Characterizing the association between regional cerebral oxygen saturation and neurological impairment after cardiac surgery
It is well-known that patients undergoing cardiac surgery are at risk of developing acute, fluctuating (i.e. delirium), and/or long-term post-operative neurological complications. Incidence of neurological impairment can be as high as 60%, and has been associated with increased length of hospital stay, mortality, and severe cognitive impairment. While the exact etiology is unknown, it has been postulated that cerebral hypoperfusion may be associated with neurological deficits after cardiac surgery. Near-infrared spectroscopy (NIRS) is a commonly used, non-invasive technique that measures regional cerebral oxygen saturation (rSO2) during cardiac surgery, acting as a surrogate marker for cerebral perfusion. However, research has demonstrated inconsistent results regarding the relationship between NIRS-derived rSO2 and post-operative neurological impairment in cardiac surgical patients. The primary objective of this thesis is to investigate whether low rSO2 levels are associated with post-operative neurological impairment in patients undergoing cardiac surgery. A systematic review was undertaken to evaluate sources of variability present in this body of research. This research project also assessed the feasibility of using robotic technology to quantify neurological functioning before and after surgery. Following an analysis of 40 patients undergoing coronary artery bypass grafting surgery, it was determined that subtle pre-existing impairment, detected by robotic technology, was a stronger indicator of post-operative dysfunction than intraoperative rSO2. Subsequent chapters describe the expansion of the study to include post-operative delirium and other cardiac surgeries, as well as relating rSO2 with other hemodynamic variables such as mean arterial pressure to quantify the degree of disturbed cerebral autoregulation. Preliminary findings also suggest that patients who experienced delirium may perform worse on the robotic tasks prior to surgery, suggesting that subtly impaired performance may be a reflection of poor cognitive reserve. In addition, I performed a pilot study that quantified neurological impairment in higher risk patients undergoing non-invasive valve replacement. The findings presented in this thesis provide evidence for the inclusion of perioperative neurological assessment as part of the standard of care for cardiac surgical patients. As well, it provides a strong framework for determining the benefits of measuring and monitoring rSO2 during cardiac surgery. Future research will include the role of delirium and pre-existing functioning to determine whether low rSO2 and/or disturbed cerebral autoregulation is associated with post-operative neurological impairment.