Intelligent Clinical Decision Automation is the Future of Clinical Decision Support

What is “intelligent decision automation”? And why is it the future? Intelligent decision automation is a term used in some industries. It combines artificial intelligence (AI) with workflow automation and has the power to change the way business is done in nearly every sector of the economy.

artificial-intelligence-503593_1280Artificial intelligence and automation are hardly new, but state of the art in these technologies has progressed significantly in recent years. Over the past few years, mainstream America has learned about “self-driving cars.” Not too long ago this idea was science fiction, and only Hollywood could get away with it. But now, it is an emerging reality — and one that will soon be realized by the average person. Initially, cars will drive on their own with human supervision, but in the coming years under computer control, as decision rules and the logic behind the controls improve, cars will do increasingly more. This same progression can happen in healthcare.

There is no industry in more desperate need for artificial intelligence and automation than health care. Famed venture capitalist Vinod Khosla has predicted: 80% of what doctors are currently doing, including diagnosing and prescribing, will be replaced in the not too distant future by intelligent systems. The transformation from an entirely human-based healthcare system, to an automated system supporting a smaller cohort of healthcare professionals will take time, but it is inevitable. There may be naysayers who doubt or even fear this change in health care — the same way I’m sure autonomous cars frighten some today. Intelligent automation in healthcare is going to happen — and when it does, many of the healthcare problems we face will be eliminated.

I believe a new class of system will take form and leverage artificial intelligence and combine workflow automation to improve how care is delivered — I’ll term this: “Intelligent Clinical Decision Automation”. This AI-powered automation will consume vast amounts of data and will automate entire processes or workflows, learning and adapting as it goes.

Clinical decision automation, powered by artificial intelligence, will complete the transformation of the practice of medicine to the science of medicine

The artificial intelligence revolution is just getting started in health care — and the machine learning applications designed today are only helping people make decisions — they are not  automating end-to-end processes. What I foresee in the near future is much different. I believe the future of health care is Intelligent Clinical Decision Automation — where decisions and workflow are automated — and where clinicians supervise machine intelligent systems that operate without human intervention.

What does this mean? It means the needed diagnostic tests and procedures are automatically ordered, the right medications are prescribed, the appropriate digital therapeutics are recommended, referrals are scheduled, prior authorizations are instantly approved, and orders are automatically placed — all based on the information you provide in advance or at the visit, which will be collected electronically and analyzed intelligently. Each of us is different, and even if two individuals present with similar complaints and symptoms, the system will personalize the care plan based on insights extrapolated from a vast data bank, encapsulating relevant experience of human clinicians.

There is a big difference between Clinical Decision Support and Intelligent Clinical Decision Automation. It is the difference between the car that warns its driver about a stopped vehicle just ahead, and the car that automatically prevents rear-ending the stopped vehicle. Today, Clinical Decision Support might warn a physician of a contraindication or may suggest a better treatment pathway. But intelligent automation will flip the paradigm: machine decisions will be the default decisions, and the physician will be there to supervise, and in rare circumstances, intervene and correct course.

At Flow Health we are working on an intelligent decision automation platform — we call it the Operating System for Value-Based Care. Our platform leverages machine learning to help health plans and providers individualize care plans and automate care team coordination. But this is just the tip of the iceberg: Intelligent Clinical Decision Automation will improve and change how healthcare is delivered for everyone.

Comments (4)

Tim Shear

Jan 24, 2016 at 9:28 AM

What is still needed is a platform for the longitudinal patient record. Searching “big data”, or worse EHRs, to pull together the patient record is an error-prone, time consuming process. I don’t see it being applied in real or near-real time decision support for the physician. Better to have pushed patient data into a single patient “record” that can be queried to produce relevant data points to apply machine learning to the specific patient’s condition. Sure, machine learning can be applied to aggregations of many patients, but it cannot result in “precision” treatment options.

Reply

The Editors

Jan 25, 2016 at 7:55 AM

Tim, We couldn’t agree more on the need for a platform. That’s why we’ve developed the Flow Health platform (see http://flowhealth.com) and the patient and provider applications needed to interact with the universal patient data layer. However, the long-term vision is to improve not only individual patients’ health through better management and understanding of each patient’s data, but to improve health on a population level through improvements in machine learning across many patient records.

Reply

Michael Hopp

Dec 23, 2017 at 1:43 AM

Yes i agree, intelligent clinical decision automation is the future of clinical decision support. An intelligent decision support system should behave like a human consultant: supporting decision makers by gathering and analyzing evidence, identifying and diagnosing problems, proposing possible courses of action and evaluating such proposed actions.

Reply

Leave a comment