AUTHOR
David Anderson
ABSTRACT
The presentation sets out a thought experiment in which a number of today’s technologies are drawn together to produce a sophisticated aid for doctors working in far flung corners of the world. By degrees, features of the example are modified having the effect of permitting ever greater degrees of autonomy for the doctor’s aid and raising questions which tease out some of the ethical issues which stem from using technology to assist medical practice. By the time of the final variation on the example a number of uncomfortable options are the only ‘reasonable’ choices left open.
Version 1
A doctor carries out an urgent medical procedure on an unconscious patient presented to the doctor by a third party.
Question
Who bears moral responsibility for the treatment?
Suggested Answer
Primarily responsibility belongs to the doctor.
Version 2
Suppose a field medical system is developed which permits a specialist to engage in two-way audio-visual communication with a doctor who is presented with a patient as in version 1. Suppose further that the specialist is able to offer advice on treatment which is based on the specialist’s superior knowledge and experience.
Question
Who bears moral responsibility for the treatment?
Suggested Answer
This is a little less clear. There is clearly a balance to be struck between the specialist and the doctor but all the responsibility seems to lie with the medical personnel (however apportioned).
Version 3
Instead of the doctor being assisted by a human specialist the advice comes from a sophisticated medical Expert System. As before, the ES is able to offer advice on treatment which is superior to the doctor who administers the treatment.
Question
Who bears moral responsibility for the treatment?
Suggested Answer
There being only one person actively involved there seems little choice but to say that the doctor is wholly responsible. But there is a problem – in version 2, the doctor (who exercised professional judgement in exactly the same way as in version 3) was thought to have less than complete moral responsibility for the treatment. Surprisingly, the use of technology seems to have increased the doctor’s personal responsibility.
Version 4
A device is developed which enables doctors and specialists not only to advise colleagues at some geographical distance on treatment regimes, but perhaps by means of a number of remote controlled instruments also permits them to carry out treatment directly. So this version removes the local doctor and considers remote treatment.
Question
Who bears moral responsibility for the treatment?
Suggested Answer
Ethically, this looks exactly like version1. The doctor carrying out the treatment bears full responsibility.
Version 5
The specialist carrying out the treatment in version 4 is replaced by a medical Expert System.
Question
Who bears moral responsibility for the treatment?
Suggested Answer
A number of uncomfortable possibilities present themselves. The ES programmers may be thought responsible but as they are not present this seems a little inappropriate. The person who authorises the use of the ES perhaps or the medical experts whose expertise is being used in the ES, or no-one. A further possibility perhaps most uncomfortable of all remains – the ES itself is responsible.
Version 6
The Expert System used in version 5 is replaced by an ES which is the result of a combination of Neural Nets and Genetic Algorithms.
Question
Who bears moral responsibility for the treatment?
Suggested Answer
Here the dilemma is sharpest of all. It seems that there are four positions which are arguable and only the last two seem to be at all tempting:
- The medical authorities are the responsible agents. No matter how sophisticated the Expert System might be, it will remain ‘just a machine’ and while it might not require human intervention, from a moral position it remains under human supervision and it is the responsibility of humans to see that it carries out it job properly.
- The responsibility belongs to the trainers and creators of the machine in an analogous way to the case of the creators of an aeroplane autopilot.
- There is no moral agent of significance in the example outlined. The Expert System is designed to act in the absence of qualified human medical personnel. Typically, lay people would present themselves or, in emergency, be presented to the ES, and from this point there is only a machine working. It is obvious that machines simply cannot be the bearer of moral predicates. When we say this is a good tractor we do not assert that it is morally estimable we mean that it is a tractor whose performance meets or exceeds reasonable expectations. So it is here. An impressive machine, to be sure but a machine not a moral agent.
- The Expert System is responsible. In many ways this is the most natural and equitable reaction of all. After all, the Expert Systems works without the intervention of human beings, indeed, it may control and direct the behaviour and actions of human beings. The Expert System is in charge, it responds dynamically and properly to the changing medical circumstances presented. It would be entirely correct to ascribe moral agency to the Expert System were it human (or even a Martian) thus to withhold such ascriptions in the current case are as naked a case of blind prejudice as has ever been documented.