The study aims to develop a novel perioperative noninvasive delivery of oxygen (DO2) and hemodynamic index in the prediction of postoperative outcomes including post-operative acute kidney injury (POAKI), perioperative neurological cognitive disorders (PONCD) and post-operative atrial fibrillation (POAF).

Our study also aims to study the effect of DO2 and hemodynamic parameters (measured noninvasively) on the development of PONCD, POAF and POAKI in cardiac surgical patients. We aim to develop a novel DO2 index and hemodynamic index predictive of PONCD, POAF and POAKI.

The secondary aim would be to develop real-time machine learning aided algorithms in the analysis of noninvasive continuous hemodynamic and DO2 data, to predict and guide perioperative treatment of POCND, POAF and POAKI in the perioperative setting.

Postoperative acute kidney injury, AF and neurocognitive disorders (POAKI, POAF and PONCD) are common complications post cardiac surgery. Postoperative neurocognitive disorders occur in up to 73% of cardiac surgical patients, while our previous studies have demonstrated the incidence of POAF and POAKI to be 14.7% and 29.9% respectively. These perioperative complications have been shown to be associated with adverse short and long term outcomes. Apart from increased hospital length of stay and mortality, their adverse effects also extend far beyond the acute perioperative period of insult. Postoperative AKI has been shown to be associated with progression to long term renal dysfunction. Postoperative atrial fibrillation was initially thought to be a benign phenomenon, however, it has been shown to be associated with increased risk of long term stroke and mortality (Hazards Ratio 1.26 and 1.2 respectively). Postoperative delirium has is also associated with accelerated cognitive decline in Alzheimer's patients, long term cognitive decline, dementia and functional decline after discharge to the community. These phenomena remain ever pertinent in the setting of high risk and aging surgical patients, and thus need to be addressed.

The need for hemodynamic optimisation and goal directed therapy remains central to current care standards in anesthesia. This highlights the need for the quantification and improvement of cardiac output and more recently, the emerging focus on perioperative oxygen delivery (DO2) in improving clinical outcomes. Oxygen delivery can be calculated based on the oxygen flux equation: DO2 = CO x (Hb x SaO2 x 1.36); this concept of optimizing systemic oxygen delivery rather than hemodynamic parameters alone has been increasingly studied in recent times. Emerging studies involving patients undergoing cardiopulmonary bypass (CPB) have demonstrated the optimization of DO2, nadir DO2 and duration below optimum DO2 to be the most specific factors of postoperative AKI . The importance of optimizing DO2 in improving clinical outcomes is also evidenced in the critical care period, where Raimundo et al recently demonstrated that patents with a mean DO2 Index of 325ml/min/m2 were at significant risk of developing AKI which progressed to chronic renal failure, as opposed to patients who maintained a mean DO2 Index of >405ml/min/m2 . The utility of optimizing DO2 in the avoidance of postoperative delirium and atrial fibrillation however remain unknown, with previous studies being retrospective, small in sample size with conflicting results.

The use of invasive hemodynamic monitoring has been used extensively in the development of perioperative goal directed therapy protocols and hemodynamic optimization. Parameters such as MAP, CI and SVV are commonplace in the operative and ICU settings, and are used to titrate fluid, vasopressor and inotrope use. However, a delicate balance between pressure (CI, MAP) and flow (SVRI) exists in systemic organ perfusion, which may be translated into clinical outcomes such as POAF, POAKI and PONCD. Separate studies have demonstrated variations in vascular resistance to be associated with increased risk of postoperative delirium and AKI, despite the maintenance of high MAP levels. These studies indicate the possible inadequacy of MAP as the sole marker of systemic organ perfusion, and the need for further hemodynamic parameters to be studied. Our study would be the first to investigate the interplay between these SVV, SVRI, CI and MAP using non-invasive hemodynamic monitoring techniques.

In the past, this monitoring of continuous hemodynamics required invasive lines, while the continuous monitoring of DO2 was previously impossible. The use of invasive lines has been associated with multiple complications including infection, thrombosis and vascular aneurysms, while continuous DO2 monitoring would involve continuous point of care testing of hemoglobin which is impractical. Therefore, the risk benefit ratio of placing invasive lines remains controversial particularly in the routine well patient who however remains at risk of perioperative complications.

The advent of new technology has allowed for the real-time noninvasive continuous individual measurement of hemodynamics and oxygen delivery. We propose the integration of 2 new devices so as to provide the continuous noninvasive monitoring of hemodynamic data and real-time noninvasive hemoglobin and oxygen saturation - the combination of which would enable us to calculate real time systemic oxygen delivery. Continuous noninvasive hemodynamic data (the Edwards Clearsight) is based on oscillonometry and the Penaz technique which has been historically established and used in clinical practice. Continuous noninvasive hemoglobin and oxygen saturation monitoring (the Masimo SET) has been validated, and is based on established principles of spectrophotometry, oximetry and the beer lambert's law . With technological improvements over the years, these noninvasive devices have been individually evaluated against clinical gold standards in multiple studies and found to be clinically applicable in predicting perioperative hemoglobin and hemodynamic trends . Therefore, we propose the novel integrated use of these two devices in deriving a real-time noninvasive DO2 and hemodynamic index predictive of postoperative atrial fibrillation, AKI and cognitive dysfunction in cardiac surgical patients.

The integrated use of a noninvasive DO2 and hemodynamic index has yet to be studied in the context of POAF, POAKI and PONCD. Data obtained through our study may facilitate the development of new integrative technology, paving the way for the development of single index which can predict organ specific perfusion. This rich source of continuous noninvasive data will be used in the development of algorithms and mathematical models in machine learning. This would allow for the provision of precise goal directed therapy to moderate or high-risk surgical patients, especially beneficial to the geriatric setting. The potential reductions in POAF, PONCD and POAKI achieved through the clinical application of our study findings confer significant perioperative benefits which extend into the community health setting due to the long-term sequelae of these diseases.

Inclusion Criteria:

• Undergoing elective cardiac surgery such as valve replacements or CABG
• Age above 21 years old
• Ability to provide informed consent

Exclusion Criteria:

• Emergency cardiac cases
• Patients on IABP postoperatively
• Patients who decline to participate in this study
• Pregnant women

Patients scheduled for elective cardiac surgery and age above 21 years old.