Predicting cognitive recovery in post resuscitation cardiac arrest patients using cerebral oximetry and CT brain imaging.


Cardiac arrest is associated with a number of cognitive outcomes including short and long term memory impairments. The immediate post resuscitation cognitive recovery of in-hospital CA remains poorly understood. Cognitive outcomes in post resuscitation CA patients were tested mainly in out-of-hospital CA, at least 3 month in average after the incident, and has not been specified with regards to memory functions. The differences in cognitive outcomes between patients may indicate improved cerebral resuscitation during cardiac arrest as Fenwick, P. demonstrated when finding higher oxygen levels in CA patients with positive cognitive functions.

Recent studies suggest that changes in brain oxygen saturation during the post resuscitation period may have the potential to indicate and predict recovery from cardiac arrest: When measured in the subsequent 24-48 hours post resuscitation, regional cerebral oxygen saturation (rSO2%) was significantly increased in the first 24h in patients who survived to discharge as opposed to those who did not (Parnia 2014). Storm C (2014) found that CA patients discharged with cerebral performance category (CPC) 1-2 had rSO2% values of 68% while those with CPC 3-5 had 58%.


This project will focus on characterising the cognitive recovery, specifically, memory functions in in-hospital cardiac arrest (IHCA) resuscitation. To do this, a battery of memory tests and overall cognitive screening will be used in conjunction with cerebral oximetry (CO) and in correlation with CT imaging. This projects aims to:


  • Assess specific memory and cognitive outcomes in post resuscitation CA patients.

  • Evaluate the CO input during and post resuscitation as a potential predictor to these outcomes.

  • Establish a data-driven identification of imaging features relevant to outcome, using whole-brain Support Vector Machine.


The aim of this project is to assess specific memory functions in IHCA patients (conscious and comatose who regain consciousness) during their hospital stay and as close as possible to their post resuscitation regain of consciousness, and long term post discharge. This will be with comparison to non-CA ICU patients’ memory functions in hospital and post discharge in order to evaluate the effect of cardiac arrest and cardio-pulmonary resuscitation (CPR) on cognitive outcomes. Indeed, the mere in-hospital medical procedure and hospitalization may affect the cognitive state of patients. For instance, in the early weeks after major noncardiac surgery, many patients, especially older ones, develop cognitive changes, impaired memory and concentration, mild personality changes, and emotional instability. These changes, which can be significant, are now commonly referred to as postoperative cognitive dysfunction (POCD).


In our study, long term [Rivermead Behavioural Memory test (RBMT)] and short term [Doors and people Test (DPT)] memory as well as a battery of cognitive functions (Hampshire’s paradigm) will be assessed in several follow ups: short term (pre hospital discharge and one week after) and long term (1, 3, 6 and 12 months afterwards) in conscious IHCA survivors. CO will be monitored during their resuscitation and for 24-48h following their arrival at the ICU.


The importance of recording CO 24-48 hours post resuscitation implies firstly for determining the optimal balance between oxygen delivery and uptake in the brain. This is since one of the major causes of death and neurological injury after cardiac arrest is delayed ischemia combined with oxygen free radical mediated reperfusion injury. CO was found to be a potential predictor for survival to discharge (Parnia 2014): The median (IQR) rSO2% during the first 24 h in patients who survived to discharge compared to those who did not survive were significantly higher 68.2% vs. 62.9%.


CT brain imaging holds promise as useful prognostic adjuncts in patients with hypoxic-ischemia brain injury caused by cardiac arrest. A distinction between gray and white matter (or its loss) has been shown to be associated with outcome. Studies have suggested that a ratio of Hounsfield units in the gray matter versus the white matter <1.18 can accurately predict poor outcome. In addition, a decrease in a whole brain Hounsfield units’ value, especially in combination with traditional clinical features, was highly specific for predicting poor outcome. Inamasu et al. studied the CT scans of comatose post-cardiac arrest patients immediately after resuscitation and found that both loss of gray and white matter distinction and the presence of sulcal effacement were correlated with poor outcome.


In this project we aim to evaluate a possible correlation between CT scans and post resuscitation survival and cognitive outcomes of CA patients.

Current Research Outline