France to investigate how many Covid-19 patients are dying at home

One of the unknowns about Covid-19 mortality in France is the number of people dying at home. With studies showing up to a third of deaths are not being reported, France’s  government said Wednesday it would provide its own estimate in June. 

A stethoscope and blood-pressure machine in a doctor's office in Bordeaux. A study by a general practitioner's union estimates one in three victims of Covid-19 may be dying at home. REUTERS - Regis Duvignau

Every day from the beginning of the Covid-19 epidemic in March, French health officials have posted data showing the number of hospitalisations, discharges and deaths due to the disease caused by the new coronavirus. 

Following weeks of rumblings about the toll the disease was taking in care homes for the elderly, officials began reporting relevant figures in mid April.

The French government now says health services will provide an estimate of how many people have died from Covid-19 at home in June. 

“We have received an initial alert… of a possible abnormally high mortality rate at home,” Health Minister Olivier Véran said Wednesday.

The estimate will be provided by the Public Health France agency and will be based on a review of death certificates, which generally state the cause and place of death.

Véran said the deaths in question were “not necessarily linked to Covid, but perhaps rather to medical complications of people who did not go to hospital”. 

A third of deaths unreported?

Two studies with widely different methologies have provided varying estimates of above average mortality and deaths outside hospital. 

A survey by MG France, a union representing general practitioners, suggested the new coronavirus was responsible 9,000 deaths at home between 17 March and 19 April.

With the official number 19,718 on 19 April, that would suggest nearly a third of Covid-19 deaths occurred in patients’ homes.  

The figure of 9,000 is extrapolated from a survey of 2,339 general practitioners, who reported 340 deaths of patients with confirmed cases of Covid-19 and who had either not gone to hospital or been discharged. 

The 2,339 respondants represent 4.1 percent of France’s general practitioners, meaning the extrapolation is based on a small sample group.

Or is it one in five?

A more mitigated estimate may be derived from a study of based on 20 years’ worth of mortality figures kept by France’s National Institute of Statistics and Economic Studies (Insee). 

Baptiste Coulmont, a sociology professor at University of Paris 8, used this data along with daily figures Insee is providing during the epidemic to visualise the toll of Covid-19. 

2020, une mortalité au profil particulier (avec les données que vient de publier @InseeFr ) https://t.co/7i8rEhq6Ov

— coulmont (@coulmont) April 24, 2020

As explained in newspaper Le Monde, the study found there were 93,324 deaths in France between 1 March and 17 April of this year, a difference of 22,198 or about 31 percent over the average 71,126 over the same period for the previous 20 years.

Coulmont notes the 2020 figures from Insee may be underestimations that will be revised over time. 

But the tentative estimate is higher than the official mortality figure of 18,681 deaths in hospitals and care homes posted on 17 April, a difference of 3,517.

While only a fraction of the MG France estimate, it suggests nearly one in five of the deaths above the normal rate are happening at home. 

Despite their differences, both analyses suggest a larger-than-normal mortality outside that reported in the currently available official figures. 

Above-average mortality in 2020

Coulmont’s analysis also shows the extent to which Covid-19 has taken an exceptional toll in France.

“From the start of March 2020 and for a month, the curve… breaks away from the average,” Coulmont wrote on his blog, noting the first two months of the year were average in terms of mortality. 

“From the start of April, about two weeks after the start of confinement, the number of deaths started to diminish,” Coulmont wrote.

The Covid-19 epidemic represents one of two spikes in 20 years, the other being the heat wave that killed an estimated 15,000 people over a three-week period in summer 2003.

The study also shows Covid-19 is far deadlier than the seasonal flu, which itself took particularly large tolls during epidemics in 2015, 2017 and 2018, without confinement measures.