Intended for healthcare professionals

Rapid response to:

Clinical Review

Care of the dying patient in the community

BMJ 2013; 347 doi: (Published 03 July 2013) Cite this as: BMJ 2013;347:f4085

Rapid Response:

Re: Care of the dying patient in the community

Laurie R Davis (this series of rapid responses, 11 July 2013) has commented ‘It can have workload implications as there appears to be a widespread belief amongst care homes that patients who might die should be visited by their GP every two weeks.’

The reason is probably the grossly inappropriate and hugely insensitive behaviour of police for EoL community deaths, when the death is not ‘expected’. In this context, expected has an obvious logical meaning: an ‘expected death’ is one which occurs after the GP has ‘promised to certify even if I cannot attend post mortem’ (with the safety provided by that promise, the coroner can hugely relax the investigative aspect of post mortem behaviour, effectively instructing suitably trained nurses/etc that ‘unless the death was obviously unnatural, just arrange for the body to be removed, try to keep the police out of it, and do not pester grieving relatives’).

There is no guidance within current community policies, which properly addresses ‘I would not be surprised if the patient dies, but I would need to attend post mortem before deciding whether to certify’ – currently police behaviour tends to treat ‘early EoL death’ as very much akin to the sudden death of a believed-to-be-healthy 25 yr old. This is absurd, to put it mildly, when you are in the position of a person who is living with the patient. The transition from considerate treatment by nurses and the GP, to ‘being treated by the police like a suspect when my loved one finally died’, is a huge shock to the just-bereaved, and very ‘disturbing’. I also think such police behaviour reinforces long-term memories of the death itself, which is a bad thing, and is being ignored by policy creators.

So the death of an elderly patient who ‘might die any time, but I (GP) cannot say whether tomorrow or in a year’s time’ is treated as a ‘sudden death’ unless the GP attends post mortem and decides to certify: and even an ‘expected death’ will become, from the perspective of the care home, ‘a death we, and the police, are not certain will be certified’ if the GP has not visited within the previous 14 days. The police have a strong tendency to ‘seek to investigate in depth’ any death until certification is effectively a certainty – this damages bereaved relatives, and is neither ‘balanced’ nor rational, if a patient is known to be ‘very elderly, final-year-of-life or has a condition which involves an ongoing but small probability of ‘dying at any time’’.

And the care home, or a relative in the patient’s own home, cannot be certain that the GP will be able to attend promptly, when the death occurs.

Moving on, David Jolley (this series of rapid responses, 18 July 2013) mentioned the Liverpool Care Pathway. The recent review of the LCP, and its follow-on releases, frequently mention communication and shared decision making, without being very clear about the meaning of either term. To put that simply, why do clinicians and ‘relatives’ have to talk to each other, is the discussion a legal requirement, and who ‘makes the decisions’ if the patient cannot make his own decisions ?

The trickiest issue with EoL behaviour is the legal aspect, and in practice this means the interpretation of the Mental Capacity Act. Clinicians do not usually agree with me about this, and most believe that ‘best interests decisions ultimately devolve to the senior clinician’ – but the Act does not state that. If there is a suitably-empowered attorney, who almost certainly will be a layman, the attorney is ‘the ranking section 4 decision maker’ because of section 6(6) of the Act. So, it is clear that ‘making a best interests decision’ does not require that one is a trained clinician: this is because the section 4 best interests test can logically only start from the concept of ‘if we could somehow ask the patient, what would he decide ?’ (see section 4(6) of the Act). As an aside, it is often clearer, to think in terms of ‘interventions being accepted or refused’ rather than of ‘treatments being accepted or refused’.

If there is not an attorney – and it is unusual for there to be an attorney – the Act does not define who can be a decision maker, it merely requires anyone who makes a best interests decision, to be able to have legitimately made it (section 4(9)). So, instead of assuming ‘the doctor is the decision maker’, let us simply assume ‘there is a decision to be made’ and look at the mechanism. The mechanism should be, that the clinicians describe the clinical outcomes with and without any offered treatments to ‘everyone who could validly have an opinion (the family and friends) about what the patient would say, if the patient could answer for himself’: then, all of the ‘family and friends’ individually answer. The question put to family and friends is ‘What would the patient want to happen’ – it is not ‘What do you want to happen’. If they all say ‘Fred would accept the treatment’, or they all say ‘Fred would refuse the treatment’, it seems simple – in such a unanimous situation, behave as if the (incapable) patient has answered directly, record who was involved in the discussions and who said what, record the best interests decision which emerged, but do not claim who ‘made the decision’.

It is more complicated, if all of the ‘proxy minds’ are not unanimous (there is not the space to elaborate, here) – but the current interpretation, by most clinicians, of the Act’s best interests requirement, isn’t logical and isn’t coherent. If anyone wishes to discuss this further, email

Competing interests: No competing interests

19 July 2013
Michael H Stone
retried non clinical
none layman Coventry