Imagine that you are a primary care physician, it is 8 p.m., and you’re just finishing your day in the office having seen 25 patients. Despite the routine nature of most of the care delivered that day, you’re acutely aware of the need to find answers to dozens of questions that you will need to provide the best care possible to your patients. You know all too well, however, that it will be challenging, if not impossible, for you to get these answers. You don’t have the time, and even if you did, finding some of the answers is like finding a needle in a haystack, even in the post-Google era.
Medicine is the definitive information-centric profession. Though nuance and clinical judgment play a critical role in medical decision-making, the most current information is the single most important variable in ensuring that correct diagnoses are made in a timely manner and that optimal, guideline-based treatments are implemented.
Our profession used to have too little information. Now it has too much.
IBM’s Jeopardy-champion computer, Watson, has the potential to solve the information glut problem. In its simplest rendition, a question-answer wunderkind, the machine can answer the questions that arise during the course of a day in clinic. The more questions we ask and get answered promptly and reliably, the more questions we will ask.
Watson can also have an impact on diagnosis. Approximately one-quarter of all medical errors involve misdiagnosis or delayed diagnosis. For example, it can take several years before symptomatic patients are correctly diagnosed with myasthenia gravis or multiple sclerosis, and many years for celiac disease. Give Watson the clues to an undiagnosed condition (double vision, difficulty swallowing, drooping eyelids) and Watson will return myasthenia gravis as a potential diagnosis instantly. While these diagnoses would be immediately apparent to specialists, they might elude detection by a primary care physician who lacks the in-depth knowledge of a specialist. Watson can bring to the primary care physician the knowledge, albeit not the experience, of the specialist.
Ultimately, Watson’s greatest value may be realized in the influence it can have on treatment. Patients, especially those with chronic disease, have a unique set of issues that need to be taken into consideration when optimizing treatment. One size doesn’t fit all. While we do not need Watson to diagnose hypertension, we will certainly need it to treat hypertension. Hypertension has many causes, some of which have been identified genetically, which call for tailored treatments, which in turn must be modified considering the patient’s age, co-morbidities, gender, other medications. Treating a 75-year-old man who has dizzy spells and is on drugs for prostate disease and arrhythmias is much more challenging than treating a 43-year-old woman with no other diseases and no medications. And even beyond the diagnosis and tailored treatment decisions, Watson could track and inform the health care provider as to which insurer requires a huge co-pay for a drug or what generic substitute might be available at a huge savings.
The combination of Watson and a nationally implemented electronic health record should have a dramatic impact on patient care and ultimately outcomes. After years of anticipation, medicine is poised to truly capitalize on information technology and enter the information age.