Can the IBM Watson team move beyond creating Paint by Number Style Medicine and grasp the disruptive innovation opportunity?



Let’s meet our main character Mrs Yamato, she is a 37 year old woman from Osako Japan who has never smoked. She goes to her Primary Care Physician with a persistent dry cough and laboured breathing. The physician prescribes a chest X Ray which shows a suspicious mass, a subsequent CT scan and biopsy confirms the likelihood of cancer. The Doctor refers Mrs Yamato to an oncologist for treatment and forwards her Electronic Medical Record or EMR.

On the day of her appointment Mrs Yamato sits in the waiting room of the Oncologist Dr Mark Norton she is understandably nervous about her situation in the few minutes he has between appointments Doctor Norton steps into his office to familiarise himself with Mrs Yamato’s case he logs into the Electronic Medical Record and instead of digging through the different sections of the EMR to try to find the relevant information for the case he clicks on the “Ask Watson” button in the bottom left (of the screen) and Watson evaluates all of the information in the EMR and analyses it against tens of thousands of documents in it’s vast core of evidence sources like medical journals, industry association guidelines, specific hospital best practices and identifies the pertinent case facts. Watson has brought forward information like the results of Mrs Yamato’s recent Chest X-Ray, CT Scan and Biopsy. Dr Norton can see a more detailed view of each fact in the EMR if he want’s to. The evidence button shows the reason why Watson pulled this infromation out as relevant to the case the advisor can also prompt Dr Norton for additional information that would result in a more complete view of the Patients condition. In this case it’s asking if Mrs Yamato has had any hamoptysis which is coughing up blood or if her hearing is normal so it’s really a two way dialogue between Dr Norton and the Watson advisor. Here he sees tests that he might consider ordering, a molecular Pathology Panel…

…Mrs Yamato explains to him (her oncologist) that it would be difficult to explain sudden hair loss to her young children and that she would like to avoid that if at all possible Dr Norton dictates this preference along with his other notes which are added to the EMR… …the first plan a three drug combination has a high level of confidence and an acceptable match with Mrs Yamato’s preference to avoid hair loss. Clearly the most effective treatment is a first priority and matching her hair loss preference is just an added bonus…

…(second consultation) at this point he (Dr Norton) feels well prepared and calls Mrs Yamato to come in. They discuss the test results and treatment options and Dr Norton asks if there have been any changes to her condition in the last ten days. She tells him that just yesterday she coughed up a small amount of blood. Knowing that this will likely alter her treatment plan Dr Norton presses on the blue button and speaks directly into the microphone on his tablet. He pushes the finish speaking button when he is done and checks to see if his speech has been correctly converted to text… …the advisor comes back with an alternate set of treatment plans…

My thoughts;

I’m clearly very optimistic about the potential for Doctors to leverage IT (I’m actually surprised that in 2013 it’s not compulsory for practising Doctors to have a connected smartphone with at least a comprehensive library of medical text and reference books saved to it’s memory) but while I’ve read a lot of optimistic news coverage of how IBM’s “Watson” question answering computing system will supposedly transform the Healthcare industry it’s never been clear to me why if this technology was so primed and ready it wouldn’t already be doing much more basic boring things like filing in the US patent office (plenty of profit to be made there crunching data, punching out documents and it could presumably put to rest the patent troll business overnight) before focusing on doing away with the need for some of the non core roles that consume healthcare budgets eg. lawyers. Perhaps the engineers who are directing Watson have fallen for the illusion that oncology is an area of medicine where the care is documented?

Some more insights into the potential can be found it this video demo showing some of the iPad interactions that an oncologist at the Memorial Sloan Kettering Cancer Centre might experience when consulting with a hypothetical patient using the IBM Watson Cancer Diagnosis and Treatment Adviser.

Let’s meet our main character Mrs Yamato, she is a 37 year old woman from Osako Japan who has never smoked

I don’t understand why they’re referring to Patients as “Main Character” but the idea that this simple line proxies as some identity for this Patient really is missing the value opportunity to capture the complexity of individuals that speech recognition and the Watson Advisor service presents.

If I was producing this video to communicate the value of Watson I wouldn’t stop pointing to the way it can capture the enormous complexity of the individual and through the use of advanced computing power enable this to be presented in ways that clinicians will find valuable at a glance and on a deep dive (if they want to) eg. “we can see here indepth information about Mrs Yamato, for example although never a smoker her father did chain smoke and so does her husband of 15 years…”

A computer is only as good as it’s programmers – Watson’s focus on the wood is making it miss the forest

Watson should digest this Mayo Clinic Proceedings Paper by Prof John Bachman MD. The biggest opportunity we have today lies firmly in listening to Patients earlier and making sense of what they are telling us. Cancer Care won’t be transformed by yet another decision support tool for Oncologists as that’s still firmly in the camp of putting faster Ambulances in the valley – when the big opportunity is obviously in getting some signs and fencing at the top of the cliff.

The failure of the programmers to appreciate the importance of listening to the Patient is also alluded to e.g. “the advisor can also prompt Dr Norton for additional information that would result in a more complete view of the Patients condition. In this case it’s asking if Mrs Yamato has had any hamoptysis which is coughing up blood or if her hearing is normal so it’s really a two way dialogue between Dr Norton and the Watson advisor

What’s wrong with Watson asking the Patient to add this very obvious piece of clinical information to their medical record – like the Instant Medical History tool we use here at 3G Doctor already does – before a rushed Doctor meets with them in the consulting room?

Perhaps the clinicians would also find that information could radically transform their Patients outlook and approach to treatment eg. this Patient’s hair loss concern didn’t emerge until in the consultation with their oncologist – as such a obvious and common concern not only is a waste of valuable time to be only talking about it at this late stage but it’s clearly caused by a failure to do enough basic listening to Patients eg. an Interactive Cancer Patient Questionnaire could’ve alluded to this Patient concern and their Family Doctor could’ve managed this concern with more time and greater understanding – perhaps by sharing with Mrs Yamato this informational video of Mario Lacouture a Dermatologist at the Memorial Sloan Kettering Cancer Centre outlining the issues and options:

Healthcare isn’t about a relationship between a computer and Doctor

The video clearly identifies that even with this Watson Advisor decision support aid the Oncologists are still going to be flying blind because they are failing to listen and document their Patients concerns:

at this point he (Dr Norton) feels well prepared and calls Mrs Yamato to come in. They discuss the test results and treatment options and Dr Norton asks if there have been any changes to her condition in the last ten days. She tells him that just yesterday she coughed up a small amount of blood

It surprises me that even in future scenarios being drawn up for the world’s oldest and largest private cancer center we still have medieval practices like this being planned for.

Radiological images and Radiologist reports aren’t facts

For me the statement that clicking on a CT image report will provide “a more detailed view of each fact” highlights a concerning lack of medical experience in the IBM Watson team. It’s critically important that we appreciate that we’re very often not looking at facts but rather the opinions of other fallible experts.

The Watson team at IBM should look to add service innovation

I get the feeling that this video will age quickly as the future of medicine isn’t just focused around the mobile devices used by Oncologists because there are already 6+ Billion citizens with these devices.

The idea that Mrs Yamato has to go and sit in a waiting room while a rushed Consultant supposedly familiarises himself with his Patient by looking at her EMR “in the few minutes he has between appointments” has me thinking that the Watson developers and producers of this video haven’t yet even considered how different things might be if Mrs Yamato had a blog and followed her Oncologist (and had 2,300 followers) on Twitter!

It surprises me that IBM haven’t focused more on the service innovation that Watson could support as it’s also going to be a lot more exciting than what they are demoing here eg. imagine if this video stated:

the next day Mrs Yamato sits down at home in Osako in front of her 54inch Samsung Smart TV to join a discussion that’s taking place between her trusted Family Doctor and the most experienced Oncologist for Patients with her symptoms at the Memorial Sloan Kettering Cancer Centre – 7,000 miles away in the United States of America! Mrs Yamato was understandably very nervous about her situation but felt great about being seen at the earliest opportunity that her Oncologist had available in her diary and of course without the need to make the trip to New York! With the discussion video recorded and automatically transcribed and filed in her secure EMR this information is now there if Mrs Yamato or any of her care team would like to review/view it at anytime. Her MSKC Oncologist was also very grateful for the opportunity to be able to schedule this appointment into a Saturday morning schedule when she finds it easier to work with her Watson Advisor from the more relaxing environment of her home office…

What do you think? How would you like to see Watson getting to work?

About 3G Doctor

The Corporate Blog of 3G Doctor
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s