Real-time audio-visual feedback with handheld nonautomated external defibrillator devices during cardiopulmonary resuscitation for in-hospital cardiac arrest: A meta-analysis

Abstract

Objective:

Restoring cardiopulmonary circulation with effective chest compression remains the cornerstone of resuscitation, yet real-time compressions may be suboptimal. This project aims to determine whether in patients with in-hospital cardiac arrest (IHCA; population), chest compressions performed with free-standing audiovisual feedback (AVF) device as compared to standard manual chest compression (comparison) results in improved outcomes, including the sustained return of spontaneous circulation (ROSC), and survival to the intensive care unit (ICU) and hospital discharge (outcomes).

Methods:

Scholarly databases and relevant bibliographies were searched, as were clinical trial registries and relevant conference proceedings to limit publication bias. Studies were not limited by date, language, or publication status. Clinical randomized controlled trials (RCT) were included that enrolled adults (age ≥ 18 years) with IHCA and assessed real-time chest compressions delivered with either the standard manual technique or with AVF from a freestanding device not linked to an automated external defibrillator (AED) or automated compressor.

Results:

Four clinical trials met inclusion criteria and were included. No ongoing trials were identified. One RCT assessed the Ambu CardioPump (Ambu Inc., Columbia, MD, USA), whereas three assessed Cardio First Angel™ (Inotech, Nubberg, Germany). No clinical RCTs compared AVF devices head-to-head. Three RCTs were multi-center. Sustained ROSC (4 studies, n = 1064) was improved with AVF use (Relative risk [RR] 1.68, 95% confidence interval [CI] 1.39–2.04), as was survival to hospital discharge (2 studies, n = 922; RR 1.78, 95% CI 1.54–2.06) and survival to hospital discharge (3 studies, n = 984; RR 1.91, 95% CI 1.62–2.25).

Conclusion:

The moderate-quality evidence suggests that chest compressions performed using a non-AED free-standing AVF device during resuscitation for IHCA improves sustained ROSC and survival to ICU and hospital discharge.

Trial Registration:

PROSPERO (CRD42020157536).

INTRODUCTION

In-hospital cardiac arrest (IHCA) is common and carries high morbidity and mortality. Data from the American Heart Association's (AHA) Get with the Guidelines-Resuscitation registry indicates that between 2008 and 2017, the U. S. annual incidence of IHCA was 292,000, or roughly 1/100 admissions.[1] The primary causes are cardiac (50%–60%), including myocardial infarction, arrhythmia, and heart failure.[2] Respiratory failure results in roughly 40% of cases and increases in frequency with longer hospital stays.[2] IHCA outcomes vary significantly globally, with return of spontaneous circulation (ROSC) rates ranging from 20% to 73%,[3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26] and North American and European rates ranging 43%–73%.[8,9,10,11,12,27] Survival to hospital discharge ranges from 1% to 42% globally,[4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,28,29] with North American and European rates ranging from 15% to 30%.[8,9,10,11,12,27] Furthermore, while outcomes have been improved over recent decades, 1-year survival rates remain low (13%), with favorable neurological outcomes reported in 20% to 85% of cases.[1]

Cardiopulmonary resuscitation (CPR) with effective chest compression remains the cornerstone of resuscitation.[3,30,31,32,33,34,35] International guidelines note the critical importance of compression components, including position, rate, force, depth, interruptions, recoil, excessive ventilation avoidance, no-flow time, and flow fraction.[3,33,34,35,36] Despite this, increasing evidence suggests that compressions administered in real-time may be suboptimal.[37,38] Some have proposed that real-time audiovisual feedback (AVF) may aid resuscitation efforts by improving the quality of delivered chest compressions,[33,39,40,41] and both the AHA and the International Liaison Committee on Resuscitation (ILCOR) have made cautious recommendations supporting their use.[27,33,35]

Several AVF devices have been developed and marketed. Some are free-standing, whereas others are linked to automated external defibrillators (AED) or other monitoring equipment. The devices are generally applied between the victim's chest and the rescuer's hands. The reliant technology ranges in complexity from a metronome to tensile springs, accelerometers, pressure sensors, and triaxial magnetic sensing [Table 1].[3,42,43,44,45,46,47,48,49,50,51] Feedback may be given in audio, visual, or tactile format. Stand-alone AVF devices provide benefits in cost and simplicity, making them potentially useful for applications both in-and outside of hospital settings. Despite an abundance of non-randomized and simulation studies [Tables 2 and 342,44,45,48,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72] data from clinical randomized controlled trials (RCT) are sparse [Table 4].[3,51,73,74] To date, there have been no < 15 non-AED compression AVF devices have been released to market [Table 1]. The bulk of available evidence is from non-randomized or crossover studies of simulated resuscitations. Seven devices [20 studies; Tables 2 and 3] have published RCT data from simulated resuscitations, none have published non-randomized clinical studies, and only two have available clinical RCT data [4 studies; Table 43,51,73,74] The results of the simulation RCTs suggest that free-standing non-AED AVF devices are associated with: (1) no improvement in correct hand position (1 of 1 study); (2) improved compression rate (12 of 15 studies; 3 no change); (3) improved compression depth (12 of 16 studies; 3 no change; 1 worse); (4) possibly improved compression release/chest recoil (4 of 7 studies; 3 no change); (5) unchanged no-flow fraction (1 of 1 study); (6) improved number inefficient compressions (2 of 2 studies); and (7) improved number of correct/error free compressions (7 of 7 studies). Moreover, there are at least five free-standing AVF devices marketed without any published studies, including Beaty (Medical Feedback Technologies Ltd.), CPR-1100 (Nihon Kohden), PrestoPatch™ (Nexus Control Systems LLC.), PrestoPush™ (Nexus Control Systems LLC.), and ПP -01 (PR-01; FactorMed Technika).

Table 1.

Description of hand-held compression feedback devices that are not linked to an automated external defibrillator or external device

Device	Manufacturer (city, country)	Commercial availability	Reliant technology	Power Source	Feedback type	Feedback method	Measurement items
Ambu CardioPump	Ambu Inc. (Columbia, MD, USA)	Available	Tensile springs	Mechanical	Auditory Visual	Metronome Visual scale	Compression depth Compression rate Recoil
Beaty	Medical Feedback Technologies ltd. (Even Yehuda, Israel)	Available	Accelerometer	Battery	Auditory	Audio tone	Adequate compression
Cardio First Angel™	Inotech (Nubberg, Germany)	Available	Tensile springs	Mechanical	Auditory Tactile	Audible click Tactile click	Compression depth Compression rate
CPR-1100 CPR Assista	Nihon Kohden Corp. (Tokyo, Japan)	Available	Accelerometer	Battery	Visual Auditory	Light indicator Metronome Verbal Cue	Compression depth Compression rate Device tilt Sinking of patient’s back
CPRCard™	Laerdal (Stavanger, Norway)	Available	Accelerometer	Battery	Visual	Digital meters	Compression depth Compression rate
CPREzy™	Health Affairs, LTD. (London, UK)	Available	Metronome Pressure sensor	Battery	Visual	Light indicator	Compression depth Compression rate
CPR-plus™	Kelly Medical Products (Princeton, USA)	Discontinued	Pressure sensor	Mechanical	Visual	Needle gauge	Compression depth
CPR PRO®b	Ivor Medical (Rijeka, Croatia)	Discontinued	Accelerometer	None	Audio Tactile Visual	Digital screen of smartphone mounted on device	Compression depth
CPRmeter™	Laerdal (Stavanger, Norway)	Discontinued	Accelerometer	Battery	Visual	Digital screen Inactivity timer	Compression depth Compression rate
CPRmeter 2™	Laerdal (Stavanger, Norway)	Available	Accelerometer	Battery	Visual	Digital screen Inactivity timer	Compression depth Compression rate
CPR RsQ Assist®	CPR RsQ Assist Inc. (Naples, USA)	Available	Metronome	Battery	Auditory	Metronome Voice	None
LinkCPR™	SunLife Science (Shanghai, China)	Available	Accelerometer	Battery	Visual	Wristbanddigital screen	Compression depth Compression rate
Pocket CPR™	Zoll Medical Corp. (Chelmsford, USA)	Available	Accelerometer	Battery	Auditory Visual	Light indicator Metronome Verbal cue	Compression depth Compression rate
TrueCPR™	Physio-Control (Redmond, USA)	Available	Electromagnetic sensors	Battery	Auditory Visual	Metronome Digital screen	Compression depth Compression rate Verbal prompt for rescue breathing
U-cpr
ПР-01 (PR-01)	FactorMed Technika (Moscow, Russia)	Available	Accelerometer	Battery	Visual Auditory	Light indicator Verbal cue	Compression depth Compression rate

^aCan communicate with a Nihon Kohden defibrillator via Bluetooth connection, ^bThe base device does not contain measuring technology itself. It serves as an ergonomic consul or mount for an electronic device (e.g., iPhone with CPR PRO application) containing an accelerometer. CPR: Cardiopulmonary resuscitation

Table 2.

Simulation randomized controlled manakin studies investigating chest compressions administered either with or without the assistance of a free-standing audiovisual feedback device (not linked to external monitor or automated external defibrillator)

Device	Author (year)	Population	Sample size	Comparison	Primary outcome and findings	Secondary outcomes and findings
CPREzy™	Beckers (2007)	Students	202	SMC	Device use associated with improved compression rate and depth	No change in full recoil or hand position
	Bielski (2018)	Lifeguards	41	SMC	Compression depth was significantly better and the standard manual compression group. No-flow fraction did not improve	Device use improved full chest release (87% vs. 68%; P=0.02)
	Noordergraaf (2006)	Hospital employees	224	SMC	Fewer ineffective compressions in the device group	Decreased time to ineffective compressions in the device group
	Yeung (2014)	Life support providers (unspecified)	101	SMCQ-CPR with metronome	CPREzy™ use improved compression depth over both comparators	CPREzy™ use improved compression rate and decreased proportion of compressions with inadequate depth compared to both comparators. Compression release was unchanged
	Veiser (2010)	Paramedics, emergency physicians	93	SMC	Device group had improved percentage of compressions with correct depth and rate, and higher rate of error free compressions
CPR PRO®	Kovic (2013)	Health care providers	24	SMC	Decreased rescuer perceived exertion and maximal HR in device group	Decreased hand and wrist pain in device group
CPRmeter™	Buléon (2013)	Students	144	SMC	Improved efficient compression rate in the device group	Improved compression rate and percentage of compressions of adequate depth in the device group
	Buléon (2016)	Health care providers	60	SMC	Improved efficient compression rate in the device group	Improved compression rate, percentage of compressions of adequate depth, and adequate release in the device group
	Calvo-Buey (2016)	Health care providers	88	SMC	Improved compression depth and complete release in device group	Higher compression rate in device group, although rates in both groups met guideline standards
	Delaunay (2015)a	Health care providers	60	SMC	Improved correct compressions in device group	No change in compression rate
	Duwat (2014)a	Paramedics	120	SMC	Decrease in compressions of adequate depth in device group	Less dispersion of compression frequency with device use
	Iskrzycki (2018)	Lifeguards	50	SMC	Improved quality of CPR score (median 69 [33-77] vs. 84 [55-93]; P<0.001)	Compression score, depth and rate improved significantly in the device group
CPR RsQ Assist®	Yuksen (2017)	Health care providers	80	SMC	Improved compression rate at 4 min in the device group	No change in compression depth at 2 min, but improved depth in controls at 4 min
LinkCPR™	Liu (2018)	Laypersons	124	SMC	Improved compression rate and depth in device group	Improved correct compressions and compression fraction in device group
Pocket CPR™	Grassl (2009)	Inexperienced laypersons	42	SMC	The device did not consistently improve compression depth or rate
	Pozner (2011)	Nurses	12	SMC	Increased compression depth and lower rate in the device group resulting increased compression rate in recommended range	Chest recoil and fatigue did not differ between groups
TrueCPR™	Al-Jeabory (2017)	Physicians	60	SMC	Increased compression depth and lower rate in the device group resulting in increased compression rate in recommended range	Decreased incorrect compressions in device group
	Grassl (2016)a	Health care providers	202	SMC	Increased percentage of correct compressions with device	Compression rate within recommended rage for both groups
	Majer (2018)	Nurses	38	SMC	Lower rate in the device group resulting in increased compression rate in recommended range	Compression depth varied greatly in both groups. Full chest recoil was improved in the device group
	Ozel (2016)	Students	83	SMC with/out metronome	Device use associated with improved rate (both groups within guideline range) and depth

^aAbstract only. IQR: Inter-quartile range; SMC: Standard manual compressions; Q-CPR: Quantitative measurement of cardiopulmonary resuscitation; HR: Heart rate

Table 3.

Simulation randomized controlled manakin studies comparing chest compressions administered either with or without the assistance of a free-standing audiovisual feedback device (not linked to external monitor or automated external defibrillator), and comparing at least two devices

Author (year)	Groups	Population	Sample size	Primary outcome and findings	Secondary outcomes and other findings
Bydzovsky (2015)a	CPREzy™ PocketCPR™ SMC	Nurses	152	Both devices improved compression quality compared to SMC controls	Direct comparisons between the devices not provided
Davis (2018)	CPR RsQ Assist® PocketCPR™ TrueCPR™ SMC	Students and healthcare providers	118	Compression depth was poor across all groups, but TrueCPR™ and PocketCPR™ demonstrated statically (not clinically) significant improvements compared to control and CPR RsQ Assist®. PocketCPR™ had the greatest % compressions with sufficient depth, while TrueCPR™ had the greatest % with adequate rate	Controls outperformed all devices in no‐flow time (P<0.001) and flow fraction (P<0.001). Full recoil was not improved by device use (P=0.31)
Dixon (2010)a	CPREzy™ Unspecified 1 Unspecified 2	Healthcare providers	21	No improvement in compression depth or compression effectiveness (depth vs. incomplete release vs. incorrect hand placement)
Kurowski (2015)	PocketCPR™ TrueCPR™ SMC	Paramedics	167	TrueCPR™ improved compression depth and effectiveness of compressions versus comparators	PocketCPR™ was the only group whose rate fell outside guideline recommendations
Schachinger (2013a)b	CPRmeter® PocketCPR™ SMC	Students	240	A significant delay in time to first compression was noted for the PocketCPR™ versus others
Schachinger (2013b)b	CPRmeter® PocketCPR™ SMC	Students	240	All groups reached recommended compression depth and rate	ECR was lower for PocketCPR™ compared to SMC. Both devices showed improvement in ECR decline.
Zapletal (2014)b	CPRmeter® PocketCPR™ SMC	Students	240	Effective compressions were significantly improved for PocketCPR™ versus CPRmeter® and SMC (others not significant)	Both devices showed improvement in ECR decline. Overall performance in the PocketCPR® group was considerably inferior to standard BLS

^aAbstract only; ^bSingle study. Data presented in 3 abstracts. Full manuscript not available. SMC: Standard manual compressions; ECR: Effective compression ratio; BLS: Basic life support

Table 4.

Prospective randomized human clinical trials of adult patients (age ≥18 years) being treated for in-hospital cardiac arrest with cardiopulmonary resuscitation including chest compressions administered either with or without the assistance of a free-standing audiovisual feedback device (not linked to external monitor or automated external defibrillator)

Device	Author (year), Citation	Methodology	Setting	Sample size	Population	Primary outcome and findings	Secondary outcomes and findings
Ambu CardioPump	Cohen (1993)	RCT	Inpatient	62	Adult patients in medical ICU, coronary care unit, cardiac-catheterization laboratory, and medical wards at 1 academic medical center	Improved sustained ROSC in the device group (62% vs. 30%; P<0.03). Improved survival for ≥24 h (45% vs. 9%; P<0.04)	No improvement in survival to hospital discharge (7% vs. 0%; P=NS). Improved neurological status (GCS) in device group (8.0±1.3 vs. 3.5±0.3; P<0.02)
Cardio First Angel™	Goharani (2019)	RCT	ICU	900	Adult patients with cardiac arrest in a mixed med-surg ICU at 8 academic medical centers	Improved sustained ROSC in the device group (66.7% vs. 42.4%, P<0.001)	Improved survival to ICU discharge (59.8% vs. 33.6%) and survival to hospital discharge (54% vs. 28.4%, P<0.001) in the device group
	Vahedian-Azimi (2016)	RCT	ICU	80	Adult patients with cardiac arrest in a mixed med-surg ICU at 4 academic medical centers	Improved sustained ROSC in device feedback group (72% vs. 35%; P=0.001)	Decrease in rib fractures (57% vs. 85%; P=0.02), but not sternum fractures (5% vs. 17%; P=0.15)
	Vahedian-Azimi (2020)	RCT	ICU	22	Adult patients with cardiac arrest in a mixed med-surg ICU at 4 academic medical centers	The incidence of ROSC was similar between groups (P=0.64)	Guideline adherence was improved in the intervention group (P=0.0005). No significant decrease in rib fractures (P=0.31) or sternum fractures (P=0.15)

RCT: Randomized controlled trial; ICU: Intensive care unit; ROSC: Return of spontaneous circulation; GCS: Glasgow Coma Scale; NS: Nonsignificant

Data comparing free-standing non-AED AVF devices head-to-head are far fewer,[42,50,68,69,70,71,72] and which devices display the highest performance remains unclear. The objective of this project is to address the following research question: in patients with IHCA (population), does chest compression performed with a free-standing non-AED AVF device as compared to standard manual chest compressions (comparison) result in improved outcomes including sustained ROSC, and survival to ICU and survival to hospital discharge (outcomes).

METHODS

Prospective human RCTs of adult patients (age ≥ 18 years) being treated for IHCA with CPR, including chest compressions, were considered for inclusion. All trials were required to compare standard manual chest compressions to chest compressions performed with a free-standing AVF device (not linked to an external monitor or AED). AFV devices linked to AEDs or automated compressors differ significantly from smaller free-standing (generally handheld) AVF devices. These devices often differ in the underlying technology, software algorithms, and intrinsic capabilities, as well as in size, cost, and the logistic practicality of deployment and maintenance in non-acute care (e.g. emergency department) or critical care environments (e.g. general medical floors). Prior meta-analyses have either excluded the free-standing AVF devices or combined their data that of devices linked to AEDs or automatic compressors.[75,76] This effort represents the first meta-analysis assessing only the free-standing AVF device subgroup.

The primary outcome was sustained ROSC, defined as ROSC lasting >30 min. The secondary outcomes were survival to intensive care unit (ICU) discharge, survival to hospital discharge, and adverse events. The project was registered with PROSPERO (CRD42020157536).

A comprehensive literature search strategy [Appendix 1] was developed in conjunction with a librarian specializing in systematic reviews of the following databases: China National Knowledge Infrastructure, Cochrane CENTRAL, CINAHL, Directory of Open Access Journals, Embase, Korean Journal Database, Latin American and Caribbean Health Sciences Literature, IEEE-Xplorer, information/Chinese Scientific Journals Database, Google Scholar, Magiran, PsycInfo, PubMed, Scopus, Scientific Electronic Library Online, Scientific Information Database, Turkish Academic Network and Information Center (TÜBİTAK ULAKBİM), Research Gate, Russian Science Citation Index, and Web of Science. The search terms included the following National Library of Medicine MeSH terms: CPR, and heart arrest. Additional non-MeSH terms included cardiac arrest, in-hospital, and the following individual device names: Beaty, Cardio First Angel, CardioPump, CPR-1100 CPR Assist, CPRCard, CPREzy, CPR-plus, CPR PRO, CPRmeter, CPR RsQ Assist, LinkCPR, Pocket CPR, PrestoPatch, PrestoPush, TrueCPR and ПP -01 (PR-01).

Only clinical RCT of free-standing compression AVF devices were included. Crossover studies and those assessing AED-linked devices were excluded. Searches were not limited by date, language, or publication status. To limit publication bias, clinical trial registries were searched including ClinicalTrials.gov, World Health Organization International Clinical Trials Registry Platform (WHO ICTRP), and the Australian New Zealand Clinical Trials Registry (ANZCTR). The bibliographies of the relevant articles were also searched. Conference proceeding from relevant disciplines in the past 5 years were also searched. Experts in the field were also contacted to inquire about possible ongoing trials. Grey literature was only eligible for inclusion if the authors responded to correspondence affirmatively with the requested information.

Risk-of-bias (RoB) was assessed using two validated tools: (1) Grading of Recommendations, Assessment, Development and Evaluations (GRADE),[78] and (2) RoB 2.0: Revised tool for Risk of Bias in randomized trials.[78] The authors considered methods of randomization and allocation, blinding (of treatment administrator, participants, and outcome assessors), selective outcome reporting (e.g., failure to report adverse events), incomplete outcome data, and sample size calculation. Each trial was graded as high, low, or unclear RoB for each of these criteria.

Statistical analysis

Model selection depended on assessments of common effect size. A fixed-effects model was used if all studies share the same true effect and the inter-study observed effect varied because of random error, or there was intra-study variation. A random-effects model was used if the true effect differed between studies due to inter-and intra-study variation. This was conducted using a restricted maximum likelihood method utilizing the “meta” code.[79]

The presence of heterogeneity and its impact on the meta-analysis was evaluated using the Cochran's Chi-square or Q test (P > 0.10) and I-Squared (I²) index (I²≥ 75% indicating considerable heterogeneity) respectively.[80] However, it is known that Cochran's Chi-square suffers from poor power when the number of collected studies is small.[81] In addition, outlying studies can have a great impact on conventional heterogeneity measures and on the conclusions of a meta-analysis.[81] For this reason, the Tau-squared (τ²) was used as a second means to determine the between-study variance.[81] In the event of significant heterogeneity between studies, we planned to do subgroup analysis or meta-regression. If significant heterogeneity did not exist, then meta-regression was not to be performed.

The common effect size was calculated as the hazard ratio (HR) and its 95% confidence interval (CI) for each main outcome. In addition, the visual assessment of the forest plots was used to find the magnitude of differences.

Publication was assessed using funnel plot analysis and the Begg and Mazumdar,[82] Egger et al.,[83] or Harbord's et al.[84] tests, where appropriate. A nonparametric “trim and fill” method of accounting for publication bias (ref) was conducted, and the modified effect size was estimated after adjusting for publication bias.[85]

We preplanned a sensitivity analysis to examine the effect of each study on the pooled effect size. All analyses were performed using STATA16 (StataCorp, College Station, Texas, USA).

RESULTS

The primary search yielded 2,766 references. Figure 1 from the Prisma flow diagram. Four clinical RCTs met the inclusion criteria. No ongoing trials were identified in Clinicaltrials.gov, WHO ICTRP, or ANZCTR. One of 4 clinical RCTs assessed the active compression active decompression Ambu CardioPump (Ambu Inc., Columbia, MD, USA),[73] whereas 3 assessed the active compression passive decompression Cardio First Angel™ [Inotech, Nubberg, Germany; Table 4].[3,51,74] No clinical RCTs compared AVF devices head-to-head. Three of 4 studies were multi-center,[3,51,74] whereas 1 was single-center.[73] One clinical RCT took place in a high-income economy (USA).[73] Three studies took place in a middle-income economy (Iran).[3,51,74]

Figure 1.

PRISMA flow diagram

The results of the trials' quality assessments are summarized in summarized in Table 5. All 4 studies were reported as low risk of selection bias. Three of 4 studies were low risk for allocation concealment,[3,51,74] whereas concealment was not described in 1 study and was thus marked unclear.[73] In none of the studies were personnel blinded as sham device use was deemed unethical or impractical; however, this may introduce performance bias. All four studies contained a control group (standard manual compressions). Each of the four included studies was low risk for attrition bias. All included studies reported their intended primary outcomes. One study did not report adverse events.[73] All included studies reported a sample size or power calculation.

Table 5.

Grade quality of evidence ratings

Certainty assessment							Summary of findings
Study number	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	Number of patients		Effect
							CC with AVF device	Standard manual CC	Relative (95% CI)	Absolute (95% CI)	Certainty
Sustained ROSC
4	RCT	Not serious	Seriousa	Not serious	Not serious	None	351/530 (66.2%)	217/534 (40.6%)	RR 1.68 (1.39-2.04)	276 more per 1000 (from 158 more to 423 more)	⨁⨁⨁◯ Moderate
Survival to ICU discharge
2	RCT	Not serious	Seriousa	Not serious	Not serious	None	278/461 (60.3%)	156/461 (33.8%)	RR 1.78 (1.54-2.06)	264 more per 1000 (from 189 more to 359 more)	⨁⨁⨁◯ Moderate
Survival to hospital discharge
3	RCT	Not serious	Seriousa	Not serious	Not serious	None	253/490 (51.6%)	132/494 (26.7%)	RR 1.91 (1.62-2.25)	243 more per 1000 (from 166 more to 334 more)	⨁⨁⨁◯ Moderate

^aRisk of bias due to lack of blinding. CC: Chest compressions; AVF: Audiovisual feedback; ROSC: Return of spontaneous circulation; RCT: Randomized controlled trial; CI: Confidence interval; RR: Relative risk

Low heterogeneity was observed between studies for all study outcomes (I² range 0–13.9, τ² range 0–0.01, all P > 0.05). The small study number limits the interpretation of the funnel plots. However, the Egger's, Begg's and Harbord's test results indicated no significant bias [Table 6]. Heterogeneity was sought within individual studies by sensitivity analysis. The results showed that the range of HR after removing a study was within the 95% CI of HR, indicating low heterogenity [Table 7].

Table 6.

Assessment of publication bias using the Begg’s, Egger’s, and Harbord’s tests

Outcomes	Begg’s test		Egger’s test		Harbord’s test
	Z	P	Z	P	Z	P
Initial rhythm asystole	−1.04	0.703	−0.17	0.865	−0.60	0.548
Initial rhythm VF or VT	1.56	0.118	1.53	0.124	0.96	0.252
Initial rhythm “other”	1.04	0.296	1.00	0.316	1.05	0.293
Intubated before arrest	0.00	1.000	0.62	0.534	0.64	0.523
Sustained ROSC	−0.34	0.865	1.21	0.226	1.24	0.215
Survival to ICU discharge	0.00	1.00	0.00	1.00	0.00	1.00
Survival to hospital discharge	1.04	0.296	0.55	0.585	0.91	0.363

VF: Ventricular fibrillation; VT: Ventricular tachycardia; ROSC: Return of spontaneous circulation; ICU: Intensive care unit

Table 7.

The results of sensitivity analyses

QOL	Sensitivity analyses results
	HR range after removing the study
Initial rhythm asystole	0.492-1.744
Initial rhythm VF or VT	0.591-1.326
Initial rhythm “other”	0.666-4.290
Intubated before arrest	0.618-1.495
Sustained ROSC	1.408-2.933
Survival to ICU discharge	0.946-4.956
Survival to hospital discharge	0.890-3.642

VF: Ventricular fibrillation; VT: Ventricular tachycardia; ROSC: Return of spontaneous circulation; ICU: Intensive care unit; QOL: Quality of life; HR: Heart rate

The results of the meta-analysis on the common effect size revealed no significant difference between device and control groups for baseline rhythm (asystole, ventricular fibrillation, ventricular tachycardia, other) and intubation before arrest (all P > 0.05).

Sustained ROSC (4 studies, n = 1064) was improved with AVF use (Relative risk [RR] 1.68, 95% CI 1.39–2.04).[3,51,74,75] Survival to hospital discharge (2 studies, n = 922) was also improved with AVF use (RR 1.78, 95% CI 1.54–2.06),[51,74] as was survival to hospital discharge (3 studies, n = 984; RR 1.91, 95% CI 1.62–2.25).[51,73,74] Although not an endpoint of our meta-analysis, one study reported the endpoint of survival ≥ 24 h and found improvement with AVF device use.[73] In addition, only one study reported on neurologic status, similarly finding improvements with AVF device use; however, this was also not an endpoint for our analysis.[73] As there was no substantial heterogeneity in the models, no meta-regression was conducted.

DISCUSSION

Many factors may influence IHCA outcomes. Hospital-level factors such as bed size, location, and academic status have all been shown to influence ICHA outcomes.[86] Other confounding factors include hospital wealth and cultural beliefs surrounding end-of-life care.[87] Reported rates of ROSC vary from as low as 20% (Iran)[4] to as high as 71% (Brazil);[7] however, a meta-analysis of studies published between 2006 and 2015 found that IHCA ROSC rates average 47%–48% with survival to hospital discharge rates averaging a dismal 14%–15%.[26]

A large gap exists between current knowledge of CPR quality and its optimal implementation, contributing to potentially preventable deaths.[74,88] Early defibrillation (when appropriate) and initiation of CPR with quality compressions remain cornerstones of resuscitation.[3,30,31,32,33,34,35,74] Real-time AVF is one strategy identified by the AHA and ILCOR as a strategy that may improve guideline adherence and IHCA outcomes.[27,33,35] Simulated studies show improved CPR quality with feedback devices [Tables 2 and 3];[42,44,45],48,,52,[53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72] however, the evidence for improvement in clinical outcomes is still very limited [Tables 4 and 5].[3,51,73,74] That said, these devices have also been shown to be easy to use by both medical professionals and lay persons[45] and remain an understudied opportunity to improve IHCA outcomes.

AFV devices linked to AED's or automated compressors differ significantly from small free-standing (generally handheld) AVF devices. Not only do they differ in the underlying technology, software algorithms, and intrinsic capabilities but also in size, cost, and the logistic practicality of deployment and maintenance in non-acute care (e.g. emergency department) or critical care environments (e.g. general medical floors). Prior meta-analyses have either excluded the free-standing AVF devices or combined their data with that of devices linked to AEDs or automatic compressors.[75,76] This effort represents the first meta-analysis assessing only the free-standing AVF device subgroup. Despite a large number of AVF devices on the market, only two devices (4 studies) have published RCT results. The remainder have only been studied in medical simulation scenarios or have no published studies. The small number of included studies is a limitation of this analysis. That said, the results suggest that real-time AVF with a free-standing AVF device (either Ambu CardioPump or Cardio First Angel™) when managing IHCA is associated with improved rates of sustained ROSC and survival to ICU and hospital discharge. Patients who received CPR with AVF device use were 1.68 times as likely to have sustained ROSC, 1.78 times as likely to survive to ICU discharge, and 1.91 times to survive to hospital discharge compared to those who received standard CPR.

These results are likely generalizable to ICU patients in middle- and high-income countries. Three of the studies took place in a middle-income economy (Iran), while one was conducted in a high-income economy (USA). Of note, Iran has a similar prevalence of cardiovascular disease risk factors as in the United States.[89] The impact on resuscitation efforts in healthcare environments in low-income economies remains unclear.

Free-standing AVF devices have the potential to improve patient outcomes following IHCA. These devices have the advantages of being inexpensive, portable, and low maintenance as compared to devices linked to AEDs or automated compressors. As such, they could be more widely available to providers outside acute care environments like the ICU, emergency department, or operating theater. However, these results should be interpreted with caution, as insufficient evidence is available to comment on long-term neurological or functional status or discharge destination (i.e. home, rehab, long-term care facility). In addition, data were not stratified by specialty or practice experience level of the provider (e.g. nurse, resident physician, attending physician) using the device. Moreover, no data were available regarding provider injuries with device use (e.g. wrist or back injuries) as some have voiced concern regarding this matter.[90,91] Greater study of these devices is needed before the widespread implementation of their use in inpatient care environments.

CONCLUSION

The existing moderate certainty evidence suggests that chest compressions performed using a non-AED free-standing AVF device during resuscitation for IHCA may improve rates of sustained ROSC and survival to ICU and hospital discharge. Data on discharge destination, level of health, and neurologic status on discharge are lacking.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Research quality and ethics statement

Data used in this article came from publicly available sources that contain aggregate, de-identified information only. Thererefore, Institutional Review Board approval was not required. Applicable EQUATOR Network (https://www.equator-network.org/) research reporting guidelines were followed.

Appendix 1: Literature Search Strategy

MEDLINE via PubMed:

(”Cardiopulmonary Resuscitation”[Mesh] OR “Cardiopulmonary Resuscitation”[tiab] OR CPR[tiab] OR “Cardio-pulmonary resuscitation”[tiab] OR “Cardio pulmonary resuscitation”[tiab] OR “Basic Cardiac Life Support”[tiab] OR “chest compression”[tiab] OR “chest compressions”[tiab])

AND

(”Equipment Design”[Mesh] OR “Mobile Applications”[Mesh] OR “real-time”[tiab] OR “real time”[tiab] OR “Cardio First Angel”[tiab] OR “Ambu CardioPump”[tiab] OR “CPR-1100 CPR Assist”[tiab] OR CPRCard[tiab] OR CPREzy[tiab] OR “CPR-plus”[tiab] OR “CPR PRO”[tiab] OR CPRmeter[tiab] OR “CPR RsQ Assist”[tiab] OR LinkCPR[tiab] OR Pocket CPR[tiab] OR TrueCPR[tiab] OR “PR-01”[tiab] OR QCPR[tiab] OR “ZOLL AED Plus”[tiab] OR “Intellisense CPR”[tiab] OR “feedback device”[tiab] OR “feedback devices”[tiab] OR “adjunct device”[tiab] OR “adjunct devices”[tiab] OR ((”hand-held”[tiab] OR portable[tiab]) AND (device[tiab] OR devices[tiab])) OR “mobile application”[tiab] OR “mobile applications”[tiab] OR “mobile apps”[tiab] OR “electronic apps”[tiab] OR “software application”[tiab] OR “software applications”[tiab] OR “software app”[tiab] OR “software apps”[tiab])

AND

(”randomized controlled trial”[pt] OR “controlled clinical trial”[pt] OR randomized[tiab] OR placebo[tiab] OR “drug therapy”[sh] OR randomly[tiab] OR trial[tiab] OR groups[tiab] NOT (”animals”[mh] NOT “humans”[mh]))

CINAHL

((MH “Resuscitation, Cardiopulmonary”) OR (MH “Advanced Cardiac Life Support”) OR TI (”Cardio-pulmonary resuscitation” OR “Cardio pulmonary resuscitation” OR “Basic Cardiac Life Support” OR “chest compression” OR “chest compressions”) OR AB (”Cardio-pulmonary resuscitation” OR “Cardio pulmonary resuscitation” OR “Basic Cardiac Life Support” OR “chest compression” OR “chest compressions”))

AND

((MH “Mobile Applications”) OR (MH “Biophysical Instruments”) OR (MH “Equipment Design”) OR (MH “Equipment and Supplies”) OR TI (”real-time” OR “real time” OR “Cardio First Angel” OR “Ambu CardioPump” OR “CPR-1100 CPR Assist” OR CPRCard OR CPREzy OR “CPR-plus” OR “CPR PRO” OR CPRmeter OR “CPR RsQ Assist” OR LinkCPR OR Pocket CPR OR TrueCPR OR “PR-01” OR QCPR OR “ZOLL AED Plus” OR “Intellisense CPR” OR “feedback device” OR “feedback devices” OR “adjunct device” OR “adjunct devices” OR ((”hand-held” OR portable) AND (device OR devices)) OR “mobile application” OR “mobile applications” OR “mobile apps” OR “electronic apps” OR “software application” OR “software applications” OR “software app” OR “software apps”) OR AB (”real-time” OR “real time” OR “Cardio First Angel” OR “Ambu CardioPump” OR “CPR-1100 CPR Assist” OR CPRCard OR CPREzy OR “CPR-plus” OR “CPR PRO” OR CPRmeter OR “CPR RsQ Assist” OR LinkCPR OR Pocket CPR OR TrueCPR OR “PR-01” OR QCPR OR “ZOLL AED Plus” OR “Intellisense CPR” OR “feedback device” OR “feedback devices” OR “adjunct device” OR “adjunct devices” OR ((”hand-held” OR portable) AND (device OR devices)) OR “mobile application” OR “mobile applications” OR “mobile apps” OR “electronic apps” OR “software application” OR “software applications” OR “software app” OR “software apps”))

China National Knowledge Infrastructure (CHKD-CNKI)

((”Cardiopulmonary resuscitation” OR “Compression” OR “CPR” OR “cardiac arrest”) AND “feedback” AND “in-hospital”) OR (”Cardio First Angel” OR “CardioPump” OR “Beaty” OR “CPR Assist” OR “CPRCard” OR “CPR-plus” OR “CPR Pro” OR “CPRmeter” OR “CRP RsQ” OR “LinkCPR” OR “Pocket CPR” OR “TrueCPR” OR “U-CPR” OR “??-01” OR “PR-01”)

Cochrane CENTRAL

(cardiopulmonary resuscitation.sh. OR (Cardio-pulmonary resuscitation or Cardio pulmonary resuscitation or Basic Cardiac Life Support or chest compression or chest compressions).mp.)

AND

(mobile applications.sh. OR equipment design.sh. OR (real-time OR real time OR Cardio First Angel OR Ambu CardioPump OR CPR-1100 CPR Assist OR CPRCard OR CPREzy OR CPR-plus OR CPR PRO OR CPRmeter OR CPR RsQ Assist OR LinkCPR OR Pocket CPR OR TrueCPR OR PR-01 OR QCPR OR ZOLL AED Plus OR Intellisense CPR OR feedback device OR feedback devices OR adjunct device OR adjunct devices OR ((hand-held OR portable) AND (device OR devices)) OR mobile application OR mobile applications OR mobile apps OR electronic apps OR software application OR software applications OR software app OR software apps).mp)

information/Chinese Scientific Journals database (CSJD-VIP)

((”Cardiopulmonary resuscitation” OR “Compression” OR “CPR” OR “cardiac arrest”) AND “feedback” OR “in-hospital”) OR (”Cardio First Angel” OR “CardioPump” OR “Beaty” OR “CPR Assist” OR “CPRCard” OR “CPR-plus” OR “CPR Pro” OR “CPRmeter” OR “CRP RsQ” OR “LinkCPR” OR “Pocket CPR” OR “TrueCPR” OR “U-CPR” OR “??-01” OR “PR-01”)

Directory of Open Access Journals (DOAJ)

((”Cardiopulmonary resuscitation” OR “Compression” OR “CPR” OR “cardiac arrest”) AND “feedback” OR “in-hospital”) OR (”Cardio First Angel” OR “CardioPump” OR “Beaty” OR “CPR Assist” OR “CPRCard” OR “CPR-plus” OR “CPR Pro” OR “CPRmeter” OR “CRP RsQ” OR “LinkCPR” OR “Pocket CPR” OR “TrueCPR” OR “U-CPR” OR “??-01” OR “PR-01”)

EMBASE

('resuscitation'/exp OR “Cardiopulmonary Resuscitation”:ab, ti OR CPR: ab, ti OR “Cardio-pulmonary resuscitation”:ab, ti OR “Cardio pulmonary resuscitation”:ab, ti OR “Basic Cardiac Life Support”:ab, ti OR “chest compression”:ab, ti OR “chest compressions”:ab, ti)

AND

('CPR feedback device'/exp OR 'mobile health application'/exp OR “real-time”:ab, ti OR “real time”:ab, ti OR “Cardio First Angel”:ab, ti OR “Ambu CardioPump”:ab, ti OR “CPR-1100 CPR Assist”:ab, ti OR CPRCard: ab, ti OR CPREzy: ab, ti OR “CPR-plus”:ab, ti OR “CPR PRO”:ab, ti OR CPRmeter: ab, ti OR “CPR RsQ Assist”:ab, ti OR LinkCPR: ab, ti OR Pocket CPR: ab, ti OR TrueCPR: ab, ti OR “PR-01”:ab, ti OR QCPR: ab, ti OR “ZOLL AED Plus”:ab, ti OR “Intellisense CPR”:ab, ti OR “feedback device”:ab, ti OR “feedback devices”:ab, ti OR “adjunct device”:ab, ti OR “adjunct devices”:ab, ti OR ((”hand-held”:ab, ti OR portable: ab, ti) AND (device: ab, ti OR devices: ab, ti)) OR “mobile application”:ab, ti OR “mobile applications”:ab, ti OR “mobile apps”:ab, ti OR “electronic apps”:ab, ti OR “software application”:ab, ti OR “software applications”:ab, ti OR “software app”:ab, ti OR “software apps”:ab, ti)

AND

('crossover procedure':de OR 'double-blind procedure':de OR 'randomized controlled trial':de OR 'single-blind procedure':de OR (random* OR factorial* OR crossover* OR cross NEXT/1 over* OR placebo* OR doubl* NEAR/1 blind* OR singl* NEAR/1 blind* OR assign* OR allocat* OR volunteer*):de, ab, ti)

Korean Journal Database (KCI)

((”Cardiopulmonary resuscitation” OR “Compression” OR “CPR” OR “cardiac arrest”) AND “feedback”) OR (”Cardio First Angel” OR “CardioPump” OR “Beaty” OR “CPR Assist” OR “CPRCard” OR “CPR-plus” OR “CPR Pro” OR “CPRmeter” OR “CRP RsQ” OR “LinkCPR” OR “Pocket CPR” OR “TrueCPR” OR “U-CPR” OR “??-01” OR “PR-01”)

Latin American and Caribbean Health Sciences Literature

(tw:(”Cardio-pulmonary resuscitation” OR “Cardio pulmonary resuscitation” OR “Basic Cardiac Life Support” OR “chest compression” OR “chest compressions” OR cpr))

AND

(tw:(”real-time” OR “real time” OR “Cardio First Angel” OR “Ambu CardioPump” OR “CPR-1100 CPR Assist” OR cprcard OR cprezy OR “CPR-plus” OR “CPR PRO” OR cprmeter OR “CPR RsQ Assist” OR linkcpr OR pocket AND cpr OR truecpr OR “PR-01” OR qcpr OR “ZOLL AED Plus” OR “Intellisense CPR” OR “feedback device” OR “feedback devices” OR “adjunct device” OR “adjunct devices” OR ((”hand-held” OR portable) AND (device OR devices)) OR “mobile application” OR “mobile applications” OR “mobile apps” OR “electronic apps” OR “software application” OR “software applications” OR “software app” OR “software apps”))

Magiran

((”Cardiopulmonary resuscitation” OR “Compression” OR “CPR” OR “cardiac arrest”) AND “feedback”) OR (”Cardio First Angel” OR “CardioPump” OR “Beaty” OR “CPR Assist” OR “CPRCard” OR “CPR-plus” OR “CPR Pro” OR “CPRmeter” OR “CRP RsQ” OR “LinkCPR” OR “Pocket CPR” OR “TrueCPR” OR “U-CPR” OR “??-01” OR “PR-01”)

Russian Science Citation Index

(”Cardiopulmonary resuscitation” OR “Compression” OR “CPR” OR “cardiac arrest”) OR (”Cardio First Angel” OR “CardioPump” OR “Beaty” OR “CPR Assist” OR “CPRCard” OR “CPR-plus” OR “CPR Pro” OR “CPRmeter” OR “CRP RsQ” OR “LinkCPR” OR “Pocket CPR” OR “TrueCPR” OR “U-CPR” OR “??-01” OR “PR-01”)

Scientific Electronic Library Online (SciELO)

((”Cardiopulmonary resuscitation” OR “Compression” OR “CPR” OR “cardiac arrest”) AND “Feedback”) OR (”Cardio First Angel” OR “CardioPump” OR “Beaty” OR “CPR Assist” OR “CPRCard” OR “CPR-plus” OR “CPR Pro” OR “CPRmeter” OR “CRP RsQ” OR “LinkCPR” OR “Pocket CPR” OR “TrueCPR” OR “U-CPR” OR “??-01” OR “PR-01”)

Scopus

(TITLE-ABS-KEY (”real-time” OR “real time” OR “Cardio First Angel” OR “Ambu CardioPump” OR “CPR-1100 CPR Assist” OR cprcard OR cprezy OR “CPR-plus” OR “CPR PRO” OR cprmeter OR “CPR RsQ Assist” OR linkcpr OR pocket AND cpr OR truecpr OR “PR-01” OR qcpr OR “ZOLL AED Plus” OR “Intellisense CPR” OR “feedback device” OR “feedback devices” OR “adjunct device” OR “adjunct devices” OR ((”hand-held” OR portable) AND (device OR devices)) OR “mobile application” OR “mobile applications” OR “mobile apps” OR “electronic apps” OR “software application” OR “software applications” OR “software app” OR “software apps”))

AND

(TITLE-ABS-KEY (”Cardio-pulmonary resuscitation” OR “Cardio pulmonary resuscitation” OR “Basic Cardiac Life Support” OR “chest compression” OR “chest compressions” OR cpr))

Tübitak Ulakbim

((”Cardiopulmonary resuscitation” OR “Compression” OR “CPR” OR “cardiac arrest”) AND “feedback”) OR (”Cardio First Angel” OR “CardioPump” OR “Beaty” OR “CPR Assist” OR “CPRCard” OR “CPR-plus” OR “CPR Pro” OR “CPRmeter” OR “CRP RsQ” OR “LinkCPR” OR “Pocket CPR” OR “TrueCPR” OR “U-CPR” OR “??-01” OR “PR-01”)

Web of Science

TS=(”real-time” OR “real time” OR “Cardio First Angel” OR “Ambu CardioPump” OR “CPR-1100 CPR Assist” OR cprcard OR cprezy OR “CPR-plus” OR “CPR PRO” OR cprmeter OR “CPR RsQ Assist” OR linkcpr OR pocket AND cpr OR truecpr OR “PR-01” OR qcpr OR “ZOLL AED Plus” OR “Intellisense CPR” OR “feedback device” OR “feedback devices” OR “adjunct device” OR “adjunct devices” OR ((”hand-held” OR portable) AND (device OR devices)) OR “mobile application” OR “mobile applications” OR “mobile apps” OR “electronic apps” OR “software application” OR “software applications” OR “software app” OR “software apps”)

AND

TS=(”Cardio-pulmonary resuscitation” OR “Cardio pulmonary resuscitation” OR “Basic Cardiac Life Support” OR “chest compression” OR “chest compressions” OR cpr)