questions were also included about age, gender and whether they had used the Living Health service that carried the NHS branded information. The best model fitted to the outcome variable identified six explanatory variables: the importance for health information of NHS Direct telephone line and health books, the user’s interest in healthy living and alternative health, the age of the subscriber and the user’s use of Living Health. Table 1 lists the variables, the estimated log odds and the number of cases. Telewest subscribers who had said that NHS Direct phone line was a very important information source were just under three times more likely to say that the NHS was a symbol of trust for them compared with those who had not phoned. This was also true of those who had used the NHSbranded information delivered via Living Health. These users were just under three times more likely to say that the NHS was a symbol of trust for them compared with those users who had not used the Living Health service. Subscribers very interested in healthy living and subscribers aged over 55 were also found to be more likely to say that the NHS was a symbol of trust for them compared with those users not interested in healthy living and younger users. Those subscribers interested in alternative health and who said that health books and


221

222

Authority of health information, David Nicholas et al. Table 1 Q: The NHS is a symbol of trust for me? Variables in model

n

Log odds (SE)

How important for health information: NHS Direct phone line? Not at all/not very Fairly Very

148 204 200

1.97† (0.27) 2.70‡ (0.31)

As a Telewest subscriber, have you ever heard of a service called ‘Living Health’? I didn’t know about it 133 I have used it 130 I know about it but haven’t used it 289

2.81‡ (0.34) 2.13‡ (0.25)

How interested are you in Healthy living? Not at all/not very Fairly Very

51 240 261

1.33 (0.39) 2.73† (0.43)

Age of subscriber? 55 and under Over 55

332 220

1.81† (0.24)

How interested are you in alternative health? Not at all/not very Fairly Very

175 222 155

1.15 (0.29) 0.58* (0.33)

How important for health information: health books/magazines? Not at all/not very Fairly Very

201 259 92

0.90 (0.26) 0.56* (0.37)

‘No’ coded as 0 (n = 109), ‘yes’ coded as 1 (n = 443), ‘missing’ (n = 171). Levels of significance (Wald’s statistic): *< 0.1, †P < 0.05, ‡P < 0.01; ***P < 0.001.

magazines were very important as a source of health information were both approximately half as likely to say that the NHS was a symbol of trust for them compared with those not very interested in these. However, the results here are only an indication and need further investigation. The relationship between the use of NHSbranded information delivered by ‘Living Health’ and trust in the NHS logo is of interest as it implies that the delivery of the information by a third party does not affect trust of the NHS logo. It argues that use of the NHS symbol on the Living Health channel does not lower the NHS as a symbol of trust. In fact, the evidence is that the presentation of NHS-branded information on DiTV, even though it is delivered by a third party, promotes trust in the NHS symbol. People who had either used the Living Health service or had heard of the service were three and two times more likely to say that the NHS was a symbol of trust for them compared with subscribers who had not

used the service. Thirty per cent of respondents who had not used the Living Health channel did not see the NHS as a symbol of trust. However, this figure fell by more than half, to 12%, for those who had used the service. Interestingly, the figure fell to 18% for those users who had heard about it but had not used it. It would seem that people are generally impressed by what they see. The positioning of the NHS logo within the information services of a third-party provider and delivered by another did not dilute the NHS as a symbol of trust. Conclusion The two studies looked at two very different aspects of the perceived authority of information. The Web study looked at some of the issues related to the trustworthiness of a commercial website that carried advertisements, while the DiTV study looked at predominantly NHS-branded

© Health Libraries Group 2003 Health Information and Libraries Journal, 20, pp.215 – 224


information and the NHS logo as a symbol of trust when this information is made available via a third party. Direct comparisons between the two studies cannot be made, although it is felt that the results can be generalised between platforms. The same issue surrounding trust of information, when advertising material is present, is thought likely to apply to a DiTV platform as well as an Internet platform. The quality and type of advertising will impact in different ways on trustworthiness. The main findings of the two studies point to a complex picture. They are: • Seventy per cent of the Web users said that they had visited more than one site. For these information seekers visiting a number of sites may be a way of checking the authority of the sites, and certainly supports the contention that authority is, indeed, ‘up for grabs’.2 Interviews with users of various systems (which the researchers are examining for the wider project) indicate that people might simply go from site to site looking for information that best fits their own preconceived ideas, thus conferring ‘authority’ on the source of the most appealing information, regardless of other criteria. Despite this, the NHS was regarded as trustworthy even among users of other sources. • Those respondents who were recommended the website were least likely to say that site trustworthiness was poor or OK, perhaps a reflection more on the trustworthiness of friends than on that of the NHS. • Web users finding the number of adverts as either poor or OK, in terms of obtrusiveness, were more likely to report that the site’s trustworthiness was also poor or only OK. This finding supports that found previously by the group19 that advertisements tend to impinge upon the credibility of the information source. This has obvious implications for the NHS in terms of the use of commercial providers. It might be advisable for the NHS or the DoH to state that where information carrying the NHS or DoH logo appears on an electronic system, the service provider must agree to not use advertising material on the same page/area. • DiTV users visiting the doctor less frequently and those less interested in health information

were less likely to accept the NHS as a symbol of trust, were less likely to recognize the NHS symbol, and were less likely to say that the NHS-branded information could be trusted. This is an example, albeit ‘in reverse’, mentioned by Mulligan,16 of those who use NHS services being more positive about the NHS generally. It may be that those who do not have direct experience are too influenced by the negative stories mentioned earlier. The fact that younger cable respondents were less likely to recognize the NHS as a symbol of trust compared with older respondents may also be a function of their relative lack of contact with the service. Of course, it may also be a lack of deference on their part. • DiTV (Telewest) subscribers who had either used the Living Health service which carried NHS-branded health information or had heard of the service were more likely to say that the NHS was a symbol of trust for them compared with DiTV subscribers who had not used the service. This is only to be expected—those who did not regard the NHS as a symbol of trust or authority would, presumably, seek their health information elsewhere. So what does this research mean for those planning digital information policy in the NHS? Firstly, the fact that the positioning of the NHS logo within the information services of a third party provider and delivered by another did not dilute the perception of the NHS as a symbol of trust. This is excellent news for the continuing rollout of health information on the back of commercial TV companies, touch-screen providers, etc. Secondly, however, the government must learn to deal with the so-called ‘promiscuous informationseeker’—the person who flits from site to site and source to source. Clearly, the NHS wants people to view it as the authority when it comes to health information. With the problems mentioned earlier surrounding the service, and the criticism that has been directed at the information provided on the NHS Direct website,17 there is some ground to make up. With the emphasis so much on patient information at the moment there are grounds for calling for more resources to go into services such as NHS Direct. The finding that those who have had to use the NHS are more likely to be positive about it is encouraging.


223

224


Much more research is needed in a number of areas before a concrete set of proposals can be offered. In particular, it may be fruitful to look at differences in perception of trust in digital information held on different platforms and environments. For instance, is the kiosk more trusted than a DiTV channel and is a kiosk located in a doctor’s surgery more trusted than the same kiosk located in Safeway? Also, is a commercial content provider, like Surgerydoor or Boots, any less trusted than an official provider like the NHS? Is the public really demanding quality kite-marking of patient information on the Internet, or do they just want some information brands they can trust? These are some of the questions that need addressing. Secondly, the issue of what factors go into a person’s evaluation of authority and trust is of great interest. An underlying assumption of this paper—which needs to be tested in future work— is that people do not analyse the information presented to them and make judgements about its quality. Instead, they take it at face value, making judgements about the source rather than the content of the information and whether the provider can be trusted or not. The question is, why do people trust certain sources? If they trust the NHS brand, what do they understand by it? Do they perceive the difference between national NHS information and information that comes from NHS agencies and Trusts? These are all critically important issues, particularly with the commitment by the government to continue rolling out digital health information to the public. This paper has only been a tentative first step in researching a fascinating and strategically vital field. References 1 Kim, P., Eng, T., Deering, M. & Maxfield, A. Published criteria for evaluating health related web sites: review. British Medical Journal 1999, 318, 647 –9. 2 Hunt, R. Lecture: Living in the digital wild west. Presented at City University, 8 May 1999. London: City University (unpublished lecture notes). 3 Impicciatore, P., Pandolfini, C., Casella, N. & Bonati, M. Reliability of health information for the public on the world wide web: systematic survey of advice on managing fever in children at home. British Medical Journal 1997, 314, 1875 –81.

4 Griffiths, K. M., Christensen, H. Quality of webbased information on treatment of depression: cross-sectional survey. British Medical Journal 2000, 321, 1511–5. 5 HONF (Health On The Net Foundation). HON Code of conduct for medical and health web sites. Health on the Net Foundation 1997. Available from: http:// www.hon.ch/HONcode/Conduct.html (accessed 22 August 2002). 6 BHIA (British Healthcare Internet Association). Quality standards for medical publishing on the web. British Healthcare Internet Association 1996. Available from: http://www.bhia.org/reference/documents/ recommend_webquality.htm (accessed 10 May 2002). 7 Hernández-Borges, A. A., Macías-Cervi, P., Gaspar-Guardado, M. A., Torres-Álvarez De Arcaya, M. L., Ruiz-Rabaza, A., Jiménez-Sosa, A. Can examination of WWW usage statistics and other indirect quality indicators help to distinguish the relative quality of medical websites? Journal of Medical Internet Research 1999, 1, e1. Available from: http://www.jmir.org/1999/1/e1/index.htm/ (accessed 30 July 2002). 8 Jadad, A. R. & Gagliardi, A. Rating health information on the Internet. Navigating to knowledge or to Babel? Journal of the American Medical Association 1998, 279, 611– 4. 9 Williams, P., Nicholas, D., Huntington, P. & McClean, F. Surfing for health: user evaluation of a health information web site. Part 1: literature review. Health Information and Libraries Journal 2002, 19, 98–108. 10 Williams, P., Nicholas, D., Huntington, P. & McClean, F. Surfing for health: user evaluation of a health information web site. Part 2: fieldwork. Health Information and Libraries Journal 2002, 19, 214–25. 11 Simon, P. The strange online death and possible rebirth of brand theory and practice. Aslib Proceedings 2001, 53, 245–9. 12 Travis, D. What drives repeat visitors to your website? 2000. Available from: http://www.system-concepts.com/articles/ forrester.html (accessed 27 January 2002). 13 Gunter, B., Nicholas, D., Williams, P. & Huntington, P. Is TV good for you? Library Association Record 2001, 103, 558–9. 14 M2 Presswire. UK-specific health website aims to capture a fifth of online health market. M2 Presswire News Release 2000. Lexis-Nexis Universe UK News database. 15 Nicholas, D., Huntington, P., Williams, P. & Blackburn, P. Digital health information and health outcomes. Journal of Information Science 2001, 27, 265–76. 16 Mulligan, J. Policy comment: what do the public think? Healthcare UK 2000. Available from: http://www.kingsfund. org.uk/eHealthSystems/assets/applets/public.pdf (accessed 8 November 2002). 17 Green, B. Electronic response to: NHS Direct. British Medical Journal 2000. Available from: http://bmj.com /cgi / eletters/321/7258/446 #9261 (accessed 4 April 2003).

© Health Libraries Group 2003 Health Information and Libraries Journal, 20, pp.215– 224