SCCM Resource Library-A Multicenter Study of Withdrawal of Therapy in Extracorporeal Cardiopulmonary Resuscitation

Video Transcription

Hello. My name is Dr. Julia Carlson, and I'm a current Senior Neurocritical Care Fellow at Massachusetts General Hospital.  Today, I will be presenting a multicenter study of withdrawal of life-sustaining therapy and extracorporeal cardiopulmonary resuscitation.  Use of extracorporeal cardiopulmonary resuscitation, eCPR, is increasing. Withdrawal of life-sustaining therapy, WLST, is an important contribution to mortality  in the broader critical care population, as well as the ECMO population specifically, because this is one of the only ways that these patients pass away while utilizing this  technology. To date, there's been very little studies regarding the ECMO population and withdrawal  of life-sustaining therapy, and even less that specifically focuses on withdrawal of life-sustaining therapy after cannulation for eCPR.  So therefore, we sought to understand the population of patients undergoing withdrawal of life-sustaining therapy after eCPR cannulation, as well as the timing of this event.  The Extracorporeal Life Support Organization is an international consortium of over 300 ECMO centers that self-report data about their patients.  We queried the database from 2007 to 2017 for adult patients who died due to family request after cannulation for eCPR.  We only included patients until 2017, because at that time, ELSO changed their survey form.  Prior to that, the form had respondents choose a mutually exclusive primary cause of death, one of which could be due to family request.  After that, more than one reason for cause of death could be selected. When working with ELSO to design this study of withdrawal of life-sustaining therapy,  their recommendation was to utilize this group of patients prior to 17 that had family requests selected as their primary cause of death.  Patients were further stratified for analysis based on a 72-hour cutoff. This was selected based on current American Heart Association and American Academy of  Neurology guidelines that recommended waiting for prognostication to occur at least 72 hours after cardiac arrest.  This is felt to be a similar patient population for our eCPR group. Early withdrawal of life-sustaining therapy was defined as early if it occurred before  72 hours, and routine if it occurred after 72 hours. First, 228 patients, over half of the 411 patients included on our study, had early  withdrawal of life-sustaining therapy in the first 72 hours of cannulation. The median eCPR trial for these patients was 24 hours.  Otherwise, there were no significant demographic differences between early and routine withdrawal of life-sustaining therapy, including age, weight, sex, or race.  This histogram shows the frequency of withdrawal of life-sustaining therapy after eCPR cannulation.  Hours of eCPR trial are on the x-axis, a number of patients that underwent withdrawal of life-sustaining therapy are on the y-axis.  This demonstrates visually, again, that the majority of patients experiencing withdrawal of life-sustaining therapy after cannulation do so within the first 72 hours.  Of those patients, most of them experienced withdrawal of life-sustaining therapy within the first 24 hours after cannulation, as demonstrated by the largest bar on the left.  We did not observe any differences in baseline support of other life-sustaining therapies,  including vasopressive agents, renal replacement therapy, and cardiac support between early and routine withdrawal of life-sustaining therapy groups.  The only difference we did observe is that pH was significantly lower in the early withdrawal of life-sustaining therapy group.  At 24 hours, we see that pH is significantly lower in the early withdrawal of life-sustaining therapy group. We also see that bicarbonate at 24 hours is lower.  With regards to hemodynamic monitoring, diastolic blood pressure and mean arterial blood pressure  are lower in the early group compared to the routine group, and lactate is significantly higher in the early withdrawal of life-sustaining therapy versus routine.  We also see increased ventilator support with regards to rate, FiO2, and peak inspiratory pressures.  We did a multivariate logistic analysis and found that when other factors were controlled for, baseline pH was no longer significant.  However, pH at 24 hours, oxygen saturation, pump flow, and peak inspiratory pressures continued to be significant influences on early withdrawal of life-sustaining therapy.  We also looked at eCPR-related complications. We used the event rate because complications are likely to accumulate over longer trial durations.  All complications were occurring at significantly higher rates in the early withdrawal of life-sustaining  therapy group in comparison to the routine withdrawal of life-sustaining therapy group. This is except for infectious complications, which had no significant difference between  groups. Over half of the withdrawal events occur within 72 hours after cannulation, and the majority of these occur within the first 24 hours.  This stands out in comparison, again, to current guidelines for post-cardiac arrest care that recommend continuing care and waiting for prognostication for at least 72 hours unless  there is evidence of severe injury, significant instability, or previously expressed goals of care.  Early withdrawal of care may miss patients who could have otherwise had an acceptable outcome.  Notably, we found that at baseline, these patients have no significant demographic differences  or utilization of other life-sustaining therapy, such as renal replacement therapy and cardiac support prior to cannulation.  This is in contrast to some of the other studies that demonstrate early withdrawal of life-sustaining therapy may be associated with age or race of patient.  What we did observe is that at baseline, early withdrawal of life-sustaining therapy patients had a lower pH. Lower pH can be seen as a sign of increased illness severity.  As well, at 24 hours, there was more evidence of increased illness severity in the early withdrawal of life-sustaining therapy group with continued low pH and having evidence  of more ventilator support. So being sicker may have contributed to earlier withdrawal of life-sustaining therapy.  Finally, we observed higher rates of complication in the early withdrawal of life-sustaining therapy group.  So it's possible that as complications accumulated, this also influenced the decision for withdrawal as well as its timing. Our study had several significant limitations.  The first was a retrospective study of voluntarily reported data, so we don't know how accurate this data may have been.  We were also limited by the way that this group was selected at the behest of the ELSO group. We also had no comparison group either of survivors or of other eCPR patients that died  for other reasons. We were only able to make within-group comparisons and make a description of the patients that  currently experience withdrawal of life-sustaining therapy after eCPR. This study, although we comment on those that had targeted temperature management,  really was not able to take into account the influence of targeted temperature management on the timing of withdrawal of life-sustaining therapy.  So in conclusion, this is the first study that specifically focused on patients experiencing withdrawal of life-sustaining therapy from eCPR.  And those patients undergoing withdrawal of life-sustaining therapy after initiation of  eCPR, many experience withdrawal of life-sustaining therapy within 72 hours after cannulation.  These patients have indications that they're sicker at 24 hours and experience more frequent complications.  I want to thank you for your time and the opportunity to present my research.

Video Summary

Dr. Julia Carlson presents a multicenter study on withdrawal of life-sustaining therapy (WLST) and extracorporeal cardiopulmonary resuscitation (eCPR). The study focuses on patients who undergo WLST after eCPR cannulation. The majority of patients experience WLST within 72 hours of cannulation, with most occurring in the first 24 hours. They found no significant demographic differences between early and routine WLST groups, but early withdrawal patients had lower pH levels and signs of increased illness severity. Complications were more common in the early withdrawal group. The study highlights the need to consider the timing of WLST in eCPR patients, as early withdrawal may miss potential positive outcomes.

Asset Subtitle

Cardiovascular, Procedures, Ethics End of Life, 2022

Asset Caption

INTRODUCTION: As the use of extracorporeal membrane oxygenation (ECMO) increases after refractory cardiac arrest, known as extracorporeal cardiopulmonary resuscitation (ECPR), understanding factors associated with withdrawal of life sustaining therapy (WLST) is critical to avoid premature withdrawal. We aimed to investigate timing and characteristics of patients with WLST after ECPR. METHODS: The Extracorporeal Life Support Organization (ELSO) database was queried from 2007-2017 for patients > 18 years who died on ECMO due to family request. Data after 2017 were excluded due to a form change. Hemodynamic and lab data was collected ±6 hours prior to cannulation and 24 hours later. Patients were stratified by timing of WLST. Early WLST was defined as < 72 hours on ECPR based on current guidelines for prognostication after cardiac arrest; WLST group included all patients with WLST > 72 hours. RESULTS: 411 ECPR patients had WLST during this period (median age 42, interquartile range 24-80; 31.7% female). WLST occurred in 55.5% (n=228) within 72 hours, with an average trial of 27.7 hrs. There were no demographic or hemodynamic differences between groups prior to cannulation. Early WLST patients had lower pre-ECMO pH (7.14 ± 0.23 vs 7.20 ± 0.17; p=0.014). Twenty-four hours from cannulation, early WLST patients had significantly lower on-ECMO pH (7.35 ± 0.13 vs. 7.42 ± 0.08; p < 0.001), elevated lactate (10.2 ± 9.8 mmol/L vs 4.3 ± 2.9 mmol/L; p=0.009), and higher oxygen requirements (FiO2 61.7% ± 26.4% vs 51.3% ±19.6%; p < 0.001). In multivariable analysis, lactate was no longer significant when other factors were controlled for. While complications accumulated in the WLST patients (infectious: 19 vs. 35%, p < 0.001; renal: 50.8 vs. 95.6%, p < 0.05), there was no significant difference in the occurrence of neurological complication (17 vs 29.5%; p=0.16). CONCLUSIONS: This is the first study to use an international database to analyze WLST in ECPR patients. A significant proportion of WLST in ECPR patients occurs prior to a standard 72-hour trial period. There were few differences prior to cannulation, and early WLST patients were more severely ill at 24 hours. Being more severely ill at that time may contribute to early WLST. These findings merit further evaluation of appropriate patient selection, shared decision making, and duration of ECPR trials prior to WLST.

Meta Tag

Content Type Presentation

Knowledge Area Cardiovascular

Knowledge Area Procedures

Knowledge Area Ethics End of Life

Knowledge Level Advanced

Learning Pathway Cardiothoracic Critical Care

Membership Level Select

Tag Extracorporeal Life Support

Tag Medical Legal Issues

Tag Extracorporeal Membrane Oxygenation ECMO

Tag Cardiothoracic Critical Care

Year 2022

Keywords

withdrawal of life-sustaining therapy

extracorporeal cardiopulmonary resuscitation

WLST

eCPR

timing of WLST

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.