These are extremely exciting times. We’re in the midst of a healthcare revolution – have you heard about it?
Medicine, in the past
Medicine is a marvellously interesting field – I love it. We’ve come along so far over these past 100 years since penicillin was invented (1928), and yet there is still a long way to go. There is still so much of disease that we do not fully understand, or even if we do, have not found the cure for.
I’ve spent a good part of my life in medicine, almost half of it now. 5 years in medical school, and another 8 years working in various public hospitals as well as in the Singapore Armed Forces.
Despite it’s rooting in hard sciences – biology, chemistry – many people say that medicine is an art, and that’s quite an apt description. As medical students, we are often told not to just ‘study your textbooks’, but to go out to the wards, and see real patients, build up your clinical experience. After graduation from medical school, our training in the various specialties is very similar to an apprenticeship – we learn from our seniors by following and observing how they treat patients, at the same time we spend a lot of time honing our clinical skills with practice, by seeing many, many patients with similar diseases and conditions. This is, in part, to build up our mental bank of clinical experiences, such that we are better able to tailor our treatments to different patient types.
For example, by working in the Endocrine department of a large hospital, I’d be managing a variety of patients with diabetes, both young and old. I’ll accumulate first hand experience at how different patient types (age, gender, race, health status etc) respond to various types of diabetic medications, such that in the future I know what treatment works well should I encounter a similar patient in the future.
Our professors, with decades of experience in their specialties, are so in tune with their patients and are able to titrate treatments with great precision, because of this bank of clinical experience from years of practice.
Is there a problem?
This has served us well so far. And yet I’d argue that there is a lot of unjustified variability in medicine. What do I mean?
Take for example a gentle 70 year old lady with multiple medical conditions such as high blood pressure, high cholesterol and diabetes. Her diabetes is not well-controlled despite being on multiple oral medications, and it’s time to move on to insulin injections.
With this patient, Doctor Lee takes a more cautious approach with the insulin dosages because he’s worried that she’d go into hypoglycemia (low blood sugar).
Doctor Tan sees a patient with a similar profile, however he manages the patient’s blood glucose in a more aggressive manner, titrating the insulin dosages closely to achieve tight glucose control, in the hope of minimize risk of complications in later years, as the patient is relatively well and has many years of expected life left.
Both approaches are reasonable, and are often related to the doctor’s clinical knowledge as well as past experiences. Perhaps Doctor Lee has had elderly patients who went into coma from low blood sugar after starting insulin, or Doctor Tan read in the latest issue of Diabetes Care that aggressive treatment, even in elderly patients, has been shown to reduce mortality rates.
Who is more right? In the past I’d say that there is no right answer. But now, I’m not so sure.
Transitioning to a world of big data
We are transitioning from a previously data-poor environment, into a data-rich new world. I’m going to quote a few well-reported statistics so you can understand the magnitude of this:
From the beginning to time to 2003, we generated 5 billion gigabytes of data. In 2014, we generate that same amount of data in 10 minutes.
By 2020, in a mere 5 years, the amount of data we have available will 20X what we have today.
Check out this cool infographic by http://Wikibon.org/bigdata
(Click photo to see in full)
Many industries have already begun to capitalise on this data revolution that we’re experiencing right now. The financial and E-commerce sectors, in particular, have done this well. With precise information about each customer’s demographic, previous activities and persona preferences, they are able to better give the customer what he/she wants, even before he/she knows it. In fact, with hyper-competition in E-commerce these days, it’s the company which has the information advantage that will prevail over its competitors.
So can we bring some of these concepts into healthcare, in order to improve clinical outcomes? Healthcare has traditionally lagged behind other industries in technology adoption, because of several factors:
1. Regulation and Legal frameworks are slow to change
For example, despite the growth and increasing usage of telemedicine in the US (Doctors on Demand) in order to use resources efficiently, it has yet to take-off in Singapore. Our medical council ethical guidelines state that doctors cannot establish a patient-doctor relationship through virtual consultation. Until our regulatory frameworks become more dynamic and keep up with the fast-changing landscape, they will remain barriers to technology adoption.
2. Patient Privacy
Unlike the data of a customer who purchased a pair of shoes from your shop, healthcare data is more personal, private and needs to be protected. I see this becoming less of an issue, because banks deal with sensitive information as well, and already have the technology to ensure security and privacy of data.
3. Industry Resistance
This is more anecdotal, but doctors are highly resistant to change. It’s something I can feel in myself as well. Perhaps, having spent so much time and hard work to perfect a particular way of doing things, we naturally get skeptical when someone tells us ‘Maybe there’s a better way to go about it’. If you’re a doctor too, feel free to disagree with me 🙂
The Future of Medicine
Once we push past these barriers, it’s not hard to imagine a world not too far in the future where we can use big data and analytics in order to improve (not replace) our clinical decision making. Our human brains are limited in the amount of information we can process before we blow up, but computers aren’t. Never before has computing power been so readily available and cheap.
Going back to the earlier problem of the 70 year old lady starting insulin, with enough information from previous patients of similar demographics, we can develop complex algorithms that can accurately predict which mode of treatment can provide the best overall outcome in terms of morbidity or mortality.
We can never get this down to a perfect science, but even a tool that can do this with 70 – 80% accurately will be extremely useful in clinical decision making.
What do you think of the data revolution in healthcare? I’d love to hear your thoughts 🙂
I’ll share more on the possibilities of transforming healthcare in a later post.