Should we perform ocular fundus in patients with severe brain trauma?

Introduction

The high incidence of intracraneal lesions of patients with brain trauma (BT) is well known. The aim of this study is to assess the incidence of an extracraneal injury: retinal haemorrhage (RH) in patients after BT.

Methods

A prospective study was performed during 6 months in a 15-bed medical–surgical ICU. RH was diagnosed by an ophthalmologist in the first 4 days after admission to the ICU. Age, personal history, ISS, RTS, SAPS, SOFA score, APACHE score, lactate levels and GCS were recorded on arrival at the ICU. Cerebral edema, brain CT findings, days under mechanical ventilation and need for neurosurgery at any time were also recorded. The statistical analysis was performed using the SPSS 11.0 program.

Results

Nineteen patients were included in the study (18 males), age 56 ± 22.56 years. Five RH were described in this group, all males. The mechanism of traumatism was fall in 80% and traffic accident in 20% in the RH group; in contrast, fall in 35.7% and traffic accident in 42.9% in the nonretinal haemorrhage group (NRH). There were no differences between groups in age, ISS, RTS, APACHE score, SAPS, SOFA score, lactate levels and days under mechanical ventilation. Eighty per cent of the patients with RH were alcoholic, and 21.4% of the NRH. The GCS at admission in the ICU was 3 in 60% of the patients with RH and 35.7% in the NRH group. Sixty per cent of the patients with RH required neurosurgery; in contrast, 7.1% in the NHR group. Eighty per cent of the patients with RH developed cerebral edema during their stay in the ICU, while this was 50% in the NRH group. Sixty per cent of the patients with RH had frontal contusion in the CT and 14.3% in the NRH group. Twenty per cent of the patients with RH had epidural haematoma; in contrast, 7.1% in the NRH. Sixty per cent of the patients with RH had epidural haematoma; in contrast, 14.2% in the NRH. All patients with RH were discharged. Two of the patients with RH had unilateral blindness at discharge from ICU.

Conclusion

Although it is not possible to draw inferences with so small a number of patients, future prospective studies are warranted to corroborate our findings. RH is a quite common finding in our study (26%). All these patients are male with an antecedent of alcoholism. There was more severe head injury in this group. GCS at arrival is lower. This group have more incidence of frontal contusion, epidural or subdural haematoma, cerebral edema and required more neurosurgical interventions.