Will computers replace doctors in diagnosing rare disease?

We have been talking about using analytics to find patients with rare disease. Whenever I tell my friends what I do they turn ashen and ask about their privacy. When I talk to friends who are doctors, they turn ashen and wonder how their patients would react to being diagnosed by a computer.

So, all you ashen friends, let me clear up a few things:

  1. Everything we do is 100% HIPAA-compliant
  2. We are not a diagnostic tool. We are not a decision-support aid. The decision-making still falls to the provider
  3. We have no interest in replacing doctors. We just want to help them

In the defense community we have shifted focus from relying on “human intelligence” and sensor collection to “activity-based intelligence”.  We rely less on people to collect data and more on the integration of data from many sources. People are insightful, but they can’t process terabytes of data. We are leveraging advanced analytics in healthcare now to collect, sort, and find patterns in the data. Activity-based intelligence takes us beyond simply looking for a needle in a haystack. It provides valuable insights into the composition of the hay, where it came from, how long has it been there, and how much it costs.

Some worry that this shift – combining analytics and data to augment the current methods of practicing clinicians -- will lead to a world with limited human interaction. How aggressive should we be in using the digital footprint left behind by patients for disease identification? Are we at the tipping point between patients finding therapies and therapies finding patients?

We believe that as more medical data are collected, the role of the healthcare professional will be transformed.  We will always need human touch when we care for the sick, but wouldn’t it be valuable if we could diagnose someone with a rare disease after three years instead of twelve? Just as the x-ray, the MRI, and the ultrasound revolutionized diagnosis, so too will medical data coupled with appropriate analytics.

This isn’t a ploy to learn about spending habits or purchasing preferences. This is about getting patients and the proper therapy together. This is about minimizing the role of luck, geography, and memory in diagnosing rare disease.

Tara Grabowsky, MD
Chief Medical Officer