Trusted doesn’t mean accurate. Safe is almost certainly stale

I am sick & tired of the advice given to people looking for medical information on the internet. All the content evaluation guidelines say the same.

Look for sponsorship, currency, factual information and audience.

A recent CNN Empowered Patient entry states

“So if you’re trying to find medical information for yourself or someone you love, and you’re not lucky enough to have access to a professional research librarian, what do you do?

“The Empowered Patient” assumes you already know the basics of good Internet searching: .gov and .edu sites are to be trusted, as are sites for major health centers (think and health organizations (such as the American Cancer Society’s

Let me tear apart some of these guidelines because they so clearly represent guidelines from the last century.

The guidelines available on the “trusted” websites were all conceived between 1995 and 1999, when information available was entirely static. That was before anyone had written about the long tail, the wisdom of crowds, community intelligence, and e-patients. They all predate the introduction of social networking sites. In short they COMPLETELY miss the point of the current use of the internet, as a disruptive technology leading the democratization of medical knowledge by opening access to countless repositories of scientific information and fostering the growth of innovative medical online communities.

Once you have opened the Pandora box by letting the world have unhindered access to PubMed you cannot expect that people will be satisfied with the safe and stale information available from “patient-oriented” information websites prepared ala “Health 0.1” by so many large medical organizations. The American e-Patients demand a lot more and they rightly deserve a lot more.

We, the e-Patients, do not want to read medical information from the trusted and vetted static websites. We need accurate and current information about our current medical problems. And with the explosion of new targeted therapies these problems are going to multiply, leaving almost any health professional unable to keep track of all the side effects of the therapies they prescribe. It is a trend I am noticing for a few years and it is clearly accelerating. In fact the acceleration is such that now it is almost a given fact that “trusted” static sites cannot produce web accessible pages containing the information we so desperately need.

An example is in order. RCC, renal cell carcinoma or kidney cancer, has an incidence of about 51,200 cases in 2007. It is a fairly uncommon cancer but it is treated by a large number of clinicians. Prognosis of RCC patients is still dire, with about 25% of patients diagnosed at stage IV and a 2-year survival for those patients at about 23%. 3 recent drugs have been added to the available treatments which until then were seriously deficient. The 3 agents are new targeted antiangiogenic agents (Sutent, Nexavar & Thorisel). 9 additional drugs are in the pipeline, in phase III trials. 5 are targeted therapies and 4 are immunotherapies.

Doctors treating RCC patients are suddenly faced with an exponential need for in-depth scientific information, even before large cohorts of patients have been treated for long periods of time. e-Patients have naturally started to fill in this growing gap. They ask daily their peers about a very wide array of side effects for each one of the newly approved antiangiogenic agents and are even able to conduct conversations about the potential benefits/risks of a treatment their doctor has recommended. These decisions are often made with the use of medical information newer than what they physicians have access to, unless they practice in a large and specialized oncology unit.

Just yesterday, a patient wrote to the ACOR group (with over 1,500 members) about a very unusual side effect. Acting on behalf of the RCC patient, I followed the advice from CNN:

There’s a lot of junk on the Internet. “It’s the wild, wild West out there,” says Alan Spielman, CEO of URAC, a company that certifies health Web sites. “You really have to be alert as you go through these sites.”

To get rid of the junk, use a search engine that looks only at reputable sites that have been vetted by health professionals. Dirline, run by the National Library of Medicine, is one such engine, as are and searches for information on government health Web sites.

I did look for information about the side effect in MedlinePlus (the current darling of those who are afraid of the unbridled medical internet). I did and couldn’t find anything. I went to the FDA site and read the entire official insert label for the antiangiogenic agent mentioned by the patient. No mention of that side effect! Of course because to be trusted and safe the information has to be written at 6 grade litteracy level! Still interested I went to the National Cancer Institute, the American Cancer Society and the American Society of Clinical Oncology websites. Searches in these 3 reputable sites didn’t produce a single result!

It turns out that only the sites who have licensed the Merimedex drug database have information about this rare side effect. And, of course, the 3 patients who responded to the inquiring patient, while I was spending my time trying to get an answer from the trusted and vetted sites. Thanks to their combined answers, the 4 now know that they must have their liver functions tested immediately following the development of this rare side effect. A prime example of the power of participatory medicine!

If presenting clear and current Health websites guidelines to the American public is important, it is the responsibility of the organizations who designed the original guidelines 1.0 to come up with the 2.0 version. HON, URAC & the NIH could lead the way. Just look at the HON Code. No social networking site could ever apply for certification. ACOR certainly cannot, since so much of the information given from patient to their peers is not authoritative in the classical way. ACOR doesn’t replace the doctor-patient relationship but unfortunately many patient to patient communications replace the non-existing communication between patients in urgent need of advice and doctors overwhelmed and unable to respond to their patient needs. But as I wrote in another blog comment, in 15 years, no one has ever reported a single “death by internet medical information”. Quite the contrary.


Posted in: trends & principles




6 Responses to “Trusted doesn’t mean accurate. Safe is almost certainly stale”

  1. Jason Wertz says:

    If I wanted to know where to get the tastiest burger in town, I don’t go to the USDA’s website, although they would be the expert on the quality of beef. I go to Yelp and I know I’ll find some very visceral, unfettered descriptions of dining experiences.

    e-Patients have the same expectation when searching for health information. Social networks provide a context that’s closer to reality than a clinical reference. There’s a value to this and it should be recognized.

    The ability to enable action, in the age of consumerism, is being accomplished by the patient networks, the technology companies that enable them, and online health companies that are 100% focused on meeting the needs of the health consumer in real time.

    So, I agree. HON or NIH need to recognize that social network sites, peer sites and health content sites compliment the traditional information sources.

    But I’m not too worked up about it. How many patients are going about checking sources? Of course, the more serious the condition, the deeper the engagement of the individual to pursue resources beyond a Google search. If a person is going to do research, I think they’ll uncover the static content plus the social content. The peer content will provide a realistic context to the static, clinical content, and obviously it’s proving out to be a wonderful thing.

  2. I agree with you, Jason.

    But I also think that, in the health information realm, social network have a much deeper function than in most other cases.

    The focused and active networks have a fundamental role in helping their participants gain a high level of health literacy. And functional health literacy is clearly associated with better health outcomes. (More about this in a forthcoming blog post).

  3. I too agree with Jason, but there’s another aspect to this.

    The underlying goal of the peer review process is to filter out crud and somehow ensure that published information is as useful as it can possibly be. As Gilles pointed out so well here, the medical peer review process is no longer guaranteed to be the best source.

    To the contrary, what I want is information reviewed by MY peers – others whose butts are on the line and who are working, full-time, to hunt down and vet every single piece of information they can find that could change their outcome.

    Remember, in my case every bit of misleading (and ultimately useless) information I found came from the best-respected medical sites. That’s an existence proof that the advice earnestly offered by et al is simply invalid and needs to be updated.

  4. A different angle on this post’s title: “Trusted is too often stale.”

    There’s also a word in your post that’s starting to irritate the crap out of me: “licensed,” as in “only the sites who have licensed the Merimedex drug database have information about this rare side effect.” Who the hell are they to have information about something like this and not make it available?

    I have no objection to someone making a buck, but something is fundamentally wrong when important medical information is kept out of the public eye.

    And, to get back to Gilles’s point, as long as the situation is like that, it’s actually harmful to tell patients only to look in places that play that game – and have no legal responsibility to keep themselves up to date.

    If they respond by saying “It’s too hard to keep up to date,” then they’ve become signatories to the e-patient tagline “because health professionals can’t do it alone.”

    Advice to “health 2.0” providers and their investors: make your money providing a better something of some sort, not by latching onto important information and keeping it private. Be assured, those walls will eventually crack open, and if you haven’t actually created something of value, your business will collapse.

    And the best way to build something of actual value for patients is, as Doc Tom Ferguson taught, to ask patients what they want.

  5. Jason Wertz says:

    Yes, “functional health literacy” is a fantastic term. Coming from the challenges of Health 2.0: How do we raise the level of functional health literacy for people who need help? What would we consider to be the standard? For me, it’s ensuring that people can a) keep themselves and family healthy(for instance, I’m considering obesity and diabetes as national epidemics right now) and b) have the tools to educate themselves to help them identify sign and symptoms to go get help when they get sick(severe disease or otherwise).

  6. Functional health literacy almost certainly means something different for different types of diseases. As the population of e-Patients grows it will become clearer that there is a real taxonomy of e-Patients just as there is a taxonomy of disease management.

    The requirements in terms of currency and accuracy of in-depth information are going to be vastly different if a chronic disease is being appropriately treated with today’s knowledge, in comparison with almost any type of cancer where we still are unable to treat the diseases as chronic. There is just too much that is unknown and it is even clearer when you deal with a rare type of cancer.

    In short there is no model ala “one size fits all” in the Health 2.0 realm. I believe that many commercial sites will realize that fact sooner than later.

Leave a Reply