Extracorporeal cardiopulmonary resuscitation at the emergency department

A retrospective patient selection evaluation

Abstract

BACKGROUND 

There is an increasing use of extracorporeal life support in refractory cardiac arrest. Recent studies highlighted the importance of an early and accurate patient selection for this invasive procedure.

OBJECTIVES 

The aim of this study was to retrospectively validate a six-criteria-screening-checklist (witnessed collapse, bystander-cardiopulmonary resuscitation/first medical contact <5 min, shockable, age <70 years, end tidal carbon dioxide >14 mmHg and pupils not anisocoric/distorted/mydriatic) as an early screening tool in patients treated with extracorporeal cardiopulmonary resuscitation (eCPR) at the emergency department.

DESIGN 

Retrospective observational study.

SETTING/PATIENTS 

All patients at least 18 years of age with nontraumatic cardiac arrest and without return of spontaneous circulation before eCPR treatment at our department between January 2013 and December 2018 were included in this retrospective observational study.

INTERVENTION 

No specific intervention was set in this observational study.

MAIN OUTCOME MEASURES 

Primary outcome was the rate of patients who fulfilled all criteria, secondary outcome was 30-day and 6-month survival in accordance with the criteria.

RESULTS 

Overall, data from a total of 92 patients were eligible for analyses. Out of these, 27 patients (29%) met all criteria. Patients, who fulfilled all criteria, showed significantly higher odds for 30-day survival [OR 6.0 (95% CI 1.78 to 20.19)] P = 0.004. Patients, who did not fulfil all criteria, showed significantly higher rates of early mortality after eCPR initiation [OR 4.57 (95% CI 1.69 to 12.37)] P = 0.003.

CONCLUSION 

Patients fulfilling all inclusion criteria showed higher rates of survival after eCPR. Our results affirm that there is a possibility and even an obvious necessity for early patient selection based on standardised criteria before eCPR treatment. Large randomised trials are urgently needed to answer this question accurately.

Introduction

Extracorporeal life support (ECLS), more specifically extracorporeal cardiopulmonary resuscitation (eCPR) when used during cardiac arrest, is an accepted option in refractory cardiac arrest with increasing use worldwide.^1,2 In 2015, guidelines recommended resuscitation concurrent with patient transport, which may facilitate in-hospital treatment options for suspected reversible causes of cardiac arrest (e.g. early percutaneous coronary intervention).³ In addition, since 2015 several studies have suggested resuscitation guidelines including the use of eCPR to prolong life with the goal of treating the underlying cause for cardiac arrest.⁴ A defined workflow and close interaction between the Emergency Medical Service (EMS) and the admitting hospital are essential for any eCPR programme, especially for the selection of potential candidates. But the question which patient may benefit most from eCPR is still not answered in a satisfactory manner. Therefore, there are no clear guidelines to determine candidacy for eCPR usage. For efficient and economical use of this expensive and limited treatment option, the process for evaluating patients should be initiated as early as possible. Our goal was the retrospective evaluation of a combination of simple criteria that would enable the inclusion/exclusion of potential patients with high accuracy. Consequently, we created a six-point checklist to facilitate an uncomplicated assessment at an early stage of EMS resuscitation effort. The aim of this study was to retrospectively validate this eCPR-checklist in patients treated with eCPR at our department.

Methods

Study design and setting

We conducted a retrospective analysis of the cardiopulmonary cerebral resuscitation registry of our Department of Emergency Medicine. This used a prospectively conceived and maintained clinical registry which included data of patients suffering from cardiac arrest (out-of-hospital and in-hospital) and who were admitted to the Department of Emergency Medicine at the General Hospital of Vienna.

In Vienna, the capital city of Austria (Europe) with 1.9 million inhabitants, EMS is operated by the Municipal Ambulance Service and supported by partner organisations. A detailed description of the Viennese EMS System has been published elsewhere.⁵ In addition to the EMS, police and firefighters are dispatched as first responders to minimise the ‘no-flow’-time. In case of a decision for potential eCPR treatment (at the discretion of the attending emergency physician at scene) the patient was transported to our emergency department (ED) with ongoing CPR. During the observation period there were no strict criteria in our department regarding eCPR. It was the attending physician's decision whether eCPR treatment was started.

The cannulation was performed by cardiac surgeons assisted by emergency physicians in most of the cases. A Seldinger technique or, if necessary, a cutdown approach, was used to cannulate the femoral artery and femoral vein. A perfusionist, a trained ED nurse or physician initialised the circulation of the extracorporeal membrane oxygenation (ECMO) device if cannulation was successful. The Cardiohelp or Rotaflow pumps, heparin-coated tubes and the Quadrox D oxygenator (Maquet, Rastatt, Germany) were commonly used. Return of spontaneous circulation (ROSC) on ECMO was defined as a pulsatile flow seen in the arterial curve or echocardiographic proof of opening of the aortic valve during systole. Early mortality was defined as never achieving ROSC during eCPR treatment. Favourable neurological outcome was rated as a cerebral performance category score of 1/2.

Study population

All adults at least 18 years of age with a nontraumatic refractory cardiac arrest admitted to the Department of Emergency Medicine between 2013 and 2018 and who received further eCPR treatment were included in this study. The collected data included demographic background information, resuscitation-specific parameters according to Utstein style criteria and outcome data, such as ROSC and 30-day mortality.⁶

Endpoints

The primary endpoint of the study represents the number of patients who met all our checklist-criteria. Furthermore, 30-day survival and 6-month survival was assessed, compared with whether or not the criteria were fulfilled.

Checklist-criteria

In the existing literature there already are well established assessment criteria concerning increased survival: an initial shockable rhythm, a witnessed collapse, bystander CPR, younger age and the time interval from collapse to possible eCPR start.^7–13 Further predictors for survival after prolonged cardiac arrests might be sustained physiological levels of end tidal carbon dioxide (ETCO₂) and signs of life during resuscitation (e.g. pupils which are reactive to light, of equal size and not excessively dilated: that is not anisocoric, unequal or mydriatic).^14–16

Therefore, we considered six criteria as most important for potential survival after eCPR: a witnessed collapse, bystander CPR or first medical contact within 5 min after collapse, an initial shockable rhythm, age less than 70 years, ETCO₂ more than 14 mmHg persistently and pupils not anisocoric, unequal or mydriatic.^7–9,14,17

Statistical analysis

The patients were stratified into two groups, namely whether all criteria were fulfilled or not. Continuous data are presented as median [interquartile range] – discrete data as counts (%). The χ²-test was used for comparison of categorical data, and for comparison of continuous variables a t test was used. For data management we used Microsoft Access and Excel (Microsoft Corporation, Redmond, Washington, USA) and statistical analyses were performed using the PASW 22.0 (IBM SPSS, Chicago, Illinois, USA). If not stated otherwise, a two-sided P value less than 0.05 was considered statistically significant.

Ethics

The study is a retrospective analysis of a registry and complies with the declaration of Helsinki. It was approved by the ethics committee of the Medical University of Vienna (Borschkegasse 8b/E06, 1090 Vienna, Austria; Chairperson Dr Singer) with the EK-Number 1814/2012.

Results

Out of 129 cardiac arrest patients with consecutive ECLS treatment during the study period, a total of 92 eCPR-patients (71%) were eligible for analyses (Fig. 1). The mean age of patients was 48 (SD 14) years, the proportion of women was 22% (n=20). Detailed baseline characteristics, including the distribution of cardiac arrest-related parameters stratified by fulfilling the criteria are presented in Table 1. Overall, 45 patients (49%) achieved ROSC after eCPR, 14 patients (15%) survived for 30 days and 12 patients (13%) survived for 6 months. Focusing on those 27 patients (29%) who met all criteria – 20 patients (74%) achieved ROSC after eCPR, nine patients (33%) survived for 30 days. The distribution of the criteria within the analysed cohort is presented in Table 2. Patients who fulfilled all criteria showed significantly higher odds for 30 day-survival [OR 6.0 (95% CI 1.78 to 20.19)] P = 0.004 and for 6-month survival [OR 4.2 (95% CI 1.19 to 14.72)] P = 0.025, whereas patients who did not fulfil all criteria showed significantly higher rates of early mortality [OR 4.57 (95% CI 1.69 to 12.37)] P = 0.003. (Table 3). Furthermore, patients who met all criteria showed statistically significant higher rates of favourable neurological outcome 6 months after resuscitation (P = 0.045). A total of five patients, who did not meet all criteria, survived. Of these, there was one patient with initial pulseless electrical activity (PEA), one patient with unwitnessed collapse and ETCO₂ less than 14 mmHg, one patient with PEA and over 70 years, one patient with PEA and ETCO₂ of less than 14 mmHg and one patient with ETCO₂ less than 14 mmHg.

Fig. 1: Inclusion-chart; ECLS, extracorporeal life support; eCPR, extracorporeal cardiopulmonary resuscitation; ROSC, return of spontaneous circulation; VV, veno-venous.

Table 1: Baseline characteristics according to checklist criteria

Table 2: Distribution of the checklist-criteria

Table 3: Binary regression model for 30-day survival

Discussion

In this retrospective analysis, we could demonstrate that during the observation period patients fulfilling all eCPR inclusion criteria showed significantly higher rates of survival and favourable neurological outcome than patients who do not fulfil the criteria. Therefore, it seems feasible to screen patients for potential eCPR treatment at an early stage of conventional CPR (cCPR).

The concept of eCPR in refractory cardiac arrest appears to be of increasing importance and was already mentioned in the latest resuscitation guidelines.³ On the one hand, this may be due to increasing availability of devices and on the other hand to encouraging outcome data.^1,4,18–20 Nevertheless, a clinically feasible patient selection to avoid futile attempts or ‘bridge to nowhere’ situations would be required.^21,22 On the contrary, results of randomised controlled eCPR trials are still missing. Therefore, no prospectively validated inclusion or exclusion criteria for eCPR are available.

With reference to the impact of the duration of cCPR before eCPR on outcome, early evaluation for inclusion seems inevitable.^23–25 In our study, the mean duration of cCPR to eCPR initiation does not differ significantly between the outcome groups. In addition, we have to consider that the start of eCPR within 82 min after collapse (in both groups) was much longer than the recommended time frame of 60 min. Other studies already showed the feasibility of a strict time management until initiation of eCPR^23–25 Despite the delayed start of eCPR, patients fulfilling the criteria showed higher survival rates. This is consistent with findings from Denmark highlighting that survival in selected patients is possible even after prolonged cCPR durations.²⁶

The developed checklist is mostly based on already verified and accepted factors leading to good outcome in cCPR and eCPR. Debaty et al. recently investigated potential patient characteristics for favourable outcome after eCPR. He described younger patient age, initial shockable rhythm, short cCPR duration to start of eCPR, higher blood pH levels and lower serum lactate levels as significantly beneficial for the primary outcome.⁸ These results seem consistent with our results and criteria selection. Lamhaut et al.¹⁵ described the presence of signs of life during CPR (breathing efforts, gasp, movements, pupils different from mydriasis) as the most potent predictors of survival in 2017. We responded to these data with the absence of anisocoric/unequal pupils that did not respond to light as a key criterion in our checklist.

In addition, we sought an alternative surrogate parameter to serum lactate and pH levels as indicators of body perfusion during cardiac arrest as these are not available in the preclinical setting and additionally have been shown not genuinely reliable for prediction of outcome after eCPR.²⁷ Paradis et al.²⁸ showed in the early 1990s that surviving a cardiac arrest depends on an adequate perfusion of the vital organs. However, direct measurement of organ blood flow during CPR is clinically not feasible. The end-tidal carbon dioxide represents an easily applicable, widely spread, noninvasive measurement of the effectiveness of CPR in terms of blood flow that is generated and the potential of successful resuscitation.¹⁴ A persistently high-end tidal carbon dioxide level, with a cut-off level of 14 mmHg was shown to be predictive for survival after conventional CPR.¹⁷

Our study showed the proportion of patients in our collective who might benefit most from eCPR. There are still some more questions to be answered before implementing our ‘screening-checklist’ as a strict guideline: Should the individual criteria gain or lose importance with increasing duration to eCPR? Should they be converted in a ‘must exist’ (e.g. bystander CPR), or a ‘should exist’ (e.g. initial rhythm)? Regarding the initial rhythm, Nakashima recently concluded that out-of-hospital cardiac arrest patients with sustained shockable rhythm may be the most promising eCPR candidates compared with patients with conversion to PEA or asystole.¹⁰ In addition to the fact that in our study all survivors not fulfilling the checklist criteria had PEA as an initial rhythm, our study group, as others, showed good survival in patients with PEA with higher electrical rate. Many PEA patients might be actually suffering from ‘Pseudo PEA’ (having cardiac contractions with low output and no detectable pulse) and could also benefit from eCPR treatment.^29,30 Further, an extension for the criteria ‘ROSC at any time during CPR’ may also be considered.

Limitations

There were no strict criteria for eCPR introduced during the observation period. Thus, our data represent a retrospective view on the current situation without ultimately portraying the meaningfulness or the level of accuracy of the respective criteria, their interactions or their individual impact. The reasons for the decision to start eCPR in individual cases are not comprehensible. The major limitation in this observation was that, in respect of our selected criteria, basically the ‘wrong’ patients were represented in the analysed collective (eCPR-treatment), resulting in a difficulty to assess the quality of the criteria regarding the outcome.

Conclusion

We were able to show that an early selection of patients eligible for eCPR is feasible and would possibly lead to higher rates of survival after eCPR. This would reduce labour and cost in primary hopeless cases. However, only a limited number of patients who received eCPR treatment fulfilled all inclusion criteria. Large prospective randomised trials are urgently needed to answer this question accurately.