Shock subtypes by left ventricular ejection fraction following out-of-hospital cardiac arrest

Background Post-resuscitation hemodynamic instability following out-of-hospital cardiac arrest (OHCA) may occur from myocardial dysfunction underlying cardiogenic shock and/or inflammation-mediated distributive shock. Distinguishing the predominant shock subtype with widely available clinical metrics may have prognostic and therapeutic value. Methods A two-hospital cohort was assembled of patients in shock following OHCA. Left ventricular ejection fraction (LVEF) was assessed via echocardiography or cardiac ventriculography within 1 day post arrest and used to delineate shock physiology. The study evaluated whether higher LVEF, indicating distributive-predominant shock physiology, was associated with neurocognitive outcome (primary endpoint), survival, and duration of multiple organ failures. The study also investigated whether volume resuscitation exhibited a subtype-specific association with outcome. Results Of 162 patients with post-resuscitation shock, 48% had normal LVEF (> 40%), consistent with distributive shock physiology. Higher LVEF was associated with less favorable neurocognitive outcome (OR 0.74, 95% CI 0.58–0.94 per 10% increase in LVEF; p = 0.01). Higher LVEF also was associated with worse survival (OR 0.81, 95% CI 0.67–0.97; p = 0.02) and fewer organ failure-free days (β = – 0.67, 95% CI – 1.28 to − 0.06; p = 0.03). Only 51% of patients received a volume challenge of at least 30 ml/kg body weight in the first 6 h post arrest, and the volume received did not differ by LVEF. Greater volume resuscitation in the first 6 h post arrest was associated with favorable neurocognitive outcome (OR 1.59, 95% CI 0.99–2.55 per liter; p = 0.03) and survival (OR 1.44, 95% CI 1.02–2.04; p = 0.02) among patients with normal LVEF but not low LVEF. Conclusions In post-resuscitation shock, higher LVEF—indicating distributive shock physiology—was associated with less favorable neurocognitive outcome, fewer days without organ failure, and higher mortality. Greater early volume resuscitation was associated with more favorable neurocognitive outcome and survival in patients with this shock subtype. Additional studies with repeated measures of complementary hemodynamic parameters are warranted to validate the clinical utility for subtyping post-resuscitation shock. Electronic supplementary material The online version of this article (10.1186/s13054-018-2078-x) contains supplementary material, which is available to authorized users.


CPC 2 -Moderate Cerebral Disability (Disabled but independent): Conscious with
sufficient cerebral function for part-time work in sheltered environment or independent activities of daily life (dress, travel by public transportation, food preparation). May have hemiplegia, seizures, ataxia, dysarthria, dysphasia, or permanent memory or mental changes.

CPC 3 -Severe Cerebral Disability (Conscious but disabled and dependent):
Conscious but dependent on others for daily support (in an institution or at home with exceptional family effort). Has at least limited cognition. This category includes a wide range of cerebral abnormalities, from patients who are ambulatory but have severe memory disturbances or dementia precluding independent existence to those who are paralyzed and can communicate only with their eyes, as in the locked-in syndrome. 4. CPC 4 -Coma / Vegetative State (Unconscious): Unconscious, unaware of surroundings, no cognition. No verbal or psychological interaction with environment. 5. CPC 5 -Brain Death (Certified brain dead or dead by traditional criteria): Certified brain death or dead by traditional criteria.
CPC on hospital discharge was determined retrospectively by reviewing each patient's medical record as previously described [11]. As available, evaluators reviewed the discharge summary, discharge referral form, and the last documented notes from the primary treatment team, neurology service, nursing, physical and occupational therapy, social worker, and case manager. Evaluations were performed blinded to left ventricular ejection fraction, hemodynamic and resuscitation data, and other illness severity measures. Two investigators, (1) a dedicated site-specific investigator for each site and (2) a study-wide investigator with access to all records at both sites, independently reviewed each chart, with discordant ratings resolved by consensus for the final dataset.
The primary outcome, specified a priori, was favorable neurocognitive outcome at hospital discharge, defined as CPC of 1 or 2. We chose to dichotomize CPC for our primary analysis to facilitate ease of understanding results, as has been done in recent high-profile cardiac arrest clinical trials [12][13][14]. To ensure results were not dependent on dichotomizing CPC, data were re-analyzed using ordinal logistic regression with CPC entered as an ordinal dependent variable.

Organ Failure-Free Days
Secondary outcomes included shock-free days, ventilator-free days, and renal, hepatic, and coagulation failure-free days. Organ failure-free days were calculated as the number of days between sustained resolution of organ failure and day 28 [8,11,15]. If a patient temporarily recovered but developed organ failure again later in the hospital course, the period of transient recovery did not count toward failure-free days. When data were missing for a given day, the last observed value was carried forward until day of live hospital discharge. Organ failures were considered resolved following live hospital discharge, except for renal failure in patients scheduled to undergo post-discharge dialysis. A value of zero failure-free days was assigned if the patient died before hospital discharge. ICU-free days and hospital-free days were also evaluated and calculated similarly.
Except as noted, organ failures were determined by applying Brussels Multiple Organ Dysfunction consensus conference definitions for clinically significant organ dysfunction [6].
Shock was defined as systolic blood pressure ≤ 90 mmHg or any vasopressor use. Pulmonary failure was defined as receipt of invasive mechanical ventilation. Renal failure was defined as creatinine ≥ 2.0 mg/dL or receipt of renal replacement therapy. Hepatic failure was defined as total bilirubin ≥ 2 mg/dL. Coagulation failure was defined as platelet count ≤ 80 x 10 3 /mm 3 . The worst values for each calendar day were used in determining if organ failure was present on a given day. Patients receiving chronic outpatient dialysis prior to admission were excluded from analysis of renal failure-free days.

Sensitivity Analyses for Primary Outcome
Multiple sensitivity analyses were performed for the primary outcome, favorable neurocognitive outcome, to ensure findings were not dependent on method of covariate adjustment or handling of the main predictor or outcome variables. Alternative covariate adjustments included replacing APACHE-II with either SOFA or number of organ failures at baseline, and adding therapeutic hypothermia as a covariate. To better reflect clinical decisionmaking and determine if results were dependent on handling LVEF as a continuous variable, LVEF was re-entered as a dichotomized variable (normal LVEF > 40% vs. low LVEF ≤ 40%) in an APACHE-II-adjusted model. In another sensitivity analysis, CPC was entered as an ordinal outcome analyzed via ordinal logistic regression. Finally, the main analysis was repeated using the expanded sensitivity cohort described above.

Baseline Characteristics
Baseline characteristics for patients not in shock are presented in Table S1, accompanied by included patient characteristics for comparison.

Sensitivity Analyses for Primary Endpoint
Multiple sensitivity analyses confirmed that the association between higher LVEF and less favorable neurocognitive outcome did not depend on method of quantifying illness severity, included covariates, or handling of the dependent and independent variables (Figure 2; Table  S2). Adjusting for alternative measures of illness severity (Day 1 SOFA, number of organ failures at baseline) again demonstrated the association between higher LVEF and less favorable neurocognitive outcome. Adding therapeutic hypothermia as a covariate in these models also did not change this association, and in no model was therapeutic hypothermia associated with neurocognitive outcome.
Reanalysis entering LVEF as a dichotomized variable and adjusting for APACHE-II similarly found that normal LVEF, compared to low LVEF, was associated with less favorable neurocognitive outcome (OR 0.36, 95% CI 0.15-0.85; p = .02). Reanalysis using ordinal logistic regression, with CPC entered as the outcome variable and adjusting for APACHE-II, confirmed that the association between higher LVEF and less favorable neurocognitive outcome was not dependent on dichotomizing CPC (OR 0.80, 95% CI 0.67-0.95 for 1-unit change in CPC per 10% increase in LVEF; p = .01; score test for proportional odds assumption p = .31).
Sensitivity analyses also were performed to address the 73 patients in shock excluded from the main analyses for lack of an LVEF measurement within one calendar day of arrest (Figure 1; Table 1). The expanded sensitivity cohort (n=235) considered LVEF assessment any time during admission and assumed that patients without an LVEF assessment during admission had normal LVEF. In this expanded cohort, normal LVEF was associated with less favorable neurocognitive outcome compared to low LVEF in unadjusted analysis (OR 0.48, 95% CI 0.25-0.90 for normal vs. low LVEF; p = .02) and APACHE-II-adjusted analysis (OR 0.33, 95% CI 0.16-0.67; p < .01).