Resuscitation 85 (2014) 664–670
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical Paper
Hemodynamics and vasopressor support in therapeutic hypothermia after cardiac arrest: Prognostic implications夽 John Bro-Jeppesen a,∗ , Jesper Kjaergaard a , Helle Søholm a , Michael Wanscher b , Freddy K. Lippert c , Jacob E. Møller a , Lars Køber a , Christian Hassager a a
Department of Cardiology, The Heart Centre, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark Department of Cardiothoracic Anaesthesia, The Heart Centre, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark c Emergency Medical Services, The Capital Region of Denmark, Copenhagen, Denmark b
a r t i c l e
i n f o
Article history: Received 10 September 2013 Received in revised form 20 December 2013 Accepted 22 December 2013 Keywords: Cardiac arrest Hypothermia Mortality Inotropic agents Hemodynamics
a b s t r a c t Aim: Inducing therapeutic hypothermia (TH) in Out-of-Hospital Cardiac Arrest (OHCA) can be challenging due to its impact on central hemodynamics and vasopressors are frequently used to maintain adequate organ perfusion. The aim of this study was to assess the association between level of vasopressor support and mortality. Methods: In a 6-year period, 310 comatose OHCA patients treated with TH were included. Temperature, hemodynamic parameters and level of vasopressors were registered from admission to 24 h after rewarming. Level of vasopressor support was assessed by the cardiovascular sub-score of Sequential Organ Failure Assessment (SOFA). The population was stratified by use of dopamine as first line intervention (D-group) or use of dopamine + norepinephrine/epinephrine (DA-group). Primary endpoint was 30-day mortality and secondary endpoint was in-hospital cause of death. Results: Patients in the DA-group carried a 49% all-cause 30-day mortality rate compared to 23% in the Dgroup, plog-rank < 0.0001, corresponding to an adjusted hazard ratio (HR) of 2.0 (95% CI: 1.3–3.0), p = 0.001). The DA-group had an increased 30-day mortality due to neurological injury (HR = 1.7 (95% CI: 1.1–2.7), p = 0.02). Cause of death was anoxic brain injury in 78%, cardiovascular failure in 18% and multi-organ failure in 4%. The hemodynamic changes of TH reversed at normothermia, although the requirement for vasopressor support (cardiovascular SOFA ≥ 3) persisted in 80% of patients. Conclusions: In survivors after OHCA treated with TH the induced hemodynamic changes reversed after normothermia, while the need for vasopressor support persisted. Patients requiring addition of norepinephrine/epinephrine on top of dopamine had an increased 30-day all-cause mortality, as well as death from neurological injury. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Targeted temperature management is an important part of the post-resuscitation care as stated in the Resuscitation Guidelines 2010 acknowledging that in-hospital interventions such as therapeutic hypothermia (TH) and emergency percutaneous coronary intervention have important implications for improving outcome after Out-of-Hospital Cardiac Arrest (OHCA).1–3 Hemodynamic instability is frequent in comatose survivors after OHCA as part
夽 A Spanish translated version of the summary of this article appears as Appendix in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2013.12.031. ∗ Corresponding author at: Department of Cardiology, The Heart Centre, Copenhagen University Hospital Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark. E-mail address: [email protected] (J. Bro-Jeppesen). 0300-9572/$ – see front matter © 2014 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.resuscitation.2013.12.031
of a post-cardiac arrest syndrome (PCAS) and may be exacerbated by induction of TH. Development of PCAS is characterized by impaired vasoregulation and myocardial stunning, which will cause hypotension necessitating vasopressors to maintain organ perfusion.4 Furthermore inducing TH can be challenging due to its impact on central hemodynamics as both cardiac index and heart rate are decreased and need for vasopressors is increased.5,6 Hypotension in the early phase of PCAS has previously been associated with increased in-hospital mortality, while induction of TH did not seem to be related to hypotension or increased need for vasopressors.7,8 However the hemodynamic impact of TH and use of vasopressors extends beyond the induction phase of TH and delayed hemodynamic instability with vasopressor requirements during and after TH may have important implications for outcome. The aim of this study was to assess the association between level of vasopressor support and mortality during and after TH and to
J. Bro-Jeppesen et al. / Resuscitation 85 (2014) 664–670
report the possible prognostic importance in consecutive comatose survivors of OHCA admitted for post-resuscitation care. 2. Materials and methods 2.1. Patient population The present study is an observational cohort study including consecutive comatose patients resuscitated from OHCA and admitted to the tertiary cardiac center at Copenhagen University Hospital Rigshospitalet in the period 1st June 2004 to 31st October 2010. Copenhagen, the capital of Denmark covers an area of 97 km2 and has approximately 600,000 inhabitants increasing during daytime by approximately 20%. Patients suffering from cardiac arrest in central Copenhagen and in the greater area of Copenhagen (population 1,200,000) are treated by the emergency medical system (EMS) staffed with physicians and referred to the nearest available hospital after ROSC. Patients suffering from OHCA in the central Copenhagen area were admitted for post-resuscitation care at Rigshospitalet. In addition, patients resuscitated from OHCA in the greater Copenhagen area were transferred for emergency coronary angiography at Rigshospitalet in case of suspected coronary occlusion. Inclusion criteria were (1) cardiac arrest with presumed cardiac cause; (2) age > 18 years; (3) sustained ROSC > 20 min; (4) Glasgow Coma Scale (GCS) ≤ 8 upon arrival in the emergency department or cardiac intensive care unit (ICU). Patients were excluded if severe cardiogenic shock was present at time of admission to the ICU. Severe cardiogenic shock was defined as need for inotropes/vasopressors to maintain systolic blood pressure above 80 and signs of organ hypoperfusion including treatment with intra-aortic balloon pump or left ventricular assist device. Patients developing cardiogenic shock during TH (n = 25) were not excluded. Patients were excluded if hypothermia treatment was not indicated, initiated or if target temperature (32–34 ◦ C) was not reached.
665
administration and adequate level of CVP dopamine was drug of first choice titrated to a maximum dose of 10 g kg−1 min−1 . If MAP remained below 65 mmHg despite volume substitution and maximum dose of dopamine, vasopressor therapy was escalated with norepinephrine and/or epinephrine at the discretion of the attending physician. Dopamine was combined with norepinephrine in 98 patients and in 20 patients with norepinephrine and epinephrine, while epinephrine was used on top of dopamine in 18 patients. 2.3. Data collection Pre-hospital data regarding the cardiac arrest including initial arrhythmia, witnessed arrest, administration of bystander CPR and time to ROSC were systematically collected upon admission according to Utstein guidelines.10 Data regarding temperature, hemodynamic parameters, sedation and infusion of vasopressors was registered on an hourly basis. Dose and type of vasopressors used (dopamine, norepinephrine and epinephrine) were registered and the highest level of vasopressor support in each phase of TH was used for analysis. Level of vasopressor support was assessed by the cardiovascular sub-score of Sequential Organ Failure Assessment (SOFA).11 The cardiovascular SOFA score was defined as: (0) MAP ≥ 70 mmHg, (1) MAP < 70 mmHg, (2) dopamine ≤ 5 g kg−1 min−1 , (3) dopamine > 5 g kg−1 min−1 or norepinephrine/epinephrine ≤0.1 g kg−1 min−1 , (4) dopamine > 15 g kg−1 min−1 or norepinephrine/epinephrine > 0.1 g kg−1 min−1 . The maximum cardiovascular SOFA registered was used for outcome analysis with the population stratified by median cardiovascular SOFA (≤3 vs. 4). The population was stratified in two vasopressor groups by use of dopamine (D-group) or use of dopamine + norepinephrine and/or epinephrine (DA-group) during the observation period. A single patient did not receive any vasopressor and was included in the D-group. The regional ethics committee approved the study with a waiver of written informed consent as all interventions were part of standard treatment and follow-up through Danish National Patient Registry does not require informed consent (H-4-2010-FSP).
2.2. Post-cardiac arrest care 2.4. Outcome All patients were admitted to the cardiac ICU for postresuscitation care with optimization of hemodynamic and metabolic parameters. General treatment goals were mean arterial pressure (MAP) above 65 mmHg, heart rate of 40–90 min−1 , central venous pressure (CVP) of 10–15 mmHg and urine output > 1.5 mL/kg/h. Patients’ hemodynamics were monitored with an arterial pressure catheter in the radial or femoral artery and a central venous line in the internal jugular or external jugular vein. TH was administered in all patients according to guidelines as soon as possible upon arrival in the emergency department or ICU.9 Hypothermia was induced by infusion of 30 mL/kg of 4 ◦ C Ringer’s solution and surface cooling was applied in the ICU using an external cooling system (Allon ThermowrapTM , MTRE, Israel or Emcools Flex.PadTM , Austria). The induction phase of TH was defined from ICU admission to reaching a core temperature below 34 ◦ C, the cooling phase was defined as hours with a target temperature (TT) of 32–34 ◦ C. The subsequent rewarming phase (targeting rewarming rate of no more than 0.5 ◦ C per hour) was defined as a core temperature above 34 ◦ C until a core temperature ≥36.5 ◦ C was reached. The following post-hypothermia phase was defined as 24 h after the end of the rewarming phase. The use of vasopressors to achieve adequate organ perfusion was primary directed by clinical parameters as MAP, CVP and urine output. Volume (crystalloids) was administered to achieve an adequate CVP. If MAP remained below 65 mmHg after volume
Vital status of patients included in the present study was obtained by linkage to the Danish National Patient Registry using the unique personal code number in May 2013 and follow-up was 100% complete. Primary endpoint was short-term survival assessed after 30 day. Long-term mortality was assessed at 1 year after OHCA. 2.5. Cause of death Cause of death in non-survivors at hospital discharge was classified in one of three categories; death related to neurological injury, death related to cardiovascular failure or death from multiorgan failure. Definitions were slightly modified from Laver et al. with regards to neurological death as death from multi-organ failure did not include signs of neurological failure in the present classification.12 Death due to neurological brain injury was defined by signs of extensive, irreversible brain damage and if active treatment was withdrawn for that reason. Classification of cause of death was assessed by two independent investigators and interobserver reliability was evaluated. 2.6. Statistics Data are presented as mean and standard deviation (SD) or proportions (%), and differences were assessed by ANOVA, Student’s t-test or 2 -test, as appropriate. For variables with a non-normal
666
J. Bro-Jeppesen et al. / Resuscitation 85 (2014) 664–670
Fig. 1. Flow chart. Flow diagram of patients admitted to Copenhagen University Hospital Rigshospitalet after OHCA in the period 2004–2010. D-group: infusion of dopamine; DA-group: infusion of dopamine with addition of norepinephrine/epinephrine. OHCA, Out-of-Hospital Cardiac Arrest; GCS, Glasgow Coma Scale.
distribution data are presented as median and 25th and 75th percentiles and differences were assessed by the Wilcoxon signed rank test. Between-groups differences for continuous variables such as vital signs and lactate were analyzed by a repeated-measurement model with vasopressor group as fixed effect. Between-group differences at specific time points (admission, TT, 12 and 24 h after TT, end of rewarming and end of post-hypothermia phase) were assessed by a t-test with Bonferroni correction to adjust for multiple comparisons. Logistic regression analysis with backward selection was used to analyze predictors of level of vasopressors. Cohen’s kappa statistics was used to assess the inter-observer reliability in the assessment of in-hospital cause of death by two independent investigators.13 Survival analysis was performed by Kaplan Meier plots and groups were compared with the log–rank test. Survival analysis was performed with all-cause mortality as outcome as well as survival free from neurological death, where patients experiencing cardiovascular deaths were censored at time of death. For multivariable modeling, Cox proportional hazards models with backward selection were applied after checking for underlying assumptions of linearity, proportionality and lack of interactions. All statistical analyses were performed using the SAS statistical software, version 9.1 (SAS Institute, Cary, NC).
3. Results 3.1. Patient population A total of 403 consecutive comatose patients with suspected cardiac cause of OHCA were admitted to Copenhagen University Hospital, Rigshospitalet in the study period, of whom 310 patients were treated with TH and reached target temperature, see Fig. 1. Patients in the D-group were younger (58 (SD 14) vs. 64 (SD 13) years, p = 0.0002) compared to the DA-group. No differences in initial rhythm, witnessed OHCA, bystander CPR or time to ROSC between the two vasopressor groups were found, Table 1. A multivariable logistic model, adjusting for variables in Table 1, showed that the DA-group was associated with higher age at arrest (ORadjusted = 1.2 95% CI: 1.0–1.3) per 5 years, p = 0.003), whereas traditional risk factors for mortality after OHCA as gender, initial rhythm, witnessed arrest, bystander CPR and time to ROSC were not associated with the DA-group.
Fig. 2. Distribution of vasopressor support. (A) Stacked box plot of distribution of D-group and DA-group in the induction, cooling, re-warming and post-hypothermia phases of therapeutic hypothermia. (B) Stacked box plot of distribution of cardiovascular SOFA score in the induction, cooling, re-warming and post-hypothermia phases of therapeutic hypothermia. SOFA: cardiovascular sub-score of Sequential Organ Failure Assessment.
3.2. Hemodynamic profile and level of vasopressor support A core temperature ≤34 ◦ C was reached with a median of 4.7 h (IQR 3.3–7.5 h) after OHCA and the duration of TH with a TT of 32–34 ◦ C was median 27 h (IQR 22–29 h). During the cooling phase 44 (14%) patients had an increase in requirement of vasopressor support with addition of norepinephrine/epinephrine to dopamine, while 28 (9%) patients and 37 (13%) had an increase in requirement of vasopressor support with addition of norepinephrine/epinephrine to dopamine in the rewarming and post-hypothermia phase, respectively. Vasopressor support with combined use of dopamine and norepinephrine/epinephrine was needed in 106 (36%) patients in the post-hypothermia phase, see Fig. 2A. During the cooling phase 225 (73%) patients had an increase in cardiovascular SOFA score compared to the induction phase and the cardiovascular SOFA score remained unchanged during the rewarming phase in 209 (67%) patients, while 55 (18%) patients had an increased cardiovascular SOFA score in the rewarming phase compared to the cooling phase. In the post-hypothermia phase 238 (80%) patients had a cardiovascular SOFA score ≥3, see Fig. 2B. There was no difference in duration of TH (26 (21–30) vs. 27 (23–29) h, p = 0.74) in the D-group versus DA-group and the overall difference in temperature between the two groups in the study period was ˇ = 0.1 ◦ C, pgroup = 0.30. The DA-group had an overall increased heart rate (ˇ = 10.4 min−1 , pgroup < 0.0001) compared to the D-group. These differences were present throughout the cooling phase, whereas no difference in heart rate was found in the rewarming and post-hypothermia phase, Fig. 3. At admission to the ICU there was no difference in MAP (74 (SD 16) vs. 77 (SD 15) mmHg, p = 0.12) in the two vasopressor groups, while at TT a significant difference in MAP (72 (SD 12) vs. 75 (SD
J. Bro-Jeppesen et al. / Resuscitation 85 (2014) 664–670
667
Table 1 Baseline characteristics and concurrent diseases of consecutive patients with Out-of-Hospital Cardiac Arrest.
Age, years, mean (SD) Male sex (%) Co-morbidities Hypertension (%) Diabetes mellitus (%) Known IHD (%) Heart failure (%) COPD (%) Renal failure (%) Previously cerebral stroke (%) Known malignancy (%) Alcohol abuse (%) Initial rhythm VF/VT (%) PEA/asystole (%) Witnessed arrest (%) Bystander CPR (%) Time to MECU arrival, min (IQR) Time to ROSC, min (IQR) Emergency coronary angiography, ≤12 h (%) Acute myocardial infarction (%) ST-elevation myocardial infarction (%)
Total n = 310
D-group n = 174
DA-group n = 136
p-Value
60 (14) 253 (82)
58 (14) 147 (84)
64 (13) 106 (78)
0.0002 0.14
81 (27) 39 (13) 74 (25) 40 (13) 25 (8) 5 (2) 25 (8) 5 (2) 36 (12)
37 (22) 19 (11) 39 (23) 18 (11) 16 (10) 0 (0) 13 (8) 2 (1) 21 (13)
44 (33) 20 (15) 35 (26) 22 (16) 9 (7) 5 (4) 12 (9) 3 (2) 15 (11)
0.04 0.35 0.58 0.15 0.38 0.01 0.70 0.48 0.71
265 (85) 45 (15) 268 (86) 188 (61) 5 (4–7) 15 (11–24) 184 (59) 169 (55) 109 (35)
150 (86) 24 (14) 156 (89) 112 (64) 5 (4–7) 15 (10–23) 107 (61) 95 (55) 58 (33)
115 (85) 21 (15) 112 (82) 76 (56) 5 (4–7) 17 (12–25) 77 (57) 74 (54) 51 (38)
0.68 0.06 0.13 0.66 0.10 0.39 0.97 0.45
Data are presented as mean and SD or median and interquartile range as appropriate. The p-value represents comparison between D- and DA-group; D-group, infusion of dopamine; DA-group, infusion of dopamine with addition of norepinephrine/epinephrine. A significance level of p < 0.05 was chosen. Abbreviations: IHD, ischemic heart disease; COPD, chronic obstructive pulmonary disease; VF/VT, ventricular fibrillation/ventricular tachycardia; PEA, pulseless electrical activity; CPR, cardio-pulmonary resuscitation; MECU, mobile emergency care unit; ROSC, return of spontaneously circulation.
11) mmHg, p = 0.03) was found between the DA-group and the D-group, respectively. This difference persisted between the two vasopressor groups throughout the cooling, rewarming and posthypothermia phases, with an estimated overall difference in MAP of ˇ = 4.1 mmHg, pgroup < 0.0001. Patients in the DA-group received smaller doses of propofol during the cooling phase compared to the D-group with an overall difference in the study period of 69.3 mg/h, p = 0.001. The lactate levels were increased in the DA-group during the cooling and rewarming phases with an estimated overall difference between the two groups of ˇ = 1.0 mmol/L, pgroup < 0.0001. 3.3. Mortality Overall 30-day mortality was 34% in the study population. Patients in the DA-group carried a 49% mortality rate compared to 23% in the D-group, plog-rank < 0.0001 (Fig. 4A) corresponding to
a hazard ratio of 2.0 (95% CI: 1.3–3.0), p = 0.001), adjusted for variables shown in Table 2. No significant interaction with development of CS and the DA-group was found (p = 0.09). One-year mortality was 51% and 26% in the DA-group compared to the D-group, respectively, plog-rank < 0.0001. Persisting use of combined dopamine and norepinephrine/epinephrine in the post-hypothermia phase was associated with increased mortality (HR = 1.8 (95% CI: 1.2–2.7), p = 0.007), whereas addition of norepinephrine/epinephrine to dopamine in the post-hypothermia phase compared to the previous phases only tended to be associated to increased mortality (HR = 1.4 (95% CI: 0.9–2.4), p = 0.15). A maximum cardiovascular SOFA score above the median was independently associated with increased mortality (HR = 1.7 (95% CI: 1.1–2.5), p = 0.01) adjusted for variables shown in Table 2. 3.4. Mortality related to neurological injury A mortality analysis assessing mortality rates due to neurological injury showed that patients in the DA-group carried a 40% 30-day mortality rate compared to 20% in the D-group, plog-rank = 0.0003 (Fig. 4B), corresponding to a hazard ratio of 1.7 (95% CI: 1.1–2.7), p = 0.02) for death due to neurological injury, adjusted for variables shown in Table 2. Similar a maximum cardiovascular SOFA score above the median was independently associated with increased mortality due to neurological injury (HR = 1.9 (95% CI: 1.2–3.1), p = 0.007). 3.5. In-hospital cause of death
Fig. 3. Hemodynamics after cardiac arrest combined with therapeutic hypothermia. Hemodynamic changes during and after therapeutic hypothermia. * Differences between the two groups at specific time points were assessed by a t-test with Bonferroni correction. T = 0 was defined at time for reaching target temperature ≤34.0 ◦ C. ; DA-group: . Temperature: —; Heart rate: - · -; Mean arterial D-group: pressure: - - -; Lactate: · · ·.
In 81 (78%) patients, cause of death was due to neurological injury, 19 (18%) died from cardiovascular causes and four (4%) died in multi-organ failure. Inter-observer reliability in assessing cause of death was excellent (kappa = 0.90 (95% CI 0.72–1.00). In patients dying due to neurological injury, life-sustaining therapy was withdrawn in 79 (97.5%) of patients due to presumed irreversible neurological injury. After discontinuation of sedation, neuroprognostication with at least one modality was performed in
668
J. Bro-Jeppesen et al. / Resuscitation 85 (2014) 664–670
Table 2 Univariable and multivariable Cox proportional hazard models predicting 30-day mortality. Univariable
DA vs. D-group (ref.) Development of cardiogenic shock during TH Sex, male Age, per 5 years Witnessed arrest Bystander CPR Initial rhythm, VF Time to ROSC, per 5 min Acute coronary angiography, ≤12 h Acute myocardial infarction No co-morbidities
Multivariable
HR (95% CI)
p-Value
HR (95% CI)
p-Value
2.5 (1.7–3.7) 6.3 (3.9–10.5) 0.5 (0.3–0.8) 1.1 (1.0–1.2) 0.5 (0.3–0.8) 0.4 (0.3–0.6) 0.4 (0.3–0.6) 1.1 (1.1–1.2) 0.8 (0.5–1.1) 1.0 (0.7–1.4) 0.6 (0.4–1.0)
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical Paper
Hemodynamics and vasopressor support in therapeutic hypothermia after cardiac arrest: Prognostic implications夽 John Bro-Jeppesen a,∗ , Jesper Kjaergaard a , Helle Søholm a , Michael Wanscher b , Freddy K. Lippert c , Jacob E. Møller a , Lars Køber a , Christian Hassager a a
Department of Cardiology, The Heart Centre, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark Department of Cardiothoracic Anaesthesia, The Heart Centre, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark c Emergency Medical Services, The Capital Region of Denmark, Copenhagen, Denmark b
a r t i c l e
i n f o
Article history: Received 10 September 2013 Received in revised form 20 December 2013 Accepted 22 December 2013 Keywords: Cardiac arrest Hypothermia Mortality Inotropic agents Hemodynamics
a b s t r a c t Aim: Inducing therapeutic hypothermia (TH) in Out-of-Hospital Cardiac Arrest (OHCA) can be challenging due to its impact on central hemodynamics and vasopressors are frequently used to maintain adequate organ perfusion. The aim of this study was to assess the association between level of vasopressor support and mortality. Methods: In a 6-year period, 310 comatose OHCA patients treated with TH were included. Temperature, hemodynamic parameters and level of vasopressors were registered from admission to 24 h after rewarming. Level of vasopressor support was assessed by the cardiovascular sub-score of Sequential Organ Failure Assessment (SOFA). The population was stratified by use of dopamine as first line intervention (D-group) or use of dopamine + norepinephrine/epinephrine (DA-group). Primary endpoint was 30-day mortality and secondary endpoint was in-hospital cause of death. Results: Patients in the DA-group carried a 49% all-cause 30-day mortality rate compared to 23% in the Dgroup, plog-rank < 0.0001, corresponding to an adjusted hazard ratio (HR) of 2.0 (95% CI: 1.3–3.0), p = 0.001). The DA-group had an increased 30-day mortality due to neurological injury (HR = 1.7 (95% CI: 1.1–2.7), p = 0.02). Cause of death was anoxic brain injury in 78%, cardiovascular failure in 18% and multi-organ failure in 4%. The hemodynamic changes of TH reversed at normothermia, although the requirement for vasopressor support (cardiovascular SOFA ≥ 3) persisted in 80% of patients. Conclusions: In survivors after OHCA treated with TH the induced hemodynamic changes reversed after normothermia, while the need for vasopressor support persisted. Patients requiring addition of norepinephrine/epinephrine on top of dopamine had an increased 30-day all-cause mortality, as well as death from neurological injury. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Targeted temperature management is an important part of the post-resuscitation care as stated in the Resuscitation Guidelines 2010 acknowledging that in-hospital interventions such as therapeutic hypothermia (TH) and emergency percutaneous coronary intervention have important implications for improving outcome after Out-of-Hospital Cardiac Arrest (OHCA).1–3 Hemodynamic instability is frequent in comatose survivors after OHCA as part
夽 A Spanish translated version of the summary of this article appears as Appendix in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2013.12.031. ∗ Corresponding author at: Department of Cardiology, The Heart Centre, Copenhagen University Hospital Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark. E-mail address: [email protected] (J. Bro-Jeppesen). 0300-9572/$ – see front matter © 2014 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.resuscitation.2013.12.031
of a post-cardiac arrest syndrome (PCAS) and may be exacerbated by induction of TH. Development of PCAS is characterized by impaired vasoregulation and myocardial stunning, which will cause hypotension necessitating vasopressors to maintain organ perfusion.4 Furthermore inducing TH can be challenging due to its impact on central hemodynamics as both cardiac index and heart rate are decreased and need for vasopressors is increased.5,6 Hypotension in the early phase of PCAS has previously been associated with increased in-hospital mortality, while induction of TH did not seem to be related to hypotension or increased need for vasopressors.7,8 However the hemodynamic impact of TH and use of vasopressors extends beyond the induction phase of TH and delayed hemodynamic instability with vasopressor requirements during and after TH may have important implications for outcome. The aim of this study was to assess the association between level of vasopressor support and mortality during and after TH and to
J. Bro-Jeppesen et al. / Resuscitation 85 (2014) 664–670
report the possible prognostic importance in consecutive comatose survivors of OHCA admitted for post-resuscitation care. 2. Materials and methods 2.1. Patient population The present study is an observational cohort study including consecutive comatose patients resuscitated from OHCA and admitted to the tertiary cardiac center at Copenhagen University Hospital Rigshospitalet in the period 1st June 2004 to 31st October 2010. Copenhagen, the capital of Denmark covers an area of 97 km2 and has approximately 600,000 inhabitants increasing during daytime by approximately 20%. Patients suffering from cardiac arrest in central Copenhagen and in the greater area of Copenhagen (population 1,200,000) are treated by the emergency medical system (EMS) staffed with physicians and referred to the nearest available hospital after ROSC. Patients suffering from OHCA in the central Copenhagen area were admitted for post-resuscitation care at Rigshospitalet. In addition, patients resuscitated from OHCA in the greater Copenhagen area were transferred for emergency coronary angiography at Rigshospitalet in case of suspected coronary occlusion. Inclusion criteria were (1) cardiac arrest with presumed cardiac cause; (2) age > 18 years; (3) sustained ROSC > 20 min; (4) Glasgow Coma Scale (GCS) ≤ 8 upon arrival in the emergency department or cardiac intensive care unit (ICU). Patients were excluded if severe cardiogenic shock was present at time of admission to the ICU. Severe cardiogenic shock was defined as need for inotropes/vasopressors to maintain systolic blood pressure above 80 and signs of organ hypoperfusion including treatment with intra-aortic balloon pump or left ventricular assist device. Patients developing cardiogenic shock during TH (n = 25) were not excluded. Patients were excluded if hypothermia treatment was not indicated, initiated or if target temperature (32–34 ◦ C) was not reached.
665
administration and adequate level of CVP dopamine was drug of first choice titrated to a maximum dose of 10 g kg−1 min−1 . If MAP remained below 65 mmHg despite volume substitution and maximum dose of dopamine, vasopressor therapy was escalated with norepinephrine and/or epinephrine at the discretion of the attending physician. Dopamine was combined with norepinephrine in 98 patients and in 20 patients with norepinephrine and epinephrine, while epinephrine was used on top of dopamine in 18 patients. 2.3. Data collection Pre-hospital data regarding the cardiac arrest including initial arrhythmia, witnessed arrest, administration of bystander CPR and time to ROSC were systematically collected upon admission according to Utstein guidelines.10 Data regarding temperature, hemodynamic parameters, sedation and infusion of vasopressors was registered on an hourly basis. Dose and type of vasopressors used (dopamine, norepinephrine and epinephrine) were registered and the highest level of vasopressor support in each phase of TH was used for analysis. Level of vasopressor support was assessed by the cardiovascular sub-score of Sequential Organ Failure Assessment (SOFA).11 The cardiovascular SOFA score was defined as: (0) MAP ≥ 70 mmHg, (1) MAP < 70 mmHg, (2) dopamine ≤ 5 g kg−1 min−1 , (3) dopamine > 5 g kg−1 min−1 or norepinephrine/epinephrine ≤0.1 g kg−1 min−1 , (4) dopamine > 15 g kg−1 min−1 or norepinephrine/epinephrine > 0.1 g kg−1 min−1 . The maximum cardiovascular SOFA registered was used for outcome analysis with the population stratified by median cardiovascular SOFA (≤3 vs. 4). The population was stratified in two vasopressor groups by use of dopamine (D-group) or use of dopamine + norepinephrine and/or epinephrine (DA-group) during the observation period. A single patient did not receive any vasopressor and was included in the D-group. The regional ethics committee approved the study with a waiver of written informed consent as all interventions were part of standard treatment and follow-up through Danish National Patient Registry does not require informed consent (H-4-2010-FSP).
2.2. Post-cardiac arrest care 2.4. Outcome All patients were admitted to the cardiac ICU for postresuscitation care with optimization of hemodynamic and metabolic parameters. General treatment goals were mean arterial pressure (MAP) above 65 mmHg, heart rate of 40–90 min−1 , central venous pressure (CVP) of 10–15 mmHg and urine output > 1.5 mL/kg/h. Patients’ hemodynamics were monitored with an arterial pressure catheter in the radial or femoral artery and a central venous line in the internal jugular or external jugular vein. TH was administered in all patients according to guidelines as soon as possible upon arrival in the emergency department or ICU.9 Hypothermia was induced by infusion of 30 mL/kg of 4 ◦ C Ringer’s solution and surface cooling was applied in the ICU using an external cooling system (Allon ThermowrapTM , MTRE, Israel or Emcools Flex.PadTM , Austria). The induction phase of TH was defined from ICU admission to reaching a core temperature below 34 ◦ C, the cooling phase was defined as hours with a target temperature (TT) of 32–34 ◦ C. The subsequent rewarming phase (targeting rewarming rate of no more than 0.5 ◦ C per hour) was defined as a core temperature above 34 ◦ C until a core temperature ≥36.5 ◦ C was reached. The following post-hypothermia phase was defined as 24 h after the end of the rewarming phase. The use of vasopressors to achieve adequate organ perfusion was primary directed by clinical parameters as MAP, CVP and urine output. Volume (crystalloids) was administered to achieve an adequate CVP. If MAP remained below 65 mmHg after volume
Vital status of patients included in the present study was obtained by linkage to the Danish National Patient Registry using the unique personal code number in May 2013 and follow-up was 100% complete. Primary endpoint was short-term survival assessed after 30 day. Long-term mortality was assessed at 1 year after OHCA. 2.5. Cause of death Cause of death in non-survivors at hospital discharge was classified in one of three categories; death related to neurological injury, death related to cardiovascular failure or death from multiorgan failure. Definitions were slightly modified from Laver et al. with regards to neurological death as death from multi-organ failure did not include signs of neurological failure in the present classification.12 Death due to neurological brain injury was defined by signs of extensive, irreversible brain damage and if active treatment was withdrawn for that reason. Classification of cause of death was assessed by two independent investigators and interobserver reliability was evaluated. 2.6. Statistics Data are presented as mean and standard deviation (SD) or proportions (%), and differences were assessed by ANOVA, Student’s t-test or 2 -test, as appropriate. For variables with a non-normal
666
J. Bro-Jeppesen et al. / Resuscitation 85 (2014) 664–670
Fig. 1. Flow chart. Flow diagram of patients admitted to Copenhagen University Hospital Rigshospitalet after OHCA in the period 2004–2010. D-group: infusion of dopamine; DA-group: infusion of dopamine with addition of norepinephrine/epinephrine. OHCA, Out-of-Hospital Cardiac Arrest; GCS, Glasgow Coma Scale.
distribution data are presented as median and 25th and 75th percentiles and differences were assessed by the Wilcoxon signed rank test. Between-groups differences for continuous variables such as vital signs and lactate were analyzed by a repeated-measurement model with vasopressor group as fixed effect. Between-group differences at specific time points (admission, TT, 12 and 24 h after TT, end of rewarming and end of post-hypothermia phase) were assessed by a t-test with Bonferroni correction to adjust for multiple comparisons. Logistic regression analysis with backward selection was used to analyze predictors of level of vasopressors. Cohen’s kappa statistics was used to assess the inter-observer reliability in the assessment of in-hospital cause of death by two independent investigators.13 Survival analysis was performed by Kaplan Meier plots and groups were compared with the log–rank test. Survival analysis was performed with all-cause mortality as outcome as well as survival free from neurological death, where patients experiencing cardiovascular deaths were censored at time of death. For multivariable modeling, Cox proportional hazards models with backward selection were applied after checking for underlying assumptions of linearity, proportionality and lack of interactions. All statistical analyses were performed using the SAS statistical software, version 9.1 (SAS Institute, Cary, NC).
3. Results 3.1. Patient population A total of 403 consecutive comatose patients with suspected cardiac cause of OHCA were admitted to Copenhagen University Hospital, Rigshospitalet in the study period, of whom 310 patients were treated with TH and reached target temperature, see Fig. 1. Patients in the D-group were younger (58 (SD 14) vs. 64 (SD 13) years, p = 0.0002) compared to the DA-group. No differences in initial rhythm, witnessed OHCA, bystander CPR or time to ROSC between the two vasopressor groups were found, Table 1. A multivariable logistic model, adjusting for variables in Table 1, showed that the DA-group was associated with higher age at arrest (ORadjusted = 1.2 95% CI: 1.0–1.3) per 5 years, p = 0.003), whereas traditional risk factors for mortality after OHCA as gender, initial rhythm, witnessed arrest, bystander CPR and time to ROSC were not associated with the DA-group.
Fig. 2. Distribution of vasopressor support. (A) Stacked box plot of distribution of D-group and DA-group in the induction, cooling, re-warming and post-hypothermia phases of therapeutic hypothermia. (B) Stacked box plot of distribution of cardiovascular SOFA score in the induction, cooling, re-warming and post-hypothermia phases of therapeutic hypothermia. SOFA: cardiovascular sub-score of Sequential Organ Failure Assessment.
3.2. Hemodynamic profile and level of vasopressor support A core temperature ≤34 ◦ C was reached with a median of 4.7 h (IQR 3.3–7.5 h) after OHCA and the duration of TH with a TT of 32–34 ◦ C was median 27 h (IQR 22–29 h). During the cooling phase 44 (14%) patients had an increase in requirement of vasopressor support with addition of norepinephrine/epinephrine to dopamine, while 28 (9%) patients and 37 (13%) had an increase in requirement of vasopressor support with addition of norepinephrine/epinephrine to dopamine in the rewarming and post-hypothermia phase, respectively. Vasopressor support with combined use of dopamine and norepinephrine/epinephrine was needed in 106 (36%) patients in the post-hypothermia phase, see Fig. 2A. During the cooling phase 225 (73%) patients had an increase in cardiovascular SOFA score compared to the induction phase and the cardiovascular SOFA score remained unchanged during the rewarming phase in 209 (67%) patients, while 55 (18%) patients had an increased cardiovascular SOFA score in the rewarming phase compared to the cooling phase. In the post-hypothermia phase 238 (80%) patients had a cardiovascular SOFA score ≥3, see Fig. 2B. There was no difference in duration of TH (26 (21–30) vs. 27 (23–29) h, p = 0.74) in the D-group versus DA-group and the overall difference in temperature between the two groups in the study period was ˇ = 0.1 ◦ C, pgroup = 0.30. The DA-group had an overall increased heart rate (ˇ = 10.4 min−1 , pgroup < 0.0001) compared to the D-group. These differences were present throughout the cooling phase, whereas no difference in heart rate was found in the rewarming and post-hypothermia phase, Fig. 3. At admission to the ICU there was no difference in MAP (74 (SD 16) vs. 77 (SD 15) mmHg, p = 0.12) in the two vasopressor groups, while at TT a significant difference in MAP (72 (SD 12) vs. 75 (SD
J. Bro-Jeppesen et al. / Resuscitation 85 (2014) 664–670
667
Table 1 Baseline characteristics and concurrent diseases of consecutive patients with Out-of-Hospital Cardiac Arrest.
Age, years, mean (SD) Male sex (%) Co-morbidities Hypertension (%) Diabetes mellitus (%) Known IHD (%) Heart failure (%) COPD (%) Renal failure (%) Previously cerebral stroke (%) Known malignancy (%) Alcohol abuse (%) Initial rhythm VF/VT (%) PEA/asystole (%) Witnessed arrest (%) Bystander CPR (%) Time to MECU arrival, min (IQR) Time to ROSC, min (IQR) Emergency coronary angiography, ≤12 h (%) Acute myocardial infarction (%) ST-elevation myocardial infarction (%)
Total n = 310
D-group n = 174
DA-group n = 136
p-Value
60 (14) 253 (82)
58 (14) 147 (84)
64 (13) 106 (78)
0.0002 0.14
81 (27) 39 (13) 74 (25) 40 (13) 25 (8) 5 (2) 25 (8) 5 (2) 36 (12)
37 (22) 19 (11) 39 (23) 18 (11) 16 (10) 0 (0) 13 (8) 2 (1) 21 (13)
44 (33) 20 (15) 35 (26) 22 (16) 9 (7) 5 (4) 12 (9) 3 (2) 15 (11)
0.04 0.35 0.58 0.15 0.38 0.01 0.70 0.48 0.71
265 (85) 45 (15) 268 (86) 188 (61) 5 (4–7) 15 (11–24) 184 (59) 169 (55) 109 (35)
150 (86) 24 (14) 156 (89) 112 (64) 5 (4–7) 15 (10–23) 107 (61) 95 (55) 58 (33)
115 (85) 21 (15) 112 (82) 76 (56) 5 (4–7) 17 (12–25) 77 (57) 74 (54) 51 (38)
0.68 0.06 0.13 0.66 0.10 0.39 0.97 0.45
Data are presented as mean and SD or median and interquartile range as appropriate. The p-value represents comparison between D- and DA-group; D-group, infusion of dopamine; DA-group, infusion of dopamine with addition of norepinephrine/epinephrine. A significance level of p < 0.05 was chosen. Abbreviations: IHD, ischemic heart disease; COPD, chronic obstructive pulmonary disease; VF/VT, ventricular fibrillation/ventricular tachycardia; PEA, pulseless electrical activity; CPR, cardio-pulmonary resuscitation; MECU, mobile emergency care unit; ROSC, return of spontaneously circulation.
11) mmHg, p = 0.03) was found between the DA-group and the D-group, respectively. This difference persisted between the two vasopressor groups throughout the cooling, rewarming and posthypothermia phases, with an estimated overall difference in MAP of ˇ = 4.1 mmHg, pgroup < 0.0001. Patients in the DA-group received smaller doses of propofol during the cooling phase compared to the D-group with an overall difference in the study period of 69.3 mg/h, p = 0.001. The lactate levels were increased in the DA-group during the cooling and rewarming phases with an estimated overall difference between the two groups of ˇ = 1.0 mmol/L, pgroup < 0.0001. 3.3. Mortality Overall 30-day mortality was 34% in the study population. Patients in the DA-group carried a 49% mortality rate compared to 23% in the D-group, plog-rank < 0.0001 (Fig. 4A) corresponding to
a hazard ratio of 2.0 (95% CI: 1.3–3.0), p = 0.001), adjusted for variables shown in Table 2. No significant interaction with development of CS and the DA-group was found (p = 0.09). One-year mortality was 51% and 26% in the DA-group compared to the D-group, respectively, plog-rank < 0.0001. Persisting use of combined dopamine and norepinephrine/epinephrine in the post-hypothermia phase was associated with increased mortality (HR = 1.8 (95% CI: 1.2–2.7), p = 0.007), whereas addition of norepinephrine/epinephrine to dopamine in the post-hypothermia phase compared to the previous phases only tended to be associated to increased mortality (HR = 1.4 (95% CI: 0.9–2.4), p = 0.15). A maximum cardiovascular SOFA score above the median was independently associated with increased mortality (HR = 1.7 (95% CI: 1.1–2.5), p = 0.01) adjusted for variables shown in Table 2. 3.4. Mortality related to neurological injury A mortality analysis assessing mortality rates due to neurological injury showed that patients in the DA-group carried a 40% 30-day mortality rate compared to 20% in the D-group, plog-rank = 0.0003 (Fig. 4B), corresponding to a hazard ratio of 1.7 (95% CI: 1.1–2.7), p = 0.02) for death due to neurological injury, adjusted for variables shown in Table 2. Similar a maximum cardiovascular SOFA score above the median was independently associated with increased mortality due to neurological injury (HR = 1.9 (95% CI: 1.2–3.1), p = 0.007). 3.5. In-hospital cause of death
Fig. 3. Hemodynamics after cardiac arrest combined with therapeutic hypothermia. Hemodynamic changes during and after therapeutic hypothermia. * Differences between the two groups at specific time points were assessed by a t-test with Bonferroni correction. T = 0 was defined at time for reaching target temperature ≤34.0 ◦ C. ; DA-group: . Temperature: —; Heart rate: - · -; Mean arterial D-group: pressure: - - -; Lactate: · · ·.
In 81 (78%) patients, cause of death was due to neurological injury, 19 (18%) died from cardiovascular causes and four (4%) died in multi-organ failure. Inter-observer reliability in assessing cause of death was excellent (kappa = 0.90 (95% CI 0.72–1.00). In patients dying due to neurological injury, life-sustaining therapy was withdrawn in 79 (97.5%) of patients due to presumed irreversible neurological injury. After discontinuation of sedation, neuroprognostication with at least one modality was performed in
668
J. Bro-Jeppesen et al. / Resuscitation 85 (2014) 664–670
Table 2 Univariable and multivariable Cox proportional hazard models predicting 30-day mortality. Univariable
DA vs. D-group (ref.) Development of cardiogenic shock during TH Sex, male Age, per 5 years Witnessed arrest Bystander CPR Initial rhythm, VF Time to ROSC, per 5 min Acute coronary angiography, ≤12 h Acute myocardial infarction No co-morbidities
Multivariable
HR (95% CI)
p-Value
HR (95% CI)
p-Value
2.5 (1.7–3.7) 6.3 (3.9–10.5) 0.5 (0.3–0.8) 1.1 (1.0–1.2) 0.5 (0.3–0.8) 0.4 (0.3–0.6) 0.4 (0.3–0.6) 1.1 (1.1–1.2) 0.8 (0.5–1.1) 1.0 (0.7–1.4) 0.6 (0.4–1.0)
