
==== Front
Resusc Plus
Resusc Plus
Resuscitation Plus
2666-5204
Elsevier

S2666-5204(24)00197-8
10.1016/j.resplu.2024.100746
100746
Clinical Paper
Prognostic value of grey-white matter ratio obtained within two hours after return of spontaneous circulation in out-of-hospital cardiac arrest survivors: A multicenter, observational study
Murakami Yuya ab
Hongo Takashi ac
Yumoto Tetsuya a
Kosaki Yoshinori a
Iida Atsuyoshi ad
Maeyama Hiroki b
Inoue Fumiya e
Ichiba Toshihisa e
Nakao Atsunori a
Naito Hiromichi naito-hiromichi@s.okayama-u.ac.jp
a⁎
a Okayama University Faculty of Medicine, Dentistry, and Pharmaceutical Sciences, Department of Emergency, Critical Care, and Disaster Medicine, 2-5-1 Shikata, Kita, Okayama 700-8558, Japan
b Department of Emergency and Critical Care Medicine, Tsuyama Chuo Hospital, Tsuyama, 1756, Tsuyama, Okayama 708-0841, Japan
c Department of Emergency, Okayama Saiseikai General Hospital, 2-25 Kokutai-cho, Okayama Kita-ku, Okayama, 700-8511, Japan
d Department of Emergency Medicine, Japanese Red Cross Okayama Hospital, 2-1-1 Aoe, Kita-ku, Okayama, Okayama, 700-8607 Japan
e Department of Emergency Medicine, Hiroshima City Hospital, 7-33 Motomachi, Naka-Ku, Hiroshima City, Hiroshima 730-8518, Japan
⁎ Corresponding author at: Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences, Department of Emergency, Critical Care, and Disaster Medicine, 2-5-1 Shikata, Kitaku, Okayama 700-8558, Japan. naito-hiromichi@s.okayama-u.ac.jp
15 8 2024
9 2024
15 8 2024
19 10074621 7 2024
28 7 2024
2 8 2024
© 2024 The Author(s)
2024
https://creativecommons.org/licenses/by-nc/4.0/ This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
Background

Grey-white matter ratio (GWR) measured by head computed tomography (CT) scan is known as a neurological prognostication tool for out-of-hospital cardiac arrest (OHCA) survivors. The prognostic value of GWR obtained early (within two hours after return of spontaneous circulation [ROSC]) remains a matter of debate.

Methods

We conducted a multicenter, retrospective, observational study at five hospitals. We included adult OHCA survivors who underwent head CT within two hours following ROSC. GWR values were measured using head CT. Average GWR values were calculated by the mean of the GWR-basal ganglia and GWR-Cerebrum. We divided the patients into poor or favorable neurological outcome groups defined by Glasgow-Pittsburgh Cerebral Performance Category scores. The predictive accuracy of GWR performance was assessed using the area under the curve (AUC). The sensitivities and specificities for predicting poor outcome were examined.

Results

Of 377 eligible patients, 281 (74.5%) showed poor neurological outcomes at one month after ROSC. Average GWR values of the poor neurological outcome group were significantly lower than those of the favorable neurological outcome. The average GWR value to predict neurological outcome with Youden index was 1.24 with AUC of 0.799. When average GWR values were 1.15 or lower, poor neurological outcomes could be predicted with 100% specificity.

Conclusions

GWR values measured by head CT scans early (within two hours after ROSC) demonstrated moderate predictive performance for overall ROSC patients. When limited to the patients with GWR values of 1.15 or lower, poor neurological outcomes could be predicted with high specificity.

Keywords

Computed tomography
Hypoxic ischemic brain injury
Cerebral edema
Neurological outcome
Post-cardiac arrest syndrome
==== Body
pmcIntroduction

Despite improvement in post-resuscitation care over the last decade, post-cardiac arrest brain injury is the major cause of coma in out-of-hospital cardiac arrest (OHCA) survivors.1, 2 In a clinical setting, clinicians caring for comatose OHCA survivors need to provide predictions of neurological outcome based on examination or their experiences during the early phase of intensive care unit (ICU) treatment.3 Neuro-prognostication greatly affects surrogates’ decisions and often leads to withdrawal of life-sustaining therapy.4 On the other hand, even for OHCA survivors who may have had functionally favorable survival, there is a possibility that the required treatment was interrupted due to an incorrect prognostic prediction, resulting in increased mortality of OHCA patients with perceived poor neurological outcomes.4 Therefore, the need for an early and accurate prediction of neurological prognosis remains as of significant importance.

A head computed tomography (CT) examination can be routinely, easily, and quickly performed in the emergency department, making head CT a widely used imaging tool for neurological prognostication.5, 6 The severity of post-cardiac arrest brain injury may be quantified and measured using the grey-white matter ratio (GWR), defined as the density (measured as Hounsfield units) ratio of grey-to-white matter.7 Lower GWR values are a prognostic indicator associated with poor neurological outcomes in severe post-cardiac arrest brain injury.8, 9 However, a past report indicated that GWR values measured from head CT performed early (within two hours following return of spontaneous circulation [ROSC]) could not accurately estimate neurological outcomes in OHCA survivors.10 Nevertheless, previous studies did not determine the specificity for poor outcomes, which are generally crucial in neuro-prediction for OHCA patients. The effectiveness of GWR values to predict poor neurologic prognosis with 100% specificity early after ROSC remains unknown.

The aim of this study was to evaluate the prognostic value of GWR measurements from head CT performed within two hours after ROSC. We specifically determined whether particular values of GWR had high (100%) specificity for poor outcome.

Methods

Study design

We conducted a multicenter, retrospective, observational study at five institutions in Japan: Okayama University Hospital, Okayama Saiseikai General Hospital, Japanese Red Cross Okayama Hospital, Tsuyama Chuo Hospital, and Hiroshima City Hospital from January 2007 to December 2019. The study was approved by the Okayama University Hospital Ethics Committee (approval number: K2101-010) and conforms to the provisions of the Declaration of Helsinki. Patient consent was waived because of the retrospective study design.

Patient population

Inclusion criteria were OHCA patients who: (1) achieved ROSC, (2) were 20 years old or older, (3) underwent head CT within two hours after ROSC, (4) were unable to obey commands after ROSC, and (5) were admitted to the ICU. Exclusion criteria were: (1) cardiac arrest due to stroke or trauma, (2) severe disability or coma before cardiac arrest defined as Glasgow-Pittsburgh Cerebral Performance Category (CPC) scores of 3 or 4,11 (3) do not attempt resuscitation (DNAR) status at ICU admission, and 4) missing neurological outcome data. Enrolled participants received post-resuscitation care based on guidelines.12, 13

Data collection

We collected the following data from emergency medical service reports and electronic medical records at each institution: age, sex, CPC score before OHCA, witness status, bystander cardiopulmonary resuscitation (CPR), initial recorded cardiac rhythm at the scene, estimated cardiac or non-cardiac origin of cardiac arrest, defibrillation, ROSC during transport or at hospital arrival, time between ROSC and head CT scan, temperature management, venous-arterial extracorporeal membrane oxygenation, laboratory test, total CPR duration, data from patient CT scans, one-month CPC after ROSC. ROSC was defined as palpable pulse sustained after resuscitation for more than 20 min.14 The CPC score was determined by the physician-in-charge one-month after cardiac arrest using hospital medical records or interviews/telephone with the patient, a surrogate, or the physician at the hospital to which patient was transferred.

GWR value measurements

A 3–5 mm slice multi-detector array CT scanner was used for our study, and the tube voltage was 120 kV in all facilities, while CT models and channel numbers were different at each institution (Supplemental Table 1). GWR values were evaluated by two physicians: an emergency physician from each institute and an interpreter from our study team who was not involved in post-resuscitation care. CT attenuation was measured in Hounsfield units (HU) by placing an elliptical region of interest (ROI) within axial sections (10–15 mm2). At the basal ganglia (BG) level, HU values were recorded bilaterally at the level of the caudate nucleus (CN), putamen (PU), corpus callosum (CC), and posterior limb of internal capsule (PLIC). Furthermore, medial cortex and medial white matter were recorded bilaterally at the level of the centrum semiovale (MC1 and MWM1, respectively) and high convexity area (MC2 and MWM2, respectively). GWR-BG was calculated as (CN + PU)/(CC + PLIC) and GWR-Cerebrum was calculated as (MC1 + MC2)/(MWM1 + MWM2). We calculated average GWR values as the mean of the GWR-BG and GWR-Cerebrum.8

Grouping and endpoints

We divided the OHCA patients into poor (CPC 3–5) and favorable (CPC 1–2) neurological outcome groups based on one-month CPC scores. The primary outcome was poor neurological outcome at one month after ROSC. We assessed the prognostic performance of GWR in predicting poor neurological outcomes.

Data analysis

Continuous variables are presented as medians and interquartile ranges. Categorical variables are presented as numbers and percentages. Categorical variables were compared using the Fisher’s exact probability test. The Mann-Whitney-U test was used to evaluate variables with non-normal distributions. A scatter plot was made, with time from ROSC to CT scan on the X-axis and average GWR values on the Y-axis with plots separately described for poor and favorable neurological outcomes. We evaluated the area under the curve (AUC) with 95% confidence interval (CI) for receiver operating characteristic (ROC) curves to determine the prediction performance of average GWR values for poor neurological outcomes. GWR values for poor neurological outcomes were determined based on Youden index, defined as the largest difference between sensitivity and 1-specificity. We calculated sensitivity, specificity, positive predictive values (PPV), negative predictive values (NPV), and false positive rate (FPR) at each average GWR value. The histograms showing proportions of both poor and favorable neurological outcome patients corresponding to each average GWR values were shown. GWR-BG and GWR-Cerebrum prediction performance for poor neurological outcomes were also evaluated. A p-value <0.05 was considered statistically significant. Statistical analysis was performed using Stata version 17 (StataCorp LP, College Station, TX).

Results

Patient characteristics

During the study period, 946 of 6,176 OHCA patients were eligible. After excluding 569 cases, 377 patients were finally included in our study cohort (Fig. 1). Patient characteristics are summarized in Table 1. Overall, the median age was 69 years old (58–78), and 249 (66.1%) patients were men. Cardiac arrest was witnessed for 184 (48.8%) patients and 138 (36.6%) patients were recorded as having initial shockable rhythm (VF/VT). Poor neurological outcomes were observed in 281 (74.5%) patients, while 96 (25.5%) patients showed favorable neurological outcomes.Fig. 1 Flow diagram of the study. CPC: cerebral performance category, CT: computed tomography, DNAR: do not attempt resuscitation, ICU: intensive care unit, OHCA: out-of-hospital cardiac arrest, ROSC: return of spontaneous circulation.

Table 1 Demographics and clinical characteristics of patients with favorable (CPC scores 1–2) and poor (CPC scores 3–5) neurological outcomes.

	All (n = 377)	CPC 1–2 (n = 96)	CPC 3–5 (n = 281)	p-value	
Patient Demographics	
 Age, year, median, (IQR)	69 (58–78)	63 (51–72)	70 (61–79)	<0.001	
 Male sex, N (%)	249/377 (66.1)	63/96 (65.6)	186/281 (66.2)	0.91	
 CPC before CA, N (%)				<0.001	
  CPC1	303/377 (80.4)	92/96 (95.8)	211/281 (75.1)		
  CPC2	74/377 (19.6)	4/96 (4.2)	70/281 (24.9)		
 Witnessed collapse, N (%)	184/377 (48.8)	61/96 (63.5)	123/281 (43.8)	0.001	
 Bystander CPR, N (%)	265/377 (70.3)	80/96 (83.3)	185/281 (65.8)	0.001	
 Initial cardiac rhythm, N (%)				<0.001	
  VF/VT	138/377 (36.6)	71/96 (74.0)	67/281 (23.8)		
  PEA/asystole	239/377 (63.4)	25/96 (26.0)	214/281 (76.2)		
 Defibrillation by EMS, N (%)	138/377 (36.6)	70/96 (72.9)	68/281 (24.2)	<0.001	
 Etiology of CA, N (%)				<0.001	
  Cardiac	208/377 (55.2)	77/96 (80.2)	131/281 (46.6)		
  Non-cardiac	169/377 (44.8)	19/96 (19.8)	150/281 (53.4)		
 ROSC while transport, N (%)	92/377 (24.4)	64/96 (66.7)	28/281 (10.0)	<0.001	
 ROSC on arrival at hospital, N (%)	98/377 (26.0)	59/96 (61.5)	39/281 (13.9)	<0.001	
 Time to ROSC, min, median, (IQR)	31 (22–44)	20 (12–30)	37 (27–48)	<0.001	
Post-arrest management	
 TM, N (%)	188/377 (49.9)	64/96 (66.7)	124/281 (44.1)	<0.001	
 V-A ECMO, N (%)	29/377 (7.7)	8/96 (8.3)	21/281 (7.5)	0.78	
Laboratory findings	
 pH, median, (IQR)	7.02 (6.9–7.19)	7.25 (7.09–7.31)	6.97 (6.87–7.09)	<0.001	
 Lac, median, (IQR)	10.5 (8.0–14.3)	8.8 (7.1–11.0)	11.6 (8.7–14.9)	<0.001	
Grey matter	
 Basal ganglia					
  CN	33.90 (31.83–36.49)	33.78 (32.12–36.18)	33.95 (31.73–36.57)	0.96	
  PU	33.95 (32.27–36.18)	34.13 (32.81–36.67)	33.80 (32.10–36.02)	0.16	
 Cerebrum					
  MC1	31.95 (29.95–34.35)	31.83 (29.65–34.35)	32.35 (30.03–34.30)	0.14	
  MC2	32.05 (29.70–35.05)	32.10 (29.65–35.28)	31.86 (29.87–34.80)	0.85	
White matter	
 Basal ganglia					
  CC	28.00 (25.65–30.80)	26.45 (24.87–29.52)	28.25 (26.00–31.17)	<0.001	
  PLIC	27.60 (25.65–29.90)	26.65 (24.66–28.73)	28.00 (26.25–30.37)	<0.001	
 Cerebrum					
  MWM1	26.00 (23.70–28.33)	24.90 (23.57–27.30)	26.45 (23.80–28.70)	0.016	
  MWM2	26.75 (23.90–29.25)	27.20 (24.35–29.62)	25.05 (23.05–27.34)	<0.001	
Grey-White matter ratio	
 GWR-BG	1.22 (1.17–1.27)	1.27 (1.24–1.31)	1.20 (1.15–1.25)	<0.001	
 GWR-Cerebrum	1.22 (1.16–1.27)	1.27 (1.24–1.32)	1.20 (1.14–1.25)	<0.001	
 Average GWR	1.22 (1.18–1.27)	1.27 (1.25–1.31)	1.20 (1.16–1.24)	<0.001	
Values for patient characteristics are expressed as median (IQR) or number (%) as appropriate. Values for Hounsfield units and GWR value are expressed mean ± standard deviation.

CPC: Cerebral Performance Category, IQR: interquartile range, CA: cardiac arrest, CPR: cardiopulmonary resuscitation, VF/VT: ventricular fibrillation/ventricular tachycardia, PEA: pulseless electrical activity, EMS: emergency medical service, ROSC: return of spontaneous circulation, TM: temperature management, V-A ECMO: venoarterial extracorporeal membrane oxygenation, Lac: lactate. CN: caudate nucleus, PU: putamen, CC: corpus callosum, PLIC: posterior limb of internal capsule, GWR: grey-white matter ratio, BG: basal ganglia.

Grey-white matter ratio

The relationship between average GWR values and time from ROSC to CT scan divided by poor (red) or favorable (blue) neurological outcomes is shown in the scatter plot (Fig. 2). Average GWR values of the poor neurological outcome group were lower than those of the favorable neurological outcome group.Fig. 2 Scatter plot diagram showing the relationship between average GWR values and time from ROSC to CT scan divided by poor (red) or favorable (blue) neurological outcomes. GWR: grey-white matter ratio, CT: computed tomography, ROSC: return of spontaneous circulation. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

ROC analysis showed that the average GWR values had moderate prognostic performance (AUC 0.799; 95% CI [0.75–0.84]) (Fig. 3). The optimal GWR ratio for predicting poor neurological outcome based on the Youden index was 1.24 with sensitivity of 74.3% [68.9–79.4], specificity of 85.4% [76.7–91.8], PPV of 93.7% [89.7–96.5], NPV of 53.2% [45.0–61.3], and FPR of 14.6% [8.2–23.3]. The histograms showing proportions of both poor and favorable neurological outcome patients corresponding to each average GWR values are described as Fig. 4. Table 2 shows the prognostic performance of various GWR values. When the average GWR values for outcome prediction was 1.15 or lower, poor neurological outcomes could be predicted with 100% specificity (AUC 0.591; 95% CI [0.56–0.61], sensitivity of 18.1% [13.8–23.2], PPV of 100% [93.0–100], NPV of 28.6% [24.6–34.7], and FPR of 0% [0.0–3.8]).Fig. 3 Receiver operating characteristic curves to determine the prediction performance of average GWR values for poor neurological outcomes.

Fig. 4 The histograms showing proportions of both poor (red) and favorable (blue) neurological outcome patients corresponding to each average GWR values. GWR: grey-white matter ratio. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 2 GWR threshold value for predicting poor (CPC scores 3–5) neurological outcomes.

	n/Na	AUC
[95% CI]	Sensitivity(%)
[95%CI]	Specificity(%)
[95%CI]	PPV(%)
[95%CI]	NPV(%)
[95%CI]	FPR(%)
[95%CI]	
GWR value 1.10	
 GWRAVERAGE	9/9	0.516
[0.50–0.52]	3.2
[1.4–5.9]	100
[96.2–100]	100
[66.4–100]	26.0
[21.7–30.9]	0
[0.0–3.8]	
GWR value 1.15	
 GWRAVERAGE	51/51	0.591
[0.56–0.61]	18.1
[13.8–23.2]	100
[96.2–100]	100
[93.0–100]	28.6
[24.6–34.7]	0
[0.0–3.8]	
GWR value 1.20	
 GWRAVERAGE	135/137	0.730
[0.69–0.76]	48.0
[42.1–54.1]	97.9
[92.7–99.7]	98.5
[94.8–99.8]	39.1
[33.0–45.7]	2.1
[0.3–7.3]	
GWR value 1.24	
 GWRAVERAGE	209/223	0.799
[0.75–0.84]	74.3
[68.9–79.4]	85.4
[76.7–91.8]	93.7
[89.7–96.5]	53.2
[45.0–61.3]	14.6
[8.2–23.3]	
GWR value 1.25	
 GWRAVERAGE	219/242	0.770
[0.72–0.81]	77.9
[72.6–82.6]	76.0
[66.3–84.2]	90.5
[86.1–93.9]	54.0
[45.3–62.7]	24.0
[15.8–33.7]	
GWR value 1.30	
 GWRAVERAGE	254/318	0.619
[0.56–0.66]	90.3
[86.3–93.6]	33.3
[24.0–43.7]	79.8
[75.0–84.1]	54.2
[40.8–67.3]	66.7
[56.3–76.0]	
95% CI: 95% confidence interval, AUC: area under curve, CPC: Cerebral Performance Category, FPR: false positive rate, PPV: positive predictive value, NPV: negative predictive value, GWR: grey-white matter ratio.

a n: pattern prevalence, N: cohort.

GWR-BG and GWR-cerebrum level analysis

A similar analysis was performed on GWR-BG and GWR-Cerebrum level. Lower GWR values showed a higher trend for poor neurological outcome in both regions (Supplemental Figs. 1 and 2). While both GWR levels had overall moderate prognostic performances, GWR values of 1.15 or lower had a high specificity of nearly 100% (Supplemental Table 2).

Discussion

In this multicenter, retrospective study, lower GWR values obtained from head CT scans within two hours after ROSC were associated with poor neurological outcomes. However, GWR demonstrated moderate predictive performance, indicating that GWR values were not definitively helpful for predicting outcomes for overall patients. This result was similar to previous study showing the poor predictability for overall patients.10 However, limited to OHCA patients with very low GWR values, poor neurological outcomes could be predicted with high specificity. Multiple GWR cutoff points were employed to explore the specificity for poor neurological outcomes, and it was determined that the GWR values for predicting poor neurological outcomes with 100% specificity was 1.15 or lower.

Lower GWR values signify severe cerebral edema, which may indicate poor neurological outcomes.16, 17 However, these guidelines did not establish a consensus on precise GWR values predictive of poor neurological outcomes.16, 17 The spectrum of GWR values utilized in predicting poor neurological outcomes across 24 prior studies varies between 1.07 and 1.26.18 Our study revealed that GWR values demonstrated a suboptimal predictive capacity for overall neurological outcomes, as evidenced by AUC of less than 0.800, based on the Youden index. This finding aligns with results from other studies demonstrating the unreliable predictive performance of GWR measurements due to low AUC.5, 19 In clinical practice, neuro-prognostication for OHCA patients should aim for a FPR of zero to prevent erroneously predicting unfavorable neurological outcomes for patients who may recover.15 GWR values of 1.15 or lower, determined to have 100% specificity in our analysis, fall within the range identified by previous research.20 Lang et al. also demonstrated that GWR values of 1.15 predicted poor outcomes with 100% specificity across all evaluated locations.21 Confirming these findings in larger cohorts may enhance the predictive performance of GWR values and validate their applicability across different scenarios.

GWR, measured from quantitative CT assessments of cerebral edema, has been found to aid in prognosticating both mortality and neurological outcomes in the early post-arrest phase. Although the 2015 American Heart Association (AHA) guidelines stated that significantly lower GWR values from head CT scans within two hours after cardiac arrest could be utilized to predict poor neurological outcomes in OHCA patients,22 this recommendation was not carried forward in the 2020 AHA and European Resuscitation Council guidelines.16, 17 Our study underscores that severe low GWR values measured from a CT scan, even two hours after ROSC, has high specificity to predict a poor neurological prognosis. Early identification of poor neurological outcomes to identify comatose patients with prospects for neurological improvement is important.23 Early recognition of OHCA patients with poor neurological prognoses may impact the reduction of medical resource utilization, facilitate organ donation, and provide surrogates with information to support shared decision-making regarding treatment strategies or withdrawal of life-sustaining therapy.15

Approximately 80% of OHCA survivors who are resuscitated remain comatose, with the majority either dying or suffering severe neurological disabilities.24 Current neuro-prognostication methods after cardiac arrest include clinical examinations, biomarkers, electroencephalography, somatosensory evoked potentials, brain CT, and brain magnetic resonance imaging (MRI).2 Recent studies have employed both qualitative and quantitative approaches to CT and MRI analysis for predicting neurological outcomes.7 MRI is noted for its higher sensitivity in predicting poor neurological outcomes by identifying post-cardiac arrest brain injuries, indicating that many OHCA patients with normal CT scans may still experience poor neurological outcomes.25 Although MRI provides valuable diagnostic information without radiation exposure, using MRI on OHCA patients who are comatose and undergoing mechanical ventilation presents acute phase challenges and suffers from the disadvantages of higher costs and lower availability compared to CT.26 Clinicians routinely perform early CT scans after ROSC to determine the cause of cardiac arrest.27 Identifying OHCA survivors with poor neurological prognoses through these CT scans is deemed more practical than using MRI.

While GWR measurements have been objective, they have been limited to visual assessments or manual ROI analysis of gray and white matter. Artificial intelligence, an advanced technology that has witnessed recent breakthroughs in critical care and resuscitation management, is evolving rapidly. Although outcome prediction after OHCA using automatically calculated GWR from CT images obtained early (within 24 h) did not have good predictive value,6 further research is warranted to explore whether a machine-learning algorithm designed to assess GWR values could effectively predict neurological outcomes in comatose OHCA patients.

Limitations

This study has several limitations. First, we cannot dismiss the possibility of selection bias for data collection due to the nature of retrospective studies. Second, our study sample may not be large enough to achieve high statistical power. Third, there might have been some information bias because we did not blind the assessment of head CTs or other clinical parameters. Fourth, this study included samples from multiple centers, and the CT scanners used at each facility differed, which may have affected the HU and GWR values. Fifth, a measurement bias (calcification, CT artifact, region) might be caused when obtaining GWR values.28 Sixth, the external validity may not be generalizable to all OHCA patients because the data on patient characteristics and medical management originated from specific regions in Japan. Finally, we unfortunately did not capture the data for withdrawal/withold of life-sustaining therapy decision. In Japan, physicians are generally reluctant to implement withdrawal of life-sustaining therapy due to sociolegal, ethical, and religious issues,29 instead treatment is often withheld for high severity patients. These decisions may have potential risk of overestimating the accuracy of GWR due to self-fulfilling prophecy.

There is no consensus on the optimal time to perform head CT and measure GWR values for predicting poor neurological outcomes in OHCA survivors, and properly designed prospective studies with more patients are needed.

Conclusion

GWR values measured by head CT scans early (within two hours after ROSC) demonstrated moderate predictive performance for overall ROSC patients. However, when limited to the patients with GWR values 1.15 or lower, poor neurological outcomes could be predicted with high specificity.

Funding source

This study was funded by Daiichi-Sankyo Funding Program (A22-0324 ).

CRediT authorship contribution statement

Yuya Murakami: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Takashi Hongo: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Tetsuya Yumoto: Writing – review & editing, Methodology, Data curation. Yoshinori Kosaki: Writing – review & editing, Methodology, Data curation. Atsuyoshi Iida: Writing – review & editing, Methodology, Data curation. Hiroki Maeyama: Writing – review & editing, Methodology, Data curation. Fumiya Inoue: Writing – review & editing, Methodology, Data curation. Toshihisa Ichiba: Writing – review & editing, Methodology, Data curation. Atsunori Nakao: Writing – review & editing, Supervision, Methodology, Funding acquisition. Hiromichi Naito: Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A Supplementary material

The following are the Supplementary material to this article:Supplementary Data 1

Acknowledgements

We thank all researchers for participating in this study. We thank Christine Burr for editing the manuscript.

Appendix A Supplementary material to this article can be found online at https://doi.org/10.1016/j.resplu.2024.100746.
==== Refs
References

1 Rossetti A.O. Rabinstein A.A.O.M. Neurological prognostication of outcome in patients in coma after cardiac arrest Lancet Neurol 15 2016 597 609 27017468
2 Sandroni C. Cronberg T.S.M. Brain injury after cardiac arrest: pathophysiology, treatment, and prognosis Intensive Care Med 47 2021 1393 1414 34705079
3 Naito H. Sakuraya M. Hongo T. Prevalence, reasons, and timing of decisions to withhold/withdraw life-sustaining therapy for out-of-hospital cardiac arrest patients with extracorporeal cardiopulmonary resuscitation Crit Care 27 2023 252 37370155
4 Elmer J. Torres C. Aufderheide T.P. Association of early withdrawal of life-sustaining therapy for perceived neurological prognosis with mortality after cardiac arrest Resuscitation 102 2016 127 135 26836944
5 Lee B.K. Jeung K.W. Song K.H. Prognostic values of gray matter to white matter ratios on early brain computed tomography in adult comatose patients after out-of-hospital cardiac arrest of cardiac etiology Resuscitation 96 2015 46 52 26232516
6 Adriaansens K.O. Jewbali L.S.D. Lemkes J.S. Routine reporting of grey-white matter differentiation in early brain computed tomography in comatose patients after cardiac arrest: a substudy of the COACT trial Resuscitation 175 2022 13 18 35378224
7 Keijzer H.M. Hoedemaekers C.W.E. Meijer F.J.A. Tonino B.A.R. Klijn C.J.M. Hofmeijer J. Brain imaging in comatose survivors of cardiac arrest: pathophysiological correlates and prognostic properties Resuscitation 133 2018 124 136 30244045
8 Metter R.B. Rittenberger J.C. Guyette F.X. Callaway C.W. Association between a quantitative CT scan measure of brain edema and outcome after cardiac arrest Resuscitation 82 2011 1180 1185 21592642
9 Cristia C. Ho M.L. Levy S. Andersen L.W. The association between a quantitative computed tomography (CT) measurement of cerebral edema and outcomes in post-cardiac arrest-a validation study Resuscitation 85 2014 1348 1353 24892267
10 Hong J.Y. Lee D.H. Oh J.H. Grey-white matter ratio measured using early unenhanced brain computed tomography shows no correlation with neurological outcomes in patients undergoing targeted temperature management after cardiac arrest Resuscitation 140 2019 161 169 30953628
11 Schmidt H. Kjaergaard J. Hassager C. Oxygen targets in comatose survivors of cardiac arrest N Engl J Med 387 2022 1467 1476 36027567
12 Field J.M. Hazinski M.F. Sayre M.R. Part 1: executive summary: 2010 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care Circulation 122 2010 S640 S656 20956217
13 Neumar R.W. Shuster M. Callaway C.W. Part 1: Executive summary: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care Circulation 132 2015 S315 S367 26472989
14 Xu J. Li C. Tang H. Pulse oximetry waveform: a non-invasive physiological predictor for the return of spontaneous circulation in cardiac arrest patients –– a multicenter, prospective observational study Resuscitation 169 2021 189 197 34624410
15 Steinberg A. Callaway C.W. Arnold R.M. Prognostication after cardiac arrest: results of an international, multi-professional survey Resuscitation 138 2019 190 197 30902688
16 Panchal A.R. Bartos J.A. Cabañas J.G. Part 3: Adult basic and advanced life support: 2020 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care Circulation 142 2020 S366 S468 33081529
17 Nolan J.P. Sandroni C. Böttiger B.W. European Resuscitation Council and European Society of Intensive Care Medicine Guidelines 2021: post-resuscitation care Resuscitation 161 2021 220 269 33773827
18 Wang W.J. Cui J. Lv G.W. Feng S.Y. Zhao Y. Zhang S.L.L.Y. Prognostic values of the gray-to-white matter ratio on brain computed tomography images for neurological outcomes after cardiac arrest: a meta-analysis Biomed Res Int 2020 2020 7949516
19 Chae M.K. Ko E. Lee J.H. Better prognostic value with combined optic nerve sheath diameter and grey-to-white matter ratio on initial brain computed tomography in post-cardiac arrest patients Resuscitation 194 2016 40 45
20 Zhou F. Wang H. Jian M. Gray-white matter ratio at the level of the basal ganglia as a predictor of neurologic outcomes in cardiac arrest survivors: a literature review Front Med (Lausanne) 15 2022 847089
21 Lang M. Nielsen N. Ullén S. A pilot study of methods for prediction of poor outcome by head computed tomography after cardiac arrest Resuscitation 179 2022 61 70 35931271
22 Callaway C.W. Donnino M.W. Fink E.L. Part 8: Post-cardiac arrest care: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care Circulation 132 2015 S465 S482 26472996
23 Belohlavek J. Smalcova J. Rob D. Effect of intra-arrest transport, extracorporeal cardiopulmonary resuscitation, and immediate invasive assessment and treatment on functional neurologic outcome in refractory out-of-hospital cardiac arrest JAMA 327 2022 737 747 35191923
24 Thomassen A. Wernberg M. Prevalence and prognostic significance of coma after cardiac arrest outside intensive care and coronary units Acta Anaesthesiol Scand 23 1979 143 148 442945
25 Lopez Soto C. Dragoi L. Heyn C.C. Imaging for neuroprognostication after cardiac arrest: systematic review and meta-analysis Neurocrit Care 32 2020 206 216 31549351
26 Prabhat A.M. Crawford A.L. Mazurek M.H. Methodology for low-field, portable magnetic resonance neuroimaging at the bedside Front Neurol 2021 12
27 Branch K.R.H. Gatewood M.O. Kudenchuk P.J. Diagnostic yield, safety, and outcomes of Head-to-pelvis sudden death CT imaging in post arrest care: the CT FIRST cohort study Resuscitation 188 2023 109785
28 Harris G.J. Barta P.E. Peng L.W. MR volume segmentation of gray matter and white matter using manual thresholding: dependence on image brightness AJNR Am J Neuroradiol 15 1994 225 230 8192065
29 Wakatake H. Hayashi K. Kitano Y. Eleven-year retrospective study characterizing patients with severe brain damage and poor neurological prognosis -role of physicians' attitude toward life-sustaining treatment BMC Palliat Care 21 2022 79 35581603
