Quantitative characterization of left ventricular function during pulseless electrical activity using echocardiography during out-of-hospital cardiac arrest

Abstract

Background:

Several prospective studies have demonstrated that the echocardiographic detection of any myocardial activity during PEA is strongly associated with higher rates of return of spontaneous circulation (ROSC). We hypothesized that PEA represents a spectrum of disease in which not only the presence of myocardial activity, but more specifically that the degree of left ventricular (LV) function would be a predictor of outcomes. The purpose of this study was to retrospectively assess the association between LV function and outcomes in patients with OHCA.

Materials and methods:

Using prospectively obtained data from an observational cohort of patients receiving focused echocardiography during cardiopulmonary resuscitation (CPR) in the Emergency Department (ED) setting, we analyzed 312 consecutive subjects with available echocardiography images with initial rhythm of PEA. We used left ventricular systolic fractional shortening (LVFS), a unidimensional echocardiographic parameter to perform the quantification of LV function during PEA. Regression analyses were performed independently to evaluate for relationships between LVFS and a primary outcome of ROSC and secondary outcome of survival to hospital admission. We analyzed LVFS both as a continuous variable and as a categorial variable using the quartiles and the median to perform multiple different comparisons and to illustrate the relationship of LVFS and outcomes of interest. We performed survival analysis using Cox proportional hazards model to evaluate the hazard corresponding to length of resuscitation.

Results:

We found a positive association between LVFS and the primary outcome of ROSC (OR 1.04, 95%CI 1.01–1.08), but not with the secondary outcome of survival to hospital admission (OR 1.02, 95%CI 0.96–1.08). Given that the relationship was not linear and that we observed a threshold effect in the relationship between LVFS and outcomes, we performed an analysis using quartiles of LVFS. The predicted probability of ROSC was 75% for LVFS between 23.4–96% (fourth quartile) compared to 47% for LVFS between 0–4.7% (first quartile). The hazard of not achieving ROSC was significantly greater for subjects with LVFS below the median (13.1%) compared to the subgroup with LVFS greater than 13.1% (p < 0.05), with the separation of the survival curves occurring at approximately 40 min of resuscitation duration.

Conclusions:

Left ventricular function measured by LVFS is positively correlated with higher probability of ROSC and may be associated with higher chances of survival in patients with PEA arrest.

Introduction

More than 350,000 out-of-hospital cardiac arrests (OHCA) occur every year in the United States.¹ Cardiac arrests are classified according to their initial electrocardiographic rhythms; which has important prognostic implications. While the survival rate to hospital discharge for OHCA victims with initial rhythms of ventricular fibrillation (VF) or ventricular tachycardia (VT) is approximately 37%, survival in patients with initial rhythm of asystole or pulseless electrical activity (PEA) is substantially lower with 11% and 12% respectively.² PEA is defined as any non-VF/VT organized rhythm without a palpable pulse. While cardiac arrest due to VF/VT has been well-studied, PEA is poorly understood and effective treatments are limited.³

In the 1990’s, it was found that patients in cardiac arrest with PEA have highly heterogeneous hemodynamics, with a substantial fraction of PEA patients retaining detectible blood pressures.⁴

Several prospective studies have demonstrated that the echocardiographic detection of any myocardial activity during PEA is strongly associated with higher rates of return of spontaneous circulation (ROSC) and survival to hospital admission. In a recent systematic review and metanalysis including 1486 patients with non-shockable rhythms, myocardial activity detected on echocardiography had a pooled sensitivity of 60.3% and specificity of 91.5% for the detection of ROSC.⁵ In patients with PEA, the sensitivity of myocardial activity for predicting ROSC was 26.1%. In the largest observational study of echocardiographic evaluation during OHCA to date, the presence of cardiac activity was the variable most strongly associated with survival at all time-points including hospital discharge.⁶ Furthermore, survival in patients with PEA may be further enhanced in patients who receive infusion of vasopressors when they have organized myocardial activity but not in those with disorganized activity.⁷ We hypothesized that PEA represents a spectrum of disease in which not only the presence of myocardial activity, but more specifically that the degree of left ventricular (LV) function would be a predictor of outcomes. The purpose of this study is to retrospectively assess the association between LV function and outcomes in patients with OHCA.

Methods

Study setting, design and patient population

This work represents a secondary analysis of prospectively obtained data from an observational cohort of patients receiving focused transthoracic echocardiography during cardiopulmonary resuscitation (CPR) in the Emergency Department (ED) setting for either OHCA or ED-based PEA or asystole cardiac arrest. The original study involved 20 sites across the United States and Canada, and was conducted between 2011 and 2014 by the Real-Time Evaluation and Assessment Sonography Outcomes Network (REASON) investigators.⁶ The source study included 796 OHCA subjects, of whom 312 subjects (39.2%) had PEA as the initial rhythm during the initial echocardiography evaluation and were included in this secondary analysis. The study protocol for this study was approved by the Institutional Review Board at University of Pennsylvania.

Data collection

Data available for this study represented a subset from the primary source investigation and included echocardiography video images, as well as standard clinical data points following Utstein guidelines of reporting cardiac arrest variables.⁸ Additional data relevant to the echocardiographic analysis included the level of experience of the clinician sonographer who obtained the images, and estimated times from onset of cardiac arrest and from initiation of resuscitation care to echocardiography (Supplementary Fig. 1).

Evaluation of quality of echocardiography images

In order to provide as much objectivity as possible to the selection of echocardiography images to be analyzed, we used a previously validated qualitative rating scale to assess the quality of echocardiography images (Supplementary Fig. 2).⁹ This resuscitation-specific rating scale classifies images from level 1 (Poor image quality ‒ Uninterpretable level) to level 5 (Excellent image quality ‒ Quantitative function level). Only images rated as good or excellent (level 4 or 5) were used to attempt measurement of left ventricular function.

Quantification of left ventricular function

Because there are no established conventions for evaluating the degree of ventricular function during cardiac arrest, we considered several echocardiographic measures as potential markers of ventricular function for analysis. While most current echocardiographic parameters of LV function require multiple views of high technical quality, the technical limitations of performing echocardiography during ongoing resuscitation represent an important challenge for any approach aiming to quantitatively measure LV function in cardiac arrest. We chose left ventricular systolic fractional shortening (LVFS), a relatively simple, unidimensional echocardiographic parameter to perform the quantification of LV function during PEA. LVFS represents the degree of shortening of the LV diameter between end-diastole and end-systole and is calculated with the formula: [LVFS = End-diastolic diameter (EDD) ‒ End-systolic diameter (ESS)/EDD]. In order to increase the accuracy of this measurement, we generated post hoc M-mode images using intra-arrest echocardiographic video clips and M. mode.ify, an open source software application designed to generate an M-mode image post hoc from any B-mode ultrasound clip.¹⁰ M. mode.ify works by breaking down an ultrasound clip into individual frames, it then rotates and crops these frames using aa set M-mode line along any possible axis. The post hoc M-mode image is created by splicing these frames together (Supplementary Fig. 3).

Two investigators trained in emergency echocardiography and blinded to the clinical data (FT, AD), conducted a step-wise review of a random sample of 32 echocardiography clips. Using the qualitative rating scale, investigators reviewed and scored images for quality. In addition to the qualitative rating of echocardiography images, investigators also performed a visual classification of the degree of LV function, using a scale developed for this study which stratifies echocardiography images into 6 levels of LV function (Supplementary Fig. 4). The purpose of this visual classification scale was to explore the performance of a semiquantitative tool that could potentially be used by clinicians in real time during resuscitation. This was an exploratory aim of our work and therefore such classification was not used to determine inclusion or exclusion of subjects from our analysis. This cohort of subjects was used to assess interrater variability between the two sonographers for both the ratings on image quality and visually estimated LV function. In a subsequent review phase, the same subset of cases was used by the investigators to perform the standardization of the methodology that would be used to measure LVFS throughout the study. During this phase, we established the minimal quality requirements needed and refined the technique to allow accurate measurement of LVFS.

After the first phase, the lead investigator (FT), who was blinded to both the clinical and outcomes data, completed the LVFS measurement and reviewed all cases following the established protocol. Fig. 1 summarizes the study workflow and cohort derivation.

Fig. 1 –

Study workflow and cohort derivation.

LV: left ventricle; LVFS: left ventricular systolic fractional shortening.

Study outcomes

The primary outcome was ROSC. Secondary outcome was survival to hospital admission.

Data analysis

Data were analyzed using standard statistical software (STATA 14.2, STATA Corp, College Station, TX). Measurements from the semi-qualitative analysis by the two investigators were assessed for inter-observer agreement using kappa statistic following the suggested guidelines of Landis and Koch to describe the strength of agreement for the classification of image quality and the visual estimate of LV function.¹¹ Descriptive statistics were used to describe demographic and baseline characteristics. Univariate linear regression was performed using Wilcoxon Mann‒Whitney tests to evaluate for relationships between continuous variables including LVFS and their distribution among dichotomous variable groups. Fisher’s exact test was used for categorical variables. Univariate analysis was first performed independently for two outcomes of 1) ROSC and 2) survival to hospital admission. Results are provided as OR (95% CI) with p values. As performed in prior investigations, variables with p < 0.2 from univariate analysis were included in initial multivariate modeling.¹² For the regression models, we analyzed LVFS both as a continuous variable and as a categorial variable using the quartiles and the median to perform multiple different comparisons and to illustrate the relationship of LVFS and outcomes of interest. Lastly, in order to evaluate the effects of total resuscitation time and LVFS, we performed survival analysis using Cox proportional hazards model to evaluate the hazard corresponding to length of resuscitation. Subjects who died ‒ defined as those who did not achieve ROSC ‒ were censored.

Results

Cohort derivation

Of the total 312 subjects with documented PEA when echocardiography was performed, 91 (29,2%) had images with enough technical quality required to perform the measurement of LVFS. From the 91 subjects included based on the minimal technical quality, 7 were found to have severe segmental motion abnormalities and paradoxical movement of the septum. These two echocardiographic findings were considered by the investigators to be significant enough to invalidate the measurement of LVFS because both affected its accuracy as a surrogate of LV systolic function. The remaining 84 subjects constituted the final cohort of subjects used to perform the primary analysis of the study.

Baseline characteristics

Baseline demographics and relevant time variables including the time from arrest to initiation of CPR (downtime), onset of arrest and onset of resuscitation to echocardiography, as well as total resuscitation length were comparable between the cohort of subjects with an obtainable LVFS and those who were excluded (Table 1). The proportion of subjects with image quality rated as excellent was 61/84 (72.6%) in the cohort of subjects with an obtainable LVFS and 39/228 (17.1%) in the group of subjects where LVFS was not obtainable (p < 0.05). The proportion of ROSC was 69/228 (30.2%) among those excluded and 46/84 (54.7%) for those with an obtainable LVFS (p < 0.05). There was also a significantly higher proportion of subjects who survived to hospital admission within the group included (35.7%) compared to those excluded (13.6%) from the LVFS analysis (p < 0.05).

Table 1 –

Demographics, outcomes and resuscitation characteristics.

	Cohort of patients with LVFS obtainable [N = 84]	Excluded [N = 228]	Comparison between groups
Age mean yr [min–max]	65 [20–94]	66 [19–95]	p = ns
Male sex – [%]	50 [59.5]	138 [60.2]	p = ns
Image quality score 5 – n [%]	61 [72.6]	39 [17.1]	p < 0.05
Sonographer experience level	31 [37.3]	98 [42.9]	p = ns
attending – n [%]
Bystander CPR – n [%]	18 [21.4]	66 [28.9]	p = ns
Outcomes
- ROSC – n [%]	47 [55.9]	69 [30.2]	p < 0.05
- Survived to admission – n [%]	30 [35.7]	31 [13.6]	p < 0.05
Time variables (min) [SD]
- Downtime	5.8 [8.5]	6.7 [10.4]	p = ns
- Onset arrest to echocardiography	29.5 [23.3]	33.0 [21.5]	p = ns
- Onset of resuscitation to echocardiography	6.9 [14.2]	5.4 [5.1]	p = ns
- Total resuscitation length	52.6 [28.7]	50.4 [26]	p = ns

ROSC: return of spontaneous circulation; CPR: cardiopulmonary resuscitation; LVFS: left ventricular systolic fractional shortening.

Echocardiography image quality and visual classification of LV function

The agreement between image reviewers in the subset of 32 randomly selected cases used for classification of images using the quality rating score was almost perfect (Kappa 0.9), and the agreement for visual classification of LV function was substantial (Kappa 0.67).

Quantification of left ventricular function during PEA

The mean values of EDD and ESD were 4.3 cm (SD 1.79) and 3.7 cm (SD 1.86) respectively. The measured values of LVFS ranged from 0 to 96%, with a mean of 16.8% and median of13.1% (SD 16.96). The mean values of LVFS in subjects who achieved and did not achieve ROSC was 21.04% (SD 19.21) and 11.67% (SD 12.12) respectively (p < 0.05). The mean values of LVFS in subjects who survived and did not survive to admission from ED was 10.54% (SD 17.2) % and 14.73% (SD 16.62) respectively (p < 0.05).

Relationship between LVFS with clinical variables

Univariate analysis showed no statistically significant association between LVFS and time variables including downtime, onset of cardiac arrest to echocardiography, onset of resuscitation to echocardiography and total resuscitation duration (Supplementary Fig. 5). There were no significant associations between LVFS or its components EDD and ESD, or other resuscitation variables including sonographer experience, initial electrocardiographic rhythm, prehospital care level and total doses of epinephrine.

Association between left ventricular function and outcomes

Multivariate regression analysis identified age and gender to be associated with ROSC and survival to admission in each model. Odds ratios for each quartile of LVFS compared to LVFS of 0 are shown in the logistic regression models for both outcomes (Table 2). We found a positive association between LVFS and the primary outcome of ROSC (OR 1.04, 95%CI 1.01–1.08), but not with the secondary outcome of survival to hospital admission (OR 1.02, 95% CI 0.96–1.08) (Fig. 2). Given that the relationship is not linear and that we observed a threshold effect in the relationship between LVFS and outcomes, we performed an analysis using quartiles of LVFS. The predicted probability of ROSC was 75% for LVFS between 23.4–96% (fourth quartile) compared to 47% for LVFS between 0–4.7% (first quartile). The probability of ROSC using predicted margins of LVFS quartiles based on the multivariate regression model is depicted in Fig. 3.

Table 2 –

Multivariate model for left ventricular function.

Outcome/variable	Odds ratio	(p value)	95% CI
ROSC
LVFS (quartiles compared to LVFS 0)
- LVFS first quartile (0.1–4.7%)	7.33	0.08	0.74–71.94
- LVFS second quartile (4.8–13.3%)	11.59	0.03	1.26–106
- LVFS third quartile (13.4–23.3%)	8.87	0.04	1.04–75.29
- LVFS fourth quartile (23.4–96%)	27.58	0.004	0.84–267.63
- Age (years)	0.97	0.09	0.95–1.00
- Gender (M vs F)	0.69	0.46	0.26–1.84
Survival to hospital admission
LVFS (quartiles compared to LVFS 0)
- LVFS first quartile (0.1–4.7%)	4.68	0.2	0.42–51.14
- LVFS second quartile (4.8–13.3%)	18.83	0.014	1.82–194.71
- LVFS third quartile (13.4–23.3%)	8.47	0.051	0.99–72.55
- LVFS fourth quartile (23.4–96%)	16.79	0.013	1.79–157.52
- Age (years)	0.95	0.07	0.93–0.98
- Gender (M vs F)	0.61	0.37	0.21–1.76

ROSC: return of spontaneous circulation; LVFS: left ventricular systolic fractional shortening.

Fig. 2 –

Estimated left ventricular function and outcomes.

PEA: pulseless electrical activity; ROSC: return of spontaneous circulation; LVFS: left ventricular systolic fractional shortening.

Fig. 3 –

Predictive margins of Left Ventricular Fractional Shortening (LVFS).

Survival analysis with LVFS

A Cox proportional hazards model was fit to study the effects of LVFS and total resuscitation duration on the probability of achieving ROSC. The hazard of not achieving ROSC was significantly greater for subjects with LVFS below the median (13.1%) compared to the subgroup with LVFS greater than 13.1% (p < 0.05), with the separation of the survival curves occurring at approximately 40 min of resuscitation duration (Fig. 4).

Fig. 4 –

Survival analysis for Left Ventricular Fractional Shortening (LVFS) and total resuscitation length.

LVFS = left ventricular systolic fractional shortening.

Discussion

In this study, we found a positive correlation between LV function, measured by LVFS and ROSC. To our knowledge, this is the first study to investigate the degree of ventricular function during OHCA resuscitation care for patients with PEA. Better understanding of the association between ventricular function and outcomes of PEA arrest has the potential to enhance our understanding of this important subtype of cardiac arrest and may be important for the development of effective interventions.

While a higher proportion of the cases had images that would have been sufficient for other purposes (e.g. identification of cardiac tamponade or detection of cardiac standstill), in order to perform the most accurate LVFS measurements possible we needed to have simultaneous visualization of both the septal and free wall of the LV throughout the cardiac cycle. These technical factors determined which ultrasound images could provide the information needed for analysis of our final cohort of subjects. Given the nature of the data used in this study (secondary analysis of previously collected data), we do not have a definitive explanation for the difference in outcomes observed between subjects included vs. excluded in our study (Table 1). One plausible explanation is that subjects excluded, which had overall worse quality of echocardiography images or where images were not obtainable at all, may have represented a group with more severe cardiac dysfunction, or were treated differently by the resuscitation team (e.g. lack of visualization of the heart could lead to lower chest compression fraction, longer CPR pauses, etc.). Future prospective studies should include resuscitation quality parameters to control for these factors.

We found a positive association between our quantitative parameter of LV function (LVFS) and ROSC. We did not find a statistically significant correlation for this outcome with the visual estimation of LV function using the PEA score, however our results do indicate a trend in such direction (p = 0.09). The small number of subjects available for analysis may have decreased our ability to see an association between LV function and survival outcomes.

The substantial agreement found between the ratings of LV function by the two investigators (Kappa 0.67), demonstrates that information provided by intra-arrest echocardiography could possibly be used at the bedside to stratify subjects with PEA according to the degree of LV function. Future research should address if such classification could be made in real time to direct therapies and treatment decisions.

In this study we used a novel image rating scale.⁹ This scale was developed independently from this study. In that scale, level 5 describes the possibility to perform a quantitative assessment, but does not specify any specific methods for this. In this study, we proposed a specific method to evaluate (LVFS). While 17% of the 228 excluded subjects (Table 1) had videoclips meeting the criteria for “Excellent Image Quality”, their images were not adequate to make measurements using this method. In the majority of these cases this was due to the absence of an uninterrupted view of the septal and LV free walls throughout the cardiac cycle. In some videoclips, there were interruptions caused by extrinsic movement of the heart, often due to the effect of ventilations. It is possible, that quantitative measurements of LV function might be possible using a different method.

In the early 1990’s, Paradis et al. undertook pioneering research to describe and define the entity of electro-mechanical dissociation (EMD), a term used to describe what is now broadly categorized as PEA.⁴ Performing invasive hemodynamic monitoring of patients in cardiac arrest, this group demonstrated that a substantial subset of patients with no palpable pulse, actually had measurable hemodynamics. They called this disease state “pseudo-EMD” which later led to the terms of PEA and pseudo-PEA, the latter describing the clinical state where despite no palpable pulse, there is presence of myocardial activity.

The use of echocardiography during cardiac arrest resuscitation has provided new insights into the heterogeneity of cardiac function during cardiac arrest. While PEA is defined by electrocardiographic findings,¹³ the echocardiographic visualization of the heart during pulse checks can show varying degrees of ventricular function. Based on previous data and our clinical observations, we hypothesized that rather than a discrete disease process, PEA represents a spectrum which begins at the end of cardiogenic shock and ends with cardiac standstill. If this hypothesis of a continuum of the arresting heart is correct, we would expect that not only the presence of cardiac activity, but that the degree of LV function should be predictive of ROSC and survival. Our results in this exploratory study support this hypothesis in that they demonstrate that the degree of LVFS as a parameter of LV function is associated with the probability of ROSC.

Particularly important, in light of our limited understanding of PEA, is that the substrate and etiology of sudden cardiac death has changed over time.¹⁴ The incidence of VF arrest has been declining and therefore the incidence of PEA, as a fraction of all cardiac arrests, has been increasing.^14–16 While in some patients with PEA an underlying cause such as pericardial tamponade or tension pneumothorax can be identified and rapid treatment can sometimes restore circulation, in the majority of patients a reversible cause is not immediately evident, and clinicians must rely on standard advanced cardiac life support. In the great majority of patients these are ineffective.¹⁷

In this context, the results of this preliminary study may have important implications. If we are able to better characterize patients with PEA by stratifying their LV function, this could become a basis of individualized prognosis and therapy. For example, it might be hypothesized that patients with better LV function, despite the lack of palpable pulse, may be physiologically more similar to a patient in cardiogenic shock and therefore amenable to similar interventions such as vasopressors, inotropes, or mechanical circulatory support, instead of or in addition to chest compressions. Prior studies have failed to discriminate among patients in PEA, thereby potentially overlooking interventions that might be effective among subgroups with this condition.

Patients with a certain degree of LV function may also benefit from having invasive hemodynamic monitoring (e.g. arterial line) to inform management decisions and allow individualized (e.g. hemodynamic-directed) treatment pathways. For example, studies demonstrate that chest compressions that are asynchronous with the ventricular systole during PEA arrest decrease ventricular filling, coronary perfusion pressures, and cardiac output.¹⁸ Similarly, recent data from pediatric cardiac arrests of bradycardia with poor perfusion, an entity similar to pseudo-PEA, suggest that early chest compressions are helpful and lead to better outcomes.¹⁹

Additionally, greater understanding of the relationship between LV function in patients with PEA and clinical outcomes could allow new and more specific definitions of cardiac arrest that represent better the patient’s physiology as well as the chances of resuscitation success. Patients currently defined and managed under the broad umbrella term of PEA, could become two or more groups with distinctive pathological or pathophysiological features (e.g. post infarction vs. toxic vs. metabolic cardiomyopathy, etc.). It is hoped that this hypothesis-generating study, might lay the groundwork for further investigations that could lead towards more tailored management of PEA.

This study has limitations. It is a secondary analysis of data collected by a previous prospective study and represents only a subset of the total patients in the original study (those with PEA and cardiac activity on echocardiography). It is possible that these patients are not representative of the population of patients in PEA cardiac arrest. Images were reviewed by two emergency physicians with significant experience in echocardiography during cardiac arrest, with sufficient time to review images and perform quantitative measurements. Their findings and interpretations may not be generalizable to other clinicians with less experience in echocardiography. Another limitation of the study is one common to many investigations of the effect of diagnostic imaging in patients in cardiac arrest: the performance of the ultrasound itself may entail treatment decisions that affect the outcome. On one side, the additional information furnished by the ultrasound may positively affect outcome by expediting appropriate interventions. On the negative side, if the images are misinterpreted or misleading, clinical decision-making might be compromised thereby leading to worse outcomes. The performance of an ultrasound also requires additional human resources and space within the resuscitation area, and the ultrasound findings themselves may lead to delays in vital interventions such as cardiac impressions or possibly affect the intensity with which interventions are implemented. It is unknown how these countervailing effects might bias the results for the patients in whom good cardiac views can be obtained. Along similar lines, although our study appears to demonstrate that image quality and the ability to obtain and LVFS measurement did not have an impact on our outcomes, in common with many clinical ultrasound studies, patient factors such as body habitus, anatomical variants, and underlying illness can all affect the ability to both obtain ultrasound images and conduct an effective resuscitation.

By definition, this cohort was a subgroup in whom high-quality could be obtained. This represents an important source of selection bias because exclusion of patients with cardiac views that would prevent LVFS measurement could have led to the skewed exclusion of some forms of cardiac pathology such as right ventricular infarction. Another limitation of the study relates to the software application (M. mode.ify) that was used to measure LVFS. This software, while visually intuitive, has never been scientifically proven to reflect fractional shortening as determined by M-mode. It might be anticipated that real M-mode might actually be more discriminating, and therefore more accurate in measuring LVFS, however this is not known.

While in our analyses we found no association between LVFS and time variables (e.g. downtime, onset of resuscitation to echocardiography, total resuscitation duration, etc.), given that the timing of echocardiography was not standardized, it is likely that that some observed differences in LVFS were due to the dynamic nature of cardiac function during the course of resuscitation. Ideally, measurement of LVFS would be standardized within the course of resuscitation, but given the exigencies of this clinical syndrome and the many variables, this may be hard to accomplish. We would note that the performance of ultrasound in the course of cardiac arrest should always occur consistent with the recommended pulse checks and pauses of ACLS protocol. Similarly, any evaluation of LV function would only occur after reversible causes of PEA have been identified and addressed.

It is likely that some of the extreme values of LVFS correspond to patients who were misclassified during the primary study from which data for this analysis was obtained. For instance, LVFS of 0% would indicate cardiac standstill (not PEA). If this is the case, then some of our data points would represent the tracking of our primary outcome, rather than a predictor. Similarly, our “LV-centric” approach on this study may not be adequately accounting for non-LV dependent or extracardiac pathology such as RV failure or tension pneumothorax, and their effect in outcomes.

Lastly, as many observational studies of cardiac arrest this study is likely affected by resuscitation time bias, where subjects that had longer resuscitation times were more likely to receive the “intervention” (in this case echocardiography). Given that longer time of arrest is known to be associated with worse outcomes, the evaluation of echocardiography may be biased toward a negative effect given that patients with shorter resuscitation times (and therefore better outcomes) were less likely to be enrolled and have echocardiography performed.

Conclusion

Left ventricular function measured by LVFS is positively correlated with higher probability of ROSC and may be associated with higher chances of survival in patients with PEA arrest. Future prospective research is needed to confirm these findings and further characterize the relationship between ventricular function and survival outcomes in cardiac arrest. Finally, it is hoped that this study, by highlighting the impact of image quality on the ability to obtain vital clinical information, will prompt equipment manufacturers as well as sonologists to prioritize image optimization in the challenging setting of cardiac arrest.