Can we predict outcome following subarachnoid haemorrhage?


By Hiren C Patel 

I would urge all readers to try and predict outcome at each stage of the journey of the patient presenting with subarachnoid haemorrhage. I say this as it not only allows one to engage the patient’s family and provide a realistic idea of the patient’s journey, but it also serves to gear the clinician’s mind to the problems that may be encountered for each individual patient.

So what of the answer to the question? In simple terms, I would argue that outcome is easy to predict on the whole. Most of the patients that we have under our care, present in good grade (70% in our institution), most undergo early intervention in a controlled setting (coiling 80%, clipping 20%), and most if not all (95%) will have a good outcome (according to us, the treating physicians) and are discharged home after a median hospital stay of eight days. In this, the majority group prediction can be easy and summarised by a friendly reassuring statement of “on balance, I will have you home in just over a week”.

Where this “simple concept” falls down is for those patients that are not in good grade as outcome in this cohort is truly unpredictable. This group of patients is heterogeneous and incorporates patients that are poor grade on account of ictus, seizures, early hydrocephalus and those that are true poor grade on account of brain injury. In this setting, many investigators, including myself, have tried to correlate outcome with presenting features, and have found no consistent association apart from pupillary response (signifying brain stem dysfunction) and age. This has an impact on the grand scheme of outcome prediction in patients that present with a subarachnoid haemorrhage because in a quarter of patients, at least, outcome cannot reliably be predicted at initial onset.

In addition to not being able to predict outcome accurately in about a quarter of patients, there are other issues which affect outcome prediction even in the good grade patients. This is partly because what, we as physicians, consider a good outcome, is not echoed by our patients.

So if as it would appear, my opinion distills down to “that in general outcome is difficult to predict” why do actively I encourage clinicians to predict outcome?

I do this in the knowledge that subarachnoid haemorrhage (even for those in similar grade) is a heterogenous and dynamic condition/disease. As a result, each patient remains unique in their underlying injury severity, radiological findings, and requirement for intervention. Each patient also has differing responses to the insult, and to treatments instituted. These differences in presentation and treatment also contribute to complications that develop, all of which contribute to outcome. I think that it is the day-to-day appreciation and prediction of how each patient is likely to behave following the ictus, respond to treatments, and develop complication(s) that ultimately determines overall outcome. It is this that we do not know and what drives my opinion that we cannot predict outcome following subarachnoid haemorrhage.

Hiren C Patel is with the Brain Injury Research Group, Greater Manchester Neurosciences Centre (GMNC), Salford Royal NHS Foundation Trust, The University of Manchester, UK.