Journalists live in fear. One fear is to make silly mistakes in situations that can wreak havoc. Before the internet, in the mid-1990s, a common and risky task in newsrooms was collecting and publishing lottery results. Just one wrong number would be enough to turn a winner into a loser (imagine him tearing up the ticket) or a loser into a winner (imagine him discovering he hasn't won anything). Caixa's website freed us from this anguish.

Mistakes do happen, but they get worse over time if they're not fixed. In the internet age, post-truth, and the like, some continue to transmit the wrong information even after corrections are made. No wonder an old institute at Folha, Corrections, is currently a standard in much of the planet of the civilized press.

If the time scale has changed and requires immediate reaction to the error, at the risk of seeing bad information exploited in networks, how corrections are made has also become vital.

These are times of fake news; we are correcting misinformation all the time. It is no longer enough just to recognize the lapse; it is necessary to try at all costs to prevent it from spreading too much. Because no matter what you do, the misunderstanding will spread.

Folha erred in this effort last week. An article published on Friday (2) about the use of expired vaccines in 1,532 municipalities became the most read text in the newspaper's history. It remained among the most accessed seven days after its publication.

Problem: The story was wrong. Under the heading "Thousands in Brazil have taken expired vaccines against Covid; see if you are one of them," it said that about 26,000 expired doses of AstraZeneca had been applied in the country. Reading the text implied that this had indeed happened, but in fact it only listed the Ministry of Health records. But is this a problem? In contact with the newspaper and on social networks, city halls and analysts stated that the official data were or could be wrong.

Will it be? "It is quite plausible, as administrative health data have many problems with entering the system, unlike what happens with financial data, which are registered more carefully because they are audited. City halls depend on them to receive funds," explained Adriano Massuda, from FG -Health, a specialist in SUS. "Municipalities make a mistake in registering deaths, let alone vaccines. There is a lack of federal coordination, management capacity. The system, a global example until 2015, has been precarious."

The newspaper was not insensitive to criticism. Also on Friday, he changed the title of the article to a more precise formulation: "Records indicate that thousands in Brazil took an expired vaccine against Covid; see if you are one of them." It also edited, in a separate text, the objections of the various health departments mentioned.

The controversy intensified over the weekend, as did the article's audience. In the second (5), this ombudsman noted in the internal criticism sent to the newsroom that the newspaper showed excessive confidence in the data and was betting high on maintaining the service motto in the article's title. It is enough to change the subject to understand the risk involved in this type of statement: "look at the passenger list of the flight that crashed". The engine of high interest in the fact is evident.

On Tuesday (6) night, finally, an Erramos was published. "A previous version of this report failed to warn that data on 26,000 doses applied after the expiration date could result from errors in the Ministry of Health's system." A third text, in which the newspaper recognizes the error, was aired at the same time.

The following morning, the website's audience revealed the inadequacy of the maneuver. The reading rate of the erroneous report was soon surpassed by that of the original article, which so far maintains the improper bait in its title and the peremptory tone that 26 thousand people received overdue doses in their arms. Only no.

"As soon as the newspaper was sure that there was an inaccuracy in the initial report, a new text was published with wide visibility reporting the error. The problem of late-term vaccination, of public utility, continued to be investigated and was the subject of published reports in the sequence," said Vinicius Mota, from the Editorial Department.

It is good for the newspaper to continue investigating the topic. It would have done better if, in place of a rushed article, it had doubted the government's figures a little and had deepened the investigation beforehand, as it doubted at the beginning of the pandemic and, in partnership with other press vehicles, took on the collection of data on cases and deaths.

The insides of the Ministry of Health are being exposed live at the CPI. Blindly trusting what comes out of there doesn't seem prudent. Letting perpetuate information that is not 100% correct even less.

_{Translated by Kiratiana Freelon}