His current research interests involve using a range of profiling techniques, including metabonomic, transcriptomic and inflammatory profiling, to identify sub-phenotypes in sepsis that could predict a patient's responsiveness to specific therapeutic strategies and lead to a personalised approach to sepsis care. This approach has lead to the finding that stratifying patients with septic shock based on the way in which their genes are expressed (sepsis response signature (SRS) transcriptomic endotypes) identified groups who respond differently to corticosteroids in this condition. Ongoing work aims to translate this finding to the bedside to use point of care tests to identify patients who are most likely to benefit from these interventions.