Everyone enjoys variety. I believe this is proven by the success of buffets across the country. I think that's also why there are so many different flavors of ice cream (though, personally I don't understand a selection other than Rocky Road).
I recently completed an annual Internal Medicine Maintenance of Certification course, where all the physicians in the room had audience response tools. What most surprised me was the amount of variability in the answers from practicing board certified physicians (including me!) that was clearly evident on the large screens.
We all want to support (when available) Evidenced Based Care while trying to decrease variability in the care we deliver to our patients. We all want sustainable quality care. I am defining quality care here as the best possible care for the best possible outcomes, with the best quality and length of life for each patient.
The difference between selecting the best care and picking your favorite ice cream is that ice cream selection is often a two-step process:
1. I want ice cream. 2. I want a particular flavor of ice cream (Rocky Road).
Delivering care to a patient has an almost infinite number of factors that can be considered—such as age, sex, race, genomic make up and even personal goals—that all play a role in the delivery of quality care.
Combine this with the fact that a well-rested, well-caffeinated physician can remember approximately 7 facts in the diagnosis and treatment of a patient. Thankfully, physicians are always well rested.
Clinical Decision Support—provided within the clinical workflow—that understands the patient remains my goal for all physicians as they care for patients.
Aside from alerts that I receive when I am ordering a prescription, that a patient may have an allergy or an unsafe interaction with another drug they are taking, my clinical decision support comes from outside sources such as Up to Date, The New England Journal of Medicine or Harrison's Textbook of Internal Medicine.
To receive this decision support, and gain the insights needed to deliver quality care, I need to open up additional programs—outside of the Electronic Medical Record—and then try and find the right content from a keyword search.
My goal would be that while I am seeing and documenting patient care, clinical decision support is being provided to me in real time. The clinical decision support tool will have read the patient’s record, my current notes, and will understand the patient's vitals, past diagnosis, medications, genomic data and even their life data (e.g. steps walked and food eaten)—all while comparing this patient against all the most current treatment options, guidelines and medical knowledge.
We, at IBM, started to tackle this challenge—of driving quality care through the removal of variability—with cancer treatments. Now patients that are seen at Memorial Sloan Kettering and even a community hospital in Manipal, India are receiving cancer treatments based upon the guidance of world-renowned oncologists and the most up to date published medical knowledge.
Being able to bring together world-class protocols, patient preferences, genomic sequencing and all of the world’s published medical knowledge—with Cognitive Computing—is helping remove variability. It is also creating a path forward for standardization and personalization in the diagnosis and treatment of not only cancer but also diabetes, sepsis, and congestive heart failure. We are all just getting started with driving an unprecedented level of standardized quality care—centered around an individual patient.
Medicine focused on "do no harm" has not historically tended to make dramatic shifts in delivery of care.
With an unsurpassed creation of knowledge (by 2020 the Library of Medicine will catalogue over 1,000,000 publications per year), a globally aging population and emerging technologies, we need a new Clinical Decision Support technology to drive quality and drive out unwanted variability. (What if hidden in one of those publications is a specific treatment for your child, your spouse or a friend?) Cognitive computing will help medicine make that dramatic leap forward.
In the future, with cognitive computing removing variation, we will be able to support a physician in selecting the correct imaging study at the right time, or when to most safely to use blood products during surgery—based upon a complete understanding of the patient, a complete understanding of the current guidelines and a complete understanding of all the current published literature.
Cognitive computing is poised to usher in a new era of quality medicine. Let's remove the variety and save the ice cream choices for after a tonsillectomy.
Michael Weiner DO MSM MSIST is the Chief Medical Information Officer for IBM Healthcare and author of Provider-Led Population Health Management: Key Healthcare Strategies in the Cognitive Era.
