Management of poor grade sub-arachnoid haemorrhage - clinical judgement v/s a formal model

The poor grade subarachnoid haemorrhage patients represent a unique cohort with lack of clear treatment protocol. Most neurosurgical units in the UK will manage them at local hospital until they make a significant recovery, this period can put them at higher risk of rebleed while with aggressive treatment a significant subset can achieve a favourable outcome. Identification of this subset is difficult and decision to treat them is associated with significant commitment of neurosurgical and ITU resources. Recent paper by Szklener has come up with a scale for prognostication in this subgroup of patients. We wanted to check the validity of this scale in our patient population and see if this scale can be used to guide early patient transfer and aggressive management at the Neurosurgical unit.

We retrospectively reviewed our referral database for all poor grade subarachnoid patients referred over 2 years. Demographic information, Fisher and WFNS scores, admitting leucocyte count and outcome information as per MRS were obtained. These were scored as per the scale suggested by Szklener.

A total of 115 poor grade subarachnoid patients were referred over the study time frame. 47 of them were accepted for admission . 18/47 patients achieved a favourable outcome (GOS4-5). Only 1 patient managed in peripheral hospital had a good outcome. There was a significant association between Szklener's score and achieving a favourable outcome p = 0.002.

A selective admission policy could work specially with current economic climate, achieving outcomes comparable to admit-all. However, to optimise outcomes for all patients an aggressive standardised management at peripheral hospitals and a uniform admission policy assisted by Szklener score may be adopted. Szklener's model predicts the outcome better than WFNS and age but more validation is needed.