Affiliations with medical care providers.

Two analytical techniques are applied to a household survey of Rhode Island residents to develop the concept of an affiliation with medical care provider. A modification of a deviant case analysis is used to examine those people who are extreme in terms of the numbers of affiliations that they have with medical care providers. Overall, 4.1% of the sample have no affiliations and 7.3% have four or more affiliations. One of the most important variables in distinguishing those people with an unusual number of affiliations is a subjective variable of health status--whether they worry about their health. Nonparticipants rarely worry about their health and are generally in good health, while those with a high number of affiliations worry a great deal, regardless of their actual health status. A multiple regression analysis reveals that the variables of number of health conditions, sex, insurance, worry about health, income, disability days, and family structure are significantly related to mean number of affiliations with an R2 of 19.5%. Discussion includes the study's implications both for greater understanding of how and why people seek care and for application in the further development of a social psychological model of health behavior.