Resuscitation of Patients with Durable Mechanical Circulatory Support with Acutely Altered Perfusion or Cardiac Arrest: A Scoping Review

Abstract

Background—

There is an increasing prevalence of durable mechanical circulatory supported patients in both the in-and-out of hospital communities. The scientific literature regarding the approach to patients supported by durable mechanical circulatory devices who suffer acutely impaired perfusion has not been well explored.

Methods—

The International Liaison Committee on Resuscitation Advanced, Basic, and Pediatric Life Support Task Forces conducted a scoping review of the literature using a population, context, and concept framework.

Results—

A total of 32 publications that included patients who were receiving durable mechanical circulatory support and required acute resuscitation were identified. Most of the identified studies were case reports or small case series. Of these, 11 (34.4%) included patients who received chest compressions. A number of studies reported upon delays in the application of chest compressions resulting from complexity due to the expected pulselessness in some patients with continuous flow left-ventricular assist devices as well as from concern regarding potential dislodgement of the mechanical circulatory support device. Three observational studies identified worse outcomes in durable mechanical circulatory support receiving patients with cardiac arrest and acutely impaired perfusion who received chest compressions as compared to those who did not, however those studies were at high risk of bias. Of 226 patients across 11 studies and two published scientific abstracts who sustained cardiac arrest while supported by durable MCS and underwent chest compressions, there were no reported instances of device dislodgement and 71 (31.4%) patients had favorable outcomes.

Conclusions:

There is a scarcity of evidence to inform the resuscitation of patients with durable mechanical circulatory support (MCS) experiencing acute impairment in perfusion and cardiac arrest. Reports indicate that delays in resuscitation often stem from rescuers’ uncertainty about the safety of administering chest compressions. Notably, no instances of device dislodgement have been documented following chest compressions, suggesting that the risk of harm from timely CPR in these patients is minimal.

Introduction:

Durable mechanical circulatory support (MCS) devices, especially left ventricular assist devices (LVAD), are increasingly being used for patients with advanced heart failure and patients supported by MCS are living longer.(1) As a result, there is a rising prevalence of durable MCS supported patients both in the hospital and in the community. The occurrence of acutely altered perfusion in patients with MCS is a medical emergency requiring immediate intervention, and first-responders must be prepared to navigate these complex clinical scenarios.

While evidence-based guidelines exist for the resuscitation of patients with cardiac arrest, the optimal approach to the resuscitation of patients with durable MCS experiencing acutely altered perfusion has not been well established. Care of the MCS supported patient with acutely altered perfusion is complicated by challenges with the identification of adequate blood flow and pressure (those with continuous flow devices do not necessarily have a palpable pulse even when clinically stable and assessment of perfusion can be difficult) and management of no-flow states (chest compressions carry a theoretical risk of device dislodgment). A 2017 statement from the American Heart Association (AHA), a 2024 British Societies LVAD Emergency Algorithm Working Group guideline, and others provide guidance for first-responders, however the literature reviews informing these recommendations did not follow formal literature review methodology.(2–5)

The International Liaison Committee on Resuscitation (ILCOR) aims to improve survival after cardiac arrest worldwide. Recognizing the rising prevalence of MCS in the community and the increasing likelihood that first responders will encounter a patient with MCS suffering from acutely altered perfusion, the ILCOR Advanced Life Support task force in coordination with the Basic and Pediatric Life Support task forces conducted a scoping review of the literature to help inform future treatment recommendations and research priorities. A scoping review methodology was selected given uncertainty regarding the depth and scope of the existing literature and to map existing peer-reviewed studies to inform future research.(6)

Materials and Methods:

Protocol Development and Approval

A formal written protocol was created by the scoping review study team and included experts from three different ILCOR task forces (Advanced Life Support, Basic Life Support, and Pediatric Life Support) as well as others with specialized expertise in cardiac surgery, cardiothoracic surgical critical care, and ventricular assist device coordination (see Data Supplement). A research librarian assisted with creation of the protocol and development of the search strategy. The finalized protocol was approved by the ILCOR Scientific Advisory Committee (SAC) prior to the execution of the search strategy.

Eligibility Criteria, Creation of a Search Strategy, and Study Screening and Inclusion

The protocol followed a Population, Concept, and Context format.(7) The population included patients of any age who were receiving durable MCS of any kind. Durable was defined by a device intended for ongoing use after hospital discharge. The concept was acutely impaired perfusion resulting in need for acute resuscitation (defined as an emergent medical or procedural intervention) and the context was both in-and-out of hospital settings.

The search strategy was created in coordination with a research librarian (ME). Two databases (PubMed and Embase) were searched to identify all potentially relevant publications. The full search terms can be found in the Data Supplement. Each identified title and abstract were reviewed by two authors and any disagreements adjudicated by consensus or a third reviewer. Full text articles were then reviewed and relevant data abstracted using a standardized form. Data abstraction was conducted by an author and reviewed by author AM—the data abstraction elements can be found in the Supplementary Table. Grey literature was searched using publicly available search engines.

Randomized controlled trials, non-randomized studies, treatment guidelines, and case series and reports were included in the initial search. Non-peer reviewed studies, unpublished studies, conference abstracts and trial protocols were sought using publicly available search engines Google and Bing. We additionally searched the references of all included studies for any missed studies. All relevant publications in any language were included so long as there was an English language abstract. Clinicaltrials.gov was searched to identify any ongoing clinical trials relevant to the scoping review. The Preferred Reporting Items for Systematic reviews and Meta-Analyses Scoping Review Extension checklist was followed (see Supplement for checklist).(8)

Results:

Results of the Search, Screening, and Full Text Review

A total of 3,556 results were identified in the initial search conducted in September 2023. After title and abstract screening, 97 (2.7%) remained and underwent full text review. Thirty-one (0.9%) studies were included after full text review and underwent data abstraction. The primary reason for exclusion of a full text study was due to incorrect concept (42 of 66 [63.6%]), often because of a focus on patients with MCS who developed stable ventricular arrhythmias without acutely altered perfusion. A repeat search using the above search strategy was performed in May 2024 with one additional study identified. See Figure 1 for complete details.

Figure 1:

Study Selection Flowchart

Studies Identified through Search Engines, Review of References, and Clinical Trial Registries

Search for additional studies using publicly available search engines was conducted on March 18^th, 2024. This search yielded two abstracts detailing cardiopulmonary resuscitation in patients with LVADs—however these abstracts do not appear to have progressed further.(9, 10) No additional studies were identified from search of the references of reviewed studies or review of clinical trial registries.

Details of Included Studies

Of the 32 studies included, 22 (68.8%) were case reports including 2 or fewer patients, 8 (25.0%) were case series including 3-10 patients, and 2 (6.3%) were retrospective cohort studies including more than 10 patients. Eleven (34.4%) studies described a patient who suffered cardiac arrest and received chest compressions. Durable MCS in all studies was a left ventricular or biventricular assist device. A complete description of included studies can be found in Table 1.

Table 1:

Details of Included Studies

Study	Publication Year	Study Type	Continent	Total number of patients with acutely altered perfusion	Population	Mechanical Support Device(s)	Chest Compressions Described
Senman et al(19)	2024	Case Series	North America	58	Both In-and-Out of hospital; Adult	LVAD	Yes
	Case series of 58 LVAD supported patients at a single institution who suffered cardiac arrest. Of these, 24 received chest compressions and 34 received no chest compressions. Per review of the notes, the most common reason for withholding of chest compressions was a perceived contraindication to chest compressions in LVAD supported patients. There were no documented cases of device dislodgement. Survival was similar between those who did and did not receive chest compressions, but neurologic outcomes were worse in patients who received chest compressions.
Sande et al(23)	2023	Case Report	North America	1	In-hospital; Adult	LVAD	No
	Case report of a patient experiencing device alarms after undergoing an ablation procedure shortly after a percutaneous LVAD placement. A bedside echo showed a large circumferential pericardial effusion with right ventricular collapse and tamponade. The patient underwent bedside pericardiocentesis with improved physiology.
Victor et al(24)	2022	Case Report	North America	1	In-hospital; Adult	LVAD	No
	Case report of a patient experiencing increasing dyspnea and hemodynamic instability 6 days after LVAD placement. LVAD flow rate adjustments and vasopressor utilization were unsuccessful, and ultrasound identified a pericardial effusion. Successful operative management was performed.
Akin et al(25)	2022	Case Report	Europe	1	In-hospital; Adult	LVAD	No
	A case report of a patient 10 days after LVAD placement involved a research study for a sublingual microcirculatory imaging tool for microvascular circulation and perfusion. The device revealed severe failure of the microcirculation, and the patient later developed hemodynamic compromise and signs of hypoperfusion. Cardiac tamponade was identified that was subsequently surgically corrected.
Ratman et al(26)	2022	Case Report	Europe	1	Out-of-Hospital; Adult	LVAD	No
	Case report of a patient admitted with low LVAD flows and multiple organ failure. Pump flows and evidence of organ injury improved with fluids.
Doita et al(27)	2022	Case Report	Asia	1	In-hospital; Adult	LVAD	Yes
	Case report of an LVAD thrombosis leading to left outflow obstruction. The clot was large enough to occupy the LVAD inflow and resulted in nearly no forward flow from the device. The patient suffered cardiac arrest. Chest compressions were administered but the patient could not be resuscitated.
Barssoum et al(21)	2022	Retrospective observational cohort	North America	578	In-hospital; Adult	LVAD	No
	Retrospective analysis of the National Inpatient Sample including LVAD patients who sustained cardiac arrest comparing outcomes of those who underwent chest compressions with those who did not. Of 578 patients, 226 (39.1%) survived to hospital discharge. Mortality was 74% for those receiving chest compressions vs. 61% for those who did not achieve chest compressions (p<0.01). This study was limited by potential misclassification as only administrative data was used and variables available for abstraction were limited.
Pokrajac et al(28)	2022	Retrospective observational cohort	North America	1	In-hospital; Pediatric	LVAD	No
	Single center, retrospective review of 54 emergency department visits in pediatric VAD patients. There were no deaths or cardiac arrests in the ED. 4 patients in the cohort died, with one experiencing cardiogenic shock and organ failure.
Esangbedo et al(29)	2022	Case Series	North America	4	In-hospital; Pediatric	LVAD; BiVAD	Yes
	Case series of 5 pediatric patients who underwent chest compressions with VAD in place. Patient 1 had had cardiac arrest due to tamponade and suffered severe neurologic injury. Patient 2 had RVAD disconnect and brief chest compressions. Patient 3 had cardiac tamponade and brief chest compressions prior to chest exploration. Patient 4 had brief chest compressions with tamponade. Of the 4 patients, 3 survived with good outcomes.
Oates et al(30)	2022	Case Report	North America	1	In-hospital; Adult	LVAD	No
	Case report of patient who deteriorated after attempt at VT ablation with hypoxemia from intratrial shunt.
Ziegler et al(31)	2021	Case Report	North America	1	Out-of-Hospital; Adult	LVAD	No
	Case report of an emergency repair using splicing of a transected driveline in a left ventricular assist device.
Iwashita et al(32)	2020	Case Report	Asia	1	Both In-and-Out of hospital; Adult	LVAD; ECMO	Yes
	Case report of cardiac arrest 2 years post LVAD placement. The device was unknown by responders and chest compressions were performed for 40 minutes. On arrival to the hospital a depleted battery was discovered and changed after 50minutes of total chest compressions. Subsequent Vt was not responsive to defibrillation and VA-ECMO was initiated. A complicated course led to patient death after ECMO was discontinued.
Eyituoyo et al(33)	2020	Case Report	North America	1	Both In-and-Out of hospital; Adult	LVAD	No
	A case report of a patient with an LVAD placed 7 years earlier who developed altered mentation and hypotension. Upon EMS arrival, an irregular rhythm was noted and presumed to be artifact from LVAD. In the emergency department, VF was noted and corrected with defibrillation. The patient developed multiorgan failure and later expired.
Saito et al(34)	2019	Case Report	Asia	1	Out-of-Hospital; Adult	LVAD	Yes
	Case report of a patient who suffered from global cerebral ischemia due to LVAD pump stoppage. Chest compressions were performed by paramedics and LVAD function was restored after hospital arrival by exchanging external cables. The patient recovered without any neurological deficit.
Harper et al(35)	2019	Case Report	North America	1	In-hospital; Adult	LVAD	No
	Case report of a patient with an LVAD placed 3 years earlier presenting to the emergency department in refractory VT and experiencing chest pain, dizziness and multiple discharges of his ICD. Received medications and external shocks and LVAD flow rate was decreased to allow better ventricular filling.
Thiele et al(36)	2018	Case Report	Europe	1	Out-of-Hospital; Adult	LVAD	Unclear
	Case report of LVAD driveline disconnect. Patient recovered with re-connecting driveline.
Ornato et al(37)	2018	Case Report	North America	1	Out-of-Hospital; Adult	LVAD	Yes
	Case report of a patient with an LVAD who suffered cardiac arrest. Patient was intubated and ETCO2 was 0 mmHg with confirmation of tube placement. Compressions started and ETCO2 rose to 28 mmHg.
Godishala et al(38)	2017	Case Series	North America	4	In-hospital; Adult	LVAD	No
	Case series of 4 patients suffering acute myocardial infarction while supported by continuous-flow LVADs. Patient 1 received shocks from ICD due to VT, attributed to electrolyte derangement; once corrected the patient was asymptomatic. Patient 2 also received shocks from ICD due to VT which was attributed to complete thrombotic occlusion of left circumflex artery; this was removed but the patient suffered complications and died following intracranial hemorrhage. Patient 3 experienced chest pressure, diaphoresis shortness of breath and presyncope due to coronary artery occlusion; once stented he remained symptom-free. Patient 4 experienced chest pain and shortness of breath due to large thrombus in aortic valve; this was removed but one month later the thrombus returned and the patient died.
Yuzefpolskaya et al(39)	2016	Case Report	North America	1	In-hospital; Adult	LVAD	Yes
	This paper presents an algorithm for assessment and management of hospitalized unresponsive LVAD patients. A case study is presented by way of rationale for the algorithm in which a patient who was post-operative day 8 after LvAD implantation developed acute altered perfusion. Chest compressions were not initially performed as the patient was recently post-operative and it was unclear whether cardiac arrest had occurred. Chest compressions were ultimately initiate 15 minutes into the event and the patient was placed on VA-ECMO. After transfer to the ICU, the patient was pronounced brain dead.
Bouchez et al(40)	2016	Case Report	Europe	2	Both in and out of hospital; Adult	LVAD	No
	Two case reports of patients with LVADs who went into VF and developed deteriorating RV function. The authors describe a “treatment protocol” that includes augmenting MAP, addressing wall tension, treating electrical storm, and defibrillation.
Plymen et al(41)	2015	Case Report	Europe	1	In-hospital; Adult	LVAD; RVAD	Yes
	Case report of a patient with LVAD who developed RV failure and arrhythmia after embolism. Patient was treated with a temporary RVAD and ultimately underwent heart transplant.
Mulukutla et al(42)	2015	Case Report	North America	1	Out-of-Hospital; Adult	BiVAD	No
	Case report of patient with BiVAD who developed sustained, unstable VT who underwent VT ablation.
Wilson et al(43)	2014	Case Report	Canada	1	In-hospital; Adult	LVAD	No
	Case report of a single patient with recurrent, brief cardiac arrest and loss of consciousness iso LVAD and fused aortic valve. Underwent aortic valve replacement with improvement.
Shinar et al(44)	2014	Case Series	North America	8	Both In-and-Out of hospital; Adult	LVAD	Yes
	Case series of 8 patients who had LVADs and underwent chest compressions with a focus on cannula dislodgement. Eight patient records were reviewed revealing no apparent dislodgement after receiving chest compressions. In all cases with return of effective circulation, post-arrest pump flows were reported as stable. Three patients underwent autopsy, with no device dislodgement found—including an autopsy for a patient who underwent 2.5 hours of chest compressions. 6 of 8 (75%) patients had return of effective circulation and 4 patients (50%) survived with good neurologic outcomes.
Cubillo et al(45)	2014	Case Report	North America	1	Out-of-Hospital; Adult	LVAD	Yes
	Case report of emergency repair of an LVAD driveline that was accidentally transected resulting in cardiac arrest. Chest compressions were initiated by a bystander and then continued by paramedics. Patient was taken to the emergency department where LVAD flows were restored, however patient had sustained substantial neurologic injury.
Garg et al(46)	2014	Case Series	North America	16	In-hospital; Adult	LVAD	No
	Case series of 16 patients with continuous-flow LVADs who suffered in-hospital cardiac arrest. 9 patients (56.3%) received chest compressions and 2 (22.2%) of those who received chest compressions survived. 4 of 9 patients (44.4%) who received chest compressions had delays of at least 2 minutes before chest compression initiation. As compared to a non-LVAD cardiac arrest cohort, time to initiation of chest compressions was substantially longer.
Haglund et al(47)	2014	Case Report	North America	1	In-hospital; Adult	LVAD	No
	Case report of a patient post-operative day 7 from LVAD implantation with acute hyperactive delirium with power source disconnection from his LVAD leading to cardiac arrest. He was found unresponsive and cyanotic. LVAD power was restored with improved perfusion, though low flow alarm continued. Chest compressions were not provided.
Duff et al(48)	2013	Case Report	North America	2	In-hospital; Pediatric	LVAD; BiVAD	No
	Case report of cardiac arrest in two pediatric patients with ventricular assist devices. Patient 1 involved LVAD failure and circulatory arrest resulting from acute pulmonary hypertension triggered by post-anesthetic hypercarbia. Patient 2 involved episodic hypoperfusion.
Schweiger et al(11)	2012	Case Report	Europe	1	Out-of-Hospital; Adult	LVAD	Yes
	Case report of 2 patients with LVADs—one of which suffered acutely altered perfusion resulting in EMS response. Paramedics unsure of whether to do CPR and wife called VAD specialist. CPR advised but patient’s wife declined.
Brenyo et al(49)	2011	Case Report	North America	1	Out-of-Hospital; Adult	LVAD	No
	Case report of patient with LVAD who suffered cardiac arrest with ventricular fibrillation and was defibrillated. He was comatose and treated with therapeutic hypothermia. After rewarming, had neurological recovery other than amnesia around the arrest event.
Rottenberg et al(50)	2011	Case Report	North America	1	In-hospital; Adult	LVAD	Yes
	Case report of patient sustaining cardiac arrest during redo sternotomy for LVAD exchange. Abdominal chest compressions were performed to avoid damage to inflow cannula.
Andersen et al(51)	2009	Case Series	Europe	3	Out-of-Hospital; Adult	LVAD	Yes
	Case series of 23 patients with HeartMate II LVADs describing the incidence of VT/VF during 266 total months of follow up. They noted an incidence of 52%, with external defibrillator or ICD shock in 8 patients and significant hemodynamic instability in 3 patients.

Identification of Altered Perfusion and Cardiac Arrest and Potential Delays in Resuscitation

Several studies highlighted the challenges of identifying durable MCS supported patients with acutely altered perfusion and cardiac arrest.(2, 3, 11–15) These studies documented complexity resulting from the expected pulselessness in continuous flow LVAD supported patients who do not have native heartrates. Other challenges described included the added difficulty of measuring blood pressure when electronic noninvasive blood pressure monitors are unreliable as well as challenges in determining adequate perfusion.

Delays in the initiation of chest compressions were documented in several reports.(13, 16–18) In one study of hospitalized patients, 4 of 9 (44.4%) patients with LVADs who suffered cardiac arrest had delays of over two minutes prior to starting chest compressions.(16) In addition, delays in chest compressions were longer for patients supported by LVADs as compared to a control cohort of cardiac arrest patients without durable MCS. In a report of 4 pediatric inpatients with durable MCS, delays in chest compressions were highlighted as a potential contributor to poor outcomes.(13) In the out-of-hospital settings, case reports highlighted delays in chest compressions due to paramedics being unsure of whether chest compressions can be administered to MCS supported patients.(11) In one study, the most common reason clinicians provided for not performing chest compressions in MCS supported patients with acutely impaired perfusion as a result of cardiac arrest was a belief that chest compressions are contraindicated in patients with LVADs.(19)

As a result of difficulty in the assessment of perfusion in durable MCS supported patients, as well as rescuer uncertainty with respect to the safety of chest compressions, a number of studies and reviews proposed algorithms for the resuscitation of patients with durable MCS.(2, 3, 13, 16, 20)

Performance of Chest Compressions vs. No Chest Compressions with Respect to Patient Outcomes

Three studies included MCS supported patients who experienced cardiac arrest and reported outcomes in those who received chest compressions and those who did not. In the largest study leveraging the United States National Inpatient Sample, 578 patients with LVADs who had a cardiac arrest were identified.(21) Compared with patients who did not receive chest compressions, patients who received chest compressions had higher in-hospital mortality (74.3% vs 55%), and receipt of CPR was associated with worse outcomes after multivariate adjustment. While this study is the largest exploration of cardiac arrest in LVAD supported patients to date, it is limited by use of administrative, in-hospital data only which may lead to incomplete capture and misclassification of events. Risk adjustment may also be incomplete as only variables available in the database could be adjusted for in the multivariate model. Patients who received chest compressions may also have had higher illness severity at baseline, including higher sepsis rates. These potential biases are noted by the study authors, who also highlight a potentially beneficial impact of chest compressions with respect to progression to brain death.

In another study of 16 patients who experienced cardiac arrest while supported by an LVAD, 9 (56.3%) received chest compressions.(16) Two of 9 (22.2%) patients who received chest compressions survived to discharge as compared to 3 of 7 (42.9%) who did not receive chest compressions. No specific comparison was made in the study between patients who did and did not receive chest compressions and adjustment for potential confounders was not performed. Both patients who received chest compressions and survived to hospital discharge had chest compressions initiated within 2 minutes of cardiac arrest, with non-survivors having longer delays to chest compressions. In a final retrospective cohort study including 58 LVAD supported patients who suffered cardiac arrest at a single center, 24 received chest compressions (41.3%). There was no difference in hospital survival (54% in those receiving chest compressions vs. 58% in those not receiving compressions), however those receiving chest compressions had longer resuscitation times and a higher rate of post-arrest neurologic injury (42% vs. 26%).(19)

Outcomes Among Patients who have a Cardiac Arrest and Receive Chest Compressions

A total of 226 patients across 11 studies and two published scientific abstracts sustained cardiac arrest while supported by durable MCS and underwent chest compressions (Table 2). Of this group, 173 (76.5%) were identified in a single study using the National Inpatient Sample, as detailed above.(21) Common causes of cardiac arrest resulting in need for chest compressions included device thrombus, cardiac tamponade, accidental disconnection, and driveline failure. Outcomes reported for patients who experienced cardiac arrest and received chest compressions were not standard across studies. Overall, 71 (31.4%) patients were reported as having a favorable outcome.

Table 2:

Studies Including Patients who Received Chest Compressions

Study	Number of Patients Receiving Chest Compressions	Device	Duration of Implantation prior to Chest Compressions	Cause of arrest	Outcome	Duration of Chest Compressions	Documentation of MCS Dislodgement or other Complication
Senman(19)	24	HVAD, Heartmate 2, Heartmate 3	See reference	See reference	Hospital survival, survival with good neurologic outcome	See reference	None
Theeuwus et al(9)	1	Heartmate 3	1.5 years	Unknown	ROSC obtained	2+ hours	None
Doita et al(27)	1	Heartware HVAD	1 year	Thrombosis	Expired in hospital after identification of hypoxemic ischemic encephalopathy	Not reported	None
Barssoum et al(21)	578	Unknown	Non-index admission	See reference	Hospital mortality	See reference	None
Esangbedo et al(29)	4	Patient 1 Heartmat e-3; Patient 2 Jarvik 2015 LVAD & PediMag RVAD; Patient 3 HeartMate 3; Patient 4 Heartware HVAD	Patient 1 10 days; Patient 2 6 days; Patient 3 9 days; Patient 4 14-days	Patient 1 cardiac tamponade; Patient 2 accidental disconnection; Patient 3 hemorrhage; Patient 4 cardiac tamponade	Patient 1 hypoxemic ischemic encephalopathy and death; Patient 2 good neurologic outcome and transplantation; Patient 3 good neurologic outcome; Patient 4 good neurologic outcome	Patient 1 15 minutes; Patient 2 4 minutes; Patient 3 2 minutes; Patient 4 2 minutes	None
Iwashita et al(32)	1	Heartmate 2	2 years	Battery depletion	Hypoxemic ischemic encephalopathy and death	120 minutes	None
Saito et al(34)	1	Jarvik 2000 LVAD	401 days	Unknown	Good neurologic outcome and transplantation	Not reported	None
Ornato et al(37)	1	Not reported	Not reported	Not reported	Return of spontaneous circulation	Not reported	None
Yuzefpolskaya et al(39)	1	Heartmate 2	8 days	Unknown	Hypoxemic ischemic encephalopathy and death	30 minutes	None
Shinar et al(44)	8	See reference					None
Cubillo et al(45)	1	HeartWare LVAD	1.5 years	Driveline transection	Hypoxemic ischemic encephalopathy and death	Not reported	None
Garg et al(46)	9	See reference					None
Retherford et al(10)	1	Heartmate 2	3 years	Fractured driveline	Good neurologic outcome	30 minutes	None

Device Dislodgement and Other Complications of Chest Compressions

No study reported dislodgement or other complications related to MCS device function after chest compressions. In one study, LVAD dislodgement after chest compressions was specifically considered through a combination of pump function assessment, imaging, and autopsy.(14) In that study including 8 patients who received chest compressions (all of whom were at least 50 days removed from LVAD implantation), no cases of dislodgement were identified. Three of the 8 patients underwent autopsy and the LVAD device was noted to be in appropriate position. This included one patient who underwent 2.5 hours of chest compressions. In one published abstract, an LVAD supported patient suffered cardiac arrest and was transported via air ambulance to an advanced care center while receiving chest compressions with a mechanical chest compression device. No dislodgement of the MCS device was identified.(9) Similarly, two patients in a case series undergoing mechanical chest compressions had no device dislodgment.(19) Five patients (4 children and 1 adult) received chest compression within 14-days of device implantation, without note of device dislodgment and all survived the initial resuscitation attempt.(13, 20)

Discussion:

In this scoping review of the acute resuscitation of durable MCS supported adults and children with acutely altered perfusion and cardiac arrest, 32 studies were identified including 11 studies in which patients received chest compressions. Studies were generally small including a number of case reports and case series. Reviewed studies highlighted challenges in the identification of durable MCS supported patients in need of acute resuscitation owing to absent pulse in continuous-flow LVAD supported patients and in complexity in blood pressure measurement and perfusion assessment. Approximately a quarter of patients who receive chest compressions while supported by MCS survive their hospitalization and no cases of device dislodgement were identified—including in patients who underwent extended resuscitation attempts. Although three retrospective, observational studies found that LVAD supported patients who suffer cardiac arrest and receive chest compressions have lower survival rates than those who do not receive chest compressions, those comparisons should be considered hypothesis generating only as a result of potentially incomplete adjustment for confounders and the possibility of misclassification wherein many of those who did not receive chest compressions suffered acutely impaired perfusion without cardiac arrest, resulting in better outcomes. Caution should be taken in the interpretation of these studies when applying their findings to clinical practice.

An AHA scientific statement regarding cardiopulmonary resuscitation in patients with mechanical circulatory support was published in 2017 and was based on a consensus processes executed in 2014 and 2015.(3) A literature search was conducted in support of the scientific statement although the methodology of the search is not fully described. The authors provide a consensus derived algorithm for the first responder approach to the patient with LVADs and total artificial hearts who suffer cardiac arrest. Since the literature search performed in support of the scientific statement was concluded, several relevant studies have been published which largely support the scientific statement recommendation to begin chest compressions in LVAD supported patients without signs of life without concern for device dislodgement. A recent guideline from the British Societies LVAD Emergency Algorithm Working Group provides recommendation for the approach to emergencies in patients with LVADs.(2) Following a narrative literature review, the authors of the guideline created recommendations by consensus of experts. This effort culminated in guidance to delay chest compressions for a maximum of two minutes while attempts at restoration of LVAD function are made.

The resuscitation algorithms proposed by the AHA and the British Societies working group, in addition to those published by device manufacturers and others (4, 5), rely on an assessment of organ perfusion, capnometry, and mean arterial pressure in guiding the approach to the durable MCS supported patient with acutely altered perfusion. These algorithms are important, as many studies in this scoping review highlighted delays in the initiation of chest compressions. Assessment and comparison of these algorithms is beyond the scope of this review; however first responders should be familiarized with a systematic approach to assessment of perfusion and resuscitation of patients with LVADs and other durable MCS to minimize delays in chest compression initiation. This is especially important given the increasing prevalence of LVADs in the community and among inpatients.

The risk of device dislodgement is an established concern among clinicians resuscitating patients with durable MCS. In survey studies, clinicians have reported substantial variability in clinical practice with respect to the administration of chest compressions to patients supported by LVADs.(22) In a case report included in this review, one patient had delayed chest compression initiation as rescuers were unsure whether chest compressions could be safely performed in an LVAD supported patient.(11) In this review of 202 durable MCS supported patients who received chest compressions, no documentation of device dislodgement or malfunction was found. This includes several studies and case reports that specifically evaluated for dislodgement. Two patients identified in this review underwent over 2 hours of chest compressions each without device dislodgement.(9, 14) Three patients received chest compressions with mechanical chest compression devices without identified dislodgement.(9, 19) In a survey study, of 62 respondents one reported a case of device dislodgement with chest compressions however no additional details are provided.(22) These findings should allay some concerns among first responders. Most patients who received chest compressions in this review were at least 50 days removed from durable MCS placement and all had LVADs or BiVADs. The safety of chest compressions among patients with other forms of MCS and those more proximal to LVAD placement is not fully elucidated by this review. A publication bias against the publicization of events where a device is in fact dislodged cannot be excluded.

Several important knowledge gaps were identified in the 2017 AHA scientific statement and are again highlighted by this scoping review. These gaps include how best to identify MCS supported patients outside of the hospital, the approach to assessment of acutely altered perfusion in durable MCS supported patients, the efficacy of chest compressions in the case of cardiac arrest, and the approach to specific populations such as pregnant patients. Given the rarity of cardiac arrest in MCS supported patients, multicenter surveillance studies will be needed to answer these critical questions. While this is the first scoping review to explore the acute resuscitation of durable MCS supported patients, the review is limited by the inclusion of only English Language literature.

Conclusions:

Evidence to guide the resuscitation of durable MCS supported patients with acutely altered perfusion and cardiac arrest is limited. Delays in resuscitation resulting from rescuer uncertainty regarding the safety of chest compressions, along with challenges in the identification of acutely altered perfusion and cardiac arrest in durable MCS supported patients have been reported. No occurrences of device dislodgement following chest compressions were identified, such that the risk of harm from timely CPR when required in these patients appears to be low.

Highlights:

• The prevalence of durable mechanical circulatory (MCS) devices in the community is increasing and rescuers will increasingly encounter MCS supported patients in need of acute resuscitation.

• The evidence guiding the approach to recognition and treatment of acutely impaired perfusion in MCS supported patients has not been previously descried.

• This scoping review provides insight into the existing literature around resuscitation for durable MCS supported patients and may support international recommendations.