• Palliative care
• end-of-life care
• double effect
• terminal sedation

Kuhse criticises our application of the doctrine of double effect to the problem of agonal respiration.¹ She points out that we encounter a problem with describing the act in question—that is, the administration of muscle relaxants during agonal respiration. We could describe the act in any number of ways, some of which would imply that it is an act of killing. She is correct to point out that killing is not a basic action and we must appeal to the agent’s intentions to determine whether the act in question is killing or something else. But this same problem occurs when we are talking about other actions, such as taking a patient off a respirator or not resuscitating a patient. In some cases, these actions would constitute killing by an act of commission or an act of omission. We justify these actions by appealing to the intentions of the actors. We can infer a person’s intentions from his or her circumstances, behaviour, and communications. If a physician weans a terminally ill patient from a respirator, we do not call this action “killing” because we infer that the physician’s intention is to reduce the patient’s suffering, not to kill the patient. We can make the same sort of inference in the agonal respiration case; we infer that the physician’s intention is to relieve suffering not to kill the patient. Obviously, intentions are subjective and therefore epistemologically problematic—how can we ever know what anyone intends to do? On the other hand, end-of-life decision making and ethics would make no sense at all if we did not assume that we can ascribe intentions to physicians, patients, family members, and other health care professionals. One does not need to subscribe to the doctrine of double effect in order to grasp this basic point. Even the well accepted practices of withholding and withdrawing care cannot be morally justified without ascribing humane intentions to the actors. So, we admit that one needs a theory of human intentions in order to apply the doctrine of double effect to the problem of agonal respiration, but this is not a fatal flaw in our approach, since one needs a theory of human intentions to make sense of virtually all medical decisions at the end of life in any case.

Kuhse argues that it might sometimes be morally justifiable to administer muscle relaxants to patients in agonal respiration, even if we do not rely on the principle of double effect. She says that administering muscle relaxants in such a case may benefit the patient (by relieving suffering) or benefit the patient’s family. But how can she maintain this view without addressing problems having to do with the intentionality of the act? If you administer muscle relaxants, your action could be construed as killing, and most people regard killing as immoral and illegal. It seems to us that you can respond to this charge in two ways. First, you can defend the act by describing it as something other than killing—that is, as relieving suffering. This is the option we prefer. Second, you could admit that the act is indeed killing but maintain that it is justifiable killing. We would prefer to avoid construing the action as killing, since this would make it a type of euthanasia and murder. Of course, one could maintain that physicians should do nothing to stop agonal respiration. They could stand by the bedside until their dying patients take their last gasp. We see no reason why patients or their families should be required to bear this burden once the patient has chosen to withdraw respiratory support and death is imminent.

Hawryluck makes a very important point when she asks whether neuromuscular blockers might actually make the patient’s suffering worse, if the patient is conscious, unable to breathe, and aware of his or her inability to breathe.² This is a valid point. We do not know what the gasping patient may be experiencing. Given this uncertainty, we should take steps to reduce the patient’s suffering.

We are advocating the use of muscle relaxants in a very small and specific group of patients. They are easy to recognise: they are terminal, have been well sedated, are unresponsive, and have entered a prolonged gasping phase. If these patients can suffer, the likely cause of suffering would be the intense muscle contraction of the prolonged gasps.

Hawryluck also points out difficulties with using the agent’s intentions to make moral distinctions between using neuromuscular blockers to palliate v using neuromuscular blockers to kill. We have discussed this issue in our paper and in our response to Kuhse. We admit that the distinction between “intending death” and “intending palliation (and foreseeing death)” can be murky and ambiguous in end-of-life decision making. Nevertheless, intentions (and wants, desires, and motives) do matter, and we must do our best to understand these subjective states of mind. In ethics, one’s mental state can make the difference between euthanasia/murder and palliation.

Finally, Hawryluck asks if our families were prepared for what was going to happen. In both of our case examples, the families received a very clear description of what the end of life would look like. The relationship was so honest that when they asked if “there was anything else we could give their children to stop the gasping” we answered “yes, but we cannot use that category of drugs”. Both families pleaded that we use the neuromuscular blocking drugs; we did not, but rather continued to escalate our already high level of sedation. One of us (RMP) is haunted by the memory of the syringe of medicine being given, which he knew would not effectively end the gasping. These families are not comforted by the fact that we acted in an “ethically permissible” manner.