The modelling presented by NPHET on the potential impact of the Delta variant and further reopening this week contains a few key assumptions and terms.
The first key term is "step change". All the numbers presented are modelled based on the assumption there will be a "step change" in social mixing on 5 July, the date indoor dining was due to reopen.
Specifically: "A step change in close social contact is introduced on 5 July 2021 and maintained constant thereafter."
Based on that, four scenarios were presented. Two scenarios assume a large "step change", two assume a "moderate step change". So in all scenarios, an increase in social mixing is assumed to happen.
That is one important input. There's one other which relates to the estimated impact of the Delta variant: Big, or bigger.
Therefore, the modelling is not providing an analysis of the potential impact of the Delta variant alone at any point, but of the potential impact of the Delta variant in combination with a "step change" increase in social contact each time.
The other key term has been used already here: "scenarios". The modelling is not forecasting or predicting an inevitable future, but presenting what could plausibly happen if - if - the assumptions in each scenario came to pass.
The four scenarios provided are labelled Optimistic, Central 1, Central 2, and Pessimistic, but they are not a bad-to-worse forecast range for the coming months.
If there was to be a moderate increase in social mixing from 5 July, combined with a moderate impact of Delta, it could result in the scenario labelled "Optimistic".
A higher increase in social mixing with a moderate Delta variant gives the "Central 1" scenario.
A moderate increase in social mixing with a higher Delta impact is Central 2", while higher social mixing combined with a higher estimate for the impact of Delta is labelled the "Pessimistic" scenario.
The four provide a range of potential cumulative cases across July to September from an optimistic 81,000 cases to a pessimistic 612,000 cases. All assume at least a moderate increase in social mixing.
The scenarios are similarly wide-ranging on hospitalisations and ICU admissions – from worse than the January peak in hospitalisations, to a quarter of the number of people we saw in January in hospital.
Yet the proposed step change on which the models are based is not due to happen anymore, since the Government has decided against reopening indoor dining on 5 July.
That decision was based on these models.
The aim of that decision is to avoid the step change on which all the scenarios are based. The scenarios therefore should not come to pass.
So what sort of situation are we looking at instead?
At the NPHET briefing on Thursday, Prime Time asked Professor Philip Nolan, Chair of the Irish Epidemiological Modelling Advisory Group, if he'd run the models without the step change in social mixing included, and what the outcome would be.
"I did, and the answer is: it’s modest," he said. "Frankly, it is anywhere between Alpha-only and the optimistic scenario."
In the modelling, an unrealistic "Alpha-only" scenario is also provided. It assumes the Delta variant hadn’t arrived, to provide a baseline for the impact of the variant alone.
On that basis, indoor dining could have gone ahead, NPHET said, as vaccination rates would have sufficiently suppressed the virus.
The optimistic scenario assumes a moderate increase in social mixing, with a conservative estimate for the likely impact of the Delta impact.
His models show that the likely impact of Delta without a change in social mixing will be somewhere between these two scenarios.
The Alpha-only scenario models that 21,000 cases of Covid-19 would have occurred across July, August and September, with 80 deaths.
The optimistic scenario says there would be 81,000 cases across those months, and 250 deaths in the same period.
It says daily cases would peak around 2,500. Thousands of people would be hospitalised, with the number of in-patients with Covid-19 peaking around 500. There would be an increase in ICU numbers, peaking at approximately 100. This would undoubtedly constitute a significant fourth wave.
At the time of writing, there are just 46 people with Covid-19 in hospital, and just 14 in ICU. There are 300 critical care beds in the hospital system, so a rise to 100 would have an impact on service.
There would be knock-on impacts on the health system due to the high rates of transmission, and potentially later also due the impact of "Long Covid" from high case numbers.
However, in the January 2021 wave, hospitalisations peaked around 2,000. There were 221 Covid-19 patients in ICU beds.
The modelling presented by NPHET indicates a January-like peak is a possibility, but far from an inevitability.
Avoiding a January-like peak would require social mixing to remain as it has been recently – for the step change not to happen, according to the modelling.
While the Government decided against reopening indoor dining this week, some other restrictions have been eased. The number of people permitted at outside events and in stadiums has been increased. Weddings already planned on the basis of 50 guests are allowed to go ahead.
At the NPHET briefing, Prime Time asked whether such changes would likely result in the "moderate" increase in social mixing modelled in the optimistic scenario. "No," Prof Nolan said.
He later told Prime Time that "the whole point of running the scenarios is to prevent them, by making the right decisions."
To summarise, if social contacts remain stable or reduce somewhat, Prof Nolan said the impact of Delta on the health system will be heavily mitigated.
Let's hope he’s right – the health service is still recovering from the scale of the last January wave.
He said he will be running more modelling early next week. That new work will consider the now-likely level of social mixing, any new data on the Delta variant, and incorporate the acceleration of the vaccine roll-out.