Press Release N° 6/2021 -Lombardy: the algorithm is correct and it functions in the same way for all the Regions. The amendment to the number of cases made by the Region has changed the RT

Rome, 23 January 2021 - As regards the statement made by President Fontana this morning on the algorithm used to calculate the RT for the monitoring of the COVID-19 epidemic, the Italian National Institute of Health specifies that:

The algorithm used by the Italian National Institute of Health is correct. It has not changed since April and it is the same for all the Regions which have used it so far without any problems.

This algorithm and the methods used to calculate the RT have been explained in detail to all the regional contacts so that they could calculate it and verify the estimates produced by the ISS. Therefore, it is accessible to everyone.

The figure is sent every week in advance to the Regions which are asked to check it and validate it with the understanding that ‘silence means assent’. So far, the Lombardy Region has never questioned this estimate.

The Institute, but also the Bruno Kessler Foundation, which developed the algorithm with the ISS to monitor the epidemic, has always been available to interact and support the Regions and Autonomous Provinces at any stage of their calculations.

With regard to the calculation of the RT, the flow of data on which it is based, namely the data on symptomatic patients, is as follows:

Each time a clinical case is detected, the relevant "clinical status" field is filled in, indicating the degree of severity of the symptoms, from less-severe to severe and, when possible, also the date of onset of the symptoms.

Presence of symptoms and date of onset are the two variables that define the individuals  to be considered for the calculation of the RT. It may happen, even if seldom, that the date of onset of the symptoms is not associated with a clinical status and, in this case, the patients are initially considered to be symptomatic because the data provided by the Region are taken to be indicative of the presence of symptoms in the absence of other information.

If, on the other hand, the "clinical status" field is not filled in and continues not to be filled in during the subsequent updates of the database until recovery or death is reported, the case is considered asymptomatic in spite of the presence of a date of onset of the symptoms. Indeed, it is not plausible that there be a date of onset of a person's symptoms without reporting any symptoms until recovery or death of the patient.

Finally, we consider as asymptomatic also the rare cases in which a date of onset of symptoms is provided but the case is declared to be in an "asymptomatic" clinical status, since the two data are in clear contradiction.

In the emergency situation in which data quality problems may occur, this conservative approach is aimed at avoiding an overestimation of the symptomatic cases on which the RT is calculated.

Since the beginning of the epidemic, Lombardy has reported a large number of cases, significantly more numerous than those observed in other Regions, indicating a date of onset of symptoms which they did not associate with a clinical status. These cases were initially considered to be symptomatic, in accordance with the procedure described above.

The ISS has brought this anomaly several times to the attention of the Lombardy region.

In addition, recently the Lombardy Region classified a large number of these cases as cured without ever having indicated what the clinical status was with regard to symptoms. Therefore, always in accordance with the procedure described above, a few weeks after being initially reported, these cases were excluded from the group of symptomatic patients because of the inconsistency of the data.

The combination of these two criticalities resulted in a recent overestimation of the number of symptomatic cases that was corrected in the previous weeks. The consequence was an incongruous increase in the difference between the number of symptomatic cases documented in the week being processed compared to the previous week, which in turn showed an over-estimated RT only for Lombardy and only in the estimates of these weeks.

Only following the amendment to the data relating to the date of onset of the symptoms and of the clinical status of the cases already reported, which took place with the data upload on January 20, and the correct identification of asymptomatic cases by the Lombardy Region, at their request, the estimates of the RT made the previous week were recalculated.