Secondary Logo

Journal Logo

Clinical Science Aspects

Shock Severity and Hospital Mortality In Out of Hospital Cardiac Arrest Patients Treated With Targeted Temperature Management

Tabi, Meir; Burstein, Barry J.; Ahmed, Abdelrahman; Dezfulian, Cameron; Kashani, Kianoush B.†,§; Jentzer, Jacob C.∗,†

Author Information
doi: 10.1097/SHK.0000000000001600



Approximately 350 to 400,000 people are treated for out-of-hospital cardiac arrest (OHCA) annually in the United States (1, 2). Patients with return of spontaneous circulation (ROSC) after cardiac arrest remain at high risk of death during hospitalization, with many survivors experiencing neurologic disability (3, 4). The poor outcomes observed in OHCA patients who remain comatose after ROSC can be improved by the use of targeted temperature management (TTM) to mitigate brain injury (5). Although hemodynamic optimization is frequently recommended after ROSC, little is known about whether this impacts outcomes for OHCA patients (6, 7).

Hypotension occurs in up to 65% of OHCA patients during the first hours after ROSC, and early hypotension is associated with higher rates of poor neurologic outcome and mortality (3, 4, 8). OHCA patients with a higher mean arterial pressure (MAP) during the first 6 to 24 h after ROSC have better outcomes, leading some experts to recommend a higher target MAP for OHCA patients (7, 9, 10). A recent randomized study in OHCA patients did not show any improvement in the extent of anoxic brain damage or neurological outcome by targeting a higher MAP goal (85–100 mm Hg) during the first 36 h in ICU (11). Other markers of shock, such as elevated blood lactate levels and vasopressor dependence, are observed in the majority of patients after OHCA and are consistently associated with adverse outcomes and increased mortality (12, 13). The use of TTM can potentially worsen these measures of shock severity, leading to harm in patients with severe shock after OHCA (14, 15).

Uncertainty remains regarding the extent to which hypotension is a potentially modifiable contributor to worse outcomes in patients treated with TTM. The objective of this study was to evaluate the combination of several measures of shock severity, including MAP, vasopressor requirements, and blood lactate levels, and if they were associated with mortality in this unique population of post-OHCA patients treated with TTM.


This study was approved by the Institutional Review Board of the Mayo Clinic (Rochester, Minn) as posing minimal risk to subjects under a waiver of informed consent. This historical cohort study was performed at Mayo Clinic Hospital, Saint Mary's campus, in Rochester, Minn. Comatose adult non-surgical patients resuscitated from OHCA who were admitted to the Cardiac Intensive Care Unit between December 2005 and September 2016 were included if they underwent targeted temperature management (TTM) using a computer-controlled cooling device. Our CICU protocol for TTM requires initiation of cooling immediately at admission and the prevalent target temperature during the study period was 33C. There were no specific institutional protocols for cardiovascular management (including a prespecified MAP target) during the study period, and all essential aspects of care were determined by the Cardiac Intensive Care Unit team. We excluded patients <18 years old, individuals whose cardiac arrest occurred in the hospital, patients who were not treated with TTM, and patients who did not have available data on MAP or vasopressor use during the first 24 h.

Demographic, vital sign, laboratory, clinical and outcome data were extracted electronically from the medical record using the Multidisciplinary Epidemiology and Translational Research in Intensive Care (METRIC) DataMart; vital sign data were imported into the METRIC DataMart every 15 min, with maximum, minimum, and mean values calculated (16). Charts were manually reviewed for clinical coronary angiography and vasopressor data. Automatically calculated severity of illness scores were obtained from the METRIC DataMart, including the Sequential Organ Failure Assessment (SOFA) and Acute Physiology and Chronic Health Evaluation (APACHE) scores (17, 18). A shockable arrest rhythm was defined as an initial rhythm of ventricular tachycardia or ventricular fibrillation; subjects with an initial rhythm of asystole or pulseless electrical activity were classified as non-shockable rhythms, even if they subsequently received a defibrillator shock. Maximum simultaneous vasopressor and inotrope doses used to calculate the Vasoactive-Inotropic Score (VIS) (19). Echocardiographic data were obtained from the Mayo Clinic database, as previously described (20). Neuron-specific enolase (NSE) levels were obtained on CICU admission and daily for 3 days per protocol, and the peak of these values was recorded.

Hospital and 30-day mortality and modified Rankin Scale at hospital discharge (mRS) were determined by manual chart review, and mRS ≤2 was considered a good neurological outcome. Continuous variables are summarized as median (interquartile range (IQR)); categorical variables are presented as number (%). Patients who received early vasopressors (dopamine, epinephrine, norepinephrine, phenylephrine, or vasopressin during the first 24 h of hospitalization) were compared to patients who did not. Comparisons of continuous variables between groups are performed using two-tailed Wilcoxon rank-sum test; categorical variables are compared using two-tailed Fisher exact test. Multivariate logistic regression was performed to identify predictors of hospital mortality, including age, APACHE-III score, CCI, arrest characteristics, use of inpatient coronary angiography, peak 24-h VIS (VIS24), initial lactate, mean 24-h MAP (mMAP24), and peak NSE; a statistical interaction term was then added for peak 24-h VIS and mean 24-h MAP. Multivariable logistic regression was repeated separately for patients who did and did not receive early vasopressors. Cox proportional-hazards analysis was performed for 30-day mortality using these same covariates. All statistical analyses were done using JMP Pro version 14.0 (SAS Institute, Cary, NC).


Study population

A total of 279 patients were admitted with OHCA and treated with TTM between December 2006 and September 2016. We excluded 10 patients who did not have available data for vital signs and one patient without available vasopressor data, leaving 268 patients in the final study population. The final study population had a median age of 63.5 (55, 71.8) years, including 72 (26.9%) females. Initial arrest rhythm was shockable in 232 (86.6%) patients, arrest was witnessed in 236 (89.1%) patients, and 168 (63.4%) patients received bystander CPR.

Measures of shock severity

A total of 201 (75.0%) of patients received vasopressors, including 160 (59.7%) patients during the first 24 h after admission (early vasopressors group) and an additional 41 patients after the first 24 h (38.0% of patients not initially on vasopressors). The median VIS24 among patients requiring early vasopressors was 15 (7, 34.3). Patients who received early vasopressors had similar demographics, comorbidities, and arrest circumstances, but higher illness severity and shock severity by all measures compared to patients who did not require vasopressors (Table 1).

Table 1 - Baseline characteristics of the final study population as well as patients who did and did not receive vasopressors during the first 24 h after admission.
Variable N with data Final study population (n = 268) Early vasopressors (n = 160) No early vasopressors (n = 108) P value
Baseline characteristics
 Age 268 63.5 (55,71.8) 64 (56.2,74) 63 (52,68) 0.03
 Male gender 268 196 (73.1%) 110 (68.8%) 86 (79.6%) 0.05
 Charlson comorbidity index 268 0 (0,2) 0 (0,2) 0 (0,2) 0.23
 Prior MI 268 33 (12.3%) 18 (11.2%) 15 (13.9%) 0.57
 Prior HF 268 21 (7.8%) 16 (10.0%) 5 (4.6%) 0.16
 Prior CKD 268 25 (9.3%) 18 (11.2%) 7 (6.5%) 0.21
 Prior DM 268 40 (14.9%) 23 (14.4%) 17 (15.7%) 0.86
 CICU length of stay (days) 267 4.7 (2.8,6.7) 4.1 (2.5,6.3) 5.0 (3.5,7.1) 0.02
 Hospital length of stay (days) 267 7.9 (4.1,11.0) 7.1 (2.9,11.2) 8.5 (5.7,10.7) 0.07
 Hospital mortality 268 104 (38.8%) 76 (47.5%) 28 (25.9%) <0.001
Arrest characteristics
 Shockable rhythm 268 282 (86.6%) 138 (86.2%) 94 (87.0%) 1.0
 Witnessed arrest 265 236 (89.1%) 142 (89.9%) 94 (87.8%) 0.69
 Bystander CPR 265 168 (63.4%) 101 (63.9%) 67 (62.6%) 0.90
 STEMI 268 90 (33.6%) 61 (38.1%) 29 (26.8%) 0.07
Coronary angiography
 Inpatient coronary angiogram 268 197 (73.5%) 116 (72.5%) 81 (75.0%) 0.67
 Coronary angiogram first 24 h 268 149 (55.6%) 93 (58.1%) 56 (51.8%) 0.32
 CAD >=50% at cath 197 156 (79.2%) 96 (82.8%) 60 (74.1%) 0.16
 Acute coronary occlusion 197 66 (33.5%) 47 (40.5%) 19 (23.5%) 0.01
 Angiographic thrombus 197 61 (31.0%) 43 (37.1%) 18 (22.2%) 0.03
 Successful PCI 197 96 (48.7%) 62 (53.4%) 34 (42.0%) 0.15
Severity of illness
 APACHE-III score 265 100 (83,122.5) 107 (92,132) 89 (76,110) <0.001
 APACHE-IV predicted mortality (%) 265 60.7 (43.1,81.0) 65.1 (50.2,85.2) 49.6 (36.2,71.5) <0.001
 Day 1 SOFA 265 8 (6,10) 9 (8,11) 6 (5,7) <0.001
 Maximum week 1 SOFA 265 9 (7,11) 10 (8,12) 7.5 (6,9) <0.001
 Mean week 1 SOFA 265 5.8 (4.4,8.5) 7.3 (5.2,9.6) 4.6 (3.7,5.5) <0.001
Vital signs data
 Mean MAP first 6 h 262 86.9 (77.4,96.5) 82.5 (73.6,91.8) 94.5 (85.0,104.4) <0.001
 Mean MAP first 24 h 268 80.2 (74.3,85.6) 77.8 (72.2,82.7) 83.3 (77.8,89.0) <0.001
 24-h urine output (L) 266 1.6 (0.8,2.6) 1.4 (0.7,2.3) 1.9 (1.1,2.8) 0.001
 24-h net in/out balance (L) 266 1.9 (0.1,3.6) 2.7 (1.1,5.3) 0.6 (-0.3,2.1) <0.001
Laboratory data
 First lactate 234 4.2 (2.5,7.1) 5.1 (3.0,8.3) 3.7 (2.0,5) <0.001
 Peak lactate 234 4.6 (2.8,7.6) 5.9 (3.4,8.9) 3.7 (2.2,5.1) <0.001
 Rising lactate (peak > first) 233 40 (17.2%) 31 (21.4%) 9 (10.2%) 0.03
 First pH 267 7.25 (7.17,7.32) 7.23 (7.15,7.30) 7.27 (7.21,7.34) <0.001
 First bicarbonate 248 20 (18,23) 20 (17,23) 22 (19,23) 0.001
 First creatinine 267 1.1 (0.9,1.3) 1.1 (0.9,1.4) 1.0 (0.8,1.3) 0.05
 Peak NSE 268 27 (17,42.8) 29 (18.2,48) 24.4 (17,36.5) 0.04
Echocardiographic findings
 LVEF (%) 214 37.8 (25,52.5) 36 (22.5,48.2) 40 (28.8,56.5) 0.04
 Cardiac output 155 4.4 (3.6,5.7) 4.2 (3.6,5.4) 4.9 (3.5,6.2) 0.04
 Cardiac index 152 2.2 (1.7,2.8) 2.1 (1.7,2.7) 2.3 (1.7,3.0) 0.09
 Stroke volume 159 68 (52,85) 61 (48,81) 72 (62,90.8) 0.003
 Stroke volume index 156 33 (26,39.8) 32 (23,38.5) 34 (29,42) 0.004
Data represented as median (interquartile range) for continuous variables and number (percent) for categorical variables. P value represents Wilcoxon rank-sum test for continuous variables and Fisher exact test for categorical variables comparing patients who did and did not receive vasopressors during the first 24 h after admission.

The median initial lactate was 4.2 (2.5, 7.1) mmol/L and the median peak lactate was 4.6 (2.8, 7.6) mmol/L; both initial and peak lactate were higher among patients receiving early vasopressors (P < 0.001). A total of 40 (17.2%) patients had rising lactate (i.e., peak > initial lactate), and this was more prevalent in patients requiring early vasopressors (21.4% vs. 10.2%, P = 0.03). The median mMAP24 was 80.2 (74.3, 85.6) mm Hg, and it was lower in patients who received early vasopressors (P < 0.001). A total of 257 (97.6%) patients had at least one systolic BP measurement <90 mm Hg or one MAP measurement <60 mm Hg during the first 24 h, including 192 (72.2%) individuals who met both criteria. The median echocardiographic LVEF was 37.8 (25, 52.5) %, and was lower in early vasopressors group patients, as were echocardiographic stroke volume and cardiac output, compared to patients who did not require vasopressors (all P < 0.05; Table 1).

Association of MAP with other markers of shock

Both initial and peak lactate correlated with mMAP24 (P < 0.01); by contrast, only peak lactate (P = 0.02) correlated with mMAP24, while initial lactate did not (P = 0.18). Peak NSE did not correlate with mMAP24 (P > 0.1). VIS24 correlated with mMAP24 (P < 0.001). The use of vasopressors increased progressively with lower mMAP24 (Supplemental Figure 1A,; P < 0.001 for trend), as did the VIS24 (Supplemental Figure 1B, and both initial and peak lactate levels (Supplemental Figure 2,

Unadjusted hospital and 30-day mortality

A total of 104 (38.8%) patients died during hospitalization, including 78 (69%) patients who died after withdrawal of life-sustaining measures due to brain injury and 26 (23%) patients who died due to refractory shock. Overall 30-day mortality occurred in 103 (38.4%) patients, 102 of whom died in the hospital. Hospital mortality was higher among patients requiring early vasopressors (47.5% vs. 25.9%, unadjusted OR 2.58, 95% CI 1.52–4.39, P < 0.001). The VIS24 was directly associated with hospital mortality (unadjusted OR 1.03 per 1 unit, 95% CI 1.02–1.04, P < 0.001; the optimal cutoff for predicting hospital mortality was 11), and hospital mortality increased progressively as a function of increasing VIS24 (Fig. 1A); all 18 patients with VIS24 ≥80 died in the hospital. Similarly, the initial lactate was directly associated with hospital mortality (unadjusted OR 1.22 per 1 mmol/L, 95% CI 1.11–1.33, P < 0.001; the optimal cut-off for predicting hospital mortality was 3.8 mmol/L), and hospital mortality increased progressively with higher initial lactate (Fig. 1B). Similar results were observed for 30-day mortality using Cox proportional-hazards analysis. Hospital mortality varied as a function of VIS24 and initial lactate (Supplemental Figure 3, Crude hospital mortality was not higher in patients with rising lactate (45.0% vs. 38.9%, P = 0.48).

Fig. 1
Fig. 1:
Crude hospital mortality as a function of maximum 24-h VIS (A) and initial lactate level (B), with 95% confidence intervals. Patients with a VIS of 0 did not receive any vasopressors or inotropes.

The mMAP24 (unadjusted OR 0.95 per 1 mm Hg, 95% CI 0.92–0.98, P < 0.001; optimal cutoff 77 mm Hg) was inversely associated with hospital mortality. Hospital mortality increased as a function of decreasing mMAP24 (Fig. 2). A notable step-up in mortality was observed in patients with mMAP24 < 75 mm Hg, especially <70 mm Hg. Patients with mMAP24 <70 mm Hg (n = 28, 10.4%) had higher hospital mortality (82.1% vs. 33.8%, unadjusted OR 9.03, 95% CI 3.31–24.63, P < 0.001). Patients requiring early vasopressors generally had higher hospital mortality at each level of mMAP24 (Fig. 2), especially at lower mMAP24 levels. When patients who did and did not receive early vasopressors were analyzed separately, mMAP24 was only associated with hospital mortality in patients who received early vasopressors (unadjusted OR 0.94 per 1 mm Hg, 95% CI 0.90–0.97, P < 0.001); P = 0.78 for patients who did not receive early vasopressors. Hospital mortality varied as a function of mMAP24 and either VIS24 (Supplemental Figure 4A, or initial lactate (Supplemental Figure 4B,

Fig. 2
Fig. 2:
Crude hospital mortality as a function of the mean 24-h MAP, in the overall population and among patients who did and did not receive vasopressors during the first 24 h.

Adjusted hospital and 30-day mortality

Significant predictors of hospital mortality identified using multivariable logistic regression are shown in Table 2. Both initial lactate (adjusted OR per 1 mmol/L 1.15, 95% CI 1.02–1.30, P = 0.03) and VIS24 (adjusted OR per 10 units 1.29, 95% CI 1.10–1.54, P = 0.003) were significant predictors of hospital mortality. The mMAP24 was not a significant predictor of mortality (P > 0.1). A significant inverse statistical interaction was observed between maximum VIS24 and mMAP24 (P = 0.02), implying a greater effect of VIS24 at lower values of mMAP24. When patients who did and did not receive early vasopressors were considered separately, mMAP24 was not associated with adjusted hospital mortality in either group (P > 0.05). After multivariable adjustment, patients with mMAP24 < 70 mm Hg had higher hospital mortality (adjusted OR 9.30, 95% CI 1.39–62.02, P = 0.02). Using Cox proportional-hazards analysis, both VIS24 (adjusted HR per 10 units 1.11, 95% CI 1.05–1.16, P < 0.001) and initial lactate level (adjusted HR per 1 mmol/L 1.11, 95% CI 1.04–1.19, P = 0.003) were associated with higher 30-day mortality, while mMAP24 was not significantly associated (adjusted HR per 10 mm Hg 0.789, 95% CI 0.582–1.060, P = 0.12).

Table 2 - Predictors of hospital mortality using multivariable logistic regression, prior to inclusion of the interaction term; final model AUROC 0.91
Variable Adjusted OR 95% CI P value
Age (per 10 years) 1.229 0.869–1.769 0.25
APACHE-III score (per 10 points) 1.072 0.923–1.248 0.36
CCI (per 1 point) 1.045 0.888–1.227 0.59
Shockable rhythm 0.623 0.196–1.982 0.42
Witnessed arrest 0.143 0.040–0.512 0.003
Bystander CPR 1.162 0.489–2.764 0.73
Initial lactate (per 1 mmol/L) 1.150 1.018–1.304 0.03
Max 24 h VIS (per 10 units) 1.287 1.101–1.542 0.003
Mean 24-h MAP (per 10 mm Hg) 0.857 0.494–1.451 0.58
Inpatient coronary angiogram 0.078 0.029–0.207 <0.001
Peak NSE (per 10 ng/L) 1.642 1.370–2.029 <0.001
VIS indicates Vasoactive-Inotropic Score.

Neurologic outcomes and long-term mortality

Among the 164 hospital survivors, the median mRS was 2, and 105 (64.0%) patients were considered to have a good neurological outcome. The mRS and good neurological outcomes varied as a function of the mMAP24 (Fig. 3), with higher mRS scores and lower rates of good neurologic outcomes at lower mMAP24 values. Higher initial lactate (AUC 0.59, P = 0.01) and VIS24 (AUC = 0.65, P < 0.001) were associated with lower risk of a good neurologic outcome on univariate analysis, but mMAP24 was not (P = 0.07). After multivariable adjustment, VIS24 remained inversely associated with good neurological outcome (adjusted OR 0.80, 95% CI 0.66–0.96, P = 0.01), but initial lactate (P = 0.82) and mMAP24 (P = 0.14) were not.

Fig. 3
Fig. 3:
Discharge Modified Rankin Scale (mRS) as a function of the mean 24-h MAP.

A total of 25 (15.2%) hospital survivors died during a median follow-up of 3.1 (1.5, 6.1) years. Patients who received early vasopressors were at higher risk of long-term mortality by Kaplan–Meier analysis (P = 0.007 by log-rank). However, when the analysis was limited only to hospital survivors, there was no longer a significant difference in post-discharge mortality (P = 0.13 by log-rank). Similarly, patients with higher initial lactate were at higher risk of long-term mortality by Kaplan–Meier analysis compared to patients with lower initial lactate (P < 0.001 by log-rank). However, when the analysis was limited only to hospital survivors, there was no longer a significant difference in post-discharge mortality (P = 0.44 by log-rank). Hospital survivors with a good neurological outcome had lower post-discharge mortality than those with a poor neurological outcome (P < 0.001 by log-rank).


In this retrospective study, we evaluated the association between shock severity, as measured by MAP, vasopressor requirements, and blood lactate levels, with mortality in OHCA patients treated with TTM. Lower mMAP24 and higher initial lactate and VIS24 were all associated with increased hospital mortality. The association between mMAP24 and hospital mortality was present only in the 60% of patients who required early vasopressors. Patients who could not maintain an mMAP24 >70 mm Hg had very high hospital mortality. After multivariable adjustment, only initial lactate and VIS24 were significant predictors of hospital mortality, while mMAP24 was not associated with mortality after adjustment. This implies that organ perfusion and the degree of pharmacologic support required to maintain MAP are more important prognostic factors after OHCA in patients receiving TTM than the MAP itself. Our study therefore builds incrementally on prior studies that have evaluated each of these variables individually, demonstrating their inter-relatedness when assessing shock severity and predicting outcomes in OHCA patients.

Shock is defined by state of cellular and tissue hypoxia, most commonly occurs when there is circulatory failure manifested as hypotension, with lactic acidosis and the need for vasopressor support (21). During the first 24 h after resuscitation, nearly all of our OHCA patients had hypotension, and vasopressor support was required in approximately 60% (plus an additional 15% beyond 24 h), similar to previous observational studies of OHCA patients treated with TTM (4, 8, 12).

OHCA patients with hypotension had higher mortality and worse neurologic outcomes in most prior observational studies (4, 8, 22, 23), although not all studies observed this association (24, 25). The presence of hypotension presumably identifies patients with higher shock severity, presumably from a more severe initial ischemic injury. A recent prospective study did not show an improvement in neurological outcome using higher MAP goals after ROSC, raising doubts about whether raising the MAP pharmacologically is beneficial and leaving uncertainty regarding optimal MAP goal to target with vasopressor therapy in patients with hypotension and shock after OHCA (11). Hypotension after OHCA could potentially induce brain ischemia in the setting of impaired cerebral blood flow autoregulation (26). Our data confirm prior observations that hypotension after OHCA is associated with worse outcomes, primarily among patients requiring vasopressors in whom MAP may be a more relevant consideration. However, the lack of an association between MAP and outcomes after adjustment for illness severity and vasopressor requirements suggests that hypotension serves primarily as a marker of greater severity of illness. We hypothesize that prior observational studies demonstrating a strong association between low MAP with mortality might not have adequately adjusted for relevant markers of shock severity (such as vasopressor load and lactic acidosis) which reflect the severity of the initial ischemic injury during OHCA.

Vasopressor and inotrope load, as quantified by the VIS, was independently predictive of mortality in our cohort, consistent with similar studies in the ICU setting (17, 27, 28). Unadjusted hospital mortality was nearly twofold higher among patients requiring early vasopressors. The majority of OHCA patients required vasoactive drug support, making quantification of vasopressor requirements by the VIS a more relevant variable than the presence or absence of vasopressor support. Since clinicians typically titrate vasoactive drugs in order to normalize systemic hemodynamics and tissue perfusion markers, the dose of these medications required by a patient to achieve the hemodynamic goals is a straightforward way to define shock severity in OHCA patients. The use and dosage of vasopressor and inotrope therapy cannot be disentangled from MAP in observational studies of patients with shock, and our data support the notion that vasoactive drug requirements may be more closely linked to outcomes than MAP itself in OHCA patients with shock. Notably, maintaining a higher MAP will require greater use of vasopressors which can lead to excess cardiovascular toxicity, to which OHCA patients are likely more susceptible based on their unstable myocardial substrate (29).

Lactate levels after OHCA integrate the duration and magnitude of the initial ischemic insult during the arrest with the severity of shock and organ hypoperfusion, explaining their prognostic value. The initial lactate level remained a significant predictor of adjusted hospital mortality in our study, consistent with previous studies in ICU patients with shock (30, 31), and in contrary to the findings of Laurikkala et al. which found that lactate level at admission was not an independent predictor of poor long-term outcome (32). As in our study, prior studies have demonstrated that an elevated lactate level is associated with an increased risk of death independent of the level of vasopressor support (33, 34). Interestingly, although lactate clearance is considered an important prognostic marker in patient critically-ill populations (35), we did not observe excess mortality in the subset of patients with a rising lactate level.

We adjusted our multivariable model for the peak NSE level during the first 72 h after ROSC, as a marker for anoxic brain injury (36). NSE has been tested for the prediction of neurological outcomes in OHCA patients with a cutoff of 33 ng/L proposed by some authors to identify patients with low likelihood of survival (37, 38). In a recent, large, observational study of OHCA patients who were treated with TTM a higher NSE cutoff (>90 ng/L) was found to be associated with almost no false positives at acceptable sensitivity and better predictive accuracy in OHCA (39). The peak NSE level in our study was a robust and independent predictor of adjusted hospital mortality.


As with all retrospective observational studies, this study has a number of limitations precluding causal inference. This cohort represents a highly selected population enriched in patients with shockable rhythms of presumed cardiac etiology, treated with TTM, so our conclusions may not apply to unselected all-comers OHCA populations. The TTM protocol used in our study included cooling to 33°C, consistent with the practice used at the period of our study, which may be associated with higher hemodynamic instability without any improvement in outcomes. We could not determine what hemodynamic goals were being targeted in our patients and a specific hemodynamic optimization protocol did not exist during the study period. Therefore, we could not distinguish patients who achieved a lower target MAP goal from patients who failed to reach a higher target MAP goal. We did not have doses of sedatives and other medications that could have influenced hemodynamics. Additionally, we did not have granular data regarding the neurological examination and other relevant prognostic data for use in our multivariable models, particularly details regarding the initial OHCA resuscitation, including arrest duration, length of resuscitation, number of defibrillation attempts, drugs and doses administered, time to ROSC, and other prognostically relevant variables. We focused on initial lactate levels, recognizing that one in six patients had a rising lactate level after admission; we did not have consistently timed serial lactate measurements to determine lactate clearance.


In OHCA patients receiving TTM, it appears that low MAP is a marker of greater overall illness severity, rather than an independent predictor of outcomes. The dosage of vasopressor support and initial blood lactate level are better predictors of mortality after OHCA than mMAP24. That being said, patients with mean mMAP24 below 70 mm Hg had very poor outcomes, suggesting that this could be a rational MAP “floor” to target during vasopressor support after OHCA. Understanding the role of shock severity markers in predicting outcomes after OHCA can assist the treating physician in prognostication. This study emphasizes the importance of randomized clinical trials for defining treatment goals in critically ill patients. Further adequately powered randomized studies are needed to determine if using higher vasopressor doses to achieve a higher MAP goal will improve neurological outcomes in OHCA patients receiving TTM without causing cardiovascular toxicity.


1. McCarthy JJ, Carr B, Sasson C, Bobrow BJ, Callaway CW, Neumar RW, Ferrer JME, Garvey JL, Ornato JP, Gonzales L, et al. Out-of-hospital cardiac arrest resuscitation systems of care: a scientific statement from the American Heart Association. Circulation 137 (21):e645–e660, 2018.
2. Kong MH, Fonarow GC, Peterson ED, Curtis AB, Hernandez AF, Sanders GD, Thomas KL, Hayes DL, Al-Khatib SM. Systematic review of the incidence of sudden cardiac death in the United States. J Am Coll Cardiol 57:794–801, 2011.
3. Laurikkala J, Wilkman E, Pettilä V, Kurola J, Reinikainen M, Hoppu S, Ala-Kokko T, Tallgren M, Tiainen M, Vaahersalo J, et al. Mean arterial pressure and vasopressor load after out-of-hospital cardiac arrest: Associations with one-year neurologic outcome. Resuscitation 105:116–122, 2016.
4. Trzeciak S, Jones AE, Kilgannon JH, Milcarek B, Hunter K, Shapiro NI, Hollenberg SM, Dellinger P, Parrillo JE. Significance of arterial hypotension after resuscitation from cardiac arrest. Crit Care Med 37 (11):2895–2903, 2009.
5. Callaway CW, Donnino MW, Fink EL, Geocadin RG, Golan E, Kern KB, Leary M, Meurer WJ, Peberdy MA, Thompson TM, et al. Post-Cardiac Arrest Care: 2015 American Heart Association Guidelines Update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 132: (18 suppl 2): S465–S482, 2015.
6. Jones AE, Shapiro NI, Kilgannon JH, Trzeciak S. Goal-directed hemodynamic optimization in the post-cardiac arrest syndrome: a systematic review. Resuscitation 77 (1):26–29, 2008.
7. Gaieski DF, Band RA, Abella BS, Neumar RW, Fuchs BD, Kolansky DM, Merchant RM, Carr BG, Becker LB, Maguire C, et al. Early goal-directed hemodynamic optimization combined with therapeutic hypothermia in comatose survivors of out-of-hospital cardiac arrest. Resuscitation 80 (4):418–424, 2009.
8. J. Hope K, Brian WR, Lisa RR, Michael EC, Alan EJ, R. Phillip D, Joseph EP, Stephen T. Early arterial hypotension is common in the post-cardiac arrest syndrome and associated with increased in-hospital mortality. Resuscitation 79:410–416, 2009.
9. Beylin ME, Perman SM, Abella BS, Leary M, Shofer FS, Grossestreuer AV, Gaieski DF. Higher mean arterial pressure with or without vasoactive agents is associated with increased survival and better neurological outcomes in comatose survivors of cardiac arrest. Intensive Care Med 39:1981–1988, 2013.
10. Sunde K, Pytte M, Jacobsen D, Mangschau A, Jensen LP, Smedsrud C, Draegni T, Steen PA. Implementation of a standardised treatment protocol for post resuscitation care after out-of-hospital cardiac arrest. Resuscitation 73:29–39, 2007.
11. Ameloot K, De Deyne C, Eertmans W, Ferdinande B, Dupont M, Palmers PJ, Petit T, Nuyens P, Maeremans J, Vundelinckx J, et al. Early goal-directed haemodynamic optimization of cerebral oxygenation in comatose survivors after cardiac arrest: the Neuroprotect post-cardiac arrest trial. Eur Heart J 40 (22):1804–1814, 2019.
12. Jentzer JC, Chonde MD, Dezfulian C. Myocardial dysfunction and shock after cardiac arrest. Biomed Res Int 2015:314796, 2015.
13. Donnino MW, Andersen LW, Giberson T, Gaieski DF, Abella BS, Peberdy MA, Rittenberger JC, Callaway CW, Ornato J, Clore J, et al. Initial lactate and lactate change in post-cardiac arrest: a multicenter validation study. Crit Care Med 42:1804–1811, 2014.
14. Annborn M, Bro-Jeppesen J, Nielsen N, Ullén S, Kjaergaard J, Hassager C, Wanscher M, Hovdenes J, Pellis T, Pelosi P, et al. TTM-trial investigators. The association of targeted temperature management at 33 and 36 (C with outcome in patients with moderate shock on admission after out-of-hospital cardiac arrest: a post hoc analysis of the Target Temperature Management trial. Intensive Care Med 40 (9):1210–1219, 2014.
15. Bro-Jeppesen J, Hassager C, Wanscher M, Østergaard M, Nielsen N, Erlinge D, Friberg H, Køber L, Kjaergaard J. Targeted temperature management at 33°C versus 36°C and impact on systemic vascular resistance and myocardial function after out-of-hospital cardiac arrest: a sub-study of the Target Temperature Management Trial. Circ Cardiovasc Interv 7 (5):663–672, 2014.
16. Herasevich V, Pickering BW, Dong Y, Peters SG, Gajic O. Informatics infrastructure for syndrome surveillance, decision support, reporting, and modeling of critical illness. Mayo Clin Proc 85:247–254, 2010.
17. Bennett CE, Wright RS, Jentzer J, Gajic O, Murphree DH, Murphy JG, Mankad SV, Wiley BM, Bell MR, Barsness GW. Severity of illness assessment with application of the APACHE IV predicted mortality and outcome trends analysis in an academic cardiac intensive care unit. J Crit Care 50:242–246, 2019.
18. Jentzer JC, Bennett C, Wiley BM, Murphree DH, Keegan MT, Gajic O, Wright RS, Barsness GW. Predictive value of the sequential organ failure assessment score for mortality in a contemporary cardiac intensive care unit population. J Am Heart Assoc 2018; 7: e008169.
19. Jentzer JC, Wiley B, Bennett C, Murphree DH, Keegan MT, Kashani KB, Bell MR, Barsness GW. Temporal trends and clinical outcomes associated with vasopressor and inotrope use in the cardiac intensive care unit. Shock 53:452–459, 2020.
20. Jentzer JC, Anavekar NS, Mankad SV, Khasawneh M, White RD, Barsness GW, Rabinstein AA, Kashani KB, Pislaru SV. Echocardiographic left ventricular diastolic dysfunction predicts hospital mortality after out-of-hospital cardiac arrest. J Crit Care 47:114–120, 2018.
21. Standl T, Annecke T, Cascorbi I, Heller AR, Sabashnikov A, Teske W. The nomenclature, definition and distinction of types of shock. Dtsch Arztebl Int 115 (45):757–768, 2018.
22. Ameloot K, Meex I, Genbrugge C, Jans F, Boer W, Verhaert D, Mullens W, Ferdinande B, Dupont M, De Deyne C, et al. Hemodynamic targets during therapeutic hypothermia after cardiac arrest: a prospective observational study. Resuscitation 91:56–62, 2015.
23. Kilgannon JH, Roberts BW, Reihl LR, Chansky ME, Jones AE, Dellinger RP, Parrillo JE, Trzeciak S. Early arterial hypotension is common in the post-cardiac arrest syndrome and associated with increased in-hospital mortality. Resuscitation 79:410–416, 2008.
24. Kim HC, Yoo J-W, Lim SY, Suh GY, Koh SO, Na S, Lim C-M, Koh Y, Choi W-I, Lee Y-J, et al. Mortality after in-hospital cardiopulmonary resuscitation: multicenter analysis in Korea. J Crit Care 28:942–946, 2013.
25. Young MN, Hollenbeck RD, Pollock JS, Giuseffi JL, Wang L, Harrell FE, McPherson JA. Higher achieved mean arterial pressure during therapeutic hypothermia is not associated with neurologically intact survival following cardiac arrest. Resuscitation 88:158–164, 2015.
26. Drummond JC. Blood pressure and the brain: how low can you go? Anesth Analg 128 (4):759–777, 2019.
27. Na SJ, Chung CR, Cho YH, Jeon K, Suh GY, Ahn JH, Carriere KC, Park TK, Lee GY, Lee JM, et al. Vasoactive inotropic score as a predictor of mortality in adult patients with cardiogenic shock: medical therapy versus ECMO. Rev Esp Cardiol (Engl Ed) 72 (1):40–47, 2019.
28. McIntosh AM, Tong S, Deakyne SJ, Davidson JA, Scott HF. Validation of the vasoactive-inotropic score in pediatric sepsis. Pediatr Crit Care Med 18 (8):750–757, 2017.
29. Asfar P, Meziani F, Hamel JF, Grelon F, Megarbane B, Anguel N, Mira JP, Dequin PF, Gergaud S, Weiss N, et al. SEPSISPAM Investigators. High versus low blood-pressure target in patients with septic shock. N Engl J Med 370 (17):1583–1593, 2014.
30. Innocenti F, Meo F, Giacomelli I, Tozzi C, Ralli ML, Donnini C, Tassinari I, Caldi F, Zanobetti M, Pini R. Prognostic value of serial lactate levels in septic patients with and without shock. Intern Emerg Med 14 (8):1321–1330, 2019.
31. Villar J, Short JH, Lighthall G. Lactate predicts both short- and long-term mortality in patients with and without sepsis. Infect Dis 2019; 12: 1178633719862776.
32. Laurikkala J, Skrifvars MB, Bäcklund M, Tiainen M, Bendel S, Karhu J, Varpula T, Vaahersalo J, Pettilä V, Wilkman E. FINNRESUSCI study group. Early lactate values after out-of-hospital cardiac arrest: associations with one-year outcome. Shock 51 (2):168–173, 2019.
33. Cocchi MN, Miller J, Hunziker S, Carney E, Salciccioli J, Farris S, Joyce N, Zimetbaum P, Howell MD, Donnino MW. The association of lactate and vasopressor need for mortality prediction in survivors of cardiac arrest. Minerva Anestesiol 77 (11):1063–1071, 2011.
34. Cocchi MN, Salciccioli J, Yankama T, Chase M, Patel PV, Liu X, Mader TJ, Donnino MW. Predicting outcome after out-of-hospital cardiac arrest: lactate, need for vasopressors, and cytochrome c. J Intensive Care Med 2019; 29: 885066619873315.
35. Düring J, Dankiewicz J, Cronberg T, Hassager C, Hovdenes J, Kjaergaard J, Kuiper M, Nielsen N, Pellis T, Stammet P, et al. Lactate, lactate clearance and outcome after cardiac arrest: a post-hoc analysis of the TTM-Trial. Acta Anaesthesiol Scand 62 (10):1436–1442, 2018.
36. Isgro MA, Bottoni P, Scatena R. Neuron-specific enolase as a biomarker: biochemical and clinical aspects. Adv Exp Med Biol 867:125–143, 2015.
37. Oksanen T, Tiainen M, Skrifvars MB, Varpula T, Kuitunen A, Castrén M, Pettilä V. Predictive power of serum NSE and OHCA score regarding 6-month neurologic outcome after out-of-hospital ventricular fibrillation and therapeutic hypothermia. Resuscitation 80:165–170, 2009.
38. Cronberg T, Rundgren M, Westhall E, Englund E, Siemund R, Rosén I, Widner H, Friberg H. Neuron-specific enolase correlates with other prognostic markers after cardiac arrest. Neurology 77:623–630, 2011.
39. Streitberger KJ, Leithner C, Wattenberg M, Tonner PH, Hasslacher J, Joannidis M, Pellis T, Di Luca E, Födisch M, Krannich A, et al. Neuron-specific enolase predicts poor outcome after cardiac arrest and targeted temperature management: a multicenter study on 1,053 patients. Crit Care Med 45 (7):1145–1151, 2017.

Blood pressure; cardiac arrest; lactate; shock; vasopressors

Supplemental Digital Content

Copyright © 2020 by the Shock Society