spinofflive
Toby Morris for The Spinoff
Toby Morris for The Spinoff

ScienceOctober 22, 2020

Siouxsie Wiles & Toby Morris: Covid-19 and the Swiss cheese system

Toby Morris for The Spinoff
Toby Morris for The Spinoff

The effort to defeat the coronavirus relies on many layers of defence. Or, let’s call them, slices of cheese.

One thing that has really irked me during Covid-19 has been the labelling of any mistake in our systems or any unexpected transmission of the virus as a failure. Yes, I understand that the media and the opposition have a responsibility to hold the government and our public services to account, but screaming FAILURE from the rooftops has everyone looking for someone to blame and massively undermines public trust in New Zealand’s pandemic response. It’s also completely out of all proportion. If you want to see failure, look to countries like the US or the UK. Identifying mistakes or holes in our response is really important as it allows us to learn and adapt our processes, something that has been key to New Zealand’s success in managing Covid-19 so far.

Twenty years ago, James Reason, a professor of psychology at the University of Manchester in the UK, published a paper in the British Medical Journal in which he described what he called the “Swiss cheese model of system accidents”. Reason was trying to move people’s thinking from treating mistakes as individual errors by “bad” people to a systems approach that accepts that humans are fallible and mistakes are to be expected. Rather than blaming individuals for failure, we should try to understand how and why the failure happened to prevent it from happening again.

But what has Swiss cheese got to do with all this?

The idea behind the systems approach is to build in layers of barriers and safeguards. In an ideal world, each of these defensive layers would be impenetrable. But in the real world, they aren’t. So Reason likened each layer to a slice of Swiss cheese – it has holes in it. To be fair to the Swiss, they have lots of different cheeses, many of which don’t have holes, so it’s probably more accurate to call Reason’s model the “Emmental Model”.

Back to the holes. They can happen for many reasons. Slips, fumbles, mistakes. People not following procedure. Or they can even be built into the system. A hole in one layer of defence isn’t necessarily a disaster if there are lots of other layers to fall back on. Then its only when all the holes line up that the defences are breached. Reason’s model is now used all over the place including in medicine, engineering, and the aviation industry.

Applying the Swiss Cheese/Emmental Model to Covid-19

I’ve seen several excellent applications of the Swiss Cheese/Emmental Model to Covid-19, like this one by virologist Ian Mackay or this one by sketchplanations. They represent each of the different public health interventions we have for Covid-19 as layers of cheese. I love how they show that we actually have quite a few defences against Covid-19 but that none of them is impenetrable.

We have the use of border controls and putting travellers into isolation/quarantine to stop the importation of cases into a country or between regions. We also have the crucial package of rapid testing, contact tracing, and isolation of infectious people and their close contacts. This test-trace-isolate strategy can be used very effectively to stop the spread of the virus. Then there’s limitations on gathering sizes or the movement of people, or on high-risk activities or places.

We also have all the things we can do as individuals. Things like physical distancing, washing our hands, practising safe cough and sneeze hygiene, wearing a mask, using contact tracing apps to keep a track of where we’ve been and who with, staying home when sick, and getting tested if we have symptoms. And, finally, there’s cleaning and ventilation. Hopefully we’ll be adding vaccination next year.

New Zealand’s swiss cheese/Emmental model for managing Covid-19

Around the world, countries are applying different layers of cheese depending on the strategy they are following to deal with the pandemic. Herd immunity? What cheese? Flatten the curve? Some slices. Elimination? All the cheese!

Here in New Zealand, at the beginning of the pandemic we applied border controls as our first slice of cheese, along with physical distancing, handwashing, and cleaning. It was soon clear these weren’t going to be enough and the virus was way ahead of us. That’s when the alert levels were brought in, a framework of escalating restrictions on movement and activities.

A good way of thinking about the different alert levels is as slices of cheese with fewer and fewer holes as you move from level one to level four. We soon applied the slice with the fewest holes, backed up by managed isolation and quarantine of travellers as well as physical distancing, handwashing, and cleaning. That minimised importation of the virus into the country, and massively restricted community transmission of the virus. It also bought us time to ramp up the country’s testing and contact tracing capabilities, including the development of an app.

Applying all those slices of cheese got us to elimination. But then one by one we started stripping back the layers as we became more and more complacent. Until all that was really left were the border controls and managed isolation and quarantine. Those slices have worked fantastically well, but we know they aren’t impenetrable.

Fast forward to August, when someone in the community with no links to the border tested positive for Covid-19. Quickly many of the slices of cheese we’d used before were applied, as well as a new one, masks. And they worked, again. We’re back to no cases of Covid-19 in the community. It’s looking like we’ve eliminated the virus again.

Elimination is something we have to keep working at

As long as people continue arriving in New Zealand from countries with widespread community transmission of Covid-19, we’re going to need to keep working at elimination. Because our systems are so good now, we are able to track rare transmission events in ways other countries can only dream of. We’ve had what seems to be transmission in managed isolation from a bin lid and a lift button. Then we’ve just had the port worker who picked up the virus despite wearing PPE.

And now we’ve had several new arrivals in Christchurch test positive in managed isolation at their routine day three test. They are from a large group of workers from Russia and Ukraine who have recently flown to New Zealand on a charter flight to take up jobs on fishing boats. They all tested negative before flying. The first thing this highlights, yet again, is that just because people test negative before they board a flight doesn’t mean they aren’t a risk. They could have taken an unreliable test, or could be incubating the virus, or could have become infected on the journey. It’s just more evidence for why our border controls and managed isolation and quarantine facilities need to stay in place. And why we need to keep using all the slices of Emmental at our disposal.

No room for complacency

I’m noticing that complacency is starting to creep back in again. The number of people using the contact tracing app is dropping, just like it did before our second outbreak in August.

At the very least, we must keep up with washing our hands, practising safe cough and sneeze hygiene, wearing a mask, using the contact tracing app, staying home when sick, and getting tested if we have symptoms. That way, if the virus does get through our managed isolation and quarantine system, we’re far less likely to need to move up the alert levels to stamp it out again. Instead we can use testing, contact tracing, and isolation.

We’re in this for the long haul, and it’s up to all of us to play our part.

Keep going!
Getty Images
Getty Images

OPINIONScienceOctober 15, 2020

Cannabis and psychosis: a referendum red herring

Getty Images
Getty Images

The connection between cannabis use and psychosis is far more murky than asserted by yesterday’s article on The Spinoff, writes Suresh Muthukumaraswamy, an expert in how psychoactive drugs modify brain and behaviour.

Yesterday, separate articles appeared on The Spinoff and on Newsroom (written by a colleague of mine) that raised concerns about potential links between cannabis and psychosis in the context of the current cannabis referendum. However, in various places evidence presented in these articles is misinterpreted, mis-stated or not properly contextualised within the referendum legislation as written. So let’s dig into the details.

It has often been contended that cannabis is associated with an increased risk of psychosis in youth users. While statistics such as “double the risk” or even “eight times the risk” appear scary, they fail to take into account low base rate risk of psychosis. A better way to present these statistics is as the Number Needed to Prevent – this is how many young people we would need to stop smoking cannabis to prevent one psychosis case emerging. Using UK data (we lack NZ data on psychosis prevalence in youth) this has been estimated as 1360 in heavily using young men and 2480 in heavily using young women. So we would need to stop about 1700 heavy cannabis users to prevent one additional psychosis case per year. Let’s be super-cautious and round this down to 1000. That would mean, if an extra 1000 young people started smoking cannabis heavily we would get one extra psychosis case.

In New Zealand there are around 500,000 young people (age 15-24) and it has been estimated 10% smoke cannabis heavily (around 50,000). So, after legalisation it would take around a 2% increase in cannabis use in this demographic to develop a single extra psychosis case. Would we see such rises? The best evidence comes from Canada which has a comparable regulatory scheme. If anything in Canada they have seen slight decreases in youth consumption. In summary, neither prohibition nor legalisation alone are effective interventions to prevent cannabis-induced psychosis because the Number Needed to Treat values are so large.

It is important to stress that no causal relationship has been established between cannabis and psychosis although numerous studies point to elevated risk. It is important to read papers carefully. Let’s take a study cited in yesterday’s Spinoff article. Deep in the text it is stated that “After psychotic symptoms at age 11 were controlled for, the risk for adult schizophreniform disorder remained higher among those who used cannabis at age 15; however, this risk was reduced by 31% and was no longer significant”. So after controlling for early psychosis symptoms before cannabis use was initiated there was no relationship between cannabis use and psychosis. It is also important to bear in mind the entire result depends on a sample size of just three cases in the category of interest – just three. Unfortunately, it is easy for even experts to misread medical statistics and sometimes authors of papers can over-state the case.

In several balanced review articles (here and here), Ian Hamilton – a UK academic specialising in the relationship between substance use and mental health – outlines the many difficulties in trying to estimate potential contributions of cannabis to psychosis development. Particularly problematic in these observational studies is “residual confounding”. Here scientists try to correct the observed relationship between psychosis and cannabis use for other variables such as age, gender, and other drug use – particularly tobacco. Typically, the strength of the cannabis-psychosis relationship decreases when corrected for these factors. Because our estimates of the confounds themselves can be vary noisy (for example, lifetime tobacco use is very hard to measure) some of the confound’s effect will remain in the data – residual confounding. Suffice to say, the fact that scientists have been studying this topic for nearly 40 years with much scrutiny does suggest that overall “at a population level the increased risk is weak and the vulnerabilities relatively rare”.

Although the overall risk of cannabis-induced psychosis is low, risk factors that have been observed linking psychosis and cannabis appear to be the strength of cannabis in terms of THC content, age of use and frequency of use. New Zealand’s Cannabis Legalisation and Control Bill attempts to ameliorate these by having caps on strength (15% THC maximum) and an age limit of 20 – which is a practical compromise from a harm reduction perspective. Perhaps the only thing missing I can see in the bill from a psychopharmacologist’s perspective is no mandate for cannabidiol (CBD) to be present in sold products. Although emerging, best evidence suggests CBD can ameliorate negative effects of strong THC products. Currently, the bill proposes extra taxes on  THC content – offsetting these taxes with CBD incentives might be one way to encourage the re-introduction of CBD into recreational cannabis from which it has been bred out by illicit growers.

Recreational users currently source drugs through an unregulated market that is only too happy to sell users more harmful drugs, such as synthetic cannabinoids and methamphetamine. The risk of psychosis with these drugs is far higher than with cannabis with a 15% THC limit cap as proposed. Not only that, since 2014 synthetics have killed over 70 Kiwis. From the disturbing toxicology and pathology resports recently published by ESR and Auckland City Hospital, the physical and mental health damages caused by synthetics drugs in non-fatal cases will likely be extensive. Not only that, we are likely to see THC levels in recreational cannabis continue to rise under black-market conditions with no ameliorating CBD content.

During the course of the referendum debate, I have observed in various commentaries a tendency to get overseas experts to weigh in on the debate. But it is entirely unclear whether these experts are even familiar with the proposed legislation and the particular demography of drug use that New Zealand faces. I would be curious to see how these experts would answer if they were asked to write legislation to reduce the potential mental health harms from a legalised cannabis regime – what features would it include? My guess is that it would end up looking pretty similar to the bill as it is written. When a referendum was announced I was on the fence. After reading the content of the bill, I moved into the “yes” column because it ticked all the public health boxes.

It may seem odd to say this, but the Cannabis Legalisation and Control Bill is not at all inconsistent with the desire to see New Zealand becoming Smokefree by 2025. The bill explicitly forbids selling of tobacco and alcohol at licensed vendors (Section 193) and similarly from consumption venues (Section 197). From a harm reduction perspective, one would hope that tax revenue could be used to subsidise cannabis-leaf vaporisers – which allow cannabinoid vapours to be inhaled avoiding concurrent tobacco use and without damaging tars and carcinogens entering the lungs. This represents the most healthy and precise way to dose cannabis. It allows users to accurately calibrate the doses they take. With product labelling on potency, users would be enabled to find their limits – avoiding excessive consumption. Also note that we will not see undercutting from foreign imports as Section 41 of the bill prohibits imports and exports.

Much of this perspective has been about risk. Risk is always relative. Did you know every year 1/100 rugby players in New Zealand will make an ACC claim for concussion/brain injury? That represents a far greater brain risk than cannabis. As a society we definitely are not going to ban rugby – and thankfully so! Instead we take a mature harm reduction approach, for example punishing dangerous tackles or implementing concussion protocols.

At this late stage in the process polls suggest that it is quite possible that the Cannabis Legalisation and Control Bill will not make it across the line. This may partly be due to the complexity of epidemiological data and legislation that have not always been accurately represented in public discourse. I hope this article helps those who are yet to vote see that passing the referendum would help improve the overall health outcomes from the use of cannabis – and that it’s unlikely to impact levels of cannabis-induced psychosis in our community.

Suresh Muthukumaraswamy is an Associate Professor in Psychopharmacology at The University of Auckland. His research and teaching speciality is on how psychoactive drugs modify brain and behaviour.

Read more: Everything you need to know about the 2020 cannabis referendum