Wednesday, 20 October 2010

Should the NHS be on Yahoo Answers?

Last week Leigh Blackall and I agreed that patients needed access to high quality information on the internet but we disagreed on how that would happen. He mentioned Yahoo Answers and I and many of those I know on twitter responded sceptically. Why would ANYONE look to Yahoo Answers??!! But my experience of looking for information online is most often centred around knowing the diagnosis. If you are trying to make sense of your symptoms then it is a whole different ballgame. So I decided to make a short screencast to show what the experience might be like.
What should we do about this? Make the existing NHS websites more user-friendly when checking symptoms? Develop better tools for symptom sorting? Make sure that the relevant NHS pages are serach optimised? Or should the NHS be patrolling Yahoo Answers?
I'd love to know your thoughts and feel free to share any stories. But remember this is a public site and others will be able to read it after you!
EDIT: After Fi's comment below I feel I should add that I think it is unlikely that patients will find a diagnosis online but they should in a sense get good triage information. Is this a serious symptom that I need to go to the doctor with, or can I leave it for a few months to see if it goes away?
I've blogged about fear of Dr Google before. We should be helping Dr Google to perform better, and we should be able to give guidance to patients on where they can get sensible information online.


  1. I still believe that diagnosis is something which should be left up to a doctor for all but the simplest of conditions. There are too many subtleties to the art of diagnosis for it to be replaced entirely by the Internet (though I think the Internet may play a role). Illnesses commonly present with an unusual cluster of symptoms, or an otherwise odd presentation, and it takes a trained and skilled physician to spot this.

    I have heard, on more than one occasion, people go "oh, I've pretty sure I've got such and such a health condition", and a bit of further probing highlights that this is based solely on a bit of Googling, that they haven't been to a doctor, and aren't seeking treatment.

    I think it's more of a problem with psychiatric conditions - so many people self-diagnose without actually understanding the difference between a mental illness, and what is just part of the human condition. It also seems to be quite common with allergies/food intolerances, though I'm sure it can occur across the board.

    Personally, I thought I had a number of different things before receiving a diagnosis for an ongoing health problem. Ironically, the only thing I refused to entertain the thought of was the one thing I was finally diagnosed with. Maybe I knew this all along and was in denial, maybe not. But I'd sat through lectures about the condition, had seen a patient with it, and was still convinced I didn't have it - That's why you need the objective opinion of a doctor.

  2. This is a great comment. I don't think that most often people are actually looking for a diagnosis online. But they are trying to figure out if what they have might be serious or not and perhaps what they could do to help themselves in the meantime. I think that the first 10 pages you hit with a symptom search on Google should give access to reputable information- with because it is a trusted site or because it is from a trusted source on a site such as Yahoo answers.
    But I'll add a little edit to my post to explain more.
    PS Hope you are making a good recovery:)

  3. In the U.S., and I think globally, the health citizen still has a lot of heavy lifting to do when self-checking symptoms online. Engaged health consumers use multiple sources to check, confirm, rule out...their "N" = 1, and they seek to personalize based on their own unique 'long tail' searches. The good news here is that people are getting more sophisticated in search; at Health 2.0, the CEO of Health Central Jeremy Shane reported that health searches used to be just a couple of words; now they are about 6.5 words on average. This helps people get to more granular results. Still, the downside is that the Pew Internet & American Life Project recently found that too few health information seekers actually check the sources of their findings online...thus, we need to educate a more digitally health-literate citizenry...more work to do for all of us at the intersection of health and IT!

  4. So no easy answers then, Jane! Thanks for your input as always.

  5. Thanks, I'm getting there. :-)

    I think in the process of identifying whether a symptom set is serious or not, there will invariably be sites giving a diagnosis. If this is right, patients will hopefully seek medical advice - great! If it's a false positive, it just perpetuates the "worried well" phenomenon. If it's a false negative, people may be lulled into a false sense that everything is OK, when really, they should get things checked out. (e.g the time I was convinced my finger was just a bit sore, when in fact it was going to end up as a rather nasty case of paronychia, which while not life threatening, was bloody painful).

    It's not something that has happened for me yet, but if and when I qualify, if a patient comes to me, I want to get a history and attempt to make a diagnosis based on that. I wouldn't want to be told from the outset what the patient thinks is wrong with them, otherwise it would bias my thoughts (red herrings etc). Equally, as a patient, I might have ideas in my mind about what I think is wrong with me, but I wouldn't express those thoughts to my doc, for the same reasons.

    Social media and the internet are great for post-dx support and info, but I think we need to be very wary about their role in replacing a visit to the nurse or doctor.

  6. Thanks again Fi. I think the notion of a 'pure' history was often an illusion. Those consulting will often have ran their symptoms by someone else- a friend/family/pharmacist/bus driver- or they will have read something in a book or newspaper or magazine that has made them think that their symptoms might mean x or y or that they should seek a particular treatment. My concern is not that patients are over consulting but that some patients aren't aware of serious symptoms. I think that all patients should know about this website:
    As someone who is working now I can tell you that it is not a problem if people are thinking about their diagnosis already- who doesn't? They are either possibly right or wrong. Either way you can usually get further faster in a consultation with a patient who had already started considering their symptoms.
    I hope some other working doctors and patients chime in on this:)
    Thanks again!

  7. I think people go online to triage, i.e., to determine if their symptoms warrant going to the ER (vs scheduling a doctor's appointment at a later time vs seeking an OTC remedy). They also go online because they are mindful that they'll have a limited amount of face-to-face time with their doctor and research will help them to focus their questions.

    In addition to general Googling, there are general health information Web sites that have symptom checkers. Sometimes these are designed to be interactive and are so complex, they're hardly useful or they just lead you from one link page to another--good for SEO marketing, bad for patients.

    I think in trying to be "sexy" (e.g., loaded with Flash) and trying to cram in every conceivable health condition (and on some sites, every possible sponsored ad), many sites become fairly unusable, particularly for people who are intimidated by internet searches.

    My wish would be to have an online patient community where people identify their online research and experts (docs, nurses, science researchers) model how to evaluate the information for credibility and show people how to use their research to create a question list for their doctors. We need more sites that teach people the basics.

  8. Thanks Dianne. I guess I am slightly puzzled by how NHS sites (and others such as the BBC) can get information on diagnoses so right but information on symptoms so wrong.
    It's interesting that you suggest that we need patient communities to address this issue. I think that most people checking out symptoms don't consider themselves patients. And are you suggesting that disease-specific patient communities could help with this?
    I think that once a diagnosis is know things get infinitely better in the UK at least.
    Thanks again,

  9. Surely NHS Direct have lots of these answers tied up in their algorithms? I'm pretty sure these are symptom-led.

    Having said that, my experience of NHS Direct isn't hugely positive!

  10. In reflecting on my colleagues' responses here, I'm reminded of the emergence of Participatory Health -- getting the patient aligned with her physician, her social network (online and offline), and the right information at the right time via the medium they want to use to access (e.g., mobile phone, tethered internet, video, printed brochure). It's all of these touchpoints we've got to get right, so they converge and enhance patient engagement.

  11. Hello Jon,

    Maybe I should go back and record more of the NHS Direct experience for you... but I was limited to 5 mins by Screenr!

    Yes, they are following an algorithm but I'm suggesting that creates quite a confusing patient experience. But it does require further exploration.


  12. Hello again Jane,
    That's a good point, well made. The websites (or Facebook pages?- that's another post!) of primary care practices should possibly contain some of this content, or at least link to good places to gain this information.
    I'll have to check out how those NHS websites look on my phone as well.

  13. Leaving it up to the NHS doctors/consultants in my case has turned into the biggest farce imaginable so much so that I have forwarded my MRI to the USA for a definitive opinion.

    One MRI prompted 5 different comments/observations and these included the references that the surgeon had liased with his Cambridge team and the other surgeon had liased with his Oxford team. My GP tells me he/she is only a GP. A further MRI produced commentary that was completely different to all the previous comments however the surgeon won't put it in writing.......I have sat up night after night surfing the net and comparing my MRI to what is available...with the upshot that I have approached the medical profession in the USA. Of course there is the option that despite the medical help being available.....I am too old to qualify for honesty or help!

  14. Hi Anne Marie,

    I read your post on my phone - so missed the Screenr part of the post. Oops, technology (and human) fail!



  15. Thanks for the generous mention in this post and screen recording. I appreciate your your genuine effort to represent my views.

    Jane Sarasohn-Kahn's comment is the most encouraging here yet, and Anonymous' story is an obvious yet not surprising concern.

    As this conversation gets more in depth (compared to the frustrating limitations of Twitter), I'm now too tempted to throw a curve ball, and link to Illich's Medical Nemesis 1975.

    No doubt that link will have lost me most of the attention I might have had from your Health colleagues, if they haven't already decided to ignore me, in disgust at the suggestion that they should edit Wikipedia and build a Yahoo Answers profile. I link to Illich as a way of expressing my own uncertainty to such a suggestion.

    Do we really want health professionals, who are so blinded to the obvious, to occupy and dominate our so-far wonderful online social information spaces?

    If a health professional sees an article on Wikipedia that is somehow faulty, and turns away without spending a few minutes of their television time making the edit, I would argue they are not a fit health professional. Wikipedia is the 7th most visited website in the world, and THE most visited information reference site (Alexa).I challenge anyone to find a more readily available, quick and easy to use reference site - ever!

    Likewise, Yahoo is the 4th most visited website in the world, a proportion of that is Yahoo Answers. If a health professional looks at YA and spots a faulty answer without spending a minute or two correcting it, they are not a fit health professional.

    Now, if this conversation was in any way around the question of whether NHS and others have the resources to "patrol" Wikipedia and Yahoo Answers, then I'd be less disagreeable. But it has not been like that so far.

    On Twitter and to some extent the comments here and elsewhere have been based on the idea that Wikipedia and Yahoo Answers (as just 2 examples of popular channels of socially constructed information) are somehow inferior, and people who rely on them lack health information literacy. What blinding snobbery and elitism! It should be turned back on them 10 fold.

    Do we want health professionals like this occupying social information spaces? Do we want them bringing their blinding prejudice and anti social thought patterns to such hopeful and optimistic spaces? I think not. People like that are our medical nemesis.

  16. Hello Leigh,

    The people who said that Yahoo Answers were not good sources of information were not doctors as far as I remember- @gfry and @mrgunn sprinf to mind.

    The fact is that I'm not continually coming across misinformation on Yahoo Answers and not responding to it because I never have reason to access Yahoo Answers. So I would have to decide to deliberately go in there and see if I could answer some queries. This is what the screencast is about- my foolish presumption that someone checking out symptoms would come across easy to use NHS websites. If I search for information on a diagnosis to find a patient-info leaflet or show a patient or students a good website I don't come across Yahoo Answers because on Google UK at least, it just doesn't rank that highly.
    The number of new queries on Yahoo Answers UK every hour is not nearly as high as the numbers who we know are searching for health information. That's why search (and the websites of the NHS) have been my focus here.

    Hopefully some of the people who originally responded to my question about Yahoo Answers will come back to respond here.

    Thanks again,

  17. Leigh,
    I feel I should also add that there is no doubt that I am a strong advocate of crowd-sourced information in patient communities. I've benefitted from being a (patient) member of one for many years. But again this is post-diagnosis.
    Some patient communities have health professionals involved but many don't (including the one I was a member of)for good reasons.
    OK- I hope that has helped to clarify my thoughts for now!

  18. Fascinating. I don't want to rush into commenting really because I want to think about things a bit more, but one framework for thought might be the following: In plain language, very user and problem dependent. In more techinical speak: Internet diagnosis, like any test, will have a sensitivity and specificity, but will also have great inter-user variability. Secondly, it is a relatively new and select group who will use the internet for this purpose, therefore their pre-test probability of having a specific condition will be different perhaps from someone seeing their GP or attending their local Emergency Department ... I'll follow the discussion with interest. Thank you

  19. Hi Anne Marie, please don't mistake my heavy calls as being directed at you. I think your post and screencast have shown a genuine willingness to engage with my position.

    Using Google Search as a measure of reach and engagement is a so so measure. I think most people with some experience in searching for online help in 'diagnosis', or just some connection to people for shared experiences and discussion on symptoms or issues, have already found their communities and trusted channels, and don't use search so much.

    My focus on Yahoo Answers and Wikipedia is just one example. There are obviously other channels, such as Google Groups, Facebook, specific forums such as Bubhug for young mums, and local forums.

    I think the online public health discussion needs to get way beyond ideas of legitimate or authorative information and ideas of "health literacy" and start talking about methods of real engagement and reach.

    There are online services that can provide a health information service with monthly statistics on where their target audiences are congregating. They could even set up facilitation services whereby they alert various organisations to opportunities to engage.

    Alternatively, and much more costly IMO, is to attempt to build community on the NHS site, mimicking the features on such spaces as Yahoo Answers and Wikipedia.. good luck with this option.

    There would be other approaches, but first, we need to get health professionals (not just doctors btw) past their high and mighty positions, and in amongst it, understanding the grammar of these spaces more.

  20. AM,

    I am glad you have brought up this issue but I am not certain my answer will appeal to the majority of non-medics out there. The issue is explaining the very important difference between using symptoms to diagnose a disease versus education about a diagnosed disease. I suspect most people have unrealistic expectations and do not understand both the art and the limitations of modern medicine.

    As we all know, the day we (in the US) declare that we are "pre-med" is the moment that, at least prior to the wide availability of online tools, we begin fielding questions from friends about medical issues. Anecdotally, these questions are usually not about explaining the pathology of a disease nor are they on the experiences of patients during treatment but are more along the lines of a quick-and-dirty diagnosis. As the internet has permeated people's lives we have seen a large number of questions searching for a diagnosis online when they do not have access (or are too embarrassed) to ask a friend or family member that is a doctor. I suspect our British counterparts have similar experiences with friends and family.

    The issue is not that doctors are selfish luddites trying to hold back progress but not improving self-diagnosis tools on the internet, participating in Yahoo! Answers, or other services where diagnosis queries are posted. The short answer is the art of diagnosis is not simply inputting a list of 10-20 symptoms into a search engine hoping to find a “hit” where someone has the exact same issues. This is not medicine. A large part about medicine is learning an immense amount of book knowledge during the basic sciences years and then honing it throw apprentice-like training with real patients under the guide of highly experienced doctors. It is at this point that we learn the types of questions we need to ask in a history and then search for clues with the various tools at our disposal while the patient is in our presence. This is not easily replicated online for many obvious reasons.

    Even in this thread a lot of people are looking for triage-like or diagnosis-like questions and not focusing on disease education. As I said earlier, even these days I field more of these types of triage questions from friends than any other type. Invariably these friends have tried searching online, ended up with bad information, and have come up with outlandishly crazy conclusions that are harmful and distressing. Not only do I have to deal with the emotional effects of bad information but I have to quickly triage these friends on the phone by asking simple questions. The most fascinating but unsurprising thing about these interactions are non-doctors often times discount the simplest symptoms which leads them astray. Interestingly even with the dire knowledge of ‘imminent death’ many of these people are not willing to go into a hospital and see a doctor.

    To answer your question directly: No, I do not believe it is appropriate for doctors to answer questions on Yahoo! Answers with limited information and wrong usage of terminology. Diagnosis in new-to-the-doctor patients should be conducted in a setting where communication is real-time and two-way. While some people advocate the idea of putting algorithms drawn in Visio online to help diagnosis, I am not in that camp because usage of such charts can be highly misleading for people without a wide knowledge base in medicine. It all comes down to this: self-directed diagnosis on the internet by lay people is not appropriate. If triage is a concern then offer people a toll-free phone number for access to on-call triage nurses. The public internet is better geared for patient education post-diagnosis and sharing of experiences with others in similar circumstances. To that end, we must provide recommendations to sites that have great information for the public. Also, the public needs to realize this is not an issue of “high and mighty” doctors but highly technically-skilled doctors that realize the limitations of these mediums.


  21. My (very brief) promised comment: I think doctors on-line should refrain from giving anything more than very general information and/or referring patients to patient-education websites where they can make their own decisions. There are too many variables for us to really assess what is going on, and we risk crossing the line to practicing medicine w/o having the necessary information.

    When asked, I'll often refer patients to or the Merck Manual Home Edition ( for more information. Both are reliable, offer overviews of conditions and symptoms, and warning signs that more information might be needed. Neither really offers much of a way for patients to evaluate their symptoms in much detail, but I think that things are safer that way.

    I don't see much of a role of physicians policing other sites (Wikipedia or Yahoo! Answers. Too much ground to cover, not enough time to do it, and some potential risk if you offer concrete advice and have a bad outcome.

  22. This is an important conversation we're having here. I'd like to add to the continuing mix the fact that there are sites where doctors are answering questions: MedHelp ( has been doing this since 1994, and the newly-launched ShareCare (from WebMD's founder, Jeff Arnold) is linking, too, with doctors to answer questions -- although ShareCare's business model has a lot to do with commerce and marketing compared with the MedHelp model. The doctors at MedHelp's site come from prestigious medical centers such as Johns Hopkins and Cleveland Clinic. ShareCare is linking up, similarly, with top doctors for Q&A with the public. This is an ever-morphing space. If health citizens prefer the MedHelp's and ShareCare's and Wikipedia's of the world to Yahoo! Answers, then they will survive based on the size of the crowd and, eventually, health outcomes. Perhaps optimistic...but a growing cadre of engaged patients are making this happen. Check out the Society for Participatory Medicine for more on this phenomenon at Thanks, Anne Marie, for hosting this conversation!

  23. I said I’d have a think. I’m glad I did. Presentation watched and all contributions so far read with interest. In my opinion, a high and mighty, technically skilled summary is required; I’ll do my best…I’ve found it quite difficult. Interestingly, the difficult bit was organising the comments which have gone before. The easy bit was typing my own comments.

    The last thing I feel is high and mighty by the way

    AMC’s original question (arising from her experience of patients consulting her with information from the internet) was how can we improve the experience of the person seeking information on the internet to make sense of their symptoms?
    • Make websites more user-friendly?
    • Better tools for symptom sorting?
    • Make sure that the relevant NHS pages are search optimised?
    • Patrolling of Yahoo Answers by NHS

    Comments and my own short additions
    • Example of a difficult diagnosis, regrettable
    • How relevant is patient’s agenda in consultation? Very. Not only does it chime with me, I hone in on it frequently (
    • Role of bias (e.g. marketing) on websites, individual must guard against this
    • People are improving their searching skills, maybe not their appraisal skills. They should try. Happy to help, subject to my own resources
    • Should the websites be used for diagnosis at all? Yes, but not at the expense of consulting other services such as NHS Direct, Walk-in-centres, highly skilled professionals such as a GPs or Emergency Care Professionals (in an A&E department for those in the UK). Reason I say this is that an engaged, educated patient must be at an advantage over a unengaged, uneducated one
    • If so, should health professionals (NHS or private) contribute to internet diagnosis or not e.g. policing of Yahoo Answers. I don’t ever use Yahoo Answers and my patients don’t generally carry internet printouts when I see them, wouldn’t mind if they did, see above
    • There is a risk of harm if poor information is acted upon and not corrected…entirely agree. If I did come across something like this, I would make a decision at the time as to the risks and benefits of getting involved, taking into account whether my professional indemnity covered me (probably not) and what might happen if something wasn’t said. I would concentrate on the general….AMC; I’ve discussed Twitter clinics with you before now and we agreed that this sort of thing was risky
    • Negative attitudes of health professionals towards public engagement…not sure that this is that real, would like to hear more about it
    • Need for ‘real engagement’. Yes. Probably resource and ‘change’ problem
    • Value of online communities/crowd sourced information (gated or open e.g. Google searches). Must be good things. Happy to engage, subject to resources and equity
    • Convergence of media to enhance patient engagement = participatory health..see above
    • Possible to build this into NHS sites? Over time and carefully. Just look at the Summary Care Record for how something that sounds so simple has to be implemented so carefully. I expect there will be a mixture of managed public websites and the private sector (much like the BBC and independent broadcasters and time will tell if they converge). This is one of the intentions of NHS Choices, I believe

    Hope this is a helpful contribution

  24. My apologies for not replying or summarising for a week! Thanks to Tom for the prompt. I think this really needs a follow-up post when I get time.

    There are very many different sites which can help provide information and support when one knows what is wrong or likely to be wrong. These are a mixture of government provided resources, private website and patient communities. But if wondering what might be wrong then the quality of information is poorer. To my mind there is at least one important reason that we might want this information on symptoms to be better: so that those with serious symptoms know to see further advice.

    I think that most of those who have responded here would agree with this.


I am reintroducing word verification to cut back on spam posts. I'm sorry if you find it frustrating,