The modelling work carried out by the modelling sub-group (IEMAG) of NPHET supports risk assessment and public health recommendations. Statistical and modelling approaches have significant and important limitations and these limitations are well understood by NPHET and key decision makers. The models used by IEMAG provided possible scenarios showing what might happen if the trends at the time continued. These scenarios are not forecasts or predictions, but instead offer a variety of “what-if” scenarios, showing the rate of infection if reproduction number has a particular value due to underlying population behaviour.
At the time the letter of 15 October was written, R was estimated at 1.4 and cases were growing at 5-6% per day, giving a doubling time of 12-14 days. The growth rate and doubling time for hospitalisation and ICU occupancy were also similar (growth rate 3-4%,doubling time 17-22 days). The letter explicitly stated that the modelling did not take into account the likely impact of Level 3 measures. It is likely that if Level 3 measures had not been implemented across the State, the case numbers indicated in the modelling would have been reached and (ii) visits between households had not been banned, nor Level 4 measures applied to Border Counties, case numbers now would be intermediate between 1,200 per day and the 1,800 projected if the reproduction number had remained at 1.4.
The IEMAG modelling also showed that for each sustained reduction in daily case numbers of 100 per day (say from 1000 cases per day to 900 cases per day) IEMAG predict that over the subsequent month alone this would prevent at least 100-140 hospitalisations, 15-20 admissions to ICU and 10-20 deaths. If case numbers are reduced from 1200 per day to 200 per day, over the subsequent month this will prevent at least 1000-1400 hospitalisations, 150-200 admissions to ICU and 100-200 deaths.
Technical notes on the approaches used by IEMAG are published on the Department of Health website (https://www.gov.ie/en/publication/dc5711-irish-epidemiology-modelling-advisory-group-to-nphet-technical-notes/ ). These are in line with best practice and associated references are available in the published technical notes.