Does the requirement of a craniotomy predict outcome? A preliminary investigation

The literature has been replete with reports that persons who require craniotomy for treatment of their traumatic brain injury have a far worse outcome. The majority of these reports have utilized the rather global Glasgow Outcome Scale as a determinant of outcome. This paper sought to evaluate the effect of craniotomy on outcome as measured by the DRS. Data was collected on 341 persons (mean age 37.7 years) with traumatic brain injury treated at the Level I trauma centre, who required inpatient rehabilitation. Surgical interventions were classified as 'no surgery', 'one cranial surgery', or 'two or more cranial surgeries'. Initial GCS scores revealed 44 persons at GCS 3-5, 102 persons at GCS 6-8, 83 persons at GCS 9-12 and 112 persons at GCS 13-15. The DRS was administered to each person at discharge from in-patient rehabilitation. Mean DRS scores were 7.07 for GCS 3-5, 6.03 for GCS 6-8, 6.53 for GCS 9-12, 5.57 for GCS 13-15 groups. A factorial ANOVA revealed an interaction between initial GCS and surgical status. Univariate ANOVA's demonstrated significant differences in the GCS 3-5 and GCS 13-15 groups, suggesting a relationship between need for surgical intervention and less favourable outcome among persons who required in-patient rehabilitation. However, no differences were demonstrated in the GCS 6-8 and GCS 9-12 groups. It appears that requiring surgical intervention is prognostic at only the extremes of the GCS categories and, thus, further investigation may reveal the limited role of need for surgical intervention injury in predicting outcome in persons with initial GCS 6-12.