Does chest compression during cardiopulmonary resuscitation provide sufficient cerebral oxygenation?

Background/aim Some of the patients suffering from cardiac arrest (CA) remain in a chronic unconscious state in intensive care units (ICUs). The primary aim of this study was to evaluate the efficacy of chest compression (CC) on cerebral oxygenation during cardiopulmonary resuscitation (CPR). As a secondary goal, we attempted to determine the effects of regional cerebral oxygen saturation (rSO2) values on consciousness and the survival rate using the Full Outline of Unresponsiveness (FOUR) scoring method. Materials and methods This observational preliminary study was carried out with 20 patients with CA who were hospitalized in ICUs. The rSO2 values measured by near-infrared spectroscopy were recorded during CA. FOUR scoring was used to determine the neurological status, severity of disease, and degree of organ dysfunction in survivors. Results Return of spontaneous circulation (ROSC) was gained in 8 (40%) of 20 patients. Maximum rSO2 values were higher in survivors than in nonsurvivors (P = 0.005). The mean FOUR score before CA was 11.50 ± 0.8 in survivors, whereas this value was 7.87 ± 0.7 for 1 week after ROSC (P < 0.0001). There was a significant positive correlation between the minimum and mean rSO2 values and the mean 1-week FOUR scores in survivors (r = 0.811, r = 0.771 and P = 0.015, P = 0.025, respectively). Conclusion Our results suggest that the maximum rSO2 values affect ROSC while the minimum and mean rSO2 values affect the post-cardiac arrest neurological outcome.


Introduction
When brain perfusion is impaired, an irreversible process begins for neurons within about 5-8 min (1). Impairment of brain perfusion during cardiac arrest (CA) almost always causes reduction of cerebral oxygenation, resulting in neurological damage (2). If the first intervention cannot be fast, or if adequate chest compression (CC) cannot be provided during cardiopulmonary resuscitation (CPR), sufficient cerebral oxygenation cannot be obtained. This is closely related to chronic unconsciousness or minimal consciousness state in cases of return of spontaneous circulation (ROSC) (3). It is defined that the threshold for ischemia is ≤50% or 20% decrease from the baseline value (4). Cerebral oximetry achieved by near-infrared spectroscopy (NIRS) technology provides information about a level of brain oxygenation and cerebral perfusion (5). NIRS is also used as a noninvasive monitoring technique that demonstrates the quality of CPR (6). Studies have shown that high regional cerebral oxygen saturation (rSO 2 ) values measured by NIRS during CA are consistent with the ROSC rate (7).
Despite frequent updates of international CPR guidelines, trained healthcare providers still perform inadequate CC with low cerebral perfusion pressure (8). Some researchers showed that the quality of CC correlated with poor neurologic outcomes (9). For this reason, some patients remain in a chronic unconscious state in intensive care units (ICUs) and the discharge rates are low (10). Although survival rates are better after CPR in hospital CA cases, a comprehensive study suggested that only 17% of these patients were discharged alive (11).
The primary aim of this study was to evaluate the efficacy of CC on cerebral oxygenation during CPR. As a secondary goal, we attempted to determine the effects of rSO 2 values on the survival rate and consciousness status of patients with ROSC using the Full Outline of Unresponsiveness (FOUR) scoring method.

Patient selection and approval
This observational preliminary study was approved by the Erciyes University Clinical Research Ethics Committee (ClinicalTrials.gov: NCT03062306) and carried out with 20 patients with CA who were hospitalized in the ICU of Nevşehir Public Hospital from May 2016 to August 2017. Informed consent was waived due to the emergency situation. All intubated patients with agonal status due to illness and ≥18 years of age who were followed in the ICU were included in the study.

NIRS application and rSO 2 recording
Procedures related to the study did not disrupt routine CPR practice. Patients with CA in the ICU were monitored with an NIRS device (INVOS 5100c, Somanetics, Troy, MI, USA) during CPR. Bilateral frontal noninvasive sensors of the device were applied on the frontal area of the patient. The device continuously recorded the right and left rSO 2 values, but we wrote down the values at 60-s intervals. The rSO 2 recording was continued until the CPR was terminated in non-ROSC patients, while it was continued for 10 min after ROSC in the patients with ROSC. The rSO 2 values used in the study were averages of bifrontal right and left rSO 2 values.

Cardiopulmonary resuscitation
CPR was performed by doctors or nurses in the ICU. All patients with CA were treated with advanced life support (ALS) in accordance with the 2015 guidelines of the European Resuscitation Council. Since the patients were followed in the ICU, they were already monitored by three-lead electrocardiography (ECG), pulse oximeter, and noninvasive blood pressure devices. All our patients had a fast onset of CPR due to follow-up in the ICU and CPR was started as soon as cardiac arrest occurred because they were being consistently monitored in the ICU. We defined the patients with ROSC as survivors.

FOUR scoring
FOUR scoring tests four neurological parameters. The number of components and the maximum score in each of the categories is four. FOUR collects all the requirements of a neurological consciousness examination and offers much more neurological detail (12). The FOUR scoring is shown in Table 1. We used FOUR scoring to determine the neurological status, the severity of disease, and the degree of organ dysfunction in patients with ROSC. FOUR scores were recorded for 1 week by a physician in the ICU once a day.

Exclusion criteria
Patients were excluded from the study if they had cranial events such as cranial trauma, intracranial hemorrhage, or cerebral ischemic vascular; pulmonary diseases that may affect oxygenation such as pneumonia or chronic obstructive pulmonary disease (COPD); FOUR scores lower than 10 before CA; or age under 18 years.

Statistical analysis
Statistical analyses were performed using SPSS for Windows 16.0 (SPSS Inc., Chicago, IL, USA). Data with normal distribution were presented as mean ± standard deviation. The paired sample t-test was used to evaluate the significance of continuous variable parameters. Pearson's correlation analysis was used for normal distribution to determine the correlation between numerical variables. Correlation coefficient (r) was evaluated as follows: 0.0001 to 0.249 is poor; 0.250 to 0.499 is moderate; 0.500 to 0.749 is strong; 0.750 to 1.000 is a very strong relationship. P < 0.05 was considered statistically significant.

Results
ROSC was gained in 8 (40%) of 20 patients who consisted of 8 (40%) males and 12 (60%) females. The mean age was 72.6 ± 4.2 years in survivors and 77.3 ± 6.5 years in nonsurvivors. The mean duration of CPR was 15.5 ± 6.6 min in survivors and 30.16 ± 6.6 min in nonsurvivors (P = 0.0001). Patient characteristics, ROSC, duration of CPR, FOUR scores before CA, and 1-week mean FOUR scores after ROSC are shown in Table 2.
There was no significant difference between survivors and nonsurvivors in terms of basal, minimum, and mean rSO 2 values, whereas maximum rSO 2 values were higher in survivors than in nonsurvivors (P = 0.005). The scatterplots of basal, minimum, maximum, and mean rSO 2 values of survivors and nonsurvivors are shown in Figure 1.
The mean FOUR score before CA was 11.50 ± 0.8 in survivors, whereas this value was 7.87 ± 0.7 for 1 week after ROSC (P < 0.0001).
There was a significant positive correlation between the minimum and mean rSO 2 values during CPR and the mean 1-week FOUR scores in survivors (r = 0.811, r = 0.771 and P = 0.015, P = 0.025, respectively). The minimum,

Discussion
Although there are many studies showing that NIRS reflects cerebral oxygenation well, we have not encountered a study that investigates the neurological outcomes with quality CC in humans. Findings in this study showed that the rSO 2 values measured by NIRS are closely related to the state of consciousness and survival of patients with ROSC.
The interval between collapse and start of CPR is one of the first and most important factors affecting the neurological status of survivors (13). It is challenging to determine this interval, especially in CA cases coming from outside of a hospital into emergency services. As the CPR duration is prolonged, optimal cerebral perfusion cannot be achieved until the underlying cardiac defect improves and consequently prolonged CPR is associated with poor neurological outcome (14,15). Kim et al. carried out a study about the optimal transition time to extracorporeal CPR for predicting good neurological outcome in patients with out-of-hospital CA and they stated that good neurological outcomes were significantly reduced 21 min after CA (16). All our patients in our study had a fast onset of CPR due to follow-up in the ICU.
The mean duration of CPR in our study was 15.5 ± 6.6 min in survivors. There was no significant correlation between CPR duration and 1-week FOUR scores (P = 0.699). Unlike in the literature, we observed that the good neurological status in survivors is mostly associated with minimum or mean rSO 2   CA: Cardiac arrest, rSO 2 : regional cerebral oxygen saturation, ROSC: return of spontaneous circulation, CPR: cardiopulmonary resuscitation, SD: standard deviation. Min: minimum, Max: maximum. 5%) than in nonsurvivors (18 ± 0.4%) (P < 0.001) (18). In our study, maximum rSO 2 values were higher in survivors (54.5 ± 2.4%) than in nonsurvivors (45.0 ± 34.7%) (P = 0.005). This may be evidence that high rSO 2 values during CPR may provide ROSC. Many scoring methods can be used to assess neurological status. The cerebral performance category (CPC) is a more effective scoring method to assess longterm neurological outcomes and survival after CA (19). Sanders et al. used the CPC score for estimating good neurological outcomes in an experimental swine model (20). They investigated the best compression-ventilation rates for a good neurological outcome. According to their study, the 100:2-CPR model had the best neurological outcome. In our study, we investigated the effect of rSO 2 values rather than compression-ventilation rates on neurological outcomes. Recently, the FOUR score has been a more commonly used scoring system to measure the depth of a coma. In our study, the mean FOUR score before CA was 11.50 ± 0.8 in survivors, whereas this value was 7.87 ± 0.7 after ROSC for 1 week (P < 0.0001). This was a natural result, since keeping the rSO 2 values above the threshold of 50% to prevent cerebral ischemia throughout CPR is quite difficult even with good CC. In the study of Kämäräinen et al., rSO 2 values were 28% (16%-33%) during high-quality CPR in hospital CA and they stated that frontal cerebral rSO 2 remained low until ROSC (4).
Quality CPR is closely related to CC being sufficiently deep and uninterrupted. If CC is frequently interrupted for checking the pulse, this will reduce minimum and mean rSO 2 values, resulting in neurological impairment. In our study, for the 7th patient, a person started CC at the 7th minute and another person took over the compression at the 12th minute. The decrease in rSO 2 over this period is indicated by the x arrow in Figure 3. In the same graph, y arrows indicate rSO 2 decreases during frequent rhythm analysis. Accordingly, Meex et al. showed that in their study, in which NIRS device technology was investigated, the rSO 2 values were increased by switching the person giving CPR (21). Whatever the reason, impairment of the quality of CC causes simultaneous decrease of rSO 2 values. In our study, survivors had a very strong positive correlation (r = 0.811, r = 0.771 and P = 0.015, P = 0.025, respectively) between the minimum rSO 2 and mean rSO 2 values and the 1-week mean FOUR scores. This result supports the close relationship between impairment of CC quality during CPR and poor neurological outcomes.
This preliminary study had some limitations. First, we had a small number of patients. It is complicated to apply such technical monitoring in ICUs. Therefore, this study does not have extensive conclusions. However, we suggest that these results can be a guide as a pilot trial for further investigations about neurological outcomes after CPR. Second, the additional illnesses of patients might have affected post-resuscitation FOUR scores. Although we registered additional diseases of our patients, we could not measure the effectiveness of the diseases on FOUR. Finally, we used FOUR scoring for survival. There are many scoring systems to evaluate neurological status, but there is no scoring system for short-term prognosis aside from the Glasgow Coma Scale (GCS). However, many shortcomings of the GCS have been recognized. First, many comatose patients in ICUs are intubated and the verbal component cannot be evaluated. Second, the GCS does not include many indicators such as changing breathing patterns and abnormal brainstem reflexes. Third, the GCS may not detect vague changes during neurological examination. Therefore, we had to use only the FOUR scoring system for 1-week.
In conclusion, although the results of this study were obtained from a limited number of patients, our results suggest that the maximum rSO 2 values affect ROSC while the minimum and mean rSO 2 values affect the post-cardiac arrest neurological outcome. Further studies including more subjects will strengthen the results of such studies. x y y y z Figure 3. The rSO 2 of patient 7 during CPR and after CPR. The caregiver who performed chest compressions for the 7th patient started the chest compressions at the 7th minute and another person took over the compression at the 12th minute. The decrease in rSO 2 over this time period is indicated by x arrow. The y arrows indicate rSO 2 decreases during frequent rhythm analysis. CPR was terminated at the point indicated by the z arrow.