Szymon Kostrzewski is a serial entrepreneur who develops tech to improve healthcare processes. Having spent countless hours in operating rooms around the world, he has seen what works – and what doesn’t. His latest venture, Evidone, helps surgeons leverage technology to better focus on their patients.

What are the challenges you see in operating rooms today?

You can think of surgery as a complex choreography. Even for a simple procedure, you need hundreds of materials to be prepared in advance. For straightforward hand surgery, for instance, 400 different items are needed in the operating room. It involves countless stakeholders, such as nurses and assistants, working in perfect synchrony.

This labour-intensive process makes the running of operating rooms very expensive, amounting to half of a hospital’s total costs. Everything – or almost – is managed manually, so medical staff must hold an incredible amount of information in their heads: how to prepare the table for a particular treatment and patient, which person to call when, and so on. This is the case in operating rooms worldwide.

We live in the digital age. Has technology not made its way to operating theatres?

The way surgical facilities operate today isn’t dissimilar to how the Mayo brothers ran their clinic a century ago. What has changed is that we’ve seen exponential growth in materials, patient throughput and treatment options – and we’re now at the point where the whole system is pushed to capacity.

And that’s why we see so much inefficiency and waste. For example, almost 10% of a surgeon’s time is spent waiting on something that should be there – but isn’t. And 70% of instruments get cleaned and sterilised without ever being used.

AI can help. Its main goal in healthcare – and especially in the operating room – should be to allow staff to focus fully on the patient, rather than thinking about whether they have a specific screwdriver, whether they’ve scanned all the disposables or how they will optimise their schedule. No, we want surgeons to be fully focused on the patient and the treatment, with technology running seamlessly in the background.

How can AI help to achieve this?

First of all, AI can bring automation. Instead of tedious note-taking, scanning and trying to liaise with multiple staff members, AI can automate these tasks. There is also much data generated in the operating room that is not being captured today. AI can help us capture it and draw relevant insights from it. Analysing big data used to be a laborious process involving Excel charts and data cleaning, but nowadays you can feed data into relevant machine models and generate insights in much less time. Once you have these insights, you start to see patterns and can use this information to improve the whole process.

You started Evidone to help solve these challenges. What will its technology bring to the operating room?

As an example, something that wastes time during surgery is finding and getting the right tools to surgeons. If the medical team forget that crucial screwdriver – and this is easily done, when so many tools are needed – it can add one hour to the surgery. Fetching a screwdriver might sound like a basic task, but it can take time if it’s in a central sterilisation department on the other side of the hospital, where it needs to be found among 20 other screwdrivers that look almost identical.

So I wanted to do something about this. Evidone is about having complete awareness of materials, people and surgery stages in the operating room so that the plan comes together smoothly. We do this by installing special cameras with computer vision directly in the operating room. The cameras observe what’s happening in real time, which instruments are in the room and who is using them. Our AI-based technology then uses this information to deduce how the surgery is going, how long it will take and anticipate what needs to be prepared next. The surgery can take hundreds of different routes, but our technology helps the medical team stay one step ahead.

What’s your biggest challenge?

To allay people’s fears around AI and new tech. All of Evidone’s analysis happens in real time in the operating room. We are not sending anything to the cloud. There is no recording, no face recognition, no retransmission outside the room. So basically, the output is a stream of events – someone took this, someone did that – and that’s it. This really helps people understand that they are not being tracked or supervised. We are simply providing a tool to let them focus on the patient.

We also have to be careful not to get carried away with the vast possibilities of AI. It would be all too easy to focus on the next exciting feature instead of what really matters – patients. With every new feature we develop, we ask ourselves the same question: will this lead to healthcare professionals spending more time with patients or with technology? If it’s the latter, we change our approach.

What are the benefits when surgeons can fully focus on their patients?

When surgeons are fully focused on their patients, it reduces surgery time. This lowers the risk of infections and post-operative pain for some types of operations. That’s one thing. The second thing is patient throughput. Depending on where you live, you may have to wait a long time for surgery. Our value proposition is to increase efficiency to the point where a surgeon can operate on one more patient per day.

Then there’s the administrative side of surgery, which is increasing for staff. This means more admin and less time with patients. For example, for any instrument that is sterilised, you need to store the temperature curve for 15 years. If we want staff to concentrate on patients, we need to release them from this burden. Embracing automation could reduce admin time, which in turns helps to lower healthcare costs.

What other possibilities might AI bring to the operating room?

I come from an orthopaedics background and can see huge potential for AI to help plan and perform surgeries in this field. Take spine procedures – traditionally, planning implant placement had to be done during surgery to work out how it would fit with the surrounding tissue. This took time. The surgeon had to step away, look at images, do the planning, then get back to their patient. With AI, much of that can now be done with images taken before surgery. In a few clicks, the surgeon has a good idea of how to do this and is well prepared for the surgery. You can imagine how much time this would save if you had 20 implants to plan, one after the other.

What will operating rooms look like in 50 years’ time?

In terms of patients, I hope they will go to the operating room feeling confident about the treatment that they are about to receive. They will be well looked after by competent staff supported by technology. They will be very well informed about the recovery and the expected outcomes.

Healthcare professionals, on the other hand, will be well prepared for surgeries. There will be efficient processes in place, including for emergencies. The medical team will have all the materials they need when they need them. They will not be disturbed by staff devices beeping away and so will be able to fully concentrate on patients, giving them the best possible care.

That’s the dream. I hope I’m still here to see it!