Clinical decision making: recognition of confusion and application of restraints

The purpose of this study was to test the effects of patients' social, behavioral and medical characteristics on nurses' clinical decision for patient confusion.

Factorial survey.

A random sample of 100 registered nurses scored 30 unique situation based vignettes on the likelihood the patient described was confused and whether restraints should be used. Every vignette contained a combination of eight independent patient variables: age; gender; race; patient affect; type, seriousness and time of patient act and medical diagnosis.

Examination of the recognition model and the intervention (applying restraints) models showed that patient variables explained approximately 40% and 43% of the variance respectively. Type of act, seriousness of act, time of act, and patient age all significantly increased nurses identifying patients as confused. The variables that were strong predictors for recognizing confusion were strong predictors for restraint use except that patients who exhibited a negative affect were more likely to be restrained than patients' who were portrayed as pleasant.

Confusion is a complex clinical phenomenon with many dimensions. Confusion tools should be standardized and routinely used.