Am Fam Physician. 2001;63(1):131-132
Computerized tomography (CT) was first used in the early 1970s, but it has remained unclear which patients with minor head trauma should undergo this method of diagnostic imagining. Two published studies reported that the selective use of CT based on clinical findings identified 96 to 98 percent of patients with significant abnormalities. However, the threshold for use of CT remains especially low in emergency department settings, where physicians are less willing to accept the risk of missing a significant abnormality not suspected on clinical evaluation alone. Haydel and colleagues performed a study to develop and validate a set of clinical criteria that could be used to determine whether patients with minor head injury need to undergo CT scan.
In the first phase of the study, 520 patients who were at least three years of age and who presented within 24 hours of a minor head injury were prospectively evaluated to determine which specific clinical findings correlated with positive findings on cranial CT scan. A minor head injury was defined as a loss of consciousness in the presence of a normal neurologic examination and a normal score (15) on the Glasgow Coma Scale. A positive CT scan included findings of a subdural, epidural or parenchymal hematoma, subarachnoid hemorrhage, cerebral contusion or depressed skull fracture. From this study, it was determined that seven clinical findings had a high positive predictive value for positive findings on CT scan (see the accompanying table).
In the second phase of the study, these seven findings were prospectively applied to a second cohort of patients who also presented with a minor head injury. Of the 909 patients between three and 94 years, 697 had one or more of the seven clinical findings; 57 had a positive finding and 640 had a negative finding on CT scan. All 57 patients had at least one of the seven findings determined by the first phase of the study. This resulted in a sensitivity of 100 percent, a negative predictive value of 100 percent and a specificity of 25 percent. There were 640 patients with one or more findings present who had normal CT scans. Finally, 212 patients with minor head injuries had none of the seven clinical criteria, and all of their CT scans were negative. Of all patients from both phases (n = 1,429), six required surgery for their injury.
The authors conclude that in patients who present with minor head injuries, the need for a cranial CT scan can accurately be determined by applying the clinical criteria validated by this study. They note that even a 10 percent reduction in the number of CT scans performed in this setting would produce a substantial decrease in health care costs.
|Finding*||Total (n = 520); no. (%)||Positive CT scan (n = 36); no. (%)||Negative CT scan (n = 484); no. (%)||P value†||Likelihood ratio‡|
|Short-term memory deficits||9 (2)||5 (14)||4 (1)||<0.001||15.0|
|Drug or alcohol intoxication||180 (35)||22 (61)||158 (33)||0.001||11.0|
|Physical evidence of trauma above clavicles||338 (65)||32 (89)||306 (63)||0.002||11.0|
|Age >60 years||42 (8)||6 (17)||36 (7)||0.05||3.0|
|Seizure||24 (5)||4 (11)||20 (4)||0.05||3.0|
|Headache||123 (24)||12 (33)||111 (23)||0.16||2.0|
|Vomiting||47 (9)||4 (11)||43 (9)||0.65||0.2|
|Coagulopathy||1 (<1)||0||1 (<1)||0.78||0.15|