So both the federal and state governments in Brazil have been...

So both the federal and state governments in Brazil have been massively underreporting the number of confirmed cases, and deaths in Brazil. This is a report from Para from late last week. That has now been officially recognised, and the report attempts to explain the cause.

Why is it important?

Because the state government of Para eased restrictions on 24th May based on massively under reported and incorrect figures!

It was absolute madness in that environment! In reality, they should have maintained the lockdown.

Minas Gerais also, under reporting by ten to one!

“We need to test more to get an idea of whatwe're up against. Without the true numbers of the epidemic, decision-making bythe authorities is based on a scenario that does not reflect the realsituation, which can lead to the failure of the measures adopted.Underreporting is a serious problem and is not exclusive to Pará. Today, inBrazil, basically, the people tested are those who arrive at hospitals withmore severe symptoms, related to Covid-19, and health professionals. For thisreason, the numbers we see reflect only a portion of the reality “, heconcludes.

PARÁ STOPPED RECORDING ALMOST A THOUSAND DEATHS 
FROM COVID-19

Pará failed to register almost athousand deaths and 14 thousand confirmed cases of Covid-19, adding up to atotal of 15 thousand underreportings that were not in the official count of thegovernment, in the last eight days. The retroactive bulletins ended up changingthe date of Covid-19's first case in Pará to March 7 - the date released wasMarch 18. According to the Secretary of Health (Sespa), Alberto Beltrame, thiswas due to a “mismatch” in the secretariat system updated by city halls, whichleft out a total of 907 deaths and 13,871 confirmed cases of the disease inthis period - that is, 43,7% of total cases and 34.8% of total deaths. The stateended Wednesday (27), with 31,671 cases and 2,605 deaths from the newcoronavirus.

In total, there are 14,778 latenotifications, including those infected and deaths, which were identified by thesecretariat. A task force was set up for the dissemination in a new format,which started on Wednesday (20). According to Beltrame, the notifications tookplace weeks ago, mainly between April and early May, and result from 122thousand rapid tests distributed among the municipalities. Recently, Rio deJaneiro also changed protocols for disclosing Covid-19 data and the new methodof counting the dead was criticized. In Rio Grande do Sul, a survey shows thatfor each official case, there are two others that are not registered. In MinasGerais, the government also admitted underreporting and estimates that there is1 confirmation of coronavirus for every 10 cases.

Altered curve

Sespa'sbulletins have been collected by G1 since the beginning of the pandemic tosurvey the Coronavirus Map, informing the progress of Covid-19 inmunicipalities, based on state departments. Even with the change in the formatof the Pará bulletin, new cases and deaths in the state continued to be countedaccording to the day of publication.

According to thebulletins, the first case of Covid-19 in Pará occurred on March 7. The patient,according to Sespa, was a 50-year-old woman from Parauapebas. The first deathoccurred on March 15 - a 60-year-old woman from Belém, according to Sespa.

Retroactivebulletins also point out that:

The informationis also disclosed on the Covid-19 panel, on the Sespa monitoring portal, andwas questioned last Tuesday (26), as it could be typos. Until this morning,they have not been corrected or denied. The bulletins have already shown casesfrom January and February, but most have already been corrected, after contactsmade by G1.

New format

The new format for the disclosure of bulletins from Covid-19, which separates cases and deaths in the last 24 hours from those occurred on previous dates, was adopted, according to secretary Beltrame, as the municipalities started to report results of rapid tests only in the Information System Mortality Rate (SIM), leaving Sespa's system out of date.

At the time, Beltrame stated that he had observed a "rate of falling from the disease, but that he was computing many deaths per day".

Four days later, the state government released a study pointing to a “downward trend”, based on Sespa's own data. The survey based the government on relaxing mandatory social isolation measures with the end of the lockdown, which came to be decided by city halls. However, the document itself mentions that "underreporting on past dates can invalidate forecast scenarios".

Understand the evolution of cases since the release of delayed notifications:

* On May 19, there were 17,177 cases and 1,554 deaths;

* On the 20th, there were 18,929 cases and 1,778 deaths, with 1,676 cases reported and 218 delayed deaths;

* On the 21st, there were 20,532 cases and 1,893 deaths, with 1,494 cases and 103 delayed deaths being reported;

* On the 22nd, there were 22,697 cases and 2,003 deaths, with 2,069 cases and 90 delayed deaths being reported;

* On the 23rd, there were 24,125 cases and 2,150 deaths, with 1,388 cases reported and 113 late deaths;

* On the 24th, there were 24,815 cases and 2,290 deaths, with 668 cases and 132 late deaths being reported;

* On the 25th, there were 27,366 cases and 2,431 deaths, with 2,442 cases and 113 late deaths reported;

* On the 26th, there were 29,882 cases and 2,522 deaths, with 2,412 cases and 65 arrears reported;

* And on the 27th, there were 31,671 cases and 2,605 deaths, with 1,722 cases and 73 delayed deaths being reported.
(Please note, the above are cumulative figures- my notes)

Recent notifications continue to delay

Underreporting causes a kind of rearrangement in the contamination curve. When he revealed the inconsistencies in the data, Beltrame went so far as to say that the peak of contamination in Pará occurred around April 20 and on May 5 there was the highest death rate in 24 hours: 84 deaths. But as notifications continue to lag, that record could be broken.

Among the delayed notifications, 893 cases and 144 deaths occurred between May 20 and 25 were identified. These most recent figures are disclosed as "on earlier dates".

On May 20, Sespa reported that only 76 new cases were confirmed and six deaths registered. Analyzing the bulletins that inserted retroactive cases in the system, it is clear that, in fact, 40 people died from Covid-19 that day, 12 of them only in the capital. The number of cases also jumps from 76 to 418.

On May 21, Sespa officially released 109 cases and 12 deaths. Adding to the delayed notifications, the number reaches 353 cases and 58 deaths. The following day, there were 96 cases and 20 deaths, but with the sum the total reaches new 239 cases and 43 deaths. On May 23, when the bulletin contained 40 cases and 34 deaths, it was 134 and 70, respectively.

Official figures remain far from city halls

During a visit to Pará, former minister Luiz Henrique Mandetta stated that the state already peaked between April 20 and the first week of May. Days earlier, he had said that Pará would be the epicenter of Covid-19. Governor Helder Barbalho then began to request proposals from businessmen to resume commercial activities.

At the time, Barbalho said he was doing an epidemiological survey to find out what the real percentage of the population with the new coronavirus is in the state. This survey was never released.

A national survey coordinated by the Federal University of Pelotas (UFPel) pointed out that Brazil has seven times more contaminated by the new coronavirus than the official data.

In Pará, the most alarming rates were that of the city of Breves, where practically a quarter of the population is or has been with Covid-19, and that of Belém, where more than 15% would have already contracted the disease.

Official figures are also still far from those recorded by city halls. On May 26, when it completes a week of the disclosure of the delayed notifications by Sespa, the sum of all bulletins from city halls reaches 6,898 cases and 217 more deaths than the official data from Sespa.

Underreporting can be 22 times greater

For the researcher Evelin Cardoso, who follows the figures of the state pandemic, there is a delay in the registration of notifications, “which makes monitoring the data confusing”.

“It is not really an underreporting, but a delayed notification. When looking at the numbers released on the day, we have an idea of the new records for the last 24 hours, but when we follow the data for the same day, over time, we notice that they go up considerably. We realized that this is not an exclusive problem in the interior of the state. There is also a delay in the numbers of the metropolitan region “.

Evelin, who is a professor at the Federal Rural University of the Amazon (UFRA) and a researcher at the Laboratory of Social Technologies (LTS), as a doctoral student at the Federal University of Pará (UFPA), participated in the development of a model that projects more realistic numbers of the disease, based on the socioeconomic and demographic reality of the metropolitan region of Belém.

“This model was made available to the Federal Public Ministry (MPF). Estimates based on data up to May 1 have proved to be quite accurate in relation to the escalation of Covid-19 in the metropolitan region. Despite the delays, we have observed the slowdown in the spread of the disease in recent weeks, but it is too early to say that this drop will be sustained, especially given the relaxation of measures in several sectors, ”he explained.

The researcher uses a recent study by the Federal University of Pelotas that points to an estimate that 15% of the population already has antibodies to the disease in Belém, which indicates that these people have already had contact with the virus. “If today we have 1.5 million inhabitants in Belém, 15.1% would be 225 thousand people who have already had contact with the virus. Which is much higher than what is notified. To date [May 27] there are around 10,000 confirmed cases in the capital. In other words, UFPel's research points out that the real number of people already infected is at least 22 times higher ”, he relates.

According to the researcher, underreporting is a national reality, which occurs due to the country's low capacity to test and process the test result in a timely manner. She also points out that underreporting is greater in cities with a high population density index.

“We need to test more to get an idea of whatwe're up against. Without the true numbers of the epidemic, decision-making bythe authorities is based on a scenario that does not reflect the realsituation, which can lead to the failure of the measures adopted.Underreporting is a serious problem and is not exclusive to Pará. Today, inBrazil, basically, the people tested are those who arrive at hospitals withmore severe symptoms, related to Covid-19, and health professionals. For thisreason, the numbers we see reflect only a portion of the reality “, heconcludes.