This article was originally published in the Jan. 23, 1970, issue of Mayovox.
A pediatrician who encounters a tough diagnostic problem might welcome an opportunity to have his or her judgment re-enforced by consultation with a number of peers and might find it equally helpful to be given references to pertinent articles in the literature.
A program designed to give such assistance is now under study in the Clinic's Section of Pediatrics. Trial of the "Computer-Assisted Diagnosis in Pediatrics" is being made at the request of Mead Johnson. This company has supplied the terminal device, which provides two-way communication with a computer center in Chicago. Mark D. Cloutier, M.D., Consultant in Pediatrics, has primary responsibility for the testing program.
Nine physicians, all imminent pediatricians, have contributed to programming the computer. The user is supplied with a manual, which contains lists of symptoms classified as pertaining to the cardiac system or general system. Each symptom has a code number.
The terminal unit is a teletypewriter, which resembles a standard typewriter with an additional control panel. Telephone connection is established with the computer center by inserting a card dialer. Completion of the connection activates the teletypewriter, and instructions for succeeding steps are transmitted via the typewriter. Information is entered by typing the appropriate numbers or symbols.
To demonstrate the system, Dr. Cloutier entered symptoms typical of a condition called Turner's syndrome. The computer replied with a diagnosis stating that there was 100% compatibility with the symptoms, and it noted that a number of additional symptoms had not been listed and gave a reference to the literature. Time of computer use was 14 seconds.
Chief Limitation Clinic Consultants have noted in the system if there is incomplete programming. When the system is completely developed, Dr. Cloutier sees it as having potential usefulness particularly in the occasional patient whose symptoms do not fall into an easily recognizable pattern. He also believes it would be an excellent teaching tool.
Developers of the system emphasize that "it is not the purpose of the computer system to make a diagnosis but rather to provide information to assist the physician in making his or her own diagnosis." The elements that determine the degree of usefulness of the system are adequacy of information entered by the programmers and acuity of the examining physician.