Correspondence

Computer-Assisted Diagnosis in Europe

N Engl J Med 1994; 331:1238November 3, 1994

Article

To the Editor:

Your thoughtful and scholarly editorial (June 23 issue)1 awarding a grade of C to computer-assisted diagnosis steers an admirable middle course between wild enthusiasm (expressed by some of our colleagues in computer-related fields) and flat rejection of all things automated (from some of our clinical colleagues).

In Europe, there has been a different emphasis on computer-aided decision making. European systems have generally addressed a specific clinical problem instead of attempting to deal with the whole of medicine, with an emphasis on teaching and guidance (and disseminating good clinical practice, where it can be determined) rather than on artificial intelligence.

This approach may not be as exciting intellectually, but it has led to rather good evidence of improvement in routine clinical performance. An eight-center trial in the United Kingdom (involving 16,737 patients with acute abdominal pain) showed approximately a 20 percent improvement in the diagnostic accuracy of the doctors, with almost a 50 percent reduction in the rates of perforation and negative laparotomy.2 Other studies have shown comparable results maintained over a 12-year period.3 Most recently, comparable results were reported in a study of 15,000 patients from 64 hospitals in the countries of the European Community.4 The results seem to have occurred by a process of leveling up because of the adoption of good practice, with the computer itself performing about as well as (but no better than) the most senior doctors in the trial. A grade of A+ is not deserved, but neither is a grade of C. Maybe a B- most appropriately reflects the European situation.

Your editorial states that there is still a long way to go. This statement is of course correct, but perhaps it raises the wrong question. If a child in grade school shows the diagnostic ability of an average doctor, the question is not what the child can do now, but what the child will do as an adult. Instead of grading current performance, perhaps we should be considering how we can best ensure that exposure to experienced clinicians and their practice (combined with currently demonstrated ability) results in a future diagnostician with an above-average capability.

The problem may seem hypothetical, but it is not. In medicine we face a compelling and massive problem. The medical course is finite. The medical faculty is finite. The amount that must be learned in order to practice “good medicine” has exploded. We need to make sure this “child” gets good grades in the future, and the child will do so only if properly nurtured and educated.

F.T. de Dombal, M.D.
University of Leeds, Leeds LS2 9LN, United Kingdom

4 References

1. 1

   Kassirer JP. A report card on computer-assisted diagnosis -- the grade: C. N Engl J Med 1994;330:1824-1825
   Full Text | Web of Science | Medline

2. 2

   Adams ID, Chan M, Clifford PC, et al. Computer aided diagnosis of acute abdominal pain: a multicentre study. BMJ 1986;293:800-804
   CrossRef | Web of Science | Medline

3. 3

   McAdam WA, Brock BM, Armitage T, Davenport P, Chan M, de Dombal FT. Twelve years' experience of computer-aided diagnosis in a district general hospital. Ann R Coll Surg Engl 1990;72:140-146
   Web of Science | Medline

4. 4

   Objective medical decision making: acute abdominal pain. In: Beneken JEW, Thevenin V, eds. Advances in biomedical engineering: results of the 4th EC Medical and Health Research Programme. Vol. 7 of Studies in health technology and informatics. Burke, Va.: IOS Press, 1993:65-87.
   

Citing Articles (3)

Citing Articles

1. 1

   P Lisboa. (2002) A review of evidence of health benefit from artificial neural networks in medical intervention. Neural Networks 15:1, 11-39
   CrossRef

2. 2

   Bharat N Nathwani, Kenneth Clarke, Thomas Lincoln, Costan Berard, Clive Taylor, Kc Ng, Ramesh Patil, Malcolm C Pike, Stanley P Azen. (1997) Evaluation of an expert system on lymph node pathology. Human Pathology 28:9, 1097-1110
   CrossRef

3. 3

   Marvin E. Gozum, Andrew S. Kanter, Dawn E. DeWitt. (1995) Benefits of computer diagnostic assistants. Journal of General Internal Medicine 10:7, 413-414
   CrossRef