Neurological prognostication

Dr Matthew Rowland

Intensivists are frequently asked to assess the validity of providing ongoing treatment in patients with the potential for significant brain damage following cardiac arrest, trauma and stroke. As a result, the holy grail would be the ability to accurately and reliably predict longer term neurological recovery early in the intensive care unit admission. This is driven by the desire for both families and clinicians to predict the neurological prognosis as well as wider concerns about intensive care resources and the health-care and societal costs associated with continuing potentially futile treatment.

The potential risks associated with inaccurately predicting a poor prognosis are huge – especially in those patients with the potential to realise good outcomes had they been provided with adequate supportive care. Furthermore, defining a “good neurological outcome” is extremely hard with huge individual variation between patients, a range of methods and techniques used and arbitrary dichotomisations into “good” and “bad” – often neglecting other aspects of recovery such as quality of life and happiness.

Despite significant advances in biomarker and imaging technology, predicting longer term neurological outcomes remains challenging in the intensive care population - especially during the era of the routine use of therapeutic hypothermia. Although a large number of different approaches to neurological prognostication are currently in use ranging from clinical assessment to blood biomarkers, somatosensory evoked potentials, EEG reactivity and neuroimaging findings, the optimum timing and methods are still unclear.

This talk will aim to summarise some of the challenges and issues around neurological prognostication in ICU patients in a range of conditions and to highlight the latest evidence to support biomarker and imaging driven approaches and potential future advances.