The ethics of doing nothing

Can we reasonably consider ‘doing nothing’ as an alternative course of action? In many cost-effectiveness analyses the intervention under consideration is compared against a ‘doing nothing’ scenario, although frequently the next best alternative is used. Ultimately the health technology assessment carried out by NICE is an informative effort and the final decision is made by the budget holder. However, NICE makes each assessment in isolation of each other and so prioritising treatments is left to the budget holder. But can the budget holder choose a ‘do nothing’ option, and should this option be considered at all in cost-effectiveness analysis?

This may come down to an issue on the role of the health care system in general. One of the principle tenets of NICE and the NHS is justice (the others being beneficence, non-maleficence, and autonomy). This NHS justice, it seems, is a sense of justice as described by John Rawls – justice as fairness. Justice as fairness is founded on two points – liberty and equality, that everyone should have the same right to basic liberties, and that inequalities should be arranged to benefit the worst off in society to ensure distributive justice. Both of these principles are satisfied by the idea of access to health care based on need and regardless of ability to pay.

We use cost-effectiveness analysis to best allocate resources, so that we all get the greatest gain for our limited resources, but that does not necessarily ensure that the worst off get priority.

In the end it comes down to a deontologism versus consequentialism debate. Deontologism dictates that there are certain moral rules that must be followed, or as Kant described them ‘categorical imperatives’, and these rules can be reached through logical reasoning and must be universal. In this case, for example, if doing nothing were universally permissible for health care professionals then it would be permissible for no-one to be treated which would negate the existence of the health care professional in the first place. So, if we say that all those with needs must be treated, this may be a deontological stance. However, we do not provide services for all those with needs, and it may be practically impossible to do so. Health care provision is proportional to need, but those with the least needs generally have to pay for their own services, unless they are sufficiently poor, for example, dentistry.

Now, if we consider health care provision to be philosophically consequentialist, can we allow a ‘do nothing’ option? Many thought experiments exist to exercise consequentialist ethics. Consider a runaway train, it’s careering down the track towards a station in which there are ten people who will die if the train gets there, you are on the train and have the option to switch tracks to divert the train away from the station. However, there are three men working on the line on the other track who will die if you pull the lever. Do you pull the lever? One argument, the utilitarian one, would say yes. The total loss would be smaller on the other track, we would therefore be maximising the total utility from the situation. Another argument may say though that not pulling the lever is the only option since if you did the deaths of the three men would be your responsibility but in doing nothing you would be morally neutral. This is a form of egoistic consequentialism. Under both these arguments a health care provider could do nothing, in the first case if utility was maximised by treating others and in the second case because the health care provider is not morally responsible for a person’s health care state in the first place.

There are objections to this line of reasoning. Peter Singer describes a situation to illustrate an objection to this. Imagine you are walking home one day. As you walk you pass a pond in which a child is drowning. The pond is not very deep and you could walk in and save the child, bearing no tangible risk to your own life. In this case the choice of inaction would lead to the child’s death, and you surely could be held responsible for that. The choice of doing nothing, then, does not negate responsibility. Moreover, if the budget holder is the government, there are certainly arguments which may attribute to them a certain responsibility for poor health in the population (consider the relationship between the macroeconomy and health).

The key issue that remains is opportunity cost. The only reasonable argument for doing nothing is that the time and resources could be better spent elsewhere, and cost-effectiveness analysis provides us with the information to know where it is best spent. However, in reality, no patient would be left to die if they turned up to a hospital and could be saved, and many adult intensive care units intervene in ways that are not cost-effective as per the NICE definition. The end of life is the most difficult to deal with, research has shown that people value a change from 0.2-0.4 QALYs more than they value a change from 0.6-0.8 QALYs. Many expensive life prolonging cancer drugs are not funded by the NHS, but there are cases of successful lobbying to have these drugs reimbursed despite their lack of cost-effectiveness. This could lead us to conclude that doing nothing is fine as long as it does not kill the patient (or allow the patient to die, depending on your stance) in which case we should always intervene. It is unfair to ask a health care professional not to act, since, as detailed, it is their responsibility if their patient dies through inaction.

For the most part, everybody is provided with the necessary treatment when they are in need. It’s really only at the end of life the problem of opportunity cost is apparent due to the high cost of interventions. Perhaps the answer lies in allowing NICE to negotiate the price of drugs, although this would not necessarily lead to price reductions since companies would be incentivised to pitch drugs at an even higher price knowing that they will be negotiated down to their acceptable price. To the contrary though it may be argued that this constitutes inaction on the part of NICE, and by negotiating (or at least trying to) they could allow more people to survive. Another issue is that the few months that are gained by (usually expensive) end of life treatment are usually in very poor quality. From an Aristotelian perspective this would not be a virtuous choice, as we would not be achieving ‘the good life’, and what’s more, Aristotle says, no-one would actually choose this state of suffering unless they were defending a philosophical position.

In the end we may defend ‘doing nothing’ as a choice as it may be necessary in the face of opportunity cost, and it is always better to know the outcomes from as many scenarios as possible when modelling it in simulation based studies. However, in practice ‘doing nothing’ may not be realisable, since the fear of death may prohibit people from accepting this option. Perhaps there is a case for allocating more resources to health care from other areas of public spending, which there certainly is a case for. What would be ideal would be a quantifiable way of measuring the benefit from all government spending and then choosing the health care budget based on this. But this is definitely a long way from reality.


4 thoughts on “The ethics of doing nothing

  1. I would argue the opposite: the ‘do-nothing’ option should be included in cost-effectiveness analyses to provide more information to the decision maker, but actually, rarely is.
    My background in health policy gives me a little bit more sympathy for including ‘do-nothing’. For any business case one needs to justify why ‘do-nothing’ is not the option that should be chosen by outlining the costs and benefits of every option. By not including ‘do-nothing’ there is no way to assess how much better the recommended option is compared to the least resource intensive alternative. It is not a question of ethics but of providing the best possible information to decision makers.
    In cost-effectiveness analyses one always calculates the costs and consequences of an option compared to a comparator. In my experience, in clinical trials that inform a cost-effectiveness analysis that comparator is rarely ‘do-nothing’, but either “treatment as usual” or placebo. For a number of the trials I work with, because we work a lot with therapeutic approaches to depression, anxiety or psychosis (as opposed to pharmaceutical approaches) having identified clinical problems that the patient might have as part of recruitment for the clinical trial their doctor is then advised of these problems. Patients that are not randomised to the treatment arm patients are managed by their doctor in line with clinical best practice. Obviously when testing a therapy like CBT one cannot provide a placebo. What we want to know is if the new treatment is better than current clinical best practice. Even in placebo randomised control trials the comparator is arguably not ‘do-nothing’ as we are all aware that there is a placebo effect. As a result, because, as you point out, it is unethical to do nothing when someone has been identified as having a clinical problem for which there is guidance about best treatment practice, we rarely have information to include in cost-effectiveness analysis on what happens when one does nothing. One could arguable obtain this information from observational studies or population based databases such as GP record databases, but this is rarely done. Instead the comparators are generally treatment as usual or placebo, not ‘do-nothing’.
    On a side point, I think I would argue that even in a universal health care system such as the NHS, not everyone is provided with the necessary treatment (and in other countries the uninsured are regularly denied access to health care treatment – but this is a separate issue). To give an example, a few years ago, so this might not still happen, we were having a discussion about routinely testing people for Hepatitis C in prisons. The problem is that because treatment for Hepatitis C is very expensive not all PCTs offer it. As a result, one could argue that it could be unethical to test people for a disease where, if they test positive, they will be aware that they have the disease but can receive no treatment for it.


    1. Great comment, Rachael.

      I think your final point is an interesting ethical issue. A similar problem arises for genetic testing, where an individual can find out their predisposition to developing a disease for which no cure exists, such as Huntington’s disease. A small cost with no apparent health benefit to the patient, but surely ignorance should never be preferred!? Who knows. I look forward to somebody submitting a blog post on the topic!

      suggested reading: Payne et al (2012)


  2. I think in the UK we rely on cost-effectiveness analysis and will accept do nothing as an option if the outcome is acceptable with regards to the cost. Compared to France, say, who provide all available treatments regardless of cost-effectiveness and end up spending a far greater amount per capita on health care without any real increase in population health outcomes (this could also be to do with their massive over prescribing).

    I think doing nothing can be included if it is a feasibly possible to actually do it in the real world, otherwise what would be the point of the information about the effects of doing nothing? For RCTs it is essential to be able to isolate an intervention’s effect so it is compared to a placebo or even doing nothing. However we do stop RCTs if there is a clear benefit to the treatment being tested which suggests that we find doing nothing unacceptable if we know a treatment that works exists.

    Also, the decision maker may have this empirical information available but I don’t think there is enough focus on whether that is a ‘good’ outcome or not. In my opinion, there is clearly a difference between an intervention that is £30,000/QALY and increases one person from 0.8-1.0 QALYs than one that saves a persons life for an extra 2 months i.e. 0.0-0.2 QALYs.

    So I would agree with you about your hep C example – the outcome there actually produces a net negative outcome if no treatment is provided.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s