Severe head injury in children : analyzing the better outcome over a decade and the role of major improvements in intensive care

• Published in: Child's nervous system 1998-04, Vol.14 (4-5), p.195-202
• Main Authors: LEVI, L, GUILBURD, J. N, BAR-YOSEF, G, ZAAROOR, M, SOUSTIEL, J. F, FEINSOD, M
• Format: Article
• Language: English
• Subjects: Adolescent; Biological and medical sciences; Child; Child, Preschool; Craniocerebral Trauma - diagnosis; Female; Glasgow Coma Scale; Health Services - standards; Humans; Infant; Infant, Newborn; Injuries of the nervous system and the skull. Diseases due to physical agents; Injury Severity Score; Intensive Care Units - organization & administration; Intensive Care Units - standards; Male; Medical sciences; Prospective Studies; Traumas. Diseases due to physical agents; Treatment Outcome; Tropical medicine
• ISSN: 0256-7040; 1433-0350
• DOI: 10.1007/s003810050210

We suggest a few possible explanations, including improvement of intensive care, as the main cause, for the improved outcome after severe head injury in children and present the predictors of outcome observed in a contemporary series. From January 1984 to June 1988 we saw 117 children (ages 0-14) with postresuscitation GCS (Glasgow Come Scale) scores of 3-8. The more recent cohort of children seen in 1994-1996 was made up of 152 patients. Apart from standard statistics we used a segmentation method called CHAID (SSPS software). Previously known predictors of outcome are found still to apply in our series. Although in the recent period there was a lower proportion of patients with GCS 3-4 (11% versus 32%), a higher percentage had suffered multiple trauma (56% versus 33%). The rates of craniotomy and of ICP monitoring were similar (66% and 61%). Comparison of the two cohorts for outcome at discharge and through 1 year shows that mortality fell from 33% to 10% and the proportion achieving improvement of neurological status increased from 24% to 56%. CHAID analysis showed that the mortality rates of patients within specific groups declined significantly over the two periods: (1) a significant reduction in mortality was seen in patients with GCS 5-7, especially those with diffuse axonal injury (DAI) (17.3% to 0%); (2) no child admitted in shock survived in the earlier period, whereas 7 with GCS 4-6 survived during the recent period. The best model for mortality prediction includes GCS, and in the GCS 4-7 subgroup, the presence of subdural hematoma. It seems that the trend toward better immediate outcome is continuous, and this is the more striking when the severity of injury is taken into consideration. Our belief is that the modern medical and surgical techniques, although incurring higher costs and necessitating ongoing intensity, are well worth the effort.