#6 Leading Practitioner Interview

We often hear from the community that practitioners are interested in getting an insight into what other health and care informatics professionals are up to. So, we are delighted to introduce you to one of our newest Leading Practitioners, Dr Vaibhav Joshi.

Vaibhav Joshi is a seasoned medical informatician focused on Data & AI in medicine.  A Founding Fellow of the Faculty of Clinical Informatics and a certified InfoSec professional A&E doctor who established the Private Healthcare Information Network to define and publicly communicate value to inform the consumer selection of healthcare in the UK.  Following on from his previous Chief Data Officer roles and recent Global Data & AI role with Avanade (Accenture), he is now in Singapore as Group Head of Data for IHH Healthcare, a multinational private healthcare chain.

1. Please tell us a little bit about yourself and your career journey so far:

I’ve had an intertwined career of over two decades ranging from hands-on medic to C-Suite executive running technology, data and AI for both healthcare and life science companies.  I’ve always been a techie and was a bedroom programmer before I got into medicine.  As things have played out, I have been fortunate enough to combine both my interests and passions.

When people ask me why, it is because I have experienced medical training and am acutely aware that it is still an apprenticeship with clinicians learning through their mistakes. 

However, these mistakes are patient harm, either physical or financial.  As a medical student, I defined the functional requirements for students to learn from electronic medical record (EMR) data rather than rely on just the haphazard exposure to learning opportunities.  But then in my decade-plus of medical practice, I despaired at seeing the same basic mistakes made time and again and this encouraged me to take a second hat into management to change things at the system level. 

Through this, I’ve pushed the systematic use of data rather than anecdotes to protect patients and improve the safety and efficiency of medical care.  Having a foot in both camps (user and admin/developer) helped me iterate the systems so that data was emergent from the normal workflows, rather than a badly timed chore for someone to feed the computer, which is the typical source of clinician frustration.

In more recent years, from my time as Chief Data Officer for a genomics company, I have come to appreciate that medicine itself is changing from principally an observational science of signs and symptoms, to one that is precision medicine, data-driven and personalised.

2. What is one thing you wish you had known when you began your career?

I wish that I had known about antipatterns and the law of unintended consequences and that security should never be an afterthought.  Sometimes the quick fixes do more harm than good, especially in the longer term.

I also would encourage undergraduate courses to teach a global view.  Reflecting on my experiences in the UK and elsewhere, we see different healthcare systems pouring endless money into reinventing the wheel only to repeat the same mistakes of their peers.

3. What advice would you give to someone who wants to advance in the profession?

You can’t become a doctor by only work-shadowing and skipping going to medical school.  Formal education is essential.  Experience will only get you so far. 

IT is often self-taught and therefore best efforts rather than best practice.  Institutions such as BCS, FEDIP and ISC2 are key to ensuring the professionalism that our patients and clients need and deserve.

Ensure that you take the time to force yourself to learn the basic foundations, especially the boring bits.  These solid foundations are key to ensure that what you (or the team you run) build is safe, and secure and won’t need sticking plasters to maintain in future years.

4. What are the best resources that have helped you along the way?

Taking a global view, medical informatics is a well-established, defined, and regulated field, with entry through examination.  Looking at those core curriculums helped shine a light on what I didn’t know that I needed to know.  Likewise in the UK, I was part of the team at the Faculty of Clinical Informatics that helped publish the Core Competency Framework (CCF) so I would encourage others to look for professional peer-reviewed resources.

Ultimately the best resources are people.  Ask, ask, and ask again.  Most problems already have solutions with understood pitfalls.  Use a wide network for your questions to ensure that you’re not in an echo chamber and avoid the classic groupthink issues.

5. What is the one common myth about your profession or field that you want to debunk?

People often confuse informatics with IT, and whilst I understand the historical basis for that, I’d like to debunk the myth that ‘data is data’ irrespective of the domain.  Unlike a credit number that can be cancelled and reissued, healthcare data and health informatics systems need to be treated with a degree of care, respect and caution that exceeds even that of the finance sectors.

6. What do you think is going to have the biggest impact on health in the next 5 years?

In any speciality, there is already more medical research being published daily than can be read, evaluated, and integrated into medical practice.   The ability of AI to process the volume of data that is multi-omics is going to transform the world.

7. How do you continue to learn to stay on top of things within your role?

I ensure that I stay broadly read but that I also experiment and am hands-on with my learning.  I’m still a programmer so have my own AI models running on my computer at home.  It’s too easy to get lost in the current waves of hype in our field and I find that testing and deploying things myself, gives me a better appreciation of where reality stops and the hyperbole begins.

At the moment, I am focused on understanding the opportunities, technical limitations and inherent risks in the latest AI solutions.  I’m concerned that no one will watch the watcher and that unsupervised, we may see poor advice from AI systems resulting in real-world harm, due to the trust that people have in computers.  There is some interesting research coming out that exposes the inner dialogue of these models versus the final output that demonstrates that it can prioritise outcome over integrity.

8. What do you think is the biggest challenge for the profession and how should we overcome it?

Last year in the UK, we discovered after a series of sudden failures, that hundreds of school buildings had been built with reinforced autoclaved aerated concrete.  Really, they should have been torn down and rebuilt many years ago.  For us, it’s a similar challenge in that we inherited a situation where a significant proportion of our leaders didn’t know what good looks like and wanted a quick fix yesterday.  Unfortunately, there are many who are happy to sell those sticky plasters.

We are also in a world where startups aim to get things working as quickly and cheaply as possible, all of which have resulted in numerous breaches of patient data.  In today’s (internet) connected world, our systems need to be secure by design.  As informatics professionals, we need to be security-conscious and empowered to speak up when things are not up to code.  Through familiarity with DevSecOps, we should ensure that we aren’t contributing to these breaches of trust.

9. Fellow Leading Practitioner Owen Powell would like to ask: What job did you think you’d be doing when you were at school?

Which school - primary, secondary or medical?  At one point, I dreamt of creating computer games for a living.  I chose A-Levels orientated towards engineering, physics and maths but as I was drawn by the challenge of medicine, the techie in me still thought I was going to be a surgeon and help our development into cyborgs.  Although probably with my mobile phone in hand, with access to the world's data and the ability to instantly speak and see (video call) anyone across the globe, we’re not that far off now.