26.8 C
Saturday 19 June 2021

Network Covid: good Draghi, but it would be good to change the algorithm

Covid: good Draghi, but it would be good to change the algorithm

Author of the content

The strategy indicated by President Draghi, summarized journalistically by the formula "vaccines for the elderly and then reopens" is a strategy that has a strong logic, but which requires a review of the classification algorithms today mainly dependent on the data of the "infections".

The theme is well known: there are two great collective interests (health protection and economic well-being) that in part conflict with each other, given that the carrying out of some productive activities (in particular commercial, tourist and cultural ones) determines the creation of gatherings and contacts that increase the risk of contagion.

To support the political decision-maker in determining the restrictions on social and economic activities, a complex data interpretation model has been developed which, through an articulated algorithm, derives the classification of risk levels (and therefore the restrictive measures) from the data relating to a large set of variables. In the model currently used, the so-called Rt (the index that should measure how many are infected by an infected person) and more generally the data relating to infections plays a crucial role.

In reality we know that not all infections are the same, given that the actual health consequences are not always the same as they are influenced (which does not mean strictly determined) by the characteristics of the subjects such as, in particular, age and health conditions. On the other hand, the truly crucial variables from the point of view of health protection are above all the number of deaths and hospitalizations.

From the data of the Ministry of Health we know that of the 754.000 cases currently (April 8) "active" over 437.000 (58%) are "asymptomatic", 279.000 (37%) are "paucisymptomatic" or "mild" and only 37.563 (the 5 %) are classified as "severe" or "critical" (most of which require hospitalization). These percentages, however, vary, as mentioned, to a significant extent according to age (see graph); in particular, the share of severe or critical cases rises to 16,0% for the over 70s who make up 50% of the most serious cases.

It is therefore conceivable that, in the face of the concentration of the vaccination campaign on older subjects, the number of cases of severe or critical clinical states is reduced earlier and more than proportionally with respect to the number of infections (even more if, as is plausible, the the elderly are less “active” speakers than the younger ones who have a greater number of contacts). The confirmation of this hypothesis should be obtained in a lowering of the curve of hospitalization flows more marked than that detectable for the flow of infections. However, the current algorithm would take into account only partially (and with delay) this eventual trend, as the weight of the “contagion” indicators is paramount. Also from considerations similar to these, I presume the proposal to include an indicator of the vaccination rate of subjects at risk in the formulation of the model and algorithm arises.

Personally, however, I believe that an even more radical solution could be adopted by identifying the number of hospitalizations as a fundamental variable for determining the level of criticality of the situation (and therefore the extent of the restrictive measures), perhaps articulated on a regional scale also according to hospital equipment according to criteria established by health experts.

On closer inspection, in fact, this is the truly relevant variable for health protection policies; paradoxically, an increase in "asymptomatic" infected people would be sanely irrelevant and that of paucisymptomatic or mild cases could also be considered an acceptable price given the seriousness of the economic conditions of hundreds of thousands of economic operators and their families.

The admissions data, however, expresses the real structural problem, that of the "pressure" on the hospital system, which when it grows beyond certain limits determines a loss of quality of care (even for non-covid patients) and, moreover, it is a fact that it has proved to be more timely and reliable than others and less susceptible to more or less involuntary alterations. Of course, such a choice should also apply in the opposite case: if, for example, a new variant were to determine a greater frequency of serious consequences on fewer infections, it would lead to more restrictive choices; in the current situation, however, it should favor conscious “re-opening” choices.

Finally, I think it makes sense to add one last consideration: setting the thresholds in terms of the number of hospitalized people means indicating clear and understandable objectives for the vast majority of citizens: one thing is to say "we will reopen only when the Rt is permanently less than 1 "" Another is to say "we will reopen when we drop below 20.000 or 10.000 hospitalized". Defining clear and understandable objectives (at least to those who want to understand) is not just a communication ploy, it is a way to make people aware of and participate in a collective effort and to isolate those who try to exploit social and economic malaise.

Of course, a more complex algorithm is - in theory - more capable of grasping the various aspects of the situation but the continuous fluctuations in classification of the past months indicate that the current one did not work very well (otherwise we would not have had "white" regions until after two weeks have turned red or orange!). 

I think it is worthwhile, therefore, to try to change the instrument, especially if we want to follow up in an organic way to what President Draghi, once again showing courage, has announced.

PS the number of hospitalized fell from 33.080 on 6 April to 31.749 on 9, we hope this is a trend that is confirmed.

We are a young editorial reality e we do not receive public funding. Our work is supported only by the contribution of the publisher (CuDriEc Srl) and advertising revenue. THE readers they are ours true wealth. Every day we try to deliver accurate, unique and true insights.
Support Moondo, support independent information!
I wish to send a free donation to Moondo (click and donate)

Yours opinion for us it is very important.
Comment on WhatsApp
Now also on Google News, click here and follow us

Sign up for our free newsletter

Stay up to date on the latest insights.
Leave your email address, select your interests and receive for free on your inbox the first page of Moondo with the most interesting news selected for you.