The Doctrine of Informed Consent and Respect for Autonomy

In this posting I want to examine the relationship, if any, between the doctrine of informed consent and respect for autonomy. In doing so I will try to answer three important questions. Firstly is the doctrine of informed consent based on respecting autonomy or is it a sui generis doctrine? The fact we talk about a doctrine of informed consent seems to suggest the latter. Secondly if informed consent isn’t based on respecting autonomy should it be so based? Lastly if informed consent is based on respecting autonomy what type autonomy should that be?

Is informed consent based on respecting patient autonomy? According to Charles Foster the doctrine of informed consent is in a mess.

“Informed consent, in practice, is a bad joke. It’s a notion created by lawyers, and like many such notions it bears little relationship to the concerns that real humans have when they’re left to themselves, but it creates many artificial, lucrative, and expensive concerns.” Practical Ethics

Let us accept that the principles underlying informed consent are unclear. Let us try to start understanding these underlying principles by examining the process of informed consent in practice. When taking informed consent a patient’s doctors propose some form of treatment and provide the patient with the information he needs in order to make a good decision. The patient then agrees or refuses to agree to this treatment. What is the purpose of this process? The purpose of the agreement part of this process seems clear. The agreement part prevents the patient being treated against his will and his doctors from being accused of assault. The purpose of the information part is different. The purpose of this part is to enable patients to make good decisions. What exactly do we mean by a good decision? A decision which is in the patient’s best interests or an autonomous decision? In a medical setting a decision which is in a patient’s best interests is one which minimises harm to the patient,  maintains or improves his physiological health. The amputation of a patient’s gangrenous leg would be an example of acting in a patient’s best interests by minimising harm. The provision of drugs to control a patient’s blood pressure would be an example of acting in a patient’s best interests by maintaining his physical health. A kidney transplant would be an example of improving a patient’s health. Autonomy is concerned with self-government and very roughly speaking an autonomous decision is one by which someone governs himself. However self-government is simply governing oneself and need not always be good government. We can govern ourselves badly. Let us assume that a patient simply wants to accept his doctors’ judgement about his treatment and doesn’t want to receive any additional information. Intuitively such a decision can be seen as a decision by which he governs himself and as a result is an autonomous decision for after all we can simply choose to follow our lawyer’s advice and most would regard such a decision as autonomous. However under The General Medical Council’s proposed guidelines such a decision might not be seen as a valid informed consent decision.

“If, after discussion, a patient insists that they don’t want even this basic information, you must explain the potential consequences of them not having it. This might include being unable to proceed if you are not confident that their consent would be valid, or if you are not confident that the proposed course of action would be beneficial to the patient overall.” (1)

In the light of the above it would appear that The General Medical Council’s guidelines are not really based on respecting autonomy and the purpose of providing information when taking informed consent is to enable patients make decisions which are in their best interests.

It might be objected that I have only briefly sketched a certain account of autonomy and that a different account might mean that the doctrine of informed consent could be based on respect for autonomy. After all in my brief account the patient in my example above who fully trusted his doctors to make a decision on his behalf might wake up to unexpectedly find himself with only one leg. It might be suggested such a decision isn’t an autonomous one and that we should adopt a more substantive account. For the sake of argument let us assume an autonomous decision must be one that concurs with some generally accepted set of norms. If these norms are to be generally accepted then they must concern what is good for people. Autonomous decisions become linked to good decisions. An autonomous decision must be a good decision and a bad decision cannot be an autonomous one. I now want to argue against accepting such a substantive account of autonomy. If an autonomous decision must be a good decision because it must concur with some accepted norms then it might be questioned whether such an account of autonomy is really doing any useful work. A substantive account of autonomy puts itself out of business because we only need to consider good and bad decisions. We don’t need to consider autonomous decisions at all. Using a substantive account of autonomy an autonomous decision must of necessity be a good decision according to some accepted norms whilst on a content neutral account of autonomy an autonomous decision can be a bad decision. Of course it is preferable that an autonomous decision, using a content neutral account of autonomy, is a good decision which benefits the decision maker but the purpose of respecting autonomy remains simply to respect the decision maker as a person and the actual decision doesn’t of necessity need to be a good one. If we accept the above then we have two options. Firstly we could abandon the pretence that underlying the doctrine of informed consent is respect for autonomy or secondly we could accept that underlying the doctrine of informed consent is respect for a content neutral or primitive account of autonomy and modify the doctrine accordingly.

If we accept the above then there is not only a difference in complexity between substantive and primitive accounts of autonomy but more importantly these different accounts have a different focus. A substantive account focusses on respecting good decisions, a primitive account focusses respecting persons by simply accepting their decisions. Adopting a primitive account requires adopting a certain degree of humility. Let us accept that if the concept of autonomy is to remain a relevant one that it must be a primitive or content neutral concept based on respecting persons. Such a concept might be classed as a Millian concept. In the rest of this posting ‘autonomy’ will refer to such a concept. Let us also accept the first of the above options that the doctrine of informed consent is a doctrine in its own right and isn’t based on respect for autonomy. The informed consent process starts with a patient’s doctors proposing a course of treatment which they believe to be in his best interests, the patient then consents or refuses to consent to the proposed treatment. I argued above that the purpose of informed consent is to get a patient to consent to treatment which his doctors believe to be in his best interests. In almost all cases the patient simply consents and no problems arise. What is interesting are cases in which the patient refuses to consent. In these cases according to the doctrine of informed consent a patient’s doctors must be sure, or take steps to ensure, that the patient knows the consequences of his actions. However let us assume that the patient maintains his refusal of consent. By what standards is his competence to refuse treatment measured by and how is it assessed? If someone takes a driving test his competence to drive is based on his actual driving. This is not true when assessing someone’s ability to give informed consent. Someone’s competence to give informed consent isn’t based on his actual decisions. His decision might of course trigger a competence assessment but his competence isn’t dependent on his actual decision. Someone’s competence to give informed consent is based on his ability to make good decisions based on his best interests. I would suggest that someone who has the ability to make good decisions based on his best interests is self-governing, is autonomous.

As mentioned above the purpose of a driving test is to measure someone’s competence to drive and the criteria by which his competence is measured is by his actual driving. The same isn’t true of informed consent. I have argued above that the purpose of informed consent is to enable patients to make good decisions. I have further argued above that the way in which a patient’s competence to make good decisions is assessed is by his ability to make autonomous decisions. Such a mismatch creates problems and perhaps is one of the reasons why the doctrine of informed consent is in a mess. Two solutions might be suggested to remedy this mismatch. First the purpose of informed consent should be to enable a patient to make a good decision in particular circumstances and the patient’s competence should be assessed by his ability to make a good decision in these circumstances. Secondly the purpose of informed consent is to enable a patient to make an autonomous decisions and his competence to do so should be assessed by his ability to make an autonomous decision.

Let us consider the first of these options. Let us accept that a good decision is one that is in the decision maker’s best interests and that a decision is a competent one only if the decision is in the decision maker’s best interests. Clearly if someone makes a decision to do something which he believes isn’t in his best interests then he isn’t making a good decision. If a patient’s competence depends on him making a good decision then he is incompetent. Perhaps he is paralysed by fear. However in practice most people make decisions which they believe to be in their best interests. Moreover in a medical setting there is usually agreement between the patient and his doctors about what these best interests are. Unfortunately in a few cases in which there might be disagreement about what is in a patient’s best interests. Let us accept that a patient who makes a decision which he believes to be in his best interests, but which is generally believed not to be in his medical interests by his doctors, is making an incompetent decision. In practice if a child or cognitively challenged adult makes a decision which others believe isn’t in his best interests his competence might be questioned and he might be treated against his will. Unfortunately if a competent decision must be a good decision then the same considerations would seem to apply to all patients. Accepting the above would mean that if any patient makes a decision which is generally regarded as not being his best interests then his decision is an incompetent one. In these circumstances the patient’s doctors might be accused of paternalism or even epistemic arrogance. Fortunately in most cases a patient’s medical best interests and what the patient believes are his overall best interests concur. However this isn’t true in all cases. In cases in which these interests don’t concur do we insist that a competent decision is based on a patient’s best medical interests or what he believes to be in his overall best interests? Practical considerations and the need to avoid the charge of paternalism suggest that we should choose the second option. However if we accept that a patient is making a competent decision when it concurs with what he believes are in his overall best interests we are back to assessing a patient’s competence to make a good decision by assessing his ability to make an autonomous one and the above mismatch remains unresolved.

The above difficulties suggest that we should choose the second option. The purpose of informed consent should be to enable patients to make autonomous decisions about their treatment and that any competence assessment should be based on their ability to make autonomous decisions. This provides an answer to the second of my three initial questions. Informed consent should be based on respect for patient autonomy. Moreover if a substantive account of autonomy makes itself redundant as I have argued above the type of autonomy underlying informed consent must be a primitive or Milliian account. This answers the third of my three initial questions. Accepting this option has consequences for the amount of information which needs to be supplied to patients when taking informed consent. In the past under the doctrine of informed consent a patient’s doctors determined what risks the patient should be made aware of. However the Montgomery ruling Montgomery ruling stated that a patient’s doctors must ensure that the patient is aware of any and all the risks involved. If patients are to make informed consent decisions based on respect for autonomy they don’t always need to be aware of any and all the risks and the Montgomery ruling seems to suggest that informed consent shouldn’t be based on respect for a primitive account of autonomy. However if we accept that if an account of autonomy is to remain a meaningful account that it must be a primitive account then if informed consent is based on respect for autonomy  it must be based on a primitive account. Accepting the above means that a patient’s doctors should have a dialogue with him about any suggested medical procedure. How this dialogue proceeds shouldn’t be preordained by some doctrine but driven by the patient’s needs. In most cases this will include informing him about any major risks and life changes he faces. However some patients might need less or more information in order to make an autonomous decision. As mentioned above an autonomous agent can make an autonomous decision to trust the advice of his lawyer or financial advisor. If someone is non-autonomous someone else might be given power of attorney to act on his behalf. Are doctors any less trustworthy than lawyers or financial advisors? It would seem to me they aren’t. Does then the context in which informed consent takes place differ from other contexts such as the law and finance in respect of an agent’s ability to make autonomous decisions? Provided the patient isn’t incapacitated by fear it isn’t. It follows if informed consent is based on respect for autonomy that a patient should be able to make a competent decision simply to take his doctors’ advice. Other patients might need more information than is usually supplied in order to make an autonomous consent decision. Doctors should make it clear that they are willing to supply more information when this is requested. For instance it might matter greatly to a Jehovah’s witness whether there would be any possibility of a blood transfusion however remote this possibility might be.

It might be objected that I have already introduced an example which shows that it would be absurd to base the doctrine of informed consent on a non-substantive or Millian account of autonomy. Let us agree that if we accept such an account that a patient can make a competent decision simply to trust his doctor. Let us recall the patient with the gangrenous leg. Let us assume that this patient simply wants to trust his doctor and refuses to listen to any information provided. He wakes up and unexpectedly finds he has only one leg. Such a scenario seems absurd. Perhaps then the basis of informed consent should be a substantive account or the doctrine of informed consent should be a self-contained doctrine. Let us assume the patient still wants to simply trust his doctors and refuses to listen to any information concerning his procedure. Should he should be forced to listen to brief details connected to his proposed treatment? Do doctors really want to force someone to listen? Is it possible to force someone to listen and digest information? Should he be left to die? In this situation it seems more likely that the patient would be judged as incompetent. He would then be treated in accordance with his best interests and his leg removed. He wakes up with only one leg. In this case adopting a substantive account of autonomy or considering the doctrine of informed consent as a sui generis doctrine changes nothing. It follows that whilst adopting a non-substantive might lead to some highly undesirable consequences in a few rare cases that it doesn’t lead absurd ones.

1. Supporting patient choices about health and care: Draft Guidance for consultation, GMC, 2019, [33]-[35]

Lying and Autonomy

  

We live in a world surrounded by fake news and lies. According to Terry Pratchett in the Truth “A lie can run round the world before the truth has got its boots on.” However even if lies sometimes spread more easily than the truth, perhaps in part because they are more palatable and we want to believe them, in the long term the truth matters because the truth has persistence which lies don’t. Perhaps the dwindling number of climate change deniers supports the above. Are there any circumstances in which lying is beneficial? Stephen Rainey writing in practicalethics  suggests that some lies “can also be a kindness, when the truth might serve no good”. Let us accept that in most circumstances lying damages both individuals by depriving them of the truth and trust. Nonetheless is Rainey correct when he suggests some lies might be beneficial in some circumstance? For instance might a government be acting beneficently if it lied about the harm done by the coronavirus to prevent panic? The fact that lying hasn’t been eliminated from human culture over time suggests it isn’t harmful in all circumstances. In this posting I want to examine what these circumstances might be. It might be thought that this is merely an interesting rather than important question, however in seeking to answer this question raises a further question about the balance between acting beneficently and respecting autonomy. I will argue that if I lie to someone and even if my lie benefits him that I also fail to respect him by failing to respect his autonomy.

Most people have no problems with defining a lie. The definition I shall adopt here is that of Sam Harris

“To lie is to intentionally mislead others when they expect honest communication.” lying

Lies so defined can explicit when liars intentionally give someone false information. Lies can also be acts of omission when the liar intentionally withholds information others expect him to provide. Perhaps passive aggression can be a form of lying if it involves withholding information. Liars can be practised by individuals, institutions or even governments.

Let us now consider whether it is ever acceptable to lie for beneficent reasons. Because lying is so prevalent our intuitions suggest that it is. Everyday examples seem to confirm our intuitions. For instance the mother who lies to her child by saying she has no money left when her child demands a second ice cream. Someone who lies to his partner about a surprise party. A government which lies to its citizens about the seriousness of some disease outbreak, such as that caused by the coronavirus, in order to prevent mass panic. Nonetheless as Harris points out we need to be extremely cautious about our intuitions. Firstly we must be sure about our motives being really beneficent. Is the mother above really worried about her child becoming obese or simply taking an easy option? Secondly we must be careful not to damage trust. The government above might avert mass panic but in the future its pronouncements on health matters might not be so readily trusted. If a lie cannot be maintained then it is better not to lie at all in order to retain trust. The Chinese government’s attempt to suppress the full extent of the coronavirus outbreak in Wuhan seems to support the above. However let us accept that there are at least some cases in which we might lie for beneficent reasons in which our motives are clear and we don’t damage trust. For instance it would be perfectly acceptable to lie to a stalker carrying a knife about the whereabouts of his ex-girlfriend.

Let us accept that in certain circumstances we can act beneficently by lying, telling white lies. Let us further accept that in some of these circumstances trust isn’t damaged. In such circumstances is Rainey correct when he suggests that lying might be an act of kindness? It certainly seems that because we are behaving beneficently that our actions might be classed as acts of kindness. However I now want to argue even if our lies serve a beneficent purpose and can be classed as acts of kindness that nonetheless it is almost always wrong to lie. Most pet owners are kind to their pets and act beneficently towards them but most people don’t want to be treated in the same way as pets however kindly. They want to be recognised as the sort of creatures who can make their own decisions. To be autonomous. Autonomous people govern themselves and this requires making decisions that matter to them. To make decisions that matters to someone requires information and lying deprives him of some of that information. It follows if we lie to someone we fail to respect his autonomy. Moreover if someone becomes aware of being lied to this lack of respect is a form of rudeness see the philosophy of rudeness.html or form of contempt. It further follows if we believe it is right to respect someone’s autonomy that we shouldn’t lie. It further follows if a government lies to or intentionally misleads the electorate over some issue such as Brexit that it not only deprives the electorate of some of the information needed to make a good decision it also shows a lack of respect for the electorate and perhaps even democracy itself. This lack of respect is similar to the lack of respect many colonialists showed for native people even if their motives were beneficent.

I now want to consider two objections to accepting the above conclusion. My first argument will be based on respecting autonomous persons and my second on respecting autonomous decisions. Firstly it might be objected that we should give precedence to acting beneficently over respecting autonomy and that sometimes lying is the only effective way to act in someone’s best interests. I now want to present two counter arguments against accepting the above. Let us accept that if we need to lie to someone in order to act in what we believe to be their best interests that there must be a clash between what we believe to be in their best interests and what they perceive to be in their best interests. It would appear that we believe we know better than the agent about what is in his best interests or should be in his best interests. In this situation we might be accused of epistemic arrogance. Secondly I would argue if we accept the above objection we don’t really understand what respect means. We can sometimes act beneficently towards someone but part time respect isn’t respect we can’t respect someone only some of the time. We can of course respect someone’s courage but not her wit. However when we respect her as a person we can’t respect her only some of the time. Part time respect just isn’t respect at all. It follows if we lie to someone in order to act beneficently we are not giving precedence to acting beneficently over respecting autonomy we are in reality failing to respect autonomy at all.

Now let us consider respecting autonomous decisions. Let accept to respect an autonomous decision means to accept it. It might be objected that any difference between respecting someone’s autonomous decisions and acting beneficently towards him is largely illusory. This objection depends on accepting a substantive account of autonomy in which autonomous decisions must concur with some accepted norms, autonomous decisions must be good decisions. Accepting such an account means that if we lie to someone, who is making a bad decision, in order to act beneficently towards him we are still respecting his autonomy. Bad decisions just aren’t autonomous decisions. However I am reluctant to accept a substantive account of autonomy. A substantive account has become more prominent recently largely in order to make respecting autonomy concur better with the doctrine of informed consent. Unfortunately as I have argued elsewhere a substantive account of autonomy puts itself out of business, see autonomy and toleration . If autonomous decisions must be good decisions in accord with some accepted norms then we can just consider good decisions and autonomy becomes a redundant concept. Of course if someone is to make an autonomous decision he mustn’t be misled or coerced but it follows that if autonomy is to remain a meaningful concept that we should adopt a primitive or Millian account. It further follows if autonomous decisions needn’t be good decisions that if we lie to a mature adult in order to protect him from a bad decision that we are failing to respect his autonomy. I would suggest accepting the above has implications for apology. Insincere apology is a form of lying. Perhaps in most circumstance apologising without really mean it is a form of white lying. Nonetheless such apologising just compounds any wrong by showing a lack of any real respect to the one lied to.

I have argued that it is always wrong to lie if this means we fail to respect someone’s autonomy even this is done for beneficent reasons. The question now arises is lying ever acceptable? If the mother of a young child who pesters her to buy another ice cream lies then her lie can be justified. The child isn’t autonomous yet and having two ice creams isn’t in her best interests. However caution is needed if we accept a primitive account of autonomy as I suggested above then slightly older children who are capable of making autonomous decisions and lying to them fails to respect their autonomy. Sam Harris uses an extreme example and asks should we lie even if with “Nazis at the door and Anne Frank is in the attic”. Of course we should lie. We should lie to protect Anne Frank and her autonomy. Lying to someone who is infringing someone else’s autonomy is perfectly acceptable, lying to protect autonomy is lying to respect autonomy.

What conclusions can be drawn from the above? Firstly we should never lie if this means failing to respect autonomy even if our lies might be in what we consider to be someone’s best interests. Even white lies are wrong. Secondly if autonomy is to remain a meaningful concept we must always give precedence to respecting autonomy over acting beneficently. Lastly lying to autonomous agents can never be kind.