Monday, May 4, 2020

NZ's Covid-19 Truth is Hiding Behind Numbers

Science and Sensibility

Long ago, in my post graduate studies and my London based operations research working career I did a lot of computer modelling of scientific systems, and built computer models. The study of Pandemics relies heavily on computer modelling. I've done some reading and produced this piece which is a contribution to New Zealand's Covid-19 public policy debate....



Each day since the start of our Covid-19 pandemic New Zealanders are given the day’s numbers at the top of the news – how many new cases, and how many deaths and recoveries. We’re also advised how many cases are in hospital and how many are critically ill. Increasingly we get information about how much has been spent subsidising workers wages, supporting businesses, and what’s needed to support the economy.

Counting and numbers is comforting science. It can be a way of confronting an onslaught as we gradually get our heads around what’s happening, and at the same time bundling it away as if it’s all under control. But what exactly are we being told each time the latest figures are announced, rising consistently, dropping slightly, increasing again? As we do what we’re told, and we wait.

Scientific analysis of an epidemic typically divides the population into three groups of people: Susceptible, Infectious and Recovered. It’s a useful way for us all to understand what’s happening. Computer models of how a pandemic progresses through a population calculate, one day at a time, how many to subtract from the Susceptible and add to the Infectious, and how many to add to the Recovered. How many die each day is also predicted. Assumptions include the average number of contacts each person has per day, the probability contact with an infected person results in infection, what proportion of the Infectious population recover, and what proportion die.

Significantly, the model shows how the Susceptible population reduces over time as the pandemic works its way through the population because people who have Recovered are assumed to be immune (the jury is still out on that assumption with Covid-19). The chance of someone getting infected after social contact drops over time because fewer people are susceptible. Most models assume that as more of the population become immune through having been infected and recovered, major outbreaks happen less and less frequently then stop, and the pandemic is said to have run its course.

Modelling is an essential tool in forming policy to guide national and global responses to a threat like Covid-19. Fine tuning of pandemic model assumptions has been enabled by data from the direct experience of frontline health workers in Wuhan, Italy and New York.

The teams that have been advising the UK government have published descriptions of their models. These are typically structured as above, but provide for more stages in the progression of an infection, divide the population into age groups and geographic areas, and model the capacity of the health system to cope – including the number of intensive care units available (ICU).

Such models have played a critical role in shaping the UK government’s response which has shifted from mitigation to suppression since the pandemic started. The London School of Hygiene and Tropical Medicine (LSHTM) model predicted that, if nothing was done to mitigate the effects of the epidemic in the UK, 85 per cent of the population would be infected, there would be 24 million clinical cases and 370,000 deaths. At its peak, 220,000 ICU beds would be required whereas only 4122 were available.

All of the UK models examined options for mitigation including case isolation, voluntary home quarantine, social distancing of the over-seventies, social distancing for everyone, and school and university closures. But modelling indicated mitigation would still lead to more than 200,000 deaths. UK researchers generally argued the only alternative options to mitigation were suppression strategies, partly because China showed they were possible, and as a reaction against the huge death toll in Italy.

The LSHTM group modelled the use of repeated lockdowns, each triggered when the number of ICU beds occupied by Covid-19 patients reached a particular number. If the threshold was set at a thousand ICU beds, the number of infections could be kept to four million and the number of deaths to 51,000. The downside of the strategy would be that 73 per cent of the time between now and December 2021 would be spent in lockdown, by which time only 11 million people would have been infected and, unless a vaccine had been found, the epidemic would still be far from over.

So far in New Zealand the number that we are not talking about is the size of our Susceptible population. That’s because it’s just about all of us, and instead we talk about elimination or eradication of the virus. However, the Covid-19 virus is alive and well all round the world and despite the best disinfection and sanitising under the sun, it is certainly present in New Zealand. Those of us who haven’t been infected yet are susceptible to infection at any moment – though those moments are minimised while we are in lockdown – and it is assumed that testing and tracing will deal with the Covid-19 viruses already in our environment.

There has been some discussion in New Zealand of second and third waves of infections (as happened in the Spanish Flu epidemic 100 years ago) which are generally though to have been caused by mutants versions of the original virus. That is not the same as recognising that modelled suppression strategies being implemented in the UK and elsewhere show the need to provide for repeated lockdowns for the rest of this year and 2021 at least. Way before then in New Zealand, based on the present rate of Government subsidy spending and the effects of lockdowns on economic activity, the treasury cupboards will be well and truly bare. Before then very difficult public policy decisions will need to be made, which go well beyond the simplistic and comforting numbers we currently see on TV.

We need to have conversations about the cost of a death, and about social inequality.

More than 32,000 people died in New Zealand last year. Of these 353 were due to motor vehicle accidents and 685 were suicides. The largest cause of death - about 30% - were cancer deaths, 15% were from heart disease, and 7% due to brain conditions like stroke.

Of the 3,912 people in the UK who died of COVID-19 in March, 91% had at least one pre-existing condition (most common was heart disease), and on average they had 2.7. Many of those who died would likely have died from pre-existing conditions. This isn’t to say those deaths don’t matter, or to forget that although significantly fewer young people have died, they too have died in numbers that in normal times would be shocking.

We can always live in hope for a vaccine, but we also need to be prepared as a country to expect a very long wait. It is easy to say that “we” are all in this together, but already the gap between the have’s and the have not’s is a chasm. It is relatively easy to ride out long lockdown periods in a pandemic with a big back garden, protected income, and substantial retirement savings. Much harder if you’re unskilled, got a young family, aged parents in care, and a mortgage.

Deaths attributable to the economic consequences of lockdowns must also be anticipated. People will vary enormously in their reactions to being confined in their homes, without regular work. Domestic violence incidents are on the rise here, and US news reports of kilometre long queues for food parcels and people openly carrying guns are frequent.

Our Government says it has “provisioned $52 billion, if necessary, to use for cushioning New Zealanders against the impacts of the virus, positioning New Zealand for recovery, and helping us to reset and rebuild our economy to support long-term recovery.” This would be an enormous debt for the country to carry.

Our death-count due to Covid-19 is now 20. How many deaths have been avoided so far is difficult to know, but we are beginning to understand what the costs of the Government’s suppression strategy are and will be, to us and for the country.

The cost of a life is routinely used in justifying the cost of a new road in New Zealand. It’s set at somewhere between $1,000,000 and $2,000,000. If a new road can avoid enough fatal motor vehicle accidents public money gets invested in the road. The benefit in lives saved exceeds the cost of the new road. Who knows what investment might reduce youth suicide, or prolong the lives of those suffering from cancer?

Part of our preparation for the rest of 2020 and 2021 will require public conversations about how the suffering caused by the pandemic should be allocated and shared, and about the very future of our society.

** I have drawn from an article by Paul Taylor, titled "Susceptible, Infectious, Recovered" contained in the 7th May issue of London Review of Books.

2 comments:

Unknown said...

When it's a big problem then the numbers can be big, overwhelmingly big. Unfortunately, Statistics are by nature impersonal. If analysis is not objective, then rational decision-making will go off the rails.
However, if we had, say, a couple of weeks of deaths every day, then had a week of just one, we'd probably say "well done". But that is little comfort to the Family of that one.
..
At some stage (soon) decisions will be made on a realistic body count versus the cost/savings. As you say, it's used in Road safety matters, probably in the purchase of Medical equipment .. ie, what price a Life?
I just hope it's not a decision with 'Anzac Cove' consequences.
But decisions DO have to be made.
..
I have concerns that these decisions need to be made in an atmosphere of ignorance. So much is unknown, or being assumed about this disease. My personal opinion is that the only group that is NOT susceptible is those that have already died.
..
In the meantime, stay safe, stay at Home if you are able.
If you need a demonstration of what breaking your Bubble does, and/or inadequate hygiene just Google a video of "mousetraps/ping-pong balls".
..
PS Reporting of the numbers is NOT a remedy, but merely the Documentation of the Event.

Anonymous said...

I think the issue that will eventually bubble to the surface will be the unspoken human suffering of those cooped up in a bubble with domestic violence, dealing with suicide and mental stress issues that will be the ultimate byproduct of the personal and family economic devastation that has yet to come. An intangible toll that will eclipse the 20 people that have died WITH the virus to date.

Monday, May 4, 2020

NZ's Covid-19 Truth is Hiding Behind Numbers

Science and Sensibility

Long ago, in my post graduate studies and my London based operations research working career I did a lot of computer modelling of scientific systems, and built computer models. The study of Pandemics relies heavily on computer modelling. I've done some reading and produced this piece which is a contribution to New Zealand's Covid-19 public policy debate....



Each day since the start of our Covid-19 pandemic New Zealanders are given the day’s numbers at the top of the news – how many new cases, and how many deaths and recoveries. We’re also advised how many cases are in hospital and how many are critically ill. Increasingly we get information about how much has been spent subsidising workers wages, supporting businesses, and what’s needed to support the economy.

Counting and numbers is comforting science. It can be a way of confronting an onslaught as we gradually get our heads around what’s happening, and at the same time bundling it away as if it’s all under control. But what exactly are we being told each time the latest figures are announced, rising consistently, dropping slightly, increasing again? As we do what we’re told, and we wait.

Scientific analysis of an epidemic typically divides the population into three groups of people: Susceptible, Infectious and Recovered. It’s a useful way for us all to understand what’s happening. Computer models of how a pandemic progresses through a population calculate, one day at a time, how many to subtract from the Susceptible and add to the Infectious, and how many to add to the Recovered. How many die each day is also predicted. Assumptions include the average number of contacts each person has per day, the probability contact with an infected person results in infection, what proportion of the Infectious population recover, and what proportion die.

Significantly, the model shows how the Susceptible population reduces over time as the pandemic works its way through the population because people who have Recovered are assumed to be immune (the jury is still out on that assumption with Covid-19). The chance of someone getting infected after social contact drops over time because fewer people are susceptible. Most models assume that as more of the population become immune through having been infected and recovered, major outbreaks happen less and less frequently then stop, and the pandemic is said to have run its course.

Modelling is an essential tool in forming policy to guide national and global responses to a threat like Covid-19. Fine tuning of pandemic model assumptions has been enabled by data from the direct experience of frontline health workers in Wuhan, Italy and New York.

The teams that have been advising the UK government have published descriptions of their models. These are typically structured as above, but provide for more stages in the progression of an infection, divide the population into age groups and geographic areas, and model the capacity of the health system to cope – including the number of intensive care units available (ICU).

Such models have played a critical role in shaping the UK government’s response which has shifted from mitigation to suppression since the pandemic started. The London School of Hygiene and Tropical Medicine (LSHTM) model predicted that, if nothing was done to mitigate the effects of the epidemic in the UK, 85 per cent of the population would be infected, there would be 24 million clinical cases and 370,000 deaths. At its peak, 220,000 ICU beds would be required whereas only 4122 were available.

All of the UK models examined options for mitigation including case isolation, voluntary home quarantine, social distancing of the over-seventies, social distancing for everyone, and school and university closures. But modelling indicated mitigation would still lead to more than 200,000 deaths. UK researchers generally argued the only alternative options to mitigation were suppression strategies, partly because China showed they were possible, and as a reaction against the huge death toll in Italy.

The LSHTM group modelled the use of repeated lockdowns, each triggered when the number of ICU beds occupied by Covid-19 patients reached a particular number. If the threshold was set at a thousand ICU beds, the number of infections could be kept to four million and the number of deaths to 51,000. The downside of the strategy would be that 73 per cent of the time between now and December 2021 would be spent in lockdown, by which time only 11 million people would have been infected and, unless a vaccine had been found, the epidemic would still be far from over.

So far in New Zealand the number that we are not talking about is the size of our Susceptible population. That’s because it’s just about all of us, and instead we talk about elimination or eradication of the virus. However, the Covid-19 virus is alive and well all round the world and despite the best disinfection and sanitising under the sun, it is certainly present in New Zealand. Those of us who haven’t been infected yet are susceptible to infection at any moment – though those moments are minimised while we are in lockdown – and it is assumed that testing and tracing will deal with the Covid-19 viruses already in our environment.

There has been some discussion in New Zealand of second and third waves of infections (as happened in the Spanish Flu epidemic 100 years ago) which are generally though to have been caused by mutants versions of the original virus. That is not the same as recognising that modelled suppression strategies being implemented in the UK and elsewhere show the need to provide for repeated lockdowns for the rest of this year and 2021 at least. Way before then in New Zealand, based on the present rate of Government subsidy spending and the effects of lockdowns on economic activity, the treasury cupboards will be well and truly bare. Before then very difficult public policy decisions will need to be made, which go well beyond the simplistic and comforting numbers we currently see on TV.

We need to have conversations about the cost of a death, and about social inequality.

More than 32,000 people died in New Zealand last year. Of these 353 were due to motor vehicle accidents and 685 were suicides. The largest cause of death - about 30% - were cancer deaths, 15% were from heart disease, and 7% due to brain conditions like stroke.

Of the 3,912 people in the UK who died of COVID-19 in March, 91% had at least one pre-existing condition (most common was heart disease), and on average they had 2.7. Many of those who died would likely have died from pre-existing conditions. This isn’t to say those deaths don’t matter, or to forget that although significantly fewer young people have died, they too have died in numbers that in normal times would be shocking.

We can always live in hope for a vaccine, but we also need to be prepared as a country to expect a very long wait. It is easy to say that “we” are all in this together, but already the gap between the have’s and the have not’s is a chasm. It is relatively easy to ride out long lockdown periods in a pandemic with a big back garden, protected income, and substantial retirement savings. Much harder if you’re unskilled, got a young family, aged parents in care, and a mortgage.

Deaths attributable to the economic consequences of lockdowns must also be anticipated. People will vary enormously in their reactions to being confined in their homes, without regular work. Domestic violence incidents are on the rise here, and US news reports of kilometre long queues for food parcels and people openly carrying guns are frequent.

Our Government says it has “provisioned $52 billion, if necessary, to use for cushioning New Zealanders against the impacts of the virus, positioning New Zealand for recovery, and helping us to reset and rebuild our economy to support long-term recovery.” This would be an enormous debt for the country to carry.

Our death-count due to Covid-19 is now 20. How many deaths have been avoided so far is difficult to know, but we are beginning to understand what the costs of the Government’s suppression strategy are and will be, to us and for the country.

The cost of a life is routinely used in justifying the cost of a new road in New Zealand. It’s set at somewhere between $1,000,000 and $2,000,000. If a new road can avoid enough fatal motor vehicle accidents public money gets invested in the road. The benefit in lives saved exceeds the cost of the new road. Who knows what investment might reduce youth suicide, or prolong the lives of those suffering from cancer?

Part of our preparation for the rest of 2020 and 2021 will require public conversations about how the suffering caused by the pandemic should be allocated and shared, and about the very future of our society.

** I have drawn from an article by Paul Taylor, titled "Susceptible, Infectious, Recovered" contained in the 7th May issue of London Review of Books.

2 comments:

Unknown said...

When it's a big problem then the numbers can be big, overwhelmingly big. Unfortunately, Statistics are by nature impersonal. If analysis is not objective, then rational decision-making will go off the rails.
However, if we had, say, a couple of weeks of deaths every day, then had a week of just one, we'd probably say "well done". But that is little comfort to the Family of that one.
..
At some stage (soon) decisions will be made on a realistic body count versus the cost/savings. As you say, it's used in Road safety matters, probably in the purchase of Medical equipment .. ie, what price a Life?
I just hope it's not a decision with 'Anzac Cove' consequences.
But decisions DO have to be made.
..
I have concerns that these decisions need to be made in an atmosphere of ignorance. So much is unknown, or being assumed about this disease. My personal opinion is that the only group that is NOT susceptible is those that have already died.
..
In the meantime, stay safe, stay at Home if you are able.
If you need a demonstration of what breaking your Bubble does, and/or inadequate hygiene just Google a video of "mousetraps/ping-pong balls".
..
PS Reporting of the numbers is NOT a remedy, but merely the Documentation of the Event.

Anonymous said...

I think the issue that will eventually bubble to the surface will be the unspoken human suffering of those cooped up in a bubble with domestic violence, dealing with suicide and mental stress issues that will be the ultimate byproduct of the personal and family economic devastation that has yet to come. An intangible toll that will eclipse the 20 people that have died WITH the virus to date.