People often ask me how I could leave clinical practice to work in clinical AI. I understand why they ask, but I’ve never regarded it as leaving. Let me explain why.

The decisions that change care aren’t made at the bedside

I’m an anaesthetist and intensivist. I chose the speciality because it means I can monitor a patient, take action and most importantly see the effect immediately. That directness is so fulfilling, but giving someone relief or save someones life is why I love it.

Early in my career in anaesthesiology, I kept running into the same wall. To improve care for the patients in front of me, my main tool at hand was to work faster. Move quicker, document quicker, get to the next patient. We could work more efficient if the logistics could be more efficient or the EHRs would not be so slow or lacking integratons.

But no matter how hard I tried, that wall remained impenetrable. I learned that the decisions determining how many patients we could take care of and how present I was when treating them, weren’t being made at the bedside at all. They were made in the another room, usually somewhere further up in the system.

So if I wanted to improve what happened at the bedside, I’d need to go beyond being solely patient-facing.

Is working in clinical AI leaving medicine

No. There’s a misconception that caring for people and making a difference is strictly a one-to-one activity. The discovery of penicillin has saved millions of lives. Nobody would describe that work as leaving medicine – but more contributing to a change on a public health scale.

The core of our profession as clinicians is to make both individuals and society benefit from our knowledge. We can be excellent one to one and still be needed in building the healthcare system that makes that possible.

If we accepted the narrow version of what care means, we’d have to say doctors shouldn’t work on health information systems. Shouldn’t work on policy. Shouldn’t advise governments or develop new treatments. That would reduce our role and function in society to something much smaller than it is. It would mean that society would not benefit from doctors knowledge, or from the investment in educating physicians at large.

Doctors have always been involved in developing drugs and that’s precisely why we have such effective ones. The same logic applies to medical technology or clinical AI in changing healthcare systems.

The pushback comes from outside the profession

The pressure to keep doctors at the bedside comes more from the system around them than from their colleagues.

I’ve heard many other stakeholders suggest it’s better for doctors to stay in the care room and not be at the table where decisions are made.

Most of my colleagues beg to differ. They’re grateful for the doctors who are making moves to ease the burden of healthcare on doctors through changing policy, to progressing technology. As doctors we’re used to seeing the result of what we do, often within minutes. In policy, that immediacy doesn’t exist, and that’s the hardest part of the trade.

Why I still go into the operating theatre

Joining Tandem Health hasn’t meant letting go of the clinical side and I don’t intend to.

When things become challenging, I go to the hospital. I put on scrubs, I go into the operating theatre, I talk to my colleagues and I absorb what’s actually happening. It’s my identity, and caring for patients is what matters in the end.

It’s also practical. You step outside the operating theatre to improve what’s happening inside it, but if you stay away too long you lose what made you credible in the first place. Knowledge and compassion.

Why the urgency around clinical AI matters

Ask doctors what takes them away from their patients and documentation comes up again and again. This is shown very well in Tandem’s Time to Care study. In intensive care I would have one system open for the patient record, another for lab results and a third screen for the quality registers. I’d then manually be moving information from one to the other, as they weren’t interconnected. Every minute I spent doing that was a minute I wasn’t giving to a patient.

It also cost more than time. Writing up one patient while being called away to another requires attention, and attention is exactly what the most challenging medical decisions need.

With clinical AI, that time goes back where it belongs. The attention a doctor brings to a difficult decision is the thing that matters most and attention spend on documentation erodes it. When a clinician is repeatedly told the systems will improve in a couple of years that’s simply not good enough.

Knowledge is only of value if it’s shared

To any doctor considering working in health technology or clinical AI, my advice would be this. Keep one foot in the clinic, keep talking to colleagues and patients because that’s where you’ll get the drive to change healthcare.

Then take that knowledge and share it with the people making the decisions. You’re the one who knows what working in a hospital operating theatre or a busy emergency room in the middle of the night looks like.

A colleague told me years ago that you don’t take power, you take responsibility and influence follows. I’ve worked that way ever since.

Clinical AI is being built right now and doctors should be part of building it.