ICT in health care
Medical systems: authorized to make decisions or not?
Medical systems, like for instance the Electronic Files on Patients (efp) , affect the way in which practitioners, insurance people and patients deal with each other. That, in turn, has ethical implications, none of which has yet been really charted.
The efp seems so useful: if you have an accident the emergency services people on duty can instantly gain access to your GP’s files so that they do not, for example, administer pain killers to which you are allergic. There is, however, another side to such information transparency. Do the emergency services people, for instance, need to know about any possible psychiatric problems you might have which are also recorded in that same file? That is merely a simple example.
‘As long as we are just talking about the contents of a few files the situation is fairly straightforward’, comments Dr. Vincent Wiegel. ‘But in reality the whole picture is far more complex. Insurers employ medical experts and they are obliged to abide by a professional code of conduct as far as confidentiality is concerned. Their assistants, by contrast, do not have to adhere to such codes. What, then, are the implications of that for the access that insurers have to information? All in all, the whole matter thus becomes difficult to control. The system therefore has to make decisions about the sharing of information.’
In this respect medical systems do not really differ much from other computer systems. Technical requirements can be reasonably objectively formulated. In the case of functional requirements, such as those linked to user-friendly interfaces, things become more complicated. But how do you gauge within such a system whether or not a person is a good mother who cares properly for her child? Still, that must somehow or other be recorded in the Electronic Child’s File, another major computer system that is being developed the aim being to allow child welfare to function better. ‘As philosophers, we try to provide software developers with tips on how they can deal with such vague, ethical kinds of notions’, Wiegel says. ‘That can range from making an extra box in the matrix of design requirements to adding aspects to the programme languages that they use.’
The standardizing of ethical concepts is difficult but it is something that has to be done if such details are to be entered into computer systems, definitely if the systems are not only going to issue such information but are also going to work with it. Robots have already made their way into the geriatric care sector and that is something that will not diminish. ‘Imagine that a robot registers that an old person is not taking his or her medicine. Must the robot then forcibly administer it, must it notify the physician on duty or must it simply warn the old patient in question? In the latter case, the onus is on the autonomy of the old person but possibly with detrimental effects for his or her health. If, however, the robot warns the doctor, the old person might come to view the robot as a perpetual spy and that is not a desirable situation either.’
One way of studying complex situations is by making use of computer simulations. In that way certain behaviour is ascribed to autonomous programmes or agents so that it subsequently becomes possible to see how the interaction of numerous agents progresses. ‘This research is still in its early stages but ultimately it should be possible to programme a large number of agents with roles that revolve around the efp’, Wiegel explains. ‘From the interaction you can then, for example, see which agents release what kind of information and whether or not that is desirable. By altering the ethical rules programmed into the agents one can then discover whether that leads to other considerations and, undoubtedly, to a better overall view.’
