A comparison of two or three radiographic views in the diagnosis of skull fractures.

The aim of the study is to determine whether a two rather than a three-view skull series is adequate for diagnosis of a skull fracture given a reliable history of the site of trauma. The radiographs of 50 patients who were diagnosed and managed as having sustained skull fractures were randomly mixed with 200 normal skull series and viewed independently by three observers. For all the film series viewed (a total of 1500 for the three observers), the diagnostic confidence level for two films was 94.4%, and for three films 94.6%. Of a total of 150 skull fracture series viewed as two films, 87 (58%) were correctly diagnosed with a confidence level of 92.7%. When viewed as three films, 92 (61.3%) were correctly diagnosed with a confidence level of 93%. Combined with analysis of false-positive and false-negative results, no statistical difference could be detected between a two or three film skull series. A two-view skull series has no statistically deleterious effect on either diagnostic accuracy or confidence of interpretation when compared with a three-view series given an accurate clinical history.