Tuesday, September 16, 2008

MHRA Report on Influence of Healthcare Blogs

Below is a link to an MHRA-sponsored report (Medicines and Healthcare products Regulatory Agency, the UK equivalent of the FDA) on the "top influencers" in healthcare thinking, including blogs such as this one, Healthcare Renewal.

The MHRA report is here:
link (PDF)

The report is based on data collected regarding seroxat. The results are possibly reflective of, or proportional to, influence on other biomedical topics as well.

It was prepared for the MHRA by Market Sentinel, a company specializing in:

... measuring, monitoring and benchmarking influence in relation to issues, brands and companies. This includes social media monitoring (blogs and messageboards) but is not limited to it. We provide clients in the public and private sector with hard facts that enable them to better understand the playing field; more effectively bring their messages to market and increase their influence ...

... in addition to services such as reputation and crisis management, and optimization of customer targeting in advertising campaigns.

The MHRA report became became publicly available after a UK Freedom of Information request, apparently by someone concerned about the seroxat controversy. The full Freedom of Information release containing the report is here (also PDF). The report begins at page 215 of the release, after a somewhat curious, identity-redacted email that states:

"Our intention is not so much to track retail opinion so to speak - the opinion of random unqualified individuals - but to track the opinions of those who matt[er], those with a demonstrated following. Our hypothesis is that it is these "influencers" whose opinions will reach the rest of the world.

We
have not so far made recommendations as to who to target amongst these stakeholders, preferring to speak first about the message you wish them to receive."

I presume the "Our" and the "We" refer to Market Sentinel, and the "you" to someone at MHRA or perhaps seroxat manufacturer GlaxoSmithKline in an attempt at "reputational repair" over the seroxat controversy, but this is unclear.

In any case, it appears the Market Sentinel Report was not prepared as a purely academic exercise.


Health care bloggers appear to be doing well, a finding that does not surprise me, aware as I am of the growing influence and power of the political blogs.

That healthcare bloggers in general appear to be doing well is a desirable observation, considering the increasing distortion of the biomedical literature by commercial interests that makes attainment of true "evidence based medicine" more difficult.

It is also desirable from the perspective of the "group think" aspect of the peer review process that makes publication of opinion, even well-documented opinion, sometimes impossible if the opinion is "politically incorrect" and/or runs contrary to the collective wisdom, collective exuberance - or the collective pocketbook.

My posts on Medical Informatics and healthcare IT problems, for example, would probably never make it through peer review in the informatics community. They challenge the dominant paradigms and what I believe to be the irrational exuberance over the technology.

I do not write about that issue lightly or without evidence. The American Medical Informatics Association, as just one example, recently decided it would not publish a book on HIT difficulty by several members of the Clinical Information Systems Workgroup in a style similar to my website on that topic (i.e., anonymized, fine-grained case examples), itself a resource that would have been impossible before the Web. The group had to go elsewhere.

I also highly doubt the posts of my colleagues about healthcare corruption and loss of core values would make it past peer review in most mainstream journals, especially in a fashion that would form an "anti anechoic effect" repository or aggregation of such cases.

Healthcare Renewal came out relatively high in influence on the MHRA-commissioned report on seroxat:

... Mainstream media accounts for 30% of the top 100 stakeholders. Media coverage of MHRA is neutral to mildly negative, with the force of emotion mainly targeted at GSK.

The second largest group is bloggers who account for 23% of the 100 most influential stakeholders. The blogs are either written by insiders in the pharmaceutical industry (Doctors, Researchers, Journalists etc) or by individuals who have experienced the effects of Seroxat first hand. Of the industry bloggers, 72% are from the USA and the rest are UK based. For the personal experience bloggers, 60% are based in the UK and the remainder is in the USA.

78% of blogs in the top 100 are written by industry insiders. The most influential industry blog is the American based Health Care Renewal blog. Contributed to by a group of health care professionals, it tackles issues which call into question the values of the health industry.

A highly influential industry blogger is Aubrey Blumsohn who writes the Scientific Misconduct Blog. Not only is he ranked high in influence, but his blog ranks the highest in terms of betweeness’ which measures how many different paths go through a specific stakeholder. Blumsohn is what Malcolm Gladwell would call a “connector”. His old blog, thejabberwock, is still often cited, despite the fact that it is no-longer regularly updated.

Blogger Bob Fiddaman is dedicated to raising the profile of the side effects of Seroxat and regularly posts comments on other blogs, which in turn increases his on-line sphere of influence. A further individual blogger who is highly influential is the author of the Seroxat secrets blog. The entire blog is dedicated to discussing and publishing any issues surrounding the drug, MHRA and GSK.

After Mainstream media and Bloggers, distribution of influence is shared between 12 groups, some of which exist to support those working in the health care industry by supplying information and support ...

Of interest is the citation analysis-like "stakeholder analysis" method used to rate the influence of various Old and New Media outlets. Here is a "stakeholder map" of the top influencers, showing connectedness and information flows (see the MHRA report linked above for a full explanation):

(click diagram to enlarge)

Healthcare Renewal is the red circle at the mid-right network border; size of each node reflects relative influence.

Some stats from the report as highlighted by colleague Roy Poses:

This blog ranked as 13 in the table of "top influencers." Other highly influential blogs included some cited by us, and/or are on our side-bar list of links. On the ranking of top influencers, Health Care Renewal outranked the Wall Street Journal, Reuters, the UK Times, Nature, Forbes the UK Telegraph, the Annals of Internal Medicine, the Canadian Medical Association Journal, and ABC News, among well known publications. On the ranking of "popular stakeholders," Health Care Renewal came in at 27. We out-ranked Reuters, the UK Times, Nature, CNN, Forbes, the UK Telegraph, and the Canadian Medical Association Journal.

Considering there were only a few posts here on the seroxat controversy, this may have to do with "trustedness" -- i.e., "goodwill" of a sort generated by the plain language, sunlight-as-best-disinfectant coverage of many other issues -- on the hyperlink-driven access patterns and information flows.

The stakeholder analysis shares some aspects of the longitudinal citation analysis methodology such as used to trace the flow of ideas here, but in a hyperlinked web context:

Citation analysis is the examination of the frequency, patterns and graphs of citations in articles and books.[1] [2] It uses citations in scholarly works to establish links to other works or other researchers. It is one of the most widely used methods of bibliometrics. Automated citation analysis has changed the nature of the research allowing millions of citations to be analyzed for large scale patterns.

[1] Rubin, Richard E. Foundations of Library and Information Science 2nd ed. New York: Neal-Schuman, 2004.
[2] Garfield, E. Citation Indexing - Its Theory and Application in Science, Technology and Humanities Philadelphia:ISI Press, 1983.

Although the MHRA-sponsored study has its limitations, blogs can indeed be quite influential. This is a lesson painfully learned by some prominent mainstream media newspeople, politicians, and others.

Perhaps HC blogs should not be dismissed as the work of pajama-clad novice journalists dabbling in their bedrooms.

(The existence of firms specializing in "social media monitoring" suggests that this is starting to be understood in some sectors.)

-- SS