This research attempted to explore the effects of three risk factors, potential patient consequences, expected physician response, and physician status, on registered nurses' communication choices and ratings across three possible phases of a situation in which a nurse believes a physician's order is not in the best interests of the patient. The situation seems to place the nurse in a bind as she attempts to satisfy both patient welfare and physician concerns and finds the extreme choices of compliance or noncompliance undesirable. Communication researchers suggest the logical response to a bind is indirect, avoidant, evasive communication--enactment of the "Doctor-Nurse Game." Altering risks associated with the patient and the physician was suspected of changing the bind-like quality of the inappropriate-order situation, and correspondingly, nurses' responses to it.

Data were collected from 298 female registered nurses at two Toledo-area hospitals. One of eight hypothetical situations (2 x 2 x 2 design) was presented to each subject who then indicated her most probable communication behavior in three sets of choices: initial handling of the situation (dominant to submissive choices); initial comments to the physician to secure change (direct to indirect choices); and whether the matter would be pursued further if an initial physician confrontation is unsuccessful.

Nurses' responses indicated that, in general, the inappropriate-order situation is more a bind than not a bind. Extreme choices of compliance and noncompliance were rejected as was the most direct comment to the physician. However, the bind seems to loosen over time, as nurses reported they would use more direct comments and more forceful action when their initial comments were unsuccessful. Risk factors, however, did loosen or tighten the nature of the bind. More avoidant initial communication was more likely when serious patient harm and, to a lesser extent, a negative physician response were expected. Physician status was a less significant factor and exhibited sample-specific effects. The results seem to reflect the changing nature of the nurse-physician relationship, but with a greater leaning toward the contemporary role relationship.