Subscriber OnlyEconomy

Gastropubs and restaurants cannot be blamed for all the regression in December

Confident assertions about role of hospitality and mutation being driven by anecdote

Chief medical officer Tony Holohan. Photograph: Colin Keegan/Collins
Chief medical officer Tony Holohan. Photograph: Colin Keegan/Collins

Occasionally during the fight against the pandemic, the inevitable fog of war has appeared heavier than it needed to due to questionable or sometimes even absent data. The public health authorities regularly publish reams of detailed epidemiological information, but sometimes important bits are missing that would help the public to better understand what is happening.

For example, the hospitality industry is squarely blamed by many as a primary source of infection, yet those who blame it cannot explain why this culpability never shows up in virus cluster data.

A related issue is that, until recently, Nphet officials bizarrely considered that the tracing of the source of infections, even when virus numbers were low and this was easily possible, was just an “academic exercise”. But it has been central to the virus response in other jurisdictions, such as Japan.

Further restrictions on business were announced by the Government on Wednesday. Photograph: Gareth Chaney/Collins
Further restrictions on business were announced by the Government on Wednesday. Photograph: Gareth Chaney/Collins

Meanwhile, the flow of reliable data about the culpability of the new UK variant for the current upsurge has been patchy. Thus the issue inevitably has become politicised in the blame game over the Government’s decision to reopen the economy and society, which immediately preceded the surge.

READ MORE

Those advising the Government on how to handle the pandemic, and the Ministers making the calls, may feel that they have sufficient information to forge ahead. But they also need to bring people with them by showing that their decisions are fully justified. It is here that the occasional (rather than chronic) lack of empirical data on how the virus spreads has had the most negative impact.

The debate around the role played by the hospitality sector on the current wave has become emotive and highly partisan. Ignorance born of patchy data is fuelling the vitriol. All of this will have an inevitable impact on the quality of the discussion when the question rolls around again in coming months of how or whether to reopen the sector. Better data would mean a better discussion.

Indoor dining

Nobody around the hospitality debate argues that the reopening of establishments for indoor dining in December had no impact on what has happened with the upsurge. But while it is surely partly to blame, the full extent of its impact should be important for us to ascertain. Yet we still don’t fully understand it.

It is pointless looking to social media for balanced discussion on this or any other issue. Yet Twitter and other online platforms are now dominated by cocksure people who, with little direct evidence, blame lobbying on behalf of gastropubs and restaurants for the totality of the regression in December.

With the data available, how could they possibly know this to be completely true? The answer is they cannot.

Last week more than 21,000 new cases of infection were reported by public health authorities, as the pandemic accelerated out of control. Yet compare this with the outbreak or cluster data published by the Health and Protection Surveillance Centre (HSPC), a part of the Health Service Executive.

The HSPC’s latest report for the seven days to the end of January 2nd shows only 10 outbreaks associated with gastropubs and restaurants that were notified in the preceding week. There were just five notified the previous week, and before that, the last hospitality clusters were notified in November.

Anecdotal evidence internationally suggests that bars and restaurants further the spread of the virus, and it is correct that we now question the wisdom of the decision to reopen in December, given where we are now. But why does it not show up at all in Irish outbreak data? It can only be because the data is flawed.

A cluster is defined as two or more associated cases. If the HSPC’s numbers are correct, then thousands of infected Irish residents were wandering in and out of restaurants and gastropubs in the run-up to Christmas, but they each infected either nobody or a maximum of one other person on each visit. If they had infected more than one other person, it ought to have shown up as a cluster.

The clearly deficient data has fuelled the industry’s sense of grievance at being lumped with all of the blame.

North’s experience

In Northern Ireland the authorities went to the trouble of sifting through the evidence to estimate the direct impact of the reopening of the hospitality sector on the virus’s propensity to reproduce, or the R number. In a document drafted for Stormont last September and published in October, the North’s health authorities estimated with “medium confidence” that the operation of the hospitality sector is responsible for 0.1-0.2 of the R number.

If you accept that the virus does not recognise borders, then it must behave the same way in the Republic as it does in the North. It then follows that if hospitality is responsible for only 0.1-0.2 of an R number that may have surged here to over 2, it is reasonable to question if it is the only driver.

UK studies put the impact of the more transmissible new variant of the virus at 0.4-0.7 of the R number, which is a multiple of the risk associated with hospitality. Yet this week, almost four weeks into this dreadful surge, it is clear that the Irish authorities still have no clear handle on its impact. Tiny, unrepresentative samples conducted here have detected it in up to 25 per cent of cases.

Dr Tony Holohan, the chief medical officer, this week wrote to the Government to say that further analysis will be done in coming weeks to get more information on the extent of the UK variant. But if we had been doing the same kind of genome sequencing as the British, its impact would have been understood here earlier and the debate over what has happened would be better informed.

Why have the HSE or the HSPC never conducted in-depth, empirical studies of the infection rates of workers in the sectors of the economy that are considered riskiest to open, such as retail and hospitality? Surely the availability of this kind of granular data would have improved the debates around the effectiveness of economic restrictions at various reopening junctures.

Hindsight heroes

Why were regional public health specialists stripped of investigative resources over the summer, when they could have helped to find out the worst sources of infection when numbers were low? This information would have helped to inform us in the winter about the most effective restrictions to implement.

No matter what those now drenched in hindsight claim on social media, nobody predicted that January cases of the virus would top 7,000 or 8,000 per day. Those who claim it was predictable should produce evidence of any contemporaneous predictions. Those predictions don’t exist.

The fact that even the State’s public health authorities did not come close to predicting the correct infection rates suggests there may be a few gaps in Nphet’s modelling and data collection. Public health officials may argue that people were prepared to go crazier over Christmas than they were expecting. So why, then, did the public research conducted on behalf of Nphet’s behavioural scientists not pick up this mood ahead of time?

None of us should be afraid to admit what we do not know, or what we may have misjudged.