Tentative diagnoses facilitate the detection of diverse lesions in chest radiographs.

The influence of clinical history upon detection has been frequently studied. In a unique investigation, detection of diverse abnormalities commonly encountered in practice was tested, and the influence of clinical history was conceptualized as suggesting a tentative diagnosis. Categorical prompts that were correct for specific abnormalities resulted in detection superior to that with unprompted reading. Prompts that were plausible but incorrect for abnormal and normal cases led to very few false-positive responses, and subsequent reports of specific abnormalities were as accurate as in unprompted reading. It is concluded that appropriate clinical history results in improved perceptual performance rather than simply improved decision-making.