An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Health Technology Review
People who experience cardiac arrest need immediate care. Cardiopulmonary resuscitation (CPR), performed manually by a human rescuer, is a first line of treatment.
Automated chest compression devices, also known as mechanical CPR devices, are designed to deliver high-quality, consistent compressions and may be of particular interest in settings with limited human rescuers or long travel times in emergency vehicles. However, whether these devices are more clinically or cost-effective than manual chest compressions is unclear.
To inform decisions about the use of automated chest compression devices for chest compressions (described as “automated chest compressions” in this report) compared to manual chest compressions, Canada’s Drug Agency (CDA-AMC) conducted a Rapid Review to identify and summarize the literature about the clinical and cost-effectiveness of automated chest compression devices in people of any age. We also aimed to identify evidence-based recommendations for the use of automated chest compression devices.
We searched key resources, including journal citation databases, and conducted a focused internet search for relevant evidence published since 2020. Two reviewers screened articles for inclusion based on predefined criteria, and 1 reviewer critically appraised the included studies and narratively summarized the findings.
We identified 4 systematic reviews (SRs) that examined the clinical effectiveness or safety of automated chest compressions via AutoPulse or Lund University Cardiopulmonary Assist System (LUCAS) devices compared to manual chest compressions, and 2 guidelines with recommendations for the use of automated chest compression devices overall. We did not find information for other devices licensed for sale in Canada. We did not find economic evaluations on the cost-effectiveness of automated chest compression devices.
Clinical evidence showed mixed results on survival, neurologic outcomes, and return to spontaneous circulation between automated chest compressions and manual chest compressions, and a potential increase in harms with the use of these devices. However, these findings are based on heterogeneous evidence of variable quality and should be interpreted with caution.
Evidence-based guidelines do not recommend the routine use of automated chest compression devices. They indicate that these devices could be applied under specific circumstances, such as when high-quality compressions are impractical or a danger to health care workers, provided professionals are trained and have experience with the device.
Evidence was largely based on studies conducted outside of Canada, making the generalizability of the evidence unclear. One primary study found in 3 of the 4 SRs had a population from Canada, and no other clinical evidence was from Canada. No evidence-based guidelines were found from Canadian organizations.
Most of the evidence did not include details about study participant demographics or dimensions of diversity or information specifically for rural, remote, territorial hospital, nurse-led hospital, small community, or tertiary care settings. The applicability of the evidence is unknown, including the potential benefits or harms in people with different sexes or genders; different ethnic, religious, educational, socioeconomic, or cultural backgrounds; or with limited access to health care services or in resource-limited settings.
Health care professionals can consider following the recommendations from evidence-based guidelines, which do not encourage the routine use of automated chest compression devices, except under specific circumstances. This aligns with the heterogeneous clinical evidence identified from the included SR.
Because there was no evidence found on cost-effectiveness or information on the clinical effectiveness for people with different sexes or genders; people from different ethnic, religious, educational, socioeconomic, or cultural backgrounds in Canada, or contexts such as rural, remote, or low-staff settings, decision-makers may wish to consider whether the potential benefits and harms from the evidence in this report are applicable to their local context before more high-quality evidence for Canadian settings is available.
Ethnicity: “A socially defined category or membership of people who may share a nationality, heritage, language, culture, and/or religion.”1
Equity-deserving: “Groups of people who have been historically disadvantaged and under-represented. These groups include but are not limited to the 4 designated groups in Canada – women, racialized groups, Indigenous Peoples, and people with disabilities — and people in the 2SLGTBQ+ community/people with diverse gender identities and sexual orientations.”2
Gender: “Gender can refer to the individual and/or social experience of being a man, a woman, or neither. Social norms, expectations and roles related to gender vary across time, space, culture, and individuals.”3
Health equity: “Equity is the absence of unfair, avoidable or remediable differences among groups of people, whether those groups are defined socially, economically, demographically, or geographically or by other dimensions of inequality (e.g., sex, gender, ethnicity, disability, or sexual orientation). Health is a fundamental human right. Health equity is achieved when everyone can attain their full potential for health and well-being. Health and health equity are determined by the conditions in which people are born, grow, live, work, play and age, as well as biological determinants. Structural determinants (political, legal, and economic) with social norms and institutional processes shape the distribution of power and resources determined by the conditions in which people are born, grow, live, work, play and age.”4
Indigenous: The term “Indigenous” is used to describe people and communities who identify with and have historical claim as “First Peoples” who have been on these lands (colonially known as Canada and the US) since time immemorial. Indigenous Peoples within Canada often refers to people who belong to First Nations, Inuit, and Métis communities; however, we acknowledge that all Indigenous communities are widely heterogenous having distinct social, economic, and political systems, as well as distinct language(s), cultures, and beliefs. In addition, we acknowledge that the term “Indigenous” may also not align with how community members self-identify and that their multilayered identities may include (but are not limited to) “nation,” “territory,” “family,” “band,” or “Native.” The term “Indigenous” is commonly used within the context of Canada, but some of the identified literature uses a variety of terms commonly accepted within their jurisdiction (i.e., the US, Australia).5,6
PROGRESS-Plus: “An acronym used to identify characteristics that stratify health opportunities and outcomes. PROGRESS refers to: place of residence, race/ethnicity/culture/language, occupation, gender/sex, religion, education, socioeconomic status, social capital. Plus refers to: (1) personal characteristics associated with discrimination (e.g., age, disability), (2): features of relationships (e.g., smoking parents, excluded from school), (3) time-dependent relationships (e.g., leaving the hospital, respite care, other instances where a person may be temporarily at a disadvantage).”7
Racialized: “A person or group of people categorized according to ethnic or racial characteristics and subjected to discrimination on that basis.”8
Racism: “Racism is the belief that a group of people are inferior based on the colour of their skin or due to the inferiority of their culture or spirituality. It leads to discriminatory behaviours and policies that oppress, ignore or treat racialized groups as ‘less than’ non-racialized groups. One result of racism is substantive inequity — a state in which racialized groups do not have equitable outcomes, or equitable opportunities, to non-racialized groups. This is systemic racism — where acceptance of these discriminatory and prejudicial practices has become normalized across our society and institutions.”9
Remote: “Statistics Canada has developed an index of remoteness that ranges from 0 (least remote) to 1 (most remote) for every municipality (census subdivision) in Canada.… The index of remoteness assigns a value to each municipality based on geographic proximity to urban areas (service and population centres) and on the population size of those urban areas. Remoteness is an important determinant of socioeconomic and health outcomes and is consequently an essential indicator for delivery of policies and programs.… Although most municipalities in the three territories are classified as remote, they are also home to urban centres, which are concentrated in 12 municipalities: Whitehorse in Yukon; Yellowknife, Hay River, Inuvik and Fort Smith in Northwest Territories; and Iqaluit, Rankin Inlet, Arviat, Baker Lake, Cambridge Bay, Kugluktuk and Gjoa Haven in Nunavut.”10
Rural: “The rural area of Canada is the area that remains after the delineation of population centres using current census population data. Within rural areas, population densities and living conditions can vary greatly. Included in rural areas are: small towns, villages, and other populated places with less than 1,000 population according to the current census; rural areas of census metropolitan areas and census agglomerations that may contain estate lots, as well as agricultural, undeveloped and non-developable lands; agricultural lands; remote and wilderness areas.”11
Sex: “The classification of people as male, female, or intersex. Sex is typically assigned at birth and is based on an assessment of one’s reproductive systems, hormones, chromosomes, and other physical characteristics.”1,3
What is the clinical effectiveness and safety of automated chest compressions compared to manual chest compression for people of any age requiring chest compressions?
What is the cost-effectiveness of automated chest compressions compared to manual chest compression for people of any age requiring chest compressions?
What are the evidence-based guidelines regarding the use of automated chest compression devices for people of any age requiring chest compressions?
Cardiac arrest is when cardiac activity stops suddenly in a person who then loses circulation, stops breathing, and becomes unresponsive.12 There are several causes of cardiac arrest, such as coronary artery disease, heart failure, a heart attack, hypertrophic cardiomyopathy (i.e., heart enlargement), blood clots in the lungs, injuries, drowning, or poisoning, including poisoning related to the ongoing drug poisoning crisis.13-15 Cardiac arrest can be fatal if care is not provided immediately by, for example, administering CPR, cardiac pacing (i.e., using a pacemaker to stimulate electric heart activity), defibrillation (i.e., sending shocks to the heart to restore regular rhythm), or cardioversion (i.e., electrical treatment for irregular heartbeats [arrythmias]).12,13,16,17 CPR is a first line of treatment performed manually by a human rescuer who compresses the patient’s chest to help blood flow to their organs.12,18
A 2024 report by the Heart and Stroke Foundation of Canada revealed that there are 60,000 cardiac arrests in Canada per year, that 10% of people survive out-of-hospital cardiac arrest (OHCA), and that doing CPR and using an automated external defibrillator can double a patient’s chance of survival.14 This report also indicated that remote, rural, and Indigenous communities in Canada have challenges accessing emergency and other medical services for cardiac arrest.14
Several health inequities exist for those who need cardiac arrest treatment, and this can depend on their geographic location, gender, race, perceived socioeconomic status (SES), body size, and education. For example, there may be fewer resources for those living in rural and remote areas, including Indigenous communities in northern Canada, where there may be no access to emergency services, fewer staff and equipment, and larger distances to travel to access health care services.19
A 2024 scoping review20 of global data (which did not clearly distinguish between sex and gender) showed that, in 59% of studies, women were less likely to receive bystander CPR in public compared to men, and more or equally likely to receive bystander CPR in residential settings; there was a reluctance to assist women in Western countries because of gender stereotypes, perceived frailty of women, pregnancy, chest exposure of women, oversexualization of women’s bodies, or assumptions that women are unlikely to have cardiac arrest.
Other studies have shown that racialized people and those with lower SES are more likely to experience OHCA and less likely to survive after being discharged from hospitals, compared to people who are white and who have higher incomes.21 Data from the US has shown that bystander CPR is less likely to be performed for those who are racialized or perceived to have lower income or education, which may in part be related to structural racism, implicit bias, cultural barriers, a lack of CPR training, poor defibrillator access, or fear of financial or legal ramifications.21,22
First developed in the 1960s, automated chest compression devices, also referred to as mechanical chest compression devices or mechanical CPR devices, were designed to deliver high-quality chest compressions to those experiencing cardiac arrest as an alternative to receiving chest compressions from a human rescuer.23,24 This may be needed if there is a lack of human personnel, human rescuers become fatigued or experience challenges in moving emergency vehicles, prolonged CPR is required, or rescuers have other tasks to perform.23 There are now devices that not only assist with chest compressions but can also monitor physiological data from patients and combine this with algorithmic artificial intelligence to make adjustments or provide feedback, and there are robots that can conduct chest compressions while also checking vital signs.25,26
The automated chest compression devices currently used in the health care system are generally categorized as load distributing band devices, which have a wide band that encircles the patient’s chest and constricts to deliver compressions, or piston devices that use a plunger to depress the sternum and in some cases include a suction cup to recoil the chest to its original position.27,28 Most of these devices are for use in adults or people who the device can fit around.29,30
As of February 2025, we identified the following automated chest compression devices that are licensed for sale in Canada as Class II medical devices:31
AutoPulse, load distributing (ZOLL Medical Corporation)32
EASY PULSE, combination of load distributing and piston33 (SCHILLER Americas Inc.)34
the Lifeline ARM, piston (Defibtech LLC)35
LUCAS devices such as LUCAS 2 and LUCAS 3, piston (Stryker Medical)36
Other devices might be licensed in Canada but were not identified.
Given that cardiac arrest is a concern in Canada with a 10th of patients surviving it out of hospital, and that there are challenges in access to emergency services for people living in rural, remote, and/or northern communities or where there is limited staff, health care workers are interested in understanding whether automated chest compression devices can provide an alternative clinical and cost-effective solution when manual CPR is difficult. A previous CDA-AMC report published in 2008 concluded that there was no evidence of benefit from the use of mechanical devices, and they did not find cost-effectiveness evidence or evidence-based guidelines.37 SRs and meta-analyses (MAs) published by others have suggested that these devices may not provide any benefit compared to manual compressions, and there is concern that their use may be associated with increased compression-induced injuries.38,39 There is substantial research on the use of these devices, but it is unclear whether this previous literature contains information specifically for rural, remote, and limited-staff settings and whether there has been new evidence published about cost-effectiveness or evidence-based guidelines released. Therefore, the current report aims to review the clinical and cost-effectiveness evidence on automated chest compression devices and summarize any recommendations on automated chest compression devices for people of any age.
To support decision-making on the use of automated chest compression devices, we conducted a Rapid Review to identify, summarize, and critically appraise available evidence on the clinical effectiveness and cost-effectiveness of automated chest compressions compared to manual compressions for people of any age. We also summarized the related guideline recommendations available for this patient population.
An information specialist conducted a customized literature search, balancing comprehensiveness with relevancy, of multiple sources and grey literature on February 5, 2025. Two reviewers screened citations and selected studies based on the inclusion criteria presented in Table 1, and 1 reviewer critically appraised and narratively summarized the included studies. Appendix 1 presents a detailed description of methods and selection criteria for included studies.
Selection Criteria.
This report includes 4 SRs40-43 with clinical effectiveness information, and 2 guidelines44,45 with recommendations for the use of automated chest compression devices. No studies were found for cost-effectiveness. The authors of 1 SR40 planned to synthesize evidence on cost-effectiveness. However, no cost-effectiveness results were presented, and the authors did not disclose the reasons (i.e., lack of identified evidence or other factors).
Appendix 1 presents the PRISMA46 flow chart of study selection. Appendix 6 includes additional references of potential interest, including an SR with insufficient quantitative information for data extraction and guidance documents (not evidence-based).
Appendix 2 contains detailed characteristics of included studies.
We identified 4 relevant SRs40-43 with MAs. Each SR searched at least 3 electronic literature databases from database inception to December 2023 in 2 SRs,40,43 to May 2023 in 1 SR,42 and to May 2020 in 1 SR41 (i.e., clinical evidence informing this report is based on evidence published before 2024). The SRs40-43 included data from 58 studies with substantial overlap of included primary studies across the SRs. One SR43 uniquely contributed approximately one-third of the included studies (21 unique studies) and 1 SR,41 which focused exclusively on safety outcomes (i.e., injuries), had very little overlap with the other SRs (10 unique studies). The remaining 2 SRs40,42 have substantial overlap with Zhu and Fu43 but have been included in this Rapid Review because they contributed unique studies and outcomes. A citation matrix illustrating the degree of primary study overlap is presented in Appendix 5.
Two SRs40,43 focused on adult patients experiencing OHCA, 1 of which was limited to nontraumatic cardiac arrest and also included 1 study with patients experiencing in-hospital cardiac arrest (IHCA).40 Another SR42 was also limited to OHCA and did not specify study eligibility limits based on participant age. The final included SR41 was limited to adult patients with cardiac arrest and no stated limits on the setting of that arrest. The studies included in this SR41 differed from those in other SRs: 9 of 11 included studies were in populations of nonsurvivors of cardiac arrest (compared to other SRs that included both survivors and nonsurvivors of cardiac arrest), and the outcomes were measured by autopsy or postmortem CT. Two SRs included studies comparing any automated chest compressions to manual chest compressions40,41 and 2 SRs were limited to specific devices: LUCAS41 and AutoPulse.40,41 All SRs40-43 included interventional studies (e.g., randomized controlled trials [RCTs]) and both prospective and retrospective observational studies. One SR43 presented subgroup analyses for different versions of LUCAS (i.e., LUCAS, LUCAS 2, LUCAS 3); the other SRs did not provide information regarding device versions.
The 58 relevant studies included in the 4 SRs40-43 were conducted in more than 20 countries across Asia, Oceania, Europe, and North America, were published between 2005 and 2023, and included from 30 to more than 30,000 participants. Three of the SRs40,42,43 included 1 primary study with participants from Canada and the US; when reported, all other studies were in populations outside of Canada. The populations reflected the review eligibility criteria, with the caveat that the 2 SRs included patients experiencing IHCA as a population of interest yet identified limited studies with this population (n = 2).40,41 No SR reported, for all included studies, the settings where automated devices were used for patients experiencing OHCA but, when reported, they included emergency departments and paramedic services (ambulance, helicopter transport). Twenty relevant included studies focused on AutoPulse, 31 focused on LUCAS, and 7 focused on both AutoPulse and LUCAS. No SRs had results for Lifeline ARM, and 1 SR43 included a primary study with EASY PULSE but did not provide device-specific outcomes.
Larik et al.42 reported the following participant characteristics, presenting this by study arm: mean age (ranging from 63 to 80 years), sex or gender (ranging from 54% to 83% males, where “males” was not specified as referring to sex or gender), patients with shockable cardiac rhythm (ranging from 7% to 64%), patients with witnessed cardiac arrest (ranging from 34% to 96%) and patients receiving bystander CPR (ranging from 0% to 57%). The other 3 SRs40,41,43 did not report these characteristics, and none of the SRs provided participant information for other PROGRESS-Plus7 criteria such as place of residence, disability, race, ethnicity, culture, language, occupation, religion, education, SES, or social capital.
No SRs40-43 extracted information on who administered the devices or conducted CPR, or the methods used to train staff for these procedures.
No SRs reported funding sources of included primary studies.
Clinical outcomes to address the research questions across the 4 SRs40-43 included:
The 2 guidelines provided recommendations for automated chest compression device use on adults.44,45 Both44,45 were from groups under the International Liaison Committee on Resuscitation (ILCOR) and had similar processes for guideline development and the involvement of ILCOR; these guidelines were for both IHCA and OHCA.
No guidelines had recommendations for rural settings, remote settings, territorial hospitals, nurse-led hospitals, small communities, or tertiary care.
Appendix 3 contains details about the strengths and limitations of the included SRs and guidelines.
We noted several strengths. SRs40-43 had clearly defined objectives, eligibility criteria, and search methods. All SRs40-43 searched at least 3 major databases, 2 SRs41,43 reported additional search methods, and all SRs provided keywords or full search strategies in the report or supplementary materials. At least 2 reviewers were involved in the study selection in all SRs, 1 SR40 reported duplicate data extraction, and another43 reported single data extraction with verification. Gao et al.41 reported funding by various grants, 3 SRs40,42,43 reported no funding for the research, and all SRs40-43 stated that the authors declared no financial or other conflicts of interest related to the research. While no SR included a list of excluded studies along with reasons for exclusion, all provided the number of studies excluded at full-text screening and broad reasons for exclusions. All SRs40-43 reported satisfactory techniques for assessing risk of bias in individual studies; this was completed by at least 2 reviewers in 2 SRs.42,43
Possible concerns were also noted across SRs. Authors of 2 SRs42,43 did not mention a review protocol and while the authors of the others40,41 mentioned registering a review protocol, in both cases the protocol was registered after review conduct began, making it difficult to confirm the degree to which methods were prespecified. Additionally, the comprehensiveness of the search strategies and search methods were unclear because SRs with similar eligibility criteria had incomplete overlap of their included primary studies. At least 2 SRs41,42 also included outcomes in their electronic search strategies, which may have limited their ability to identify relevant studies, especially for AEs, because these are inconsistently reported in titles and abstracts.
The degree, or lack of clarity about the degree, of consistency across study populations and the potential effect of heterogeneity is a primary concern across these SRs. The authors of each SR extracted and presented key variables, including study design, numbers of participants, and broad settings. However, other than 1 SR,42 reviews did not report important variables of included primary studies such as age, gender or sex distributions, or the proportion of patients with a witnessed cardiac arrest or shockable rhythm. Authors also presented little information on health equity variables, including the settings (e.g., urban or rural) or participants’ dimensions of diversity based on PROGRESS-Plus criteria.7 This lack of information makes it difficult to assess the internal validity of the MA results and the applicability of the results to varied populations. No SR presented the sources of funding for included primary studies, inhibiting the ability to assess the risk of sponsorship bias across the evidence base.
Importantly, MAs frequently combined controlled trials and observational studies, had studies that were deemed to be at a low and high risk of bias, and had inconsistent outcome definitions across studies (i.e., outcome definitions for complications and neurological outcomes were not provided for each included study). As evidenced by the study-level data from Larik et al.,42 imbalances in key factors often existed between study arms, and no SR with MAs reported using adjusted effect measures from observational studies to account for imbalances. Furthermore, many of the MAs noted high levels of heterogeneity across their included studies (both clinical and statistical). While subgroup and sensitivity analyses were conducted (by device, study design, risk of bias), statistical heterogeneity often remained moderate or high and largely unexplained. Finally, 3 of the SRs40,41,43 investigated the potential presence and impact of publication bias, but concerns exist with some of these assessments. Overall, while there appear to be strengths in the conduct of these SRs, the validity and applicability of their MA results is unclear.
Both guidelines44,45 clearly outlined their scope and purpose, indicated who the target users and intended population were, sought the views of the target population, explained how SRs were conducted to inform the guideline, described how the guideline was validated, provided the methods for forming recommendations, provided specific recommendations that were easy to identify with options for managing the health issue, provided explicit links between the recommendations and the supporting evidence, noted resource implications of applying the recommendations, described facilitators and barriers to its application, and addressed conflicts of interest of the guideline development group members.
One guideline45 clearly described the criteria for selecting evidence, and 1 guideline44 did not report this information. One guideline44 clearly described the strengths and limitations of the body of evidence that linked to the strength of the recommendations, and the other45 did not. A procedure for updating the guideline or monitoring the criteria was provided for 1 guideline44 and was unclear for the other.44
Neither of the guidelines provided tools to support the application of recommendations in practice. It was unclear whether any of the guideline funding bodies influenced its content.
Appendix 4 presents the main study findings.
Four SRs40-43 provided information on the clinical effectiveness and safety of automated chest compressions versus manual compression. There was considerable overlap in the primary studies that were included in these SRs; the pooled estimates from separate reviews thus contain much of the same data (refer to Appendix 5 for details regarding overlap).
Three SRs40,42,43 compared survival outcomes between people who experienced cardiac arrest and received automated chest compressions with those who received manual compressions. MAs, often with a high degree of unexplained heterogeneity and including studies at both a high and low risk of bias, provided mixed results across survival outcomes but suggested that AutoPulse may be associated with similar survival at discharge and potentially increased survival at other time points whereas LUCAS may be associated with a similar or decreased survival at discharge compared to manual compressions. The following outcomes were reported:
Survival to hospital admission: Two SRs40,43 examined device-specific survival to hospital admission — 1 SR43 with MAs and 1 SR40 with primary study results. All but 1 relevant primary study focused on adults with OHCA (in the other, adolescents were eligible for inclusion, but study participant age ranges were not reported), using manual chest compression as the comparator. The overview of the results follows and should be interpreted with consideration of the limitations:
Automated chest compression devices (mix of AutoPulse and LUCAS within the studies):
There was no statistically significant difference in the odds of survival to admission (1 SR with MA of 2 studies).43
AutoPulse:
LUCAS:
Survival to hospital discharge: Three SRs,40,42,43 2 with MAs,42,43 included evidence for this outcome.
Automated chest compression devices (mix of AutoPulse and LUCAS within the studies):
There was no statistically significant difference in the odds of survival to discharge (1 SR with MA of 4 studies).43
AutoPulse:
LUCAS:
There were mixed results for odds of survival to discharge with LUCAS — lower odds (1 SR with MA of 9 studies)43 and no statistically significant difference (1 SR with MA of 7 studies).42
There was no statistically significant difference in the odds of survival to discharge with the use of LUCAS 2 (1 SR with 1 primary study).43
There was no statistically significant difference in the odds of survival to discharge with the use of LUCAS 3 (1 SR with MA of 2 studies).43
30-day survival:
Automated chest compression devices:
No SRs included results for this outcome for studies of mixed devices.
AutoPulse
There were higher odds of survival to 30 days with the use of AutoPulse (1 SR with results from 1 retrospective observational study deemed to be at a high risk of bias).40
LUCAS
No SRs included results for this outcome.
Other survival outcomes:
Two SRs40,42 compared neurologic outcomes between people who experienced cardiac arrest and received automated chest compressions with those who received manual compressions. Results from 1 MA in 1 SR42 were not extracted because it included 1 large ineligible study; thus, the results of the 12 relevant primary studies from that SR are discussed. The SRs defined this outcome as a “favourable neurologic outcome” in 1 SR42 and survivors with “cerebral performance category 1” and “overall performance category 1” in the other SR40 (specific scale is not referenced in the SR) based on results from 1 relevant primary study. Neither SR presented the time points for outcome measurements.40,42 Given the variability of the study designs, outcome measurements, and risk of bias, it is difficult to draw conclusions regarding neurologic outcomes. The following is a summary:
Automated chest compression devices (AutoPulse, LUCAS, or mixed):
AutoPulse:
There were mixed results, ranging from increased to decreased odds of a favourable neurologic outcome with the use of AutoPulse (4 primary studies reported in 1 of 2 SRs).40,42 One of these primary studies (included in 1 SR42) was an RCT with 4,231 participants and was rated at a low risk of bias: it reported no statistically significant difference in the odds of a favourable neurologic outcome with AutoPulse in adults experiencing OHCA.
LUCAS:
Three SRs40,42,43 with MAs compared the odds of ROSC with automated chest compressions with that of manual chest compressions for patients experiencing cardiac arrest; due to variable risk of bias across included studies and high statistical heterogeneity, the MA results should be interpreted with caution.
Automated chest compression devices (AutoPulse, LUCAS, or mixed):
AutoPulse:
LUCAS:
There was no statistically significant difference in the odds of ROSC with the use of LUCAS (2 SRs with MAs of 13 studies in which the device version is not defined,42 and 1543 in which results for LUCAS 2 and LUCAS 3 are considered separately).
There was no statistically significant difference in the odds of ROSC with the use of LUCAS 2 (1 SR with 1 retrospective observational study).43
There was no statistically significant difference in the odds of ROSC with the use of LUCAS 3 (1 SR with MA of 2 studies).43
Two SRs40,41 compared the compression-induced injuries of people who experienced cardiac arrest and received automated chest compressions with those who received manual compressions. One SR41 contributed most of the results for safety outcomes; it was limited to primary studies that compared AutoPulse or LUCAS to manual compressions, and 9 out of 11 included studies were conducted in populations of nonsurvivors of cardiac arrest. The safety outcomes included:
Life-threatening compression-induced injuries41
Skeletal fractures41
Visceral injuries41
Other soft tissue injuries.41
Overall and life-threatening compression-related injuries: Both SRs40,41 included an MA of the odds of compression-related injuries, and Gao et al.41 included an MA of the odds of all life-threatening injuries (not otherwise defined).
Automated chest compression devices (AutoPulse, LUCAS, or mixed):
There were higher odds of any compression-related injury with the use of automated chest compression devices (1 SR with MA of 4 studies).41
There was no statistically significant difference in the odds of a life-threatening injury with the use of automated chest compression devices, but there was high imprecision in the odds ratio estimate (95% confidence interval [CI], 0.53 to 53.16) (1 SR with MA of 3 studies).41
AutoPulse:
There was no statistically significant difference in the odds of any compression-related injury with the use of AutoPulse (1 SR with MA of 4 RCTs).40
LUCAS:
There were no results included in the SR for this outcome for LUCAS.
Skeletal fractures: One SR41 reported outcomes of skeletal fractures.
Automated chest compression devices (AutoPulse, LUCAS, or mixed):
There was no statistically significant difference in the odds of sternal fractures (MA of 11 studies; high statistical heterogeneity) or vertebral fractures (MA of 4 studies; wide CIs) with the use of automated chest compression devices.41
There was a higher odds of posterior rib fractures with the use of automated chest compression devices (MA of 5 studies).41
AutoPulse:
There were no statistically significant differences in the odds of sternal fractures (MA of 4 studies), anterolateral fractures (MA of 3 studies), or vertebral fractures (MA of 2 studies; wide 95% CI) with the use of AutoPulse.41
There were higher odds of posterior rib fractures with the use of AutoPulse (MA of 3 studies).41
LUCAS:
There was no statistically significant difference in the odds of vertebral fractures with the use of LUCAS (MA with 2 studies; wide 95% CI).41
There were higher odds of sternal fractures (MA of 8 studies), rib fractures (MA of 7 studies), and multiple (≥ 3) rib fractures (MA of 3 studies) with the use of LUCAS.41
Visceral injuries: One SR41 with MAs reported outcomes of visceral injuries. As with other analyses, 95% CIs were often wide and the heterogeneity high, even within subgroup analyses.
Automated chest compression devices (AutoPulse, LUCAS, or mixed):
There were no statistically significant differences in the odds of visceral injuries (MA of 3 studies), lung lesions (MA of 8 studies), spleen lesions (MA of 3 studies), or kidney and perirenal lesions (MA of 4 studies) with the use of automated compression devices.41
There were higher odds of heart lesions (MA of 8 studies), liver lesions (MA of 8 studies), and pneumothorax (MA of 9 studies) with the use of automated chest compression devices.41
AutoPulse:
There were no statistically significant differences in the odds of heart lesions (MA of 2 studies), liver lesions (MA of 2 studies), spleen lesions (1 retrospective cohort study), or kidney and perirenal lesions (1 retrospective cohort study; wide CI) with the use of AutoPulse.41
There were higher odds of pneumothorax with the use of AutoPulse (MA of 3 studies).41
LUCAS:
There were no statistically significant differences in the odds of lung lesions (MA of 7 studies), spleen lesions (MA of 2 studies), kidney and perirenal lesions (MA of 3 studies), or pneumothorax (MA of 7 studies) with the use of LUCAS.41
There were higher odds of heart lesions (MA of 6 studies), liver lesions (MA of 7 studies), and lesions of major vessels (MA of 4 studies) with the use of LUCAS.41
Other soft tissue injuries: Gao et al.41 also reported the odds of other soft tissue injuries.
Automated chest compression devices (AutoPulse, LUCAS, or mixed):
There were no statistically significant differences in the odds of hemothorax (MA of 6 studies; high heterogeneity) or retrosternal bleeding (MA of 5 studies) with the use of automated chest compression devices.41
There were higher odds of hemoperitoneum (MA of 3 studies) and skin lesions (MA of 5 studies) with automated chest compression devices.41
AutoPulse:
There were no statistically significant differences in the odds of hemothorax (1 retrospective cohort study) or retrosternal bleeding (1 retrospective study) with the use of AutoPulse.41
There were higher odds of hemoperitoneum (1 retrospective study) and skin lesions (1 retrospective study) with the use of AutoPulse.41
LUCAS:
There were no statistically significant differences in the odds of retrosternal bleeding (MA of 4 studies) or mediastinal hemorrhage (MA of 4 studies) with the use of LUCAS.41
There were higher odds of hemothorax (MA of 5 studies), hemoperitoneum (MA of 2 studies), and skin lesions (MA of 4 studies) with the use of LUCAS.41
In general, both guidelines,44,45 which were based on evidence that was reported as weak44 or for which the overall quality of the evidence was not clearly reported,45 did not recommend the routine use of automated chest compressions, and advised considering using these devices under specific circumstances, such as:
When high-quality manual chest compression is not practical or can be dangerous for the provider44,45
When prolonged CPR is needed for patients in cardiac arrest who have hyperkalemia45
When a patient has coronary thrombosis and no sustained ROSC, and resuscitation is not futile45
When resuscitating and treating possible causes in a catheterization laboratory.45
In special circumstances for which automated chest compression devices were warranted, the guidelines recommended the following considerations for staff:
Using only trained teams who are familiar with the device to minimize interruptions while the device is being used45
Having the provider limit CPR interruptions while using and removing the device.44
These recommendations for automated chest compression devices were often considered in parallel with other health technologies or guidance not relevant for this report; the complete guidance with recommendations for all devices, settings, and circumstances can be found in the guideline publications.44,45
Neither of the guidelines had specific information for rural settings, remote settings, territorial hospitals, nurse-led hospitals, small communities, or tertiary care. However, 1 guideline44 used information from supporting evidence that listed situations with limited personnel or moving ambulances as a risk to providers.
Neither of the guidelines had specific recommendations for different sexes or genders or people with different ethnic, religious, educational, socioeconomic, or cultural backgrounds.
In this Rapid Review, we found evidence on clinical effectiveness and recommendations from evidence-based guidelines; however, we did not find evidence published since 2020 on the cost-effectiveness of automated chest compression devices, representing a gap in knowledge on the value of these devices for the health care system in the current economic climate. The clinical evidence we found focused on AutoPulse and LUCAS devices; there is a gap in SR evidence for other devices such as EASY PULSE and Lifeline ARM. The included guidelines had recommendations for automated chest compression devices overall rather than specific recommendations about individual devices; it is unclear whether recommendations could differ by device type.
Within the clinical evidence, the MAs within the SRs combined primary studies that had diverse study designs, that had a range of low to high risk of bias in the primary studies, and whose effect estimates lacked adjustments for key confounders. The MAs often had high heterogeneity that could not be explained after the SR authors conducted various sensitivity analyses. The SRs primarily focused on effectiveness outcomes, and the AEs reported in the SRs were primarily based on primary studies with nonsurvivors of cardiac arrest; therefore, the risk of AEs in a population of both survivors and nonsurvivors is unclear. Across the 4 SRs,40-43 there was also limited or no information on 2 outcomes that the Core Outcome Set for Cardiac Arrest initiative47 recommends reporting for adults: health-related quality of life or survival status after 30 days. Overall, the lack of high-quality evidence and heterogeneity across primary studies make the findings difficult to interpret conclusively.
Most included SRs did not report information for various PROGRESS-Plus7 criteria (e.g., age, sex, gender, race, ethnicity, culture, language, occupation, religion, education, SES, social capital, discrimination [e.g., disability-based], or relationships), specific settings (e.g., rural settings, remote settings, territorial hospitals, nurse-led hospitals, small communities, tertiary care) or on variables that could have affected the results (e.g., patients with shockable cardiac rhythm; events during which cardiac arrest was witnessed, patients who received bystander CPR). The included guidelines44,45 also did not have recommendations that considered PROGRESS-Plus criteria. No SRs extracted information on who administered the devices or conducted CPR or the methods used to train staff for these procedures. Altogether, it is unclear whether any of these contextual factors could have explained any effects of the interventions on health outcomes or provided additional knowledge about unique settings. For example, it is unclear whether automated chest compressions could be used on people who experience frailty because of their age or disability status. Similarly, it is unclear from the evidence found in this Rapid Review whether people with different body sizes may experience automated chest compression devices differently. Because previous research20-22 has shown that whether a bystander chooses to perform CPR can be biased against a patient’s gender, race, perceived income, or educational status or a bystander’s culture, training, or fear of consequences, it is unknown whether these factors could have played a role in the research studies and understanding of the evidence.
Further, having information on the use of automated devices in settings that are remote or where there are fewer health care providers would be important for understanding if there are unique scenarios that may require training for health care providers on device use (e.g., if there are longer travel times for emergency medical services, if the division of tasks with fewer personnel is challenging, in rural or remote areas where there are potentially greater risks of AEs because of difficult terrain or smaller hospitals with fewer resources to address AEs). The lack of cost-effectiveness information for the devices overall and for specific settings also makes it difficult to know whether there may be value in providing access to these devices in low-resource settings and whether the additional costs of the devices could be worth the potential benefits to patients.
The SRs40-43 in this Rapid Review included evidence from more than 20 countries across Asia, Oceania, Europe, and North America, with 1 primary study reported by SRs as including participants from Canada; the generalizability of the findings to settings in Canada is unknown. Three40,41,43 of the 4 SRs and the 2 guidelines44,45 focused on adults with cardiac arrest; 1 SR42 did not specify eligibility based on participant age.42 Thus, it is unclear whether the evidence found in this Rapid Review could apply to other age groups.
Because no SRs reported on rural settings, remote settings, territorial hospitals, nurse-led hospitals, small communities, or tertiary care; on how health care teams were trained on device use; or on patient race, ethnicity, culture, language, occupation, religion, education, SES, or social capital, it is unclear what the effect of using automated chest compression devices is in these contexts and in the context of potential health inequities.
We conducted a Rapid Review of the evidence on the clinical effectiveness and cost-effectiveness of automated chest compressions for people of any age; we also searched for guidelines with recommendations for these populations. We found 4 SRs40-43 published between 2021 and 2024 with global evidence for automated chest compression compared to manual chest compression, and 2 evidence-based guidelines44,45 published between 2020 and 2021, from American and European organizations, with recommendations for any automated chest compression device. We did not find any eligible economic evaluations with cost-effectiveness information.
This Rapid Review found heterogeneous evidence based on studies with varying risk of bias (from low to high) for the clinical effectiveness of automated chest compressions compared to manual compressions. The identified clinical evidence on survival, neurologic outcomes, and ROSC is mixed and suggests a potential increase in compression-induced injuries.
Survival: Based on heterogeneous evidence40,42,43 that includes studies with a high risk of bias, AutoPulse may be beneficial or have no effect on survival, whereas LUCAS may be harmful or have no effect on survival.
Neurologic outcomes: There is mixed evidence40,42 on the effect of relevant devices (i.e., AutoPulse, LUCAS) on neurologic outcomes. Given the variability of the study designs, outcome measurements, and risk of bias, it is difficult to draw conclusions for this outcome.
ROSC: Based on heterogeneous evidence40,42 that includes studies with a high risk of bias, AutoPulse may be beneficial or have no effect on ROSC, whereas LUCAS may have no effect on ROSC.
AEs: Based on evidence on nonsurvivors of cardiac arrest, the use of AutoPulse or LUCAS devices may increase the risk of overall and specific compression-induced injuries and may have no effect on the risk of life-threatening injuries, and complication-induced injuries may vary by specific device type.
The evidence-based guidelines44,45 did not recommend the routine use of automated chest compression devices. They indicated that the use of these devices, often alongside other health technologies or guidance, could be considered in specific circumstances such as when high-quality chest compressions are impractical or dangerous for the rescuer; when prolonged CPR is needed for patients in cardiac arrest who have hyperkalemia; if a patient has coronary thrombosis, no sustained ROSC, and resuscitation is not futile; or in a catheterization laboratory when resuscitating and treating possible causes. Under these circumstances, guidelines recommend having trained professionals familiar with the device to limit CPR interruptions when using and removing the device. Although we did not include consensus statements in this Rapid Review, we found 148 published by the Australasian College for Emergency Medicine that had similar recommendations, namely that automated CPR devices should only be used if staff are trained and if the devices are available so that fewer staff need to be present during compressions.
This Rapid Review compares to the previous CDA-AMC report in the following ways:
Similar to previous SRs and MAs,38,39 we found evidence of harms with the use of these devices, including AEs such as compression-induced injuries.
Different from the previous CDA-AMC report,37 which did not find evidence-based guidelines, we found guidelines that recommend not using these devices routinely.
Different from previous SRs and MAs and the previous CDA-AMC report,37-39 which suggested no evidence of benefits, we found mixed and heterogenous evidence of the clinical effectiveness of these devices.
Similar to the previous CDA-AMC report,37 we found no cost-effectiveness information.
Across the included evidence in this Rapid Review, there was evidence from 1 primary study across 3 included SRs40,42,43 that had participants in Canada; however, generalizability of the evidence to the context in Canada is unclear. Most SRs did not report age, sex, or gender, and it is unclear whether the findings can be extrapolated to different populations without this information. Similarly, because there was limited information on other PROGRESS-Plus criteria,7 it is unclear whether the results are applicable to different settings (e.g., urban, rural, remote, under-staffed, and territorial hospital settings and other settings in Canada) or for equity-deserving groups based on dimensions of diversity such as age, gender, SES, education, or ethnic, religious, or cultural backgrounds.
To address the identified issues, further studies are needed that include the following: improvements on risk-of-bias issues reported in current primary literature (e.g., robust prospective studies); measurement of outcomes such as 30-day survival and quality of life, which are important considerations in cardiac arrest studies;47 and evidence for populations in different settings (e.g., urban, rural, remote, and under-staffed, and territorial hospital settings and other settings in Canada) or for equity-deserving groups based on dimensions of diversity such as age, gender, SES, education, or ethnic, religious, or cultural backgrounds. Previous research suggests that the longer distances that rural and remote patients need to travel results in greater travel times and potential delays in receiving the care that they need, in addition to the disparities in survival between urban and rural settings in Canada;49,50 thus, understanding the effectiveness of mechanical CPR devices in rural and remote transport contexts is important. Another research consideration in rural and remote settings in Canada, including in northern Canada, is how colder temperatures can affect automated chest compression devices and battery performance.51 Battery life may have unique considerations in this context because, for longer travel times, longer-lasting batteries or a greater supply of batteries may be needed.
Because most of the identified literature was for AutoPulse and LUCAS devices, future primary research could also address the clinical effectiveness of other devices, such as EASY PULSE and Lifeline ARM.
Although there is a large body of evidence on automated chest compression devices, including published overviews of reviews, the included SRs in this Rapid Review had substantial heterogeneity within their MAs. There is a need to reanalyze the existing literature to explore reasons for this heterogeneity and consider using adjusted analyses or subgroup analyses from existing primary data (e.g., considering variables such as whether cardiac arrest was witnessed, the proportion of patients with shockable cardiac rhythm, results by age or gender). Future SRs should be high-quality and report the funding sources of primary studies, include the rationale for how the MAs are conducted (e.g., how primary study results are combined), provide separate results for RCTs and observational studies, and extract variables that may give insights on health equity such as place of residence (e.g., urban, rural, or remote), race, ethnicity, occupation, religion, education, SES, social capital, or disability.
Although we found 2 guidelines44,45 that provided recommendations for chest compression devices overall, it is unclear whether different devices require different recommendations. Future guideline developers could consider providing recommendations for different devices, if appropriate.
Because we did not find economic evaluations published since 2020, research on cost-effectiveness is needed, taking the current economic climate into consideration.
Based on this Rapid Review, evidence-based guidelines do not recommend the routine use of automated chest compression devices except in specific scenarios. The heterogeneous clinical evidence does not provide sufficient information about patient demographics or for certain settings, there are risks of AEs from these devices, and information about the cost-effectiveness of the devices was not found. Decision-makers can use this evidence to help understand whether using automated chest compression devices provides potential clinical benefits or harms in their local context.
Clinicians and researchers may also wish to consider the following in practice: whether there are certain populations that have limited access to emergency services overall or limited access to automated devices; how health care staff can get additional specialized training on using automated devices, especially in settings with long distances and difficult terrain; how participants are selected into primary research studies and which population types are included in research based on dimensions of diversity (e.g., including women, people of different ethnic and socioeconomic backgrounds); how patient quality of life and experience outcomes are collected; and the balance of any potential benefits to patients and providers given known safety risks.
This document was externally reviewed by the following content expert, who has granted permission to be cited:
Dr. Brodie Nolan, MD, MSc, FRCPC; Emergency Physician and Trauma Team Leader (St. Michael’s Hospital), Scientist (Li Ka Shing Knowledge Institute), Assistant Professor (University of Toronto), Transport Medicine Physician and Chair, Research (Ornge)
adverse event
confidence interval
cardiopulmonary resuscitation
in-hospital cardiac arrest
International Liaison Committee on Resuscitation
Lund University Cardiopulmonary Assist System
meta-analysis
out-of-hospital cardiac arrest
randomized controlled trial
return of spontaneous circulation
socioeconomic status
systematic review
Please note that this appendix has not been copy-edited.
An information specialist conducted a literature search on key resources including MEDLINE via Ovid, Scopus, the Cochrane Database of Systematic Reviews, the International HTA Database, the websites of health technology assessment agencies in Canada and major international HTA agencies, as well as a focused internet search. The search approach was customized to retrieve a limited set of results, balancing comprehensiveness with relevance. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. Search concepts were developed based on the elements of the research questions and selection criteria. The main search concept was automated chest compression devices. The search was completed on February 5, 2025 and limited to English-language documents published since January 1, 2020. Search strategies available on request.
Two reviewers screened citations and selected studies. In the first level of screening, they independently screened titles and abstracts of all retrieved citations for relevance following a liberal-accelerated approach, whereby a single reviewer was required to include a study and exclusion by both reviewers was needed to exclude a study. Full texts of titles and abstracts that were judged to be potentially relevant by at least 1 reviewer were retrieved and independently assessed by 2 reviewers for inclusion based on the inclusion criteria presented in Table 1. Discrepancies between reviewers at the full-text level were discussed until consensus was reached. Figure 1 presents the PRISMA flow chart of the study selection.
Articles were excluded if they did not meet the selection criteria outlined in Table 1 or were duplicate studies. Studies were also excluded if they were not published in English, simulations in mannequins, the intervention of interest was aimed at elevating the human body during chest compressions, the intervention was an audiovisual feedback device to help with performing chest compressions, the study results were included in an already included SR, the study did not have quantitative results comparing intervention and control, or the study was withdrawn from a journal. Guidelines with unclear methodology were also excluded.
One reviewer critically appraised the included studies using the following tools as a guide: A MeaSurement Tool to Assess systematic Reviews 2 (AMSTAR 2)52 for SRs and the Appraisal of Guidelines for Research and Evaluation (AGREE) II instrument53 for guidelines. Summary scores for the included studies were not calculated; rather, the strengths and limitations of each included study were described narratively in this report.
One reviewer extracted data directly into standardized tables created in Microsoft Word, which were modified as necessary. The extracted information included study characteristics, methodology (e.g., study design), population, intervention, comparator, and results regarding the outcomes of interest. One reviewer extracted information from the included studies using the PROGRESS-Plus7 tool to describe different population groups. Each included study was checked to determine if PROGRESS-Plus7 criteria were reported by study authors to describe the participants. Detailed characteristics, if available, were then extracted and reported in tables in Appendix 2. The main PROGRESS-Plus7 criteria include place of residence, race/ethnicity/culture/language, occupation, gender/sex, religion, education, SES, and social capital. As part of report writing, we discuss these characteristics across the evidence, if they are available, when presenting results within the text.
When reporting on sex, gender, race, or ethnicity in this Rapid Review, we planned to retain the language used by the original study authors, and, whenever possible, we referred to these groups based on guidance from the CDA-AMC Style Guide54 at the time this Rapid Review was conducted, with an understanding that language is constantly evolving.
Selection of Included Studies.
Characteristics of Included Systematic Reviews.
Characteristics of Included Guidelines.
Please note that this appendix has not been copy-edited.
Strengths and Limitations of Systematic Reviews Using AMSTAR 2.
Strengths and Limitations of Guidelines Using AGREE II.
Please note that this appendix has not been copy-edited.
Summary of Findings by Outcomes — Survival.
Summary of Findings by Outcomes — Neurologic Outcomes.
Summary of Findings by Outcomes — ROSC.
Summary of Findings by Outcomes — Overall Rate of Compression-Induced Injuries.
Summary of Findings by Outcomes — Life-Threatening Injuries.
Summary of Findings by Outcomes — Skeletal Fractures.
Summary of Findings by Outcomes — Visceral Injuries.
Summary of Findings by Outcomes — Other Soft Tissue Injuries.
Summary of Recommendations in Included Guidelines.
Please note that this appendix has not been copy-edited.
Overlap in Relevant Primary Studies Between Included SRs.
Please note that this appendix has not been copy-edited.
Canada’s Drug Agency (CDA-AMC) is a pan-Canadian health organization. Created and funded by Canada’s federal, provincial, and territorial governments, we’re responsible for driving better coordination, alignment, and public value within Canada’s drug and health technology landscape. We provide Canada’s health system leaders with independent evidence and advice so they can make informed drug, health technology, and health system decisions, and we collaborate with national and international partners to enhance our collective impact.
Disclaimer: CDA-AMC has taken care to ensure that the information in this document was accurate, complete, and up to date when it was published, but does not make any guarantee to that effect. Your use of this information is subject to this disclaimer and the Terms of Use at cda-amc.ca.
The information in this document is made available for informational and educational purposes only and should not be used as a substitute for professional medical advice, the application of clinical judgment in respect of the care of a particular patient, or other professional judgments in any decision-making process. You assume full responsibility for the use of the information and rely on it at your own risk.
CDA-AMC does not endorse any information, drugs, therapies, treatments, products, processes, or services. The views and opinions of third parties published in this document do not necessarily reflect those of CDA-AMC. The copyright and other intellectual property rights in this document are owned by the Canadian Agency for Drugs and Technologies in Health (operating as CDA-AMC) and its licensors.
Questions or requests for information about this report can be directed to [email protected].
Copyright © 2025 - Canadian Agency for Drugs and Technologies in Health. Except where otherwise noted, this work is distributed under the terms of a Creative Commons Attribution-NonCommercial- NoDerivatives 4.0 International licence (CC BY-NC-ND).
Your browsing activity is empty.
Activity recording is turned off.
See more...
