Reports about the latest Covid-19 modelling published by NPHET have topped the news agenda in recent days.

Two scenarios were presented to the Cabinet sub-committee on Covid-19, covering the period to late January.

The optimistic scenario showed cases peaking around 5,000 per day in mid-December. That's about a 15% to 25% increase on the current case rate.

A peak of around 12,000 is modelled in the pessimistic scenario.

In both scenarios, a peak in hospitalisations and ICU occupancy rates would occur around Christmas.

But it's important to understand the assumptions underpinning the modelling, and also that they were flagged as particularly uncertain to decision makers.

This is the first modelling run done done by Professor Philip Nolan's Epidemiological Modelling Advisory Group that attempts to include the impact of waning vaccine immunity.

One issue is that the science and data about waning immunity is still emerging.

Additionally, the current pace of the spread of the virus is difficult to identify exactly, as test positivity and case numbers are particularly high. For both those reasons and others, the models come with a red flag of "uncertain".

Prof Philip Nolan chairs NPHET's Epidemiological Modelling Advisory Group

It's also worth underlining that the graphs shown are scenarios, not forecasts, projections or predictions.

In modelling, "scenarios" are considered plausible future outcomes based on the assumptions made. The word "forecast" is used for outcomes that are considered probable.

The two scenarios are therefore what the team appear to have considered plausible - possible - futures at the time the modelling was run.

The modelling is based on the assumptions made about what could happen in the future and, importantly, the modelling team's understanding of what has happened in the past.

These scenarios were presented to ministers to inform the decision-making process before Tuesday's announcement.

The government announcement that followed included various measures designed to stem the spread of the virus, including asking more people to work from home; requiring late bars and nightclubs to close at midnight; and saying household close contacts of a confirmed case should restrict their movements for five days even if fully vaccinated.

In other words, the modelling covered plausible scenarios for a future in which nothing changed after they were presented. Since then, decisions have been taken to try - try - to change what happens in the future.

Regardless, the assumptions that went into the modelling provide an insight into NPHET's view of key aspects relating to the of spread of the virus. So what are the assumptions?

Social contact

The first assumption is that the level of social contact in society does not change from the level it was at in the second week of November.

Therefore, the modelling doesn't account for any fall in social contact due to the concern currently being expressed publicly by the Government, NPHET, and hospital staff. Nor does it assume social contact will increase to any specific level in the run up to Christmas.

Child transmission

The next assumption is that children transmit the virus less than adults. NPHET members have been saying for months that the evidence internationally, and their own analysis of the data in Ireland, shows that children under 12 years old do not transmit like adults. The modelling team has built that conclusion into the models.

Some prominent health experts, including Anthony Staines, Professor of Health Systems in Dublin City University, believe other evidence shows this assumption – among others in the modelling – is misplaced.

Immunity in the population

After that, the only input that differs between the two scenarios is described.

In the optimistic scenario, this input is described as: "The fraction of infections not detected is set at 60%."

For the pessimistic scenario, the number is 40%.

This assumption is important. It explains the difference between the two peaks.

It's an attempt to estimate the number of people still in the population who have no immunity through either vaccination or infection. If there are more people with no immunity, there is more potential for a higher case peak.

The percentages mentioned relate to the number of confirmed cases over the course of the pandemic.

More coronavirus stories

About 500,000 positive cases have been detected so far through testing. An unknown number of other cases will have occurred but not been confirmed through testing, yet the individual infected will have gained some immunity.

If the 500,000 represents 40% of all infections which occurred, then 1.23m total infections are assumed to have happened. This is the optimistic scenario. Paradoxically, in the optimistic scenario the system picked up fewer cases, so a bigger level of immunity exists which hasn't been confirmed by tests.

If 500,000 represents 60% of all infections, then it's assumed 833,000 infections occurred. This is the pessimistic scenario because there's assumed to be less immunity in the population, so the potential exists for a comparably bigger surge in the future.

The pessimistic scenario shows cases peaking around 12,000 then falling off, the optimistic one peaks around 5,000.

In both, the fall off then occurs because enough people in the population have some form of immunity at that point in time to stop the virus spreading. Population immunity - at least for a period - according to the models.

However, with more people assumed to be unprotected in the population under the pessimistic scenario, the impact on hospitals is far higher. Hospitalisations peak at 4,000, with the number in ICU between 400 and 500.

In the optimistic scenario, hospital occupancy peaks around 1000, with 200 people in ICU with Covid-19.

As of Thursday evening, there were 643 people in hospital with Covid-19 and 117 in intensive care.

Booster roll-out

The modelling also assumes that the booster roll-out reaches everyone over the age of 60 by mid-January. About 80% of over-80s have already received a booster. In the models, it's estimated that most 70-79 year olds will have had one by 1 January.

On Tuesday, Minister for Health Stephen Donnelly extended vaccine booster eligibility to 50-59 year olds. As that decision hadn't been announced when the models were run, this age cohort isn't boosted in the modelling.

The boosters contribute to the eventual dip in cases in each scenario.

On how the waves are assumed to end, Prof Nolan tweeted: "This wave of infection does come to an end, because many unvaccinated adults become infected and are then immune, many vaccinated adults get breakthrough infections, boosting their immunity, and many adults over 50 will receive booster doses."

The modelling is the first to include the impact of waning vaccine immunity

The outcome of this scale of infections occurring is the estimated impact on the hospital system.

He continued: "We do not want this to happen, it is not a strategy or a policy, we would rather protect people through further vaccination, but if we do not reduce transmission, this is what will happen."

If that were to happen, the question would then become 'how long does this level of population immunity last?' because it's now clear that immunity wanes.

Vaccine effectiveness

Waning immunity for vaccinated people is included in the modelling, and considered to be occurring among all vaccinated age groups. However, older people's immunity wanes more than younger people.

The modelling assumes all people who are fully vaccinated reach a peak level of immunity 28 days after their most recent dose.

At that peak, it's assumed vaccines are 70% to 80% effective against symptomatic infection. They're further considered 80% to 90% effective against severe disease – in others words, hospitalisation.

This means the risk a vaccinated person has of being infected and getting symptoms at the vaccine's protective peak is considered to be 30% or 20% of whatever level of risk they had before they were fully vaccinated. The likelihood of hospitalisation is further reduced.

As many of us now know, the baseline level of risk from Covid-19 is heavily linked to age.

Therefore, with vaccines, older people who are vaccinated have a reduced risk, but it's a reduction from a higher level than in younger people.

Waning vaccine-gained immunity

In the modelling, on day 29 after full vaccination, vaccine-gained immunity begins to wane.

When it comes to the pace of waning, the modelling assumes that the protection gained against symptomatic infection among older people falls by 60% over the course of a year or more, but half that fall occurs in the first 90 days.

For younger age groups, it's assumed it falls in a similar way, but by 40%.

When it comes to protection against severe disease, it falls to 90% from the peak in younger age groups, and 80% in older age groups.

This is especially important, because it results in the total number of expected hospitalisations in older age groups more than doubling.

If 1,000 unvaccinated people were infected, about 50 could be expected to be hospitalised. If that same group were all vaccinated with vaccines that are 90% effective against severe disease, just five would be expected to be hospitalised.

However, if the effectiveness of the vaccines against severe disease wanes to 80% of its peak, then the number of hospitalisations in the group would be 14.

When thousands of infections occur each day, that shift from 5 to 14 is significant.

Infection-gained immunity

The final assumption is a crude one. Infection-gained immunity is considered permanent. This is not supported by the available evidence, which shows that reinfection does occur.

This assumption may result in the level of immunity in the population being somewhat overestimated, albeit it's difficult to quantify, as many people who were infected will also have been vaccinated.

Members of the NPHET modelling team told Prime Time it's likely the effect of waning infection-gained will be added to the modelling in the next iteration.