How Two “Citizen Detectives” Revealed the Real Impact of the Pandemic in Mexico

1,832. That was the total death toll by coronavirus in Mexico DF during the first wave, according to official data.

The first death had been announced in mid-March 2020 and this figure was published in late May.

But, among some, raised doubts.

“We had a very high mortality rate, around 8% compared to 1% in the rest of the world,” economic consultant Laurianne Despeghel explained to the BBC. data.

That high death rate was not because the coronavirus was more dangerous in Mexico, but because very little testing was done.

That May, Mexico was near the bottom of the range for the number of tests among Latin American countries, with only 99 tests per 100,000 inhabitants.

“So there were probably a lot of people who were dying from coronavirus and they weren’t being tagged as such”.

Excess deaths

Since there wasn’t much testing done, it was very difficult to figure out how many people were actually being affected by COVID-19, but an event from the past gave Despeghel an idea to find out what he needed to know.

“When I was young in France, there was a heat wave and many people died and the only way to know the extent of the phenomenon was comparing that year with others”.

It’s simple, if you have the data: what you do is take the death toll for the current period and compare it with the typical numbers in previous years. The difference is known as “excess deaths.”

Because many countries have found it difficult to carry out sufficient testing – in fact, almost all countries in the first wave – the excess deaths allows experts to see the impact of the pandemic.

And it is a figure that includes not only direct deaths, but also indirect ones.

In search of the numbers

The problem was that Mexico only publishes official mortality figures in October following a full calendar year. Therefore, the excess deaths in March and April 2020 only would be reported until October 2021.

Despeghel wasn’t going to give up, and mutual friends introduced her to software engineer Mario Romero, another geek who was intrigued by the subject.

Together they decided to determine for themselves the excess mortality figures in the country’s capital, Mexico City.

“Laurianne found a website where anyone could request an official copy of a birth or death certificate,” says Romero.

“We started playing with it, basically trying to figure out how it worked” … and all of a sudden, they realized something that would be very useful to them.

The way they numbered the certificates was very simple: “the first death certificate registered at the beginning of the year was certificate number one for that registry office, and then two, three or four and so on.”

And, although there were more than 50 registry offices throughout the city, they all used the same system.

That meant that if they could determine which certificate was the highest number, they would have the death toll in that area so far this year.

But to know what it was, they had to ask the website for one certificate after another until it replied that the last one they had requested did not exist, which would imply that the previous one was the most recent.

A long task that was left to a program that Romero created, “a binary search algorithm, which made increasingly refined guesses.”

They let the program run, going through each registry office.

“Then we added everything together and it gave us what we thought was a death count for 2020.”

To be on the safe side, they tested the program by looking at the officially published data from previous years, “2017 and 2018, at that time, to verify if our numbers matched the officially published numbers, and they did.”

8,000 people

Despeghel and Romero estimated that while the total official death toll on May 20 was 1832, 8,000 more people had died than would have been expected during this period.

If the government’s figures were accurate, that meant that in Mexico City 25% of the excess deaths during the first wave were due to the coronavirus.

To put that in perspective, the official number of deaths from covid-19 in the UK in the same period was 80% of the total excess deaths. In Italy, about 70%.

Or to put it another way, in Mexico there had been four times more excess deaths than official deaths from coronavirus during the two months in which the first wave peaked.

To be precise…

That excess of deaths had not necessarily been directly by covid-19, but Despeghel and Romero believed that it was at least an indirect result of the pandemic.

“The way to think about it is to say that this is the death toll from the pandemic”, Clarifies Despeghel.

“It includes people who died directly from covid-19, as well as people who, due to the coronavirus crisis, could not get a bed in the hospital, or were afraid to go to a hospital for another illness, fearing contracting covid.”

They wrote their findings and published them on a data blog for the Mexican magazine Nexos. Your report caused quite a stir and their numbers were picked up by other media.

A few months later, authorities began compiling their own excess mortality figures that, to Despeghel and Romero’s relief, were very similar to what the pair of data detectives had produced.

And they were pleased that their efforts had been taken over by the authorities.

“We managed to change the debate a bit from the official death toll from coronavirus to excess deaths that is not only directly related to covid-19, but also indirectly,” the economist told the BBC.

What they found ended up being part of a similar pattern in other Latin American countries, like Peru and Bolivia.

Excess death figures have shown that the effect of the pandemic on mortality is much greater than indicated by the confirmed number of deaths from coronavirus.