I met Herb Sherman socially at a time when his wife had just undergone mitral valve surgery at one of the Harvard teaching hospitals. Herb told me that she had had her valve replaced with a porcine valve, and he had just come from the library where he looked at the experience with porcine valves. He said that he found a shocking paucity of experience in animals as well as in people, but in animals before the valves were used in people, and in people before the valves were widely advertized as an important advance. At that time Herb was in charge of the communications satellite program at the Lincoln Laboratories. He said to me, 'You know, I, I'm responsible for the team that puts up communications satellites'. He said, 'If I did as little research or if as a little research were done on a... a satellite as I find done in medicine on heart valves, if I really were to launch a satellite with as little basis on... in research', he said, 'I would lose my job without question'. We began a friendship at that time. Within a short period of time he took a leave of absence from the Lincoln Laboratories. He and his colleague Barney Reifen were convinced that computers had something to offer to medical care, and they came, fully funded by Lincoln Laboratories, as they worked with me at Beth Israel Hospital to explore the potential usefulness of communication satellites... potential usefulness of computers to... to medical care.

Our first experiment was to determine how useful a computer might be in making a specific medical diagnosis, and what Herb decided to look at was the accuracy with which doctors diagnose streptococcal... streptococcus as the cause of sore throats. It took very little time before Herb decided that a computer would not be nearly as useful as a pencil and a piece of paper for carrying out the experiment that he proposed. What the experiment involved looking at... training a group of high school graduates as to the symptoms and signs that are consonant with the diagnosis of strep throat, and then comparing what a team of people that he and Barney trained, where the... the success of that team as compared with doctors and nurses who were looking at the same group of patients. They proved to have an accuracy of something of the order of, of 80 or 85%. The doctors were closer to 60 or 65%. This is before the, of course, the tests were done to ascertain whether or not strep was the cause of the sore throat. But that was the first study that Herb and Barney were involved in, and subsequently they remained at Beth Israel and came to the School of Public Health with me. When we set up the Center for the Analysis of Health Practices, they were invaluable in terms of their work with physicians like Howard Frazier and Don Berwick, and with colleagues, other colleagues, at the School of Public Health.