Do we Value Life?

23 January 2021

by David Cook PhD

In 2017 the National Institute for Health and Care Excellence (NICE) rejected the drug nivolumab for use in the NHS to treat patients with advanced head and neck cancers. The reason given was that, despite the drug showing positive benefits, it was judged to be too expensive based on the cost per ‘quality adjusted life year’ (QALY). For patients with this disease (and clinicians treating them) this was a hugely disappointing decision and although subsequently nivolumab has been approved for use, at the point of this judgement it must have felt to these patients that their lives were somehow being deemed to be less valuable than those of other patients.

Let’s wind forward to today and Lord Sumption discussing the impact of lockdown on society and apparently suggesting something similar, namely, that some lives are less valuable than others.

But in both of these cases is this what was actually meant? Are we really assigning a value to a life? Are we really judging that some lives are more valuable than others and so more worthy of saving?

To answer these questions, let’s focus on QALYs because these seem to be highly culpable in the crime of ‘life valuation’.

Quality Adjusted Life Years (QALYs) are not used to assess the quality of a life and they are certainly not used to make a judgement on its value.

The reason for this is because QALYs are used to assess the impact and value of an intervention. The judgement as to the quality of someone’s life is something that only the individual can make, but regardless of how they feel about it as a whole, they would certainly be able to tell if it had improved or got worse after some kind of treatment. If I whack you on the hand with a ruler has this improved your quality of life? What if I now kiss it better?

This is the fundamental point, QALYs are always used comparatively: did this treatment or intervention improve or reduce the quality of life?

In assessing the value of new therapies, QALYs are used to try and produce an objective view of their (hopefully positive) impact. A good example of the challenges of this kind of assessment and why QALYs are so helpful is if we think about how we would assess the value of a new analgesic or pain treatment. Such a treatment may have no effect on life expectancy and so its whole impact is on quality of life. But how do you assess this impact when pain is such a personal experience? The only way is to actually ask the individual patient. As a result, a major part of the assessment of the benefit of such medicines is done through use of questionnaires and asking how the individual feels; did the treatment improve your quality of life? Then, by aggregating all of these individual responses together, we can start to assess whether overall the treatment was beneficial or not. You can see that at no point are we making a judgement of the quality or value of the patient’s life. The assessment we are making is of the value of the treatment.

In practice, QALYs usually score the quality of life between zero and one, with one being full health and zero death. For everything else in between, we look to the individual to help us define this number.

Quality of life is not the only feature of QALYs, time of effect is also important (the ‘Y’ is years after all). You might think it obvious that extending life is always a good thing, but this might not be the case if the quality of life is substantially reduced. This is why some patients decide to discontinue or refuse life-extending treatment and why consent for treatment is so important. It is also why both time and quality are features of QALYs – a treatment producing a shorter time of high-quality life may generate more QALYs, and so be more valuable, than one producing a longer time of lower-quality life.

The fact that time is a feature of QALYs also means, that if all other things are equal, then improvements in quality of life for patients nearing the end of their lives produce lower amounts of treatment value than equivalent improvements in quality of life for patients with more years ahead of them. This is categorically NOT saying that individuals towards the end of their lives are less valuable than those who are not. It just recognises that some people have longer to live than others. However, it is easy to see how a looser use of language can easily lead to the misunderstanding that the value we are discussing is of the life itself.

So, let’s have a look at lockdowns from a QALY perspective.

Lockdowns do not treat COVID-19 in any medical sense, their entire purpose is to avoid individuals catching coronavirus and developing the disease. This means gains in QALYs for lockdowns are actually due to avoiding the losses associated with contracting COVID-19. It is therefore extremely difficult to estimate these gains because if lockdowns do have a significant impact on reducing the number of people who catch COVID-19 then it means counting things that don’t happen!

This is always the challenge in preventative measures – the best outcome is that nothing untoward happens. So, how can we possibly assess the positive impact of lockdown? This is, where the modellers and their predictions come in. We estimate the value of the lockdown treatment as avoiding the predicted deaths of doing nothing or doing less. In a very real sense, we are impacting on real lives to avoid virtual deaths.

It is also where an understanding of the demographics of the disease has influence and where we risk being misunderstood using this kind of approach to estimate the value of lockdowns. This is because COVID-19 has a disproportionately more significant impact on older and sicker people and so, the value of lockdown as an intervention is more limited than if COVID-19 had the same impacts across the entire population. This is simply because saving an older life generates fewer QALYs than a younger one. It is not because older lives are less worth saving than younger ones.

David Miles and his colleagues have done an extensive analysis of the economic value of lockdown using QALYs. In this they clearly demonstrate that even in the most extreme modelling scenarios, where the number of COVID-19 deaths avoided is at its maximum, based on the normal NICE criteria lockdown is not cost effective: if lockdown had been nivolumab then it would have been rejected as a treatment. The reason for this is, in part, due to the age of those who benefit limiting the number of QALYs generated. This is an excellent piece of analysis and so I’m not going to reiterate it here. Instead, I’m going to park the costs and focus on thinking about the broader impacts of lockdowns purely from a quality of life perspective.

Unlike the predicted benefits of lockdown on QALYs, the reduction in QALYs to the population under lockdown are here and now, real and not computer simulated, and, if we so desired, to an extent measurable. This is for the simple reason that lockdown requires societal changes to be effective and these changes have negative consequences. To steal some drug terminology; there is no margin between the beneficial effects and the negative (side) effects; there is no safe dose of lockdown.

Because of the very nature of lockdown, everyone’s quality of life will be reduced to some extent due to the loss of normal social activities. The loss of these things on an individual level is minimal compared to the impact of a severe case of COVID-19, and so they only produce a more modest reduction in the quality of life. The problem is that a lot of people suffer this reduction and so those small losses soon add up a large number of QALYs.

Some might classify such losses as ‘trivial’, but it doesn’t mean we shouldn’t acknowledge them. Of course, to understand them properly we would need to ask everyone (or a good representative sample) to assess the impact of lockdown on their individual quality of life. If we were to do this then I suspect we would find that for many people the reduction in their quality of life is far from trivial, particularly for groups like students, or those who have seen their livelihoods destroyed.

Unfortunately, as is being documented more and more frequently, lockdown also produces reductions in QALYs that are due to significant decreases in health, both mental and physical. There is no doubt about whether lockdown causes these effects (although they are sometimes misattributed to being due to COVID-19, rather than our response to it), the question is how large will the losses of QALYs be? Have we potentially swapped one health crisis for another? Sadly, we will have years to count the costs and find out.

Finally, there is one group of individuals who are worth focusing on, and that is those at the forefront of treating COVID-19 patients: doctors, nurses and carers. From a quality of life perspective, how does lockdown affect this group? On one hand, a reduction in COVID-19 patients is clearly beneficial as it will reduce the stress and anxiety of dealing with large numbers of sick individuals. In fact, reducing the burden on healthcare professionals is one of the main rationales behind lockdown. However, these individuals are also subjected to the same lockdown restrictions as everyone else, which potentially removes a large number of normal coping mechanisms used to deal with the stresses of being on the frontline; a good moan about work down the pub, a trip to the cinema for a bit of escapism, visiting some friends on the well-earned day off etc. etc. etc. A paper by Kim and colleagues highlights this: although nurses found treating patients with COVID-19 stressful and it created a level of anxiety, it was also a positive experience. In contrast, quarantining due to COVID-19 restrictions was associated with moderate/severe depression. The very mechanisms designed to ease the burden of COVID-19 on health workers, could simultaneously make their quality of life worse.

It can feel very cold and calculating, and even somehow inhuman, to talk about lives and deaths using things like QALYs. But the whole point of doing this is precisely to treat all lives as equal and, in doing so, make the best decisions on interventions for patients as a whole without any judgement as to whether one life has more value than another. Resources within any healthcare system, be they financial or human (doctors and nurses), are always finite. Without adopting approaches similar to this, how else should we distribute these in a way that is fair to everyone?

Coming back to nivolumab, the decision to not initially approve was because approval would have meant spending a disproportionate amount of a limited budget on one patient group, to the disadvantage other patients. It may be of very little comfort to those affected by the decision, but it is precisely because there was no judgement as to whether one life was worth more than another that the decision was what it was.

SARS-CoV-2 is a significant new human pathogen and COVID-19 a potentially fatal disease. The aim of effective public health measures in the face of the pandemic must therefore be to reduce the loss of QALYs across the population. The ‘sweet spot’ for the best intervention is probably one with some societal restrictions but well short of full lockdown. This would blunt the impact of COVID-19 but without the associated loss of QALYs due severe societal restrictions and so minimise the loss of QALYs to the population as a whole. But, as Miles et al. point out, doing this probably means accepting a greater number of COVID-19 related deaths.

It is of course this final point which is where the issue lies. With such a single-minded focus on COVID-19 infections and deaths, any policy which doesn’t prioritise this is, in the current climate, politically untenable. Yet, ironically, by raising one group of patients or one disease to special status, we step over the very line we accuse Lord Sumption of crossing, and we stop thinking about the value of the intervention and start valuing some lives more than others. Something to bear in mind when we consider our response to COVID-19.

Dr. David Cook is senior scientist with over 20 years experience in drug research and development.

Disclaimer
This article represents the opinion of the author and does not represent the opinions of any business or entity for which the author works, has worked or is affiliated to.

Social media & sharing icons powered by UltimatelySocial