The proposed research will examine a set of factors, in addition to the occurrence of an untoward event, that is, an undesirable outcome of medical care, that may affect the probability of malpractice claims. Untoward events include poor outcomes of surgery - including death, miscarriages, babies born with malformations and adverse reactions to drugs. As predictors or causal factors leading to claims of malpractice, the research will focus on the characteristics of the interactions among the patient-provider-delivery setting, provider vulnerability to suit, salience to the patient or family and the presence of programmatic efforts to manage high risk care situations and the occurrence of untoward events. The question is: Can these factors explain why, for patients experiencing the same poor outcome and with the same level of merit to a potential claim, some lead to malpractice cases and others do not? The study rests on comparing similar untoward events where claims do and do not arise for five types of untoward events occurring in five HMOs during a five year period. It is significant that the study will be based on enumeration of the population of untoward events; this is made possible because of the defined populations and record keeping systems of the HMOs participating in the study. Other studies generally have employed claims; problems have existed in defining, either for the patient or physician, the population at risk and the cases not brought. This study will also be able, because of access to enrollment, medical and personnel records, to characterize the patient, physician and system and their relationships in ways not employed in any existing study. It is also significant that the study will incorporate professional assessments of the probable merit of a potential claim of malpractice. Because all of the research sites are HMOs and share certain characteristics - group practice, prepayment, comprehensive benefits, large, urban - a number of other variables hypothesized to influence malpractice will be controlled.
