Digitalizing healthcare: better or worse? What can go wrong

In brief

On 1 September 2026, a national health portal opens. Enthusiasm is high, and questions are few. This article takes up the other side of the discussion.

It is not an argument against digitalization. Estonia and Denmark have worked this way for years, and paper has its own failures, which are less visible only because we have got used to them.

It is an argument for a better question: not "digital or paper", but "this system, at this moment, for the person who has no way of getting into it".

Judge for yourself. Below are the arguments against, the arguments for, and five concrete signs you can watch for this autumn.

The Electronic Health Record (Dosarul Electronic de Sănătate, or DES) was planned from 2012 and implemented in 2014, with about 81.7 million euros of European funds. From November 2016 it became mandatory for all healthcare providers.

What followed, in short: the system was offline for more than a year, from January 2018. More than 13 million records were created, but by June 2018 only 10,676 patients had ever accessed their record. Laboratories, most hospitals, emergency departments and ambulance services did not send data. The Constitutional Court found some provisions unconstitutional, in a 2018 decision.

The practical result is the one family doctors (GPs) describe: a record that exists, but has nothing in it, and that no one opens because it is not worth the time.

Ten thousand people out of nineteen million. That is not a communication problem. It is what happens when a system is launched to meet a date rather than to be useful.

This does not guarantee that the new portal will fail in the same way. But anyone who tells you that "it is different now" owes you an explanation of what exactly has changed.

Here we are not talking about the past, but about last month.

The Court of Auditors (Curtea de Conturi) audited the health insurance IT platform and found, in a report from October 2025, that the investment had exceeded 1.1 billion lei — about 920 million for design and construction, and 191 million for maintenance, licenses and support. The finding that matters to you: in recent years, the system has been out of service at least once a month. The hardware and software infrastructure is described as outdated and lacking the necessary updates.

The Cluj Association of Family Doctors reported, at the end of October 2025, an outage of more than 24 hours, with an error rate they estimated at over 95%. During those hours it was not possible to check whether you were insured, services were not validated and electronic prescriptions were not issued. The doctors also pointed out that their obligation to validate services within 72 hours stays in force even when the platform is down.

What do doctors do when the system goes down? They write prescriptions on paper, offline, and enter them later.

Now put the two things side by side. The system goes down every month. The backup solution is paper. And the announced reform is that paper will no longer be printed.

No one has explained publicly what happens to a patient who is at the pharmacy on the day the platform is down and has nothing printed in hand. For someone waiting for an anti-nausea medicine after a treatment cycle, that day is not a statistic.

According to Eurostat data for 2025, Romania is in last place in the European Union for basic digital skills: 31.8% of citizens. Even among young people aged 16 to 24, the figure is 53.3%, second to last in the EU.

Almost two in three adults do not have the skills considered basic. People with cancer are, on average, older than the general population, so the real proportion in oncology waiting rooms is probably worse than the average.

The official line is that anyone who uses WhatsApp will be able to use the platform. Maybe. But WhatsApp does not ask you to photograph your ID card without glare, film yourself turning your head in three directions, and read a text aloud for voice verification.

There is also a route without a phone — an account set up at the doctor's, with your ID card. It is the good news of the whole story. The problem is that it is the least publicized, and the people who need it most are exactly the ones who do not read about it online.

The portal shows, essentially, what has been reported and reimbursed through the public system.

For a cancer patient in Romania, that leaves out large pieces: consultations paid out of pocket, investigations done privately without reimbursement, a second opinion sought in another country, what was discussed in a committee without ending up in a reportable document.

The risk is not that data is missing. The risk is that a good-looking but incomplete record starts to be treated as the complete record. A doctor in a hurry who looks in the portal and sees nothing may draw the wrong conclusion that nothing has happened.

An empty record is obviously empty. A half-full record is dangerous, because it looks like a complete one.

The law says nothing about a family member's access to someone else's record. CNAS (the National Health Insurance House) has announced that the patient will decide who sees their information and at what level, but it has not explained publicly where this control is or how it works.

Families will not wait. They will do what they have always done: share the password.

The consequence is paradoxical. A system without delegated access produces, in practice, less confidentiality than paper. On paper, you show your daughter one result and not the others. With a shared password, you give her everything, forever, with no trace in the system that someone else has logged in.

An access mechanism with levels and with revocation is not a luxury. It is what keeps the family inside the rules, instead of pushing it outside them.

The president of CNAS explained that an appointment can be made by a family member, a neighbor or even a stranger, from their own account, without access to your medical data.

For a parent without internet, that is exactly the right solution. For the system, it is a new point of exposure. Someone can take a slot in your name, an appointment you did not ask for can be made, slots can be blocked without anyone showing up.

That is not a reason not to do it. It is a reason for the account to have a place where you can clearly see all the appointments made in your name and who made them. If that place does not exist, the objection stands.

The platform was supposed to be up and running in the second quarter of 2025. The deadline moved toward the end of 2026 — about a year and a half of delay, flagged publicly by doctors who asked for information about the project and received vague answers.

A project funded from European funds, with targets to hit on fixed dates, is under pressure that has nothing to do with how ready it is. The launch on 1 September is a political and administrative date. It is not, in itself, proof that the system is ready.

The practical corollary: in September and October you are, without having signed anything, in a testing period.

It would be dishonest to stop here.

• The trips for paperwork are real. A cancer patient with five doctors spends whole days carrying documents between them. If the medical letter reaches the family doctor on its own, you have gained a day of your life.
• A single place for prescriptions and history matters. The more doctors you have, the more it matters.
• Transparency about reimbursements is a tool. You can see what has been paid in your name. That is a form of control paper never gave you.
• Appointments visible in real time beat calling twelve numbers. For someone tired after a treatment cycle, the difference is enormous.
• Other countries have succeeded. Digitalizing healthcare is not a wrong idea. It is an idea that is hard to implement.

The most common mistake in this discussion is comparing imperfect digital systems with an imaginary paper system that never makes mistakes.

Paper gets lost. Handwriting gets misread. A lost result means a repeated investigation, with extra radiation and lost weeks. The file stays behind in the other city on exactly the day of an emergency admission. No one counts these failures, because they do not appear in any audit report.

The right question is not whether digitalization is good. It is whether this transition keeps a safety net in place while the system is still unreliable, and whether it leaves a way in for those who cannot get in on their own.

You do not have to take anyone's word for it. Look at these:

1. Is there a place in the account where you can give a family member access, with levels and the option to withdraw it? If yes, the system was designed for real people. If not, families will use shared passwords.
2. How much history can you see on day one? An empty account in September is the DES warning sign, all over again.
3. What happens at the first system outage? Is there a written procedure communicated to patients, or is it improvised at the front desk?
4. Will the family doctor still print things if you ask? If the answer is "that is no longer possible", the safety net was cut before it was needed.
5. How many providers are actually in the appointment system in October? The obligation starts in the fourth quarter. An appointment system with half the practices in it is a list, not a system.

If four out of five look good, things are better than before. If three out of five look bad, you have the right to be skeptical without being accused of opposing progress.

The good part is that you do not depend on the answer.

• Create your account. One more account costs you nothing, and if the platform works, you gain.
• Do not throw away the paper file. Not until you can see in the account everything you need for your next consultation.
• Ask for printed copies of what is important. In the first months, one extra sheet of paper costs less than a wasted trip.
• Keep your own list of medicines and doses. It is the only record that never goes down and does not depend on anyone.
• Photograph important documents with your phone. A result saved in your phone's gallery has rescued more consultations than any platform.
• When something is missing from the portal, ask for it from whoever issued it. Its absence from the system does not cancel your right to the document.

A good digital system does two things at once: it makes life easier for the person who can get into it, and it does not make life harder for the person who cannot.

The first part is easy and will show quickly, in usage figures that someone will announce with satisfaction. The second part is hard to measure and does not appear in any press release, because the people left out do not fill in feedback forms.

So no, we do not know yet whether it is better or worse. We only know where to look to find out: not at the people who log in, but at the people who cannot. If in December the 74-year-old in a rural village gets their prescription as simply as in August, digitalization has succeeded. If they have to make an extra trip, it has failed for them — no matter how good the platform looks for everyone else.