Between March 1 and May 12, Spain recorded 43,295 more deaths than what would be considered normal for this time of the year, based on past mortality rates. This is up 52% from the expected deaths for the period.
The figure includes 27,302 confirmed fatalities from Covid-19, but there are an additional 15,993 deaths that show up on civil registries but are not recorded as coronavirus victims, even though many of them probably are.
On Wednesday, a system update with new data sent in by civil registries turned up 12,000 more excess deaths than were previously known. Of these, 7,300 were in Catalonia (mostly in Barcelona), 2,500 in the Madrid region and 800 in Castilla-La Mancha.
The excess fatality figures may also include deaths not directly caused by Covid-19
The Covid-19 death toll provided by the Health Ministry only counts cases where the deceased were tested for the virus. But there are infected people who died without getting tested, and who may have passed away outside a hospital – inside their homes or at care facilities – and do not show up in the official statistics.
But just how many are they?
The best estimates are provided by the civil registries, which send their figures to the Mortality Monitoring System (MoMo) run by the Carlos III Health Institute.
In late April, using consolidated data, there were over 185,000 deaths, compared with expected figures of around 145,000 to May. And that is despite the fact that the entire population had been mostly confined to their homes since March 15. Without these mobility restrictions, the death tally would be even higher.
Excess deaths are calculated using a mathematical model that provides the number of expected deaths under normal conditions. There are typically spikes during heat waves or flu season. But the coronavirus crisis is eclipsing all previous events.
The data provided by MoMo reveals that the excess deaths during the current crisis are not comparable to anything that has happened in the last two years. In the summer of 2018 there were around 700 excess deaths due to a heat wave, a 2% rise from the expected number. And in January 2019 there were approximately 3,000 excess deaths, very likely from the flu, up 4% from expected figures.
The excess death figures are suffering from reporting delays. One of the problems with regional civil registry records is that deaths are notified two to three days late. And during the coronavirus crisis, these time frames have become even longer. Also, some civil registries have not yet digitalized their records, which creates even greater delays. The present analysis takes this into account, and focuses on deaths that occurred until May 12.
Data collected by MoMo shows that the cost of the coronavirus crisis is not the same across Spain’s regions. Madrid and Castilla-La Mancha have registered twice as many deaths as expected. In Madrid, the figures on excess deaths suggest that over 6,000 people probably died from Covid-19 yet are not showing up in the official statistics.
In La Rioja, which recorded one of Spain’s first outbreaks, deaths were up 48% from expected figures in recent weeks. In Navarre there was a spike in deaths a week ago, up 56% from what would be considered normal. And in the Basque Country there were 40% more deaths than expected.
At the other end of the spectrum, the least affected regions are the Canary Islands, Murcia, Galicia and Andalusia, where observed deaths were less than 15% above the expected figure. The Valencia region recorded 1,197 deaths to April 10, around 20% more than expected.
Another factor to take into account is that the mathematical models for predicting deaths were not created for lockdown situations. These models probably factor in fatalities from events like car accidents, which are currently at historically low levels due to the confinement measures. This means that the number of deaths caused by Covid-19 could be even greater.
The excess fatality figures may also include deaths not directly caused by Covid-19, such as people who died of other causes when the health system was overstretched and could not provide the same kind of care as under normal conditions. Confinement measures and fear of going to the hospital may have also delayed ill people from visiting a doctor.
The true impact of the coronavirus on mortality will only be more precisely known at a future date, when it becomes possible to analyze registered deaths by various causes; the National Statistics Institute (INE) releases this information on an annual basis.
Our observed and expected death figures were calculated using MoMo’s figures, available on its website, but they do not coincide completely. This is because we aggregated the number of weekly deaths, and also because we raised the figures for some regions to adjust for the lack of digitalized information in some civil registries. We did this on the basis of 2019 definitive death records provided by the INE.
English version by Susana Urra.