Letter: Computerised predictions of mortality

Click to follow
The Independent Online
Sir: I refer to the press coverage of David J. Bihari's proposed intention to use a computer program to make decisions as to who should, and should not, be treated in the Intensive Care Unit at Guy's Hospital. Dr Bihari has now made clear (letter, 31 August) that such a program will not be solely responsible for making these decisions. He has also made clear that his threat may have been motivated more by the wish to highlight attention on budget cuts rather than the desire to use the program directly in clinical management.

I believe that the moral issues raised by this complex interplay of the increasing cost of medical care capable of (in certain circumstances) delaying death and prolonging suffering indefinitely, and the role of machines in the primary medical objective of saving lives, deserve closer consideration.

Life and death decisions are constantly made by doctors, who are trained to start treatment, not stop it. Sometimes, despite considerable experience, they get it wrong, and the law recognises the concept of an error of judgement in the professional person.

I suggest that, after proper scientific assessment, computer prediction of likely mortality may become a valuable asset, assuming that any potential clinical use is within a framework of strict and informed ethical guidelines. If these conditions are satisfied, and, for the sake of argument, it can be shown that computers consistently outperform doctors, then we should not be afraid to entrust at least part of the decision-making process to technology, just as we do when flying commercial aircraft.

Intensive care medicine is a science, and Sir Peter Medawar, the Nobel laureate, said that a science does not truly become mature until it develops predictive capability. The public reaction to Dr Bihari's particular stance should not impede the progress of intensive care research, whose ultimate purpose is to save, not cast aside, life.

Computers may help us with this quest. I doubt whether computers could ever reliably predict the moment at which it becomes unethical to spend large sums of money on individual patients.

Yours sincerely,


Consultant and Lecturer in

Infection and Immunity

Division of Microbiology

Guy's and St Thomas's

Medical and Dental School

London, SE1

31 August