Almost since the beginning of the covid-19 pandemic, there has been a recurring discussion about the number of deaths caused by this disease. Another issue discussed is whether or not the number of deceased persons represents an excess of mortality. Immediately afterwards, a tagline is added about transparency in information.
The debate is inevitable because it is not usually considered how the information is obtained, nor its different purpose and, of course, it does not inquire about the required process or the systems used.
The sources of mortality data
One system is the National Epidemiological Surveillance Network (RENAVE), managed in Spain by the regional health services and the National Epidemiology Center (CNE) and from which the Ministry of Health draws. It is directed to immediate action. For this reason, it works with information very close to the moment of the event, and improves its quality later. It needs, at first, to identify all possible deaths from the disease (that is, it tries to be a system with great “sensitivity”). Subsequently, it seeks to increase its specificity (trying to identify those who, according to the available information, have actually died from the disease).
For this reason, the surveillance systems have enabled processes to contrast the cases of people considered as cases of covid-19 who have died and seek additional information, through the consultation of medical records or similar mechanisms, to be able to confirm the death and attribute it or not to the disease.
Herein lies the crux of the matter. Because sometimes the information has been, or is, scarce or non-existent. This makes it difficult to assign death to disease. It is not the same to die “with” covid-19 than to die “from” covid-19. And both circumstances can occur, especially in a disease with greater affectation in vulnerable people who, prior to suffering it, already had a greater probability of dying than other people.
From RENAVE, the Ministry of Health and the CNE provide the daily number of deaths from covid-19 by date of death, also including access to the MoMo system, which analyzes excess mortality from all causes (which was created to analyze the effects of heat waves).
The second system is the one documented by the medical death certificate (CMD). In it, when someone dies, a doctor, with the clinical information available to him, records, among other data, the cause of death. Its main purpose is to document the death and allow the burial of the deceased person. It is sent to the corresponding Civil Registry, and the death will be recorded there for legal purposes.
Together with the CMD, the Death Statistical Bulletin (BED) is also filled out. It includes the causes of death, distinguishing between: immediate cause, antecedent or intermediate cause (disease or injury, if any, that contributed to the immediate) and initial or fundamental cause. It is the responsibility of the National Institute of Statistics (INE), which has agreements with the Autonomous Communities so that the causes of death are coded in each BED and the so-called “basic cause of death” is selected. That is the one that will appear in the “death statistics according to cause of death.”
For covid-19, the World Health Organization (WHO) published, in April 2020, the Guide for the coding of covid-19 as a cause of death. It established two “emergency” codes: “U07.1 COVID-19, identified virus”, which is assigned to the diagnosis confirmed by laboratory test, and “U07.2 COVID-19, unidentified virus” assigned when there is a clinical or epidemiological diagnosis but laboratory confirmation is inconclusive or not available.
This distinction is important, since in the case of code U07.2, there is a non-negligible probability that it could include deaths that would have also occurred without the existence of COVID, or include cases even with negative laboratory tests and nonspecific clinic. This can lead to an overestimation of deaths attributed to this cause.
With this, the INE provides two analyzes:
- The estimate of the number of weekly deaths during the covid-19 outbreak (experimental statistics) that basically come from civil registries, without distinguishing causes.
- The causes of death that have been codified, and of those that have recently been published those that occurred during the first months of the pandemic (January 2020-May 2020).
The results of the mortality analyzes
So we have different sources that can be used. What is observed is that the year 2020, without having finished yet, is the one with the most deaths in the 21st century. Deaths from infectious diseases have increased in 2020 by 1 687.7% compared to the same period of the previous year and, within this group, those coded with U07.1 represent 67.5%, with a mortality rate of 68 8 per 100,000 inhabitants.
Until May 2020, inclusive, the figures provided by the INE of deaths with the code U07.1, are 32 652, and those identified by RENAVE are 29 156: a difference of 3 496. This difference is not territorially homogeneous, since 10 autonomous communities and cities have more deaths in INE than in RENAVE (concentrating the main difference, almost 90%, in 2 autonomous communities), while in 8 of them there are more deaths in RENAVE than in INE with this code.
Differences are being highlighted without seeing matches. Among them, it should be noted that a significant volume of cases has been identified by all systems, one of them, RENAVE, working with very strong pressure and scarce resources. And that when discrepancies have occurred, they have been in both directions, since there are also cases identified by RENAVE that do not have a code U07.1 assigned in the CMD / BED, and sometimes neither U07.2, and no one asks stop considering them as cases of people who died “due to” covid-19.
Conclusions
- At the national level there has been an excess of mortality.
- The excess can be attributed mainly to covid-19.
- The surveillance systems in most of the territories have identified the most important part of the cases that have later been confirmed through the CMD / BED using the code U07.1 (there have been autonomous communities that have even identified more).
- One of the main discrepancies lies in the treatment that must be given to the cases that have the code U07.2 assigned, since there it will be difficult to conclude whether they should be included as covid-19 cases or not.
But the main conclusion should be that the real discussion should not focus only on the numbers, since behind it there are, above all, people, pain and grief.
Óscar Zurriaga, Associate Professor. Department of Preventive Medicine and Public Health (UV). Serv. Epidemiological Studies and Statistic. Sanit. (Valencian generalitat). Unit Joint Research on Rare Diseases UV-FISABIO. CIBER Epidemiology and Public Health, University of Valencia
This article was originally published on The Conversation. Read the original.
