John R. Clarke, MD, FACMI

Year Elected: 
Drexel University

Currently: Drexel University

John Clarke received his BA at Wesleyan University in 1965 and his MD at the University of Pennsylvania in 1968. He then trained in general surgery at Presbyterian-St. Luke’s Hospital in Chicago and St. Joseph Mercy Hospital in Ann Arbor from 1968 to 1975. After a fellowship in clinical trauma and trauma research at Boston City Hospital and Boston University in 1975 to 1977, he began a career in academic trauma surgery that led to his current position as Professor of Surgery at Drexel University in Philadelphia. A fellowship in cognitive science at Penn in 1982 and then a sabbatical with Dr. F. T. deDombal on computer-aided (Bayesian) diagnosis of acute abdominal pain in 1984 were turning points in Dr. Clarke’s career. One of the few surgeons who saw the promise of informatics and decision sciences at that time, he was a founding member of the Society for Medical Decision Making and forged collaborations with computer scientists who were seeking a rich domain for work on clinical decision support. Today, in addition to his surgical appointment at Drexel, he is an adjunct professor in the Department of Computer and Information Science at the University of Pennsylvania.

Dr. Clarke has shown us that surgical decision making can be descriptively represented by decision analysis and that the use of decision analysis improves the appropriateness of surgical decisions. Furthermore, through his research, he has demonstrated that computer-based decision support can provide real-time quality assurance for trauma resuscitations, a clinical situation that is time-sensitive and involves multiple interacting diagnoses. Multiple disparate computer-based recommendations can be coordinated without conflict by representing the actions as goals, to be subsequently coordinated by an appropriate planning algorithm. Graphic representation of information, such as bullet paths in trauma surgery, can be integrated with categorical and continuous variables by converting to probabilities within a Bayesian belief network. His TraumAID project, in collaboration with Dr. Bonnie Webber, Dr. Omolola Ogunyemi, and others, has shown us the relevance of informatics methods in surgical diagnosis and planning.