Healthcare needs doctors. And engineers

Many doctors of the new generation leave medical school with solid training in clinical care and an eye on the technologies that are transforming healthcare. Yet they enter the market poorly prepared for something that rarely appears in the curriculum: navigating a system that was not designed for them.

This mismatch, which begins in the first years of practice, bothers me. It causes fatigue, discouragement and burnout in many of these professionals, not because they are fragile, but because of the exhaustion of a poorly designed system piled on top of the natural tiredness of an intense routine. The tiredness of the routine is recoverable. The tiredness of the system corrodes.

I do not write as someone who sees the system from the outside. I started out reading chest X-rays and CT scans, imagining that my career would fit inside a dark room, in front of images. I learned, as a doctor, that understanding the patient in front of me was only part of the job.

The other part, just as significant as the first, was understanding the system that patient was part of, and how often that model worked against everyone: doctors, patients, health plans, companies. Today, in medical management, I see the whole bill: even the best clinical decision can be made unworkable by a hospital operation without the infrastructure to sustain it

The human cost of inefficiency

There is a truth medicine finds hard to say out loud: much of healthcare professionals’ time today is not spent caring for people. It is spent sustaining a bureaucracy that should be invisible. Procedure coding. Justifications for health plans. Manual denial disputes. Audit forms. These are clinical hours consumed by an apparatus that has nothing to do with what these professionals studied to do. It is a doctor filling out forms when they should be looking the patient in the eye. It is a nurse answering audits when they should be at the bedside.

The figure that captures this is almost absurd. In 2025, private hospitals saw initial denials reach more than 15% of billing. From there, an operational army spends months disputing, reconciling and proving what was already right, with payments arriving much later, in the best-case scenario. Time that should be serving care becomes the working capital of an operational dispute.

The problem is not the professionals, but the way the healthcare operation was designed. When I started looking at this from the inside, what struck me most was not the waste itself. It was the disproportion between the complexity of a hospital and the engineering capacity dedicated to sustaining it.

What other sectors have already understood

A large hospital runs nonstop, coordinates dozens of critical departments in real time and manages contracts with multiple health plans, each with its own tables, deadlines and rules. In the volume and sensitivity of its data, it rivals a financial institution. And in much of the sector it is still run on spreadsheets, legacy systems that do not talk to each other and processes that depend on the memory of whoever has been there the longest.

What is most striking, however, is not the size of the denials. It is the size of the technology team relative to the complexity it has to sustain. The technical area is usually a tiny fraction of the staff. Where a bank or a fintech would put technology at the center of the operation, the hospital still treats it as support. The disproportion between the problem to be solved and the capacity dedicated to solving it is what jumps out.

In healthcare, this gap has a consequence that few sectors face so directly: it reaches care. The chain is simple. When a hospital fails to receive part of the revenue it was entitled to, that money does not become a bed, does not become medication, does not become a nurse’s salary, does not become a working ICU. Operational efficiency, here, is not an administrative matter. It is a matter of care.

Saying that healthcare needs engineers does not diminish those who provide care. Brazilian medicine delivers extraordinary results under conditions that would be unacceptable in many other sectors. But entrusting the operation to individual effort is not a model. It is a bet that is renewed every day.

Engineers in the broad sense: people who think about architecture, about scale, about how things work underneath. People who look at processes with recurring waste, deadlines that drag on and unreadable contracts, and see what others do not.

Banking, retail, logistics: they all reached a point where they realized they could not grow while treating technology as support rather than as the core. Healthcare is now at that threshold. The transformation does not come from more software on top of the wrong processes. It comes from replacing the manual, reactive layer with infrastructure that operates in real time, reduces avoidable waste and gives back to healthcare professionals what belongs to them: time dedicated to care.

When I talk with students and early-career doctors, I always come back to the same point: the healthcare you are going to practice will depend more and more on well-built systems. That calls for a different relationship with technology. You do not need to become an engineer. You need to understand enough to demand, collaborate and recognize when a system was built to help and when it was built to get in the way.

The Rivio manifesto has a sentence that stayed with me: “efficient hospitals save more lives.” It says simply what we sometimes forget in the middle of the operation. Every real recovered from an improper denial goes back into care. Every hour given back to the doctor by automating a process goes back to the patient. Every decision made with real data is a better decision.

My obligation, and that of everyone who works in healthcare in Brazil, is to make sure the system the next doctors inherit is worthy of what they learned. Healthcare does not have to choose between being human and being efficient. It needs to be both. And for that, it needs everyone: doctors, nurses, managers. And engineers.