Resuscitation 88 (2015) 132–137
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical Paper
Efficacy of diffusion-weighted magnetic resonance imaging performed before therapeutic hypothermia in predicting clinical outcome in comatose cardiopulmonary arrest survivors Jung Soo Park a , Suk Woo Lee a , Hoon Kim a , Jin Hong Min a,∗ , Jun Ho Kang a , Kyung Sik Yi b , Kyung Hye Park c , Byung Kook Lee d a
Department of Emergency Medicine, College of Medicine, Chungbuk National University Hospital, Cheongju, Republic of Korea Department of Radiology, College of Medicine, Chungbuk National University Hospital, Cheongju, Republic of Korea c Department of Emergency Medicine, Inje University, Busan, South Korea d Department of Emergency Medicine, Chonnam National University Hospital, 42, Jebong-ro, Donggu, Gwangju, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 4 September 2014 Received in revised form 13 November 2014 Accepted 28 November 2014 Keywords: Heart arrest Hypothermia Coma Prognosis
a b s t r a c t Aim of the study: To develop a clinically relevant and qualitative brain magnetic resonance imaging (MRI) scoring system for acute stage comatose cardiac arrest patients. Methods: Consecutive comatose post-cardiopulmonary arrest patients were prospectively enrolled. Routine brain MRI sequences were scored by two independent and blinded experts. Predefined brain regions were qualitatively scored on diffusion-weighted imaging (DWI) sequences according to the severity of the abnormality on a scale from 0 to 4. The mean score provided by the raters determined poor outcome defined under the Cerebral Performance Categories 3, 4, or 5. DWI scans were repeated after therapeutic hypothermia (TH). The same qualitative scoring system was applied and results were compared to the initial scores. Results: Out of 24 recruited patients, 19 with brain MRI scans were included. Of the 19 included patients, seven showed a good outcome at hospital discharge and 12 patients showed poor neurologic outcome. Median time from the arrest to the initial DWI was 166 min (IQR 114–240 min). At 100% specificity, the overall, cortex, and cortex plus deep grey nuclei scores predicted poor patient outcome with a sensitivity of 91.7–100% (95% CI). Follow-up DWI scans after TH showed worse results than initial scans. Conclusion: A qualitative MRI scoring system effectively assessed the severity of hypoxic-ischaemic brain injury following cardiopulmonary arrest. The scoring system may provide useful prognostic information in comatose cardiopulmonary arrest patients. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Improvements in intensive care medicine have led to increased bed ridden vegetative states and severe neurological deficits in cardiopulmonary arrested survivors (CPAS). In the past, only 10–30% of comatose CPAS demonstrated functional recovery.1 However, these percentages improved with increasing use of therapeutic hypothermia (TH).2,3 Post-cardiopulmonary arrest brain injury is a common cause of morbidity and mortality. Despite the use of TH, 45–70% of cardiac arrest victims still experience severe neurological deficits due to anoxic reperfusion cerebral injury.4,5 Conversely, premature withdrawal of life support from patients who may have a
∗ Corresponding author. Fax: +82 43 269 7396.. E-mail address: [email protected] (J.H. Min). http://dx.doi.org/10.1016/j.resuscitation.2014.11.031 0300-9572/© 2015 Elsevier Ireland Ltd. All rights reserved.
chance of functional recovery represents an ethical dilemma. Thus, early and accurate identification of patients who will not likely recover is an important healthcare issue. Many studies have focused on early identification of comatose CPAS expected not to regain consciousness.6–8 Historically, the most specific early predictors for poor outcome after cardiopulmonary arrest include the following: absence of brain stem or extensor reflexes, absence of motor response at post-arrest day 3, absence of cortical responses by somatosensory evoked potentials (SSEP) at 24 h post arrest, serum neuron specific enolase (NSE) levels >33 g/L in the first 3 days, and early myoclonic status epilepticus.9 These predictors have substantive limitations, a low sensitivity for poor outcome, and may not be valid in patients treated with hypothermia.10 The first limitation in the predictors is that they identify only a subset of poor-outcome patients with high specificity. Second, neurological examinations and results of
J.S. Park et al. / Resuscitation 88 (2015) 132–137
Brain regions Frontal lobe Supratentorial Cortex Parietal lobe Temporal lobe Occipital lobe Insula Hippocampus Deep grey nuclei Caudate Putamen Globus pallidus Thalamus White matter Frontal lobe Parietal lobe Temporal lobe O c c i p i t al l o b e Corpus callosum Midbrain Infratentorial Brainstem Pons Medulla Cerebellum Cortex White matter Dentate nuclei
133
Scoring system 0 = normal 1 = possibly abnormal 2 = abnormal, mild 3 = abnormal, moderate 4 = abnormal, severe
Fig. 1. The 21 scored brain regions and DWI scoring system.
electroencephalography may be difficult to interpret in the presence of sedative agents or metabolic derangements. Third, serum markers are potentially susceptible to false-positive test results and are not yet readily available in many hospitals.9,11 Several studies have reported brain magnetic resonance imaging (MRI) as a useful tool in imaging comatose CPAS.12–24 Typically, the presence of extensive and severe cortical signal abnormalities on MRI is associated with poor outcome.12–15,18–22 MRI offers the advantage of providing an objective measure of cerebral injury and may be particularly useful in patients who have received sedative agents. Furthermore, the method is useful in patients who have metabolic derangements that render neurological examinations unreliable and patients treated with TH. Recent developments in brain imaging through MRI include diffusion-weighted MRI (DWI). DWI is a novel, fast (10 min), and potentially powerful marker of early global ischaemic brain injury.21 Additionally, patient monitoring devices are available during DWI. Preliminary studies have shown widespread abnormalities detected by DWI in poor-outcome patients at 1 week post-cardiac arrest.21 However, obtaining a brain MRI in critically ill patients with potential cardiac instability is challenging. We used a clinically applicable brain MRI scoring system that is qualitative and has potential for broad use and application. This is an important tool for facilities without access to qualitative MRI analysis or technically inadequate scanners. This qualitative MRI scoring system has been described in previous studies as a tool to predict outcome following perinatal asphyxia.24–27 The overall aim of this study was to determine the feasibility and prognostic utility of DWI before TH in comatose CPAS in a prospective study.
approved by the institutional review board, and written consent from a legally authorized representative was required for study participation. Clinical parameters were obtained in a prospective and standardized fashion at predefined time points. Neurological examinations (including a Glasgow Coma Scale score and an assessment of brainstem function) were performed daily during the first 3 days and at 1 and 2 weeks after cardiopulmonary arrest. Functional outcomes were determined by the Cerebral Performance Categories (CPC) at discharge. The first MRI examinations were conducted before TH within at least 6 h. Additional MRI examinations were conducted within the first week after the arrest if the treating physicians agreed the patient could safely undergo MRI after TH. Reasons for patients not undergoing an MRI and adverse events during patient transport to or from the MRI suite and during imaging acquisition were documented. 2.2. Therapeutic hypothermia protocol
2. Methods
In all patients, TH was applied according to a written TH protocol. TH was induced with ice packs, intravenous cold saline, and cooling devices (Blanketrol II, Cincinnati Subzero Products, Cincinnati, USA; Artic Sun Energy Transfer Pads, Medivance Corp, Louisville, USA; COOLGARD 3000 Thermal Regulation System, Alsius Corporation, Irvine, USA). The target temperature of 33 ± 0.5 ◦ C was maintained for 24 h. Upon completion of the TH maintenance phase, patients were rewarmed to 36.5 ◦ C at a rate of 0.25–0.5 ◦ C h−1 . During TH, the temperature was monitored using a bladder temperature probe. Midazolam and atracurium were used for sedation and shivering control. All patients received standard intensive care according to our institutional intensive care unit protocol.
2.1. Study design and population
2.3. Magnetic resonance imaging and scoring
Consecutive CPAS who remained comatose after successful resuscitation were prospectively enrolled over a 2-year period at Chungbuk National University Hospital, a university-affiliated 600bed hospital in Chungju, Korea, from March 2012 to March 2014. Patients were enrolled if they met the following inclusion criteria: ≥16 years of age, status post resuscitation for in- or out-of-hospital cardiac arrest, and persistent coma defined as the inability to open eyes to voice and inability to follow commands. The study was
All patients with technically adequate brain DWIs obtained before TH within at least 6 h after the arrest are included in this study. MRIs were obtained with a 1.5 T system (Achieva 1.5 T; Philips Medical System, The Netherlands). For DWI, whole-brain axial plane, single-shot spin-echo planar imaging was acquired by applying diffusion-sensitizing gradients along three orthogonal directions with a diffusion weighting factor b = 1000 s/mm2 plus one reference scan with b = 0. The section thickness was 5 mm and
134
J.S. Park et al. / Resuscitation 88 (2015) 132–137
Fig. 2. Scoring of sample representative MRI scans.
the section gap was 1 mm. Patients were routinely scanned once within the 6-hour period post arrest before TH and all MRIs were included in the analyses. Images were assessed by a certified neuroradiologist and an emergency physician. The adjudicators were blinded to patient information, outcome, and readings of other adjudicators. Images were scored with the use of an MRI scoring system created to assess the severity of cerebral abnormalities due to cardiopulmonary arrest in predefined brain regions (Fig. 1). The adjudicators were instructed to score only MRI abnormalities that could be attributed to acute global hypoxic–ischaemic brain injury. The 21 brain regions scored individually were as follows: cortical grey matter and subcortical white matter in the frontal, parietal, temporal, and occipital lobes, hippocampus, insular cortex, corpus callosum, deep grey nuclei (caudate, putamen, globus pallidus, and thalamus), cerebellum (cortex, white matter, and dentate nucleus) and the brainstem (midbrain, pons, and medulla). All brain regions were scored using only DWI sequences according to the extent and severity of the signal abnormalities on a scale from 0 to 4. A score of 0, 1, 2, 3, and 4 indicated no abnormality, possibly abnormal, mildly abnormal, moderately abnormal, and severely abnormal, respectively (Fig. 2). The “overall score” consisted of all points given to all brain regions on DWI and the sum score was used during analyses. Follow up DWI scans were performed within 48 h of completion of TH. The same qualitatively scored system was applied and follow up scores were compared with the initial scores. Follow-up DWI scans were only performed when agreed upon by legal representatives. 2.4. Outcome Good outcome was defined as a CPC of 1 or 2 (good or moderate disability), and poor outcome was defined as a CPC of 3, 4, or 5 (severe disability, vegetative state, or brain death) at discharge. A CPC of 3 was considered a poor outcome because it was not sufficient to conduct common functions. 2.5. Statistical analysis Continuous variables were reported as the median and interquartile range (IQR) (non-normal distribution) or the mean and standard deviation (normal distribution). The Mann–Whitney test (non-paired result) or the Wilcoxon signed rank test (paired result) was conducted for comparisons of continuous variables. Categorical variables are represented as frequencies and percentages. The receiver operating characteristic (ROC) was analysed with the corresponding area under curve. Cut-off values with 100% specificity were calculated for predicting poor neurologic outcome at discharge. The agreement degree for DWI scoring between the neuroradiologist and emergency physician was measured using the kappa index and related 95% confidence intervals. Data were analysed using PASW/SPSSTM software, version 18 (IBM Inc., Chicago, USA). ROC curves were calculated and compared with MedCalc version 12.4.0 (free trial, MedCalc Software, Mariakerke, Belgium). Significance was set at p < 0.05.
3. Results 3.1. Clinical characteristics Over a 2-year study period, 288 patients who experienced cardiac arrest were admitted to our hospital and 30 patients survived. Of these, 24 patients met the inclusion criteria and 19 were included in the analyses. The participants were not included in the analysis if they were not treated with TH (n = 3) or were too unstable to undergo MRI (n = 2). The clinical characteristics of patients included in the analyses are shown in Table 1. According to the neurologic outcome at hospital discharge, seven patients (36.8%) were assigned to the good outcome group and 12 patients (63.2%) were assigned to the poor outcome group. A DWI was obtained at 166 min (114–240 min) after arrest and 121 min (88.01–99 min) after return of spontaneous circulation.
Table 1 Clinical characteristics of comatose patients after cardiopulmonary arrest with brain MRIs before therapeutic hypothermia. Characteristic Gender, N (%) Age, years, mean ± SD Witness arrest, N (%) First monitored rhythm VF/pulseless VT, N (%) Asystole, N (%) PEA, N (%) Location of arrest Out-of-hospital, N (%) In-hospital, N (%) Bystander CPR Aetiology Cardiac origin, N (%) Respiratory origin, N (%) Absolute anoxic duration, min, median (IQR) Time from arrest to ROSC, min, mean ± SD Duration of BLS, min, mean ± SD Duration of ACLS, min, median (IQR) Time from arrest to MRI, min, median (IQR) Time from ROSC to MRI, min, median (IQR) Time from arrest to TH, min, mean ± SD Time from ROSC to TH, min, mean ± SD Outcome Good outcome, N (%) Poor outcome, N (%) Hospital stay, days, mean ± SD
N = 19 Male
16 (84.2) 54.6 ± 18.7 12 (63.2) 5 (26.3) 8 (42.1) 6 (31.6) 5 (26.3) 14 (73.7) 7 (36.8) 9 (47.4) 10 (52.6) 5.0 (0.0–9.0) 33.4 ± 15.5 9.3 ± 7.9 15.0 (8.0–25.0) 166.0 (114.0–240.0) 121.0 (88.0–199.0) 388.2 ± 126.9 348.5 ± 120.5
7 (36.8) 12 (63.2) 31.7 ± 15.0
Note: Return of spontaneous circulation (ROSC); basic life support (BLS); advanced cardiac life support (ACLS); magnetic resonance imaging (MRI); therapeutic hypothermia (TH).
J.S. Park et al. / Resuscitation 88 (2015) 132–137
135
Table 2 Qualitative DWI scores in seven patients with a good outcome and 12 patients with a poor outcome. Score (possible max.)
Overall (84) Cortex (24) DGN (16) Cortex + DGN (40) White matter (20) Brainstem (12) Cerebellum (12)
Good outcome
Poor outcome
Median (IQR)
Median (IQR)
3 (0.5–4.5) 3 (0.5–4.5) 0 3 (0.5–4.5) 0 0 0
21 (11–37) 12 (6.5–21.5) 6 (2.5–11) 18.5 (9.5–30.5) 0 0 (0–0.5) 3 (0–6.5)
p-value
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical Paper
Efficacy of diffusion-weighted magnetic resonance imaging performed before therapeutic hypothermia in predicting clinical outcome in comatose cardiopulmonary arrest survivors Jung Soo Park a , Suk Woo Lee a , Hoon Kim a , Jin Hong Min a,∗ , Jun Ho Kang a , Kyung Sik Yi b , Kyung Hye Park c , Byung Kook Lee d a
Department of Emergency Medicine, College of Medicine, Chungbuk National University Hospital, Cheongju, Republic of Korea Department of Radiology, College of Medicine, Chungbuk National University Hospital, Cheongju, Republic of Korea c Department of Emergency Medicine, Inje University, Busan, South Korea d Department of Emergency Medicine, Chonnam National University Hospital, 42, Jebong-ro, Donggu, Gwangju, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 4 September 2014 Received in revised form 13 November 2014 Accepted 28 November 2014 Keywords: Heart arrest Hypothermia Coma Prognosis
a b s t r a c t Aim of the study: To develop a clinically relevant and qualitative brain magnetic resonance imaging (MRI) scoring system for acute stage comatose cardiac arrest patients. Methods: Consecutive comatose post-cardiopulmonary arrest patients were prospectively enrolled. Routine brain MRI sequences were scored by two independent and blinded experts. Predefined brain regions were qualitatively scored on diffusion-weighted imaging (DWI) sequences according to the severity of the abnormality on a scale from 0 to 4. The mean score provided by the raters determined poor outcome defined under the Cerebral Performance Categories 3, 4, or 5. DWI scans were repeated after therapeutic hypothermia (TH). The same qualitative scoring system was applied and results were compared to the initial scores. Results: Out of 24 recruited patients, 19 with brain MRI scans were included. Of the 19 included patients, seven showed a good outcome at hospital discharge and 12 patients showed poor neurologic outcome. Median time from the arrest to the initial DWI was 166 min (IQR 114–240 min). At 100% specificity, the overall, cortex, and cortex plus deep grey nuclei scores predicted poor patient outcome with a sensitivity of 91.7–100% (95% CI). Follow-up DWI scans after TH showed worse results than initial scans. Conclusion: A qualitative MRI scoring system effectively assessed the severity of hypoxic-ischaemic brain injury following cardiopulmonary arrest. The scoring system may provide useful prognostic information in comatose cardiopulmonary arrest patients. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Improvements in intensive care medicine have led to increased bed ridden vegetative states and severe neurological deficits in cardiopulmonary arrested survivors (CPAS). In the past, only 10–30% of comatose CPAS demonstrated functional recovery.1 However, these percentages improved with increasing use of therapeutic hypothermia (TH).2,3 Post-cardiopulmonary arrest brain injury is a common cause of morbidity and mortality. Despite the use of TH, 45–70% of cardiac arrest victims still experience severe neurological deficits due to anoxic reperfusion cerebral injury.4,5 Conversely, premature withdrawal of life support from patients who may have a
∗ Corresponding author. Fax: +82 43 269 7396.. E-mail address: [email protected] (J.H. Min). http://dx.doi.org/10.1016/j.resuscitation.2014.11.031 0300-9572/© 2015 Elsevier Ireland Ltd. All rights reserved.
chance of functional recovery represents an ethical dilemma. Thus, early and accurate identification of patients who will not likely recover is an important healthcare issue. Many studies have focused on early identification of comatose CPAS expected not to regain consciousness.6–8 Historically, the most specific early predictors for poor outcome after cardiopulmonary arrest include the following: absence of brain stem or extensor reflexes, absence of motor response at post-arrest day 3, absence of cortical responses by somatosensory evoked potentials (SSEP) at 24 h post arrest, serum neuron specific enolase (NSE) levels >33 g/L in the first 3 days, and early myoclonic status epilepticus.9 These predictors have substantive limitations, a low sensitivity for poor outcome, and may not be valid in patients treated with hypothermia.10 The first limitation in the predictors is that they identify only a subset of poor-outcome patients with high specificity. Second, neurological examinations and results of
J.S. Park et al. / Resuscitation 88 (2015) 132–137
Brain regions Frontal lobe Supratentorial Cortex Parietal lobe Temporal lobe Occipital lobe Insula Hippocampus Deep grey nuclei Caudate Putamen Globus pallidus Thalamus White matter Frontal lobe Parietal lobe Temporal lobe O c c i p i t al l o b e Corpus callosum Midbrain Infratentorial Brainstem Pons Medulla Cerebellum Cortex White matter Dentate nuclei
133
Scoring system 0 = normal 1 = possibly abnormal 2 = abnormal, mild 3 = abnormal, moderate 4 = abnormal, severe
Fig. 1. The 21 scored brain regions and DWI scoring system.
electroencephalography may be difficult to interpret in the presence of sedative agents or metabolic derangements. Third, serum markers are potentially susceptible to false-positive test results and are not yet readily available in many hospitals.9,11 Several studies have reported brain magnetic resonance imaging (MRI) as a useful tool in imaging comatose CPAS.12–24 Typically, the presence of extensive and severe cortical signal abnormalities on MRI is associated with poor outcome.12–15,18–22 MRI offers the advantage of providing an objective measure of cerebral injury and may be particularly useful in patients who have received sedative agents. Furthermore, the method is useful in patients who have metabolic derangements that render neurological examinations unreliable and patients treated with TH. Recent developments in brain imaging through MRI include diffusion-weighted MRI (DWI). DWI is a novel, fast (10 min), and potentially powerful marker of early global ischaemic brain injury.21 Additionally, patient monitoring devices are available during DWI. Preliminary studies have shown widespread abnormalities detected by DWI in poor-outcome patients at 1 week post-cardiac arrest.21 However, obtaining a brain MRI in critically ill patients with potential cardiac instability is challenging. We used a clinically applicable brain MRI scoring system that is qualitative and has potential for broad use and application. This is an important tool for facilities without access to qualitative MRI analysis or technically inadequate scanners. This qualitative MRI scoring system has been described in previous studies as a tool to predict outcome following perinatal asphyxia.24–27 The overall aim of this study was to determine the feasibility and prognostic utility of DWI before TH in comatose CPAS in a prospective study.
approved by the institutional review board, and written consent from a legally authorized representative was required for study participation. Clinical parameters were obtained in a prospective and standardized fashion at predefined time points. Neurological examinations (including a Glasgow Coma Scale score and an assessment of brainstem function) were performed daily during the first 3 days and at 1 and 2 weeks after cardiopulmonary arrest. Functional outcomes were determined by the Cerebral Performance Categories (CPC) at discharge. The first MRI examinations were conducted before TH within at least 6 h. Additional MRI examinations were conducted within the first week after the arrest if the treating physicians agreed the patient could safely undergo MRI after TH. Reasons for patients not undergoing an MRI and adverse events during patient transport to or from the MRI suite and during imaging acquisition were documented. 2.2. Therapeutic hypothermia protocol
2. Methods
In all patients, TH was applied according to a written TH protocol. TH was induced with ice packs, intravenous cold saline, and cooling devices (Blanketrol II, Cincinnati Subzero Products, Cincinnati, USA; Artic Sun Energy Transfer Pads, Medivance Corp, Louisville, USA; COOLGARD 3000 Thermal Regulation System, Alsius Corporation, Irvine, USA). The target temperature of 33 ± 0.5 ◦ C was maintained for 24 h. Upon completion of the TH maintenance phase, patients were rewarmed to 36.5 ◦ C at a rate of 0.25–0.5 ◦ C h−1 . During TH, the temperature was monitored using a bladder temperature probe. Midazolam and atracurium were used for sedation and shivering control. All patients received standard intensive care according to our institutional intensive care unit protocol.
2.1. Study design and population
2.3. Magnetic resonance imaging and scoring
Consecutive CPAS who remained comatose after successful resuscitation were prospectively enrolled over a 2-year period at Chungbuk National University Hospital, a university-affiliated 600bed hospital in Chungju, Korea, from March 2012 to March 2014. Patients were enrolled if they met the following inclusion criteria: ≥16 years of age, status post resuscitation for in- or out-of-hospital cardiac arrest, and persistent coma defined as the inability to open eyes to voice and inability to follow commands. The study was
All patients with technically adequate brain DWIs obtained before TH within at least 6 h after the arrest are included in this study. MRIs were obtained with a 1.5 T system (Achieva 1.5 T; Philips Medical System, The Netherlands). For DWI, whole-brain axial plane, single-shot spin-echo planar imaging was acquired by applying diffusion-sensitizing gradients along three orthogonal directions with a diffusion weighting factor b = 1000 s/mm2 plus one reference scan with b = 0. The section thickness was 5 mm and
134
J.S. Park et al. / Resuscitation 88 (2015) 132–137
Fig. 2. Scoring of sample representative MRI scans.
the section gap was 1 mm. Patients were routinely scanned once within the 6-hour period post arrest before TH and all MRIs were included in the analyses. Images were assessed by a certified neuroradiologist and an emergency physician. The adjudicators were blinded to patient information, outcome, and readings of other adjudicators. Images were scored with the use of an MRI scoring system created to assess the severity of cerebral abnormalities due to cardiopulmonary arrest in predefined brain regions (Fig. 1). The adjudicators were instructed to score only MRI abnormalities that could be attributed to acute global hypoxic–ischaemic brain injury. The 21 brain regions scored individually were as follows: cortical grey matter and subcortical white matter in the frontal, parietal, temporal, and occipital lobes, hippocampus, insular cortex, corpus callosum, deep grey nuclei (caudate, putamen, globus pallidus, and thalamus), cerebellum (cortex, white matter, and dentate nucleus) and the brainstem (midbrain, pons, and medulla). All brain regions were scored using only DWI sequences according to the extent and severity of the signal abnormalities on a scale from 0 to 4. A score of 0, 1, 2, 3, and 4 indicated no abnormality, possibly abnormal, mildly abnormal, moderately abnormal, and severely abnormal, respectively (Fig. 2). The “overall score” consisted of all points given to all brain regions on DWI and the sum score was used during analyses. Follow up DWI scans were performed within 48 h of completion of TH. The same qualitatively scored system was applied and follow up scores were compared with the initial scores. Follow-up DWI scans were only performed when agreed upon by legal representatives. 2.4. Outcome Good outcome was defined as a CPC of 1 or 2 (good or moderate disability), and poor outcome was defined as a CPC of 3, 4, or 5 (severe disability, vegetative state, or brain death) at discharge. A CPC of 3 was considered a poor outcome because it was not sufficient to conduct common functions. 2.5. Statistical analysis Continuous variables were reported as the median and interquartile range (IQR) (non-normal distribution) or the mean and standard deviation (normal distribution). The Mann–Whitney test (non-paired result) or the Wilcoxon signed rank test (paired result) was conducted for comparisons of continuous variables. Categorical variables are represented as frequencies and percentages. The receiver operating characteristic (ROC) was analysed with the corresponding area under curve. Cut-off values with 100% specificity were calculated for predicting poor neurologic outcome at discharge. The agreement degree for DWI scoring between the neuroradiologist and emergency physician was measured using the kappa index and related 95% confidence intervals. Data were analysed using PASW/SPSSTM software, version 18 (IBM Inc., Chicago, USA). ROC curves were calculated and compared with MedCalc version 12.4.0 (free trial, MedCalc Software, Mariakerke, Belgium). Significance was set at p < 0.05.
3. Results 3.1. Clinical characteristics Over a 2-year study period, 288 patients who experienced cardiac arrest were admitted to our hospital and 30 patients survived. Of these, 24 patients met the inclusion criteria and 19 were included in the analyses. The participants were not included in the analysis if they were not treated with TH (n = 3) or were too unstable to undergo MRI (n = 2). The clinical characteristics of patients included in the analyses are shown in Table 1. According to the neurologic outcome at hospital discharge, seven patients (36.8%) were assigned to the good outcome group and 12 patients (63.2%) were assigned to the poor outcome group. A DWI was obtained at 166 min (114–240 min) after arrest and 121 min (88.01–99 min) after return of spontaneous circulation.
Table 1 Clinical characteristics of comatose patients after cardiopulmonary arrest with brain MRIs before therapeutic hypothermia. Characteristic Gender, N (%) Age, years, mean ± SD Witness arrest, N (%) First monitored rhythm VF/pulseless VT, N (%) Asystole, N (%) PEA, N (%) Location of arrest Out-of-hospital, N (%) In-hospital, N (%) Bystander CPR Aetiology Cardiac origin, N (%) Respiratory origin, N (%) Absolute anoxic duration, min, median (IQR) Time from arrest to ROSC, min, mean ± SD Duration of BLS, min, mean ± SD Duration of ACLS, min, median (IQR) Time from arrest to MRI, min, median (IQR) Time from ROSC to MRI, min, median (IQR) Time from arrest to TH, min, mean ± SD Time from ROSC to TH, min, mean ± SD Outcome Good outcome, N (%) Poor outcome, N (%) Hospital stay, days, mean ± SD
N = 19 Male
16 (84.2) 54.6 ± 18.7 12 (63.2) 5 (26.3) 8 (42.1) 6 (31.6) 5 (26.3) 14 (73.7) 7 (36.8) 9 (47.4) 10 (52.6) 5.0 (0.0–9.0) 33.4 ± 15.5 9.3 ± 7.9 15.0 (8.0–25.0) 166.0 (114.0–240.0) 121.0 (88.0–199.0) 388.2 ± 126.9 348.5 ± 120.5
7 (36.8) 12 (63.2) 31.7 ± 15.0
Note: Return of spontaneous circulation (ROSC); basic life support (BLS); advanced cardiac life support (ACLS); magnetic resonance imaging (MRI); therapeutic hypothermia (TH).
J.S. Park et al. / Resuscitation 88 (2015) 132–137
135
Table 2 Qualitative DWI scores in seven patients with a good outcome and 12 patients with a poor outcome. Score (possible max.)
Overall (84) Cortex (24) DGN (16) Cortex + DGN (40) White matter (20) Brainstem (12) Cerebellum (12)
Good outcome
Poor outcome
Median (IQR)
Median (IQR)
3 (0.5–4.5) 3 (0.5–4.5) 0 3 (0.5–4.5) 0 0 0
21 (11–37) 12 (6.5–21.5) 6 (2.5–11) 18.5 (9.5–30.5) 0 0 (0–0.5) 3 (0–6.5)
p-value
