Spain currently has 124,300 active cases of COVID-19 according to the mathematical model developed by the UPC and IGTP

- Research

A multi-disciplinary team which includes reseachers from the Germans Trias i Pujol Research Institute (IGTP) has analyzed the active cases of COVIDA-19 in the different autonomous regions in Spain. The analysis shows that there re 124,300 active cases which means that the virus can continue to propagate. By autonomous regions, the most affected would be Madrid and Catalonia, which have accumulated 28% and 27% of the active cases respectively.

The prediccions produced by the researchers Martí Català of the IGTP, and Enric Alvarez and David Pino of the Computational Biology and Complex Systems research group (BIOCOM-SC) of the Universitat Politècnica de Catalunya · BarcelonaTech (UPC) has been able to calculate the active cases per autonomous region and how many people would actually have the disease. The data shown in the following table indicate that the autonomouis regions with more active cases continue to be Madrid (with 35,000 active cases, 28% of the total) and Catalonia (with 34,000 active cases, 27% of the total), followed by Castile-León (13,000 active cases) and Castile La Mancha (11,000 active cases). On the other hand, the Canary Islands (500 active cases), Cantabria (600 active cases), La Rioja (600 active cases), Murcia (600 active cases) and Asturias (700 active cases) would be the autonomous regions with the lowest figures.

The model is based on considering the lethality of COVID-19 as 1% in those regions with a significant penetration of the disease. This value is obtained from an analysis of the studies of sero-prevalance carried out in other countries and of the evolution of the Case Fatality Rate (proportion of deaths in respect to detected cases) by age for the countries with a high detection rate, such as South Corea. In this context, one aspect to be borne in mind is that the number of deaths in residences for the elderly is a good indicator of the penetration of the disease. The UPC researcher Enric Alvarez, leader of the study explains:

"In the areas where the pandemic has penetrated in the population of over-70s, there are a significant number of cases in residences for the elderly. Due to the poorer prognosis for this group, the average lethality for the population can rise to 1.2%, depending on the final number of deaths due to COVID-19 in residences for the elderly. This figure will be difficult to check, given the lack of confirmation in many cases." In areas with lower penetration and affectation, lethality can even be below 0.7%, in this situation the number of real cases is far from the estimations of the researchers. If lethality is lower, the real cases without symptoms, or with mild symptoms will be higher than those estimated.

From lethality, the delay in diagnosis is analyzed, this means the time that passes between developing symptoms and being introduced into the database as a positive case and also depends on the capacity of the country to detect. In Spain approximately between 8-10% of cases are detected, explains the IGTP researcher Martí Català:

"Diagnoses in Spain are done by PCR, which, until now, has basically been carried out in hospitals. This gives us an evolution of the severe cases. Tests in primary care centres are being planned to increase the percentage of diagnoses. This will also reduce the interval between the appearance of symptoms and entry into the official figures. At the moment this delay is of between 10 and 12 days, according to our model."

With this last value, from reported data for autonomos regions the real number of cases for the last fourteen days can be established. These cases are considered active by the European Centre for Disease Prevention and Control (ECDC) as a percentage of them, determined by the level of symptoms, can infect others. The active cases are the number of people potentially contagious, which doesn't mean that they all are.

On target for the data on seropositive people

The model developed by the team of researchers correctly predictived the figures in the results of the first round of the sero-epidemiological study for COVID-19 that the Ministry of Health made public (ENE-COVID). The model is not only able to obtain a similar result to the the study for global sero-prevalence in Spain (5%), but also reproduced the variablity in each autonomous region. The model does not agree with the numbers for the study principally in La Rioja and to a lesser extent for Andalusia and the Canary Islands.

In these last two regions, the prevalence observed has been higher than that estimated by the model (assuming a lethality of 1%). In these cases the lethality is clearly less than 1%, probably because the groups over 60 years of age have been less affected than in other regions.

In La Rioja, the model assigns a much higher prevalence than that reflected in the ENE-COVID study. David Pino points out, "We believe that the development of the disease in this area in particular could have affected the results. In this region there are areas of basic healthcare that represent a little more than 10% of the population of La Rioja but registered 30% of the the cases on 18 May and probably a similar percentage of deaths. By contrast, Logroño and neighbouring municipalities, which have more than 50% of the population have registered 40% of the cases of COVID-19 in La Rioja. As this is a study of sero-postivity, it has been carried out in terms of the population of the region, it is possible that the figures in the ENE-COVID are lower than in reality. The over-estimation of the model in these autonomous regions would be less than we see on the graph."

A multi-disciplinary team

The team is made up of Daniel López-Codina, Sergio Alonso, Miguel Marchena and Enrique Álvarez, of the Computational Biology and Complex Systems Group (BIOCOM-SC) at the UPC; David Pino, of the Fluid Dynamics Group of the Physics Department of the UPC and Martí Catalá and Pere-Joan Cardona of the Centre for Comparative Medicine and Bioimage (CMCiB-IGTP), coordinated by Clara Prats (BIOCOM-SC / CMCiB-IGTP). The Physics Engineering students Tomà Urdiales and Pablo Palacios of the Barcelona School of Telecommunications Engineering (ETSETB) of the UPC also collaborated on the work. The graphics have been elaborated by Ferran Caymel. The project is supported by the "la Caixa" Foundation.

You can get daily updates of the predictions and reports from the BIOCOM-SC research team from their Twitter account

See the original paper on this work here