Blog post

Automation in Healthcare- Part 1

By Anurag Gupta | September 15, 2015 | 0 Comments

Automation

(This is a 2 part blog: This is Part 1 of 2)

There is significant talk about doomsday, when machines take over a majority of our jobs. Look at the some of the recent business best sellers books or even Hollywood; and many are hypothesizing about the attack on our livelihood.

Healthcare Delivery with its archaic way of doing things (the human body has changed very little in last 40,000 years or so; & so have the paper based medical records rooms, as some may argue), siloed mentality and a physician centric model (before you say it, yes, population health & integrated care, is still WIP) is often spoken of, as a ripe target for the forces of automation. Advances in Artificial Intelligence (AI), more especially machine learning are often cited as important driving factors for this upcoming revolution; where up to 80% of what the doctors do could be replaced by machines.

Really? Is that the truth? And what should healthcare institutions and technology firms do about it?

Before we delve any further, it’s important to understand an important concept. With all the hype surrounding AI, we tend to forget it’s another close cousin: Intelligence Augmentation (IA) or Intelligence Amplification as some may call it. In most simplistic terms, whereas AI tries to create a computer which mimics being the smartest human (along with cognitive capabilities); IA only tries to make a human smart, assisted by machine/computer.

It is little secret that doctors spend an enormous amount of time initially in medical schools and then on their jobs, accumulating vast ‘libraries of cases in their heads’, on which they base their diagnosis skills. However, just like advances in technology, biology is also having its ‘aha moment’ now; albeit surely aided by technology. This means that clinical interventions & best practices change rapidly and it is almost impossible to keep abreast of latest development. This applies to almost all ‘ologies (oncology, gastroenterology, nephrology, cardiology etc..) in the hospital. This is where IA can help; and it should be not be confused with applying AI in healthcare. IA can assist the doctors and even nurses (many outpatient cases are minor cases and do not require a specialist person to attend). In short the objective is to assist the human decision making, not replace it (atleast, not for now). This perfectly complements the years of experience and the trained eye of the clinical personnel and enhances the decision making through computers. Computers are very good in tasks that humans are not too good at; and vice versa.

To put it in a simple analogy: In complex decision making tasks: man + computer are better than a smartest computer alone

If you are a CxO, then consider 2 aspects: Introducing IA in your organisation and Making sure that your clinical staff learns to use them. And if you are a technology firm; then understand that IA will provide the most benefits in the area of ‘knowledge’ work for your clients.

So, where should AI be explored then? We will explore that question in the next post: Part 2.

What are your thoughts?

Leave a Comment