a b s t r a c t

Aim: The incidence of compression-Associated injuries from using the CLOVER3000, a new mechanical cardiopul- monary resuscitation (CPR) device, is not well studied in the context of out-of-hospital cardiac arrest (OHCA). Thus, we aimed to compare compression-associated injuries between CLOVER3000 and Manual CPR.

Methods: This single-center, retrospective, cohort study used data from the medical records of a tertiary care center in Japan between April 2019 and August 2022. We included adult Non-survivor patients with non-traumatic OHCA who were transported by emergency medical services and underwent post-mortem com- puted tomography. Compression-associated injuries were tested using logistic regression models adjusting for age, sex, bystander CPR performance, and CPR duration.

Results: A total of 189 patients (CLOVER3000, 42.3%; manual CPR, 57.7%) were included in the analysis. The over- all incidence of compression-associated injuries was similar between the two groups (92.5% vs. 94.54%; adjusted odds ratio (AOR), 0.62 [95% confidence interval (CI), 0.06-1.44]). The most common injury was anterolateral rib fractures with a similar incidence between the two groups (88.7% vs. 88.9%; AOR, 1.03 [95% CI, 0.38 to 2.78]). The second most common injury was sternal fracture in both groups (53.1% vs. 56.7%; AOR, 0.68 [95% CI, 0.36-1.30]). The Incidence rates of other injuries were not statistically different between the both groups.

Conclusion: We observed a similar overall incidence of compression-associated injuries between the CLOVER3000 and manual CPR groups on small sample size.

(C) 2023

1. Introduction

high-quality cardiopulmonary resuscitation (CPR) is crucial for the survival chain of cardiac arrest [1]. However, high-quality Manual compression for out-of-hospital cardiac arrest (OHCA) is limited by many factors, such as insufficient emergency medical service (EMS) manpower, rescuer fatigue, ambulance movement, requirements for airway management and intravenous line establishment, and infectious disease exposure [2]. In such cases, trained personnel must consider using Mechanical CPR devices (e.g. LUCAS(R) and Autopulse(R)) for im- proved CPR quality. However, previous randomized controlled studies [3-5] and a guideline [6] reported that mechanical CPR does not im- prove survival rate. In addition, a previous systematic review reported that mechanical CPR has a significantly higher overall rate of compression-associated injuries compared with manual CPR [7].

* Corresponding author at: Department of Emergency Medicine, Fukui Prefectural Hospital, 2-8-1 Yotsui, Fukui City, Fukui, Japan.

E-mail address: [email protected] (M. Hayashi).

In 2016, a new mechanical CPR device (CLOVER3000; KOHKEN Medical Co., Ltd., Tokyo, Japan) [8] tailored to the Japanese physique was released in Japan. In general, the appropriate compression force of a device varies: a stiffer chest requires a more forceful compression than a softer chest and an increasing number of compressions is needed for decreased chest stiffness [9]. Given a strong compression force, prolonged compressions may increase compression-associated injuries. The CLOVER3000 can be adjusted for CPR compression force through the adjustment knob. Therefore, the incidence of compression- associated injuries from using CLOVER3000 may be lower than that of manual CPR.

Little is known about the survival rate and the incidence of compression-induced injuries between mechanical CPR using CLOVER3000 and manual CPR for OHCA. The hypothesis of this study was that the incidence of compression-associated injuries from using CLOVER3000 might be lower than that of manual CPR. Thus, in this study, we aimed to compare compression-associated injuries between mechanical and manual CPR in non-survivor patients of OHCA using post-mortem computed tomography (PMCT).

https://doi.org/10.1016/j.ajem.2023.03.032

0735-6757/(C) 2023

Fig. 1. CLOVER3000 (reprinted with permission from KOHKEN Medical Co., Ltd.) A: CLOVER3000 is an active compression-decompression CPR device. Its optimal compression force can be adjusted to a depth of 5 cm through the adjustment valve. B: CLOVER3000 can perform mechanical ventilation using compressed oxygen gas.

1. Methods
   1. CLOVER3000 [8] (Fig. 1)

CLOVER3000 performs chest compressions at a depth of 5 cm and a rate of 100/min. It performs CPR at a cycle of 30 compressions and 2 breaths for synchronized ventilation and a rate of 10 breaths per minute for non-synchronized ventilation. It is an active compression- decompression CPR device similar to the LUCAS(R) Chest Compression System (Jolife AB, Lund, Sweden) [10]. However, the two devices have several differences. First, the optimal compression force of CLOVER3000 can be adjusted to a depth of 5 cm through the adjustment valve. Hence, even with thoracic deformity due to fractures associated with chest compression, an appropriate depth of chest compression is possible. Second, it provides both CPR and ventilation using compressed oxygen gas in a single device, requiring little manpower. Third, it is not a suction cup but an angle-adjustable silicone cup. Therefore, parallel chest compression is possible even with a round back. Finally, it can be used for patients with a chest thickness of >12 cm, while LUCAS(R) can only be used for patients with a chest thickness of >17 cm. Thus, compared with Lucas(R), CLOVER3000 can be used in patients of smaller size.

• 1. Data source

This single-centre, retrospective, cohort study used data from the medical records of Fukui Prefectural Hospital between April 2019 and August 2022. Fukui Prefectural Hospital is a tertiary care centre with approximately 22,000 emergency department visits annually. Fukui City EMS began to use CLOVER 3000 in April 2021, and thereafter prioritized the use of CLOVER3000 over manual CPR. The inclusion criteria for the recruited patients were as follows: (1) non-survivor patients who were 18 years or older, (2) had non-traumatic OHCA,

(3) underwent CPR by Fukui City EMS, (3) were transported by the Fukui City ambulance, and (4) underwent PMCT. The exclusion criteria were as follows: (1) underwent CPR using devices other than CLO- VER3000 during transport or emergency department and (2) priori malformations. We obtained the following data from the medical records: age, sex, witness presence, bystander CPR performance, initial rhythm, CPR duration, pre-hospital airway management, use of CPR de- vices, and cause of death. CPR duration was defined as the time from EMS contact to CPR termination. Two emergency physicians (MH and ST) independently assessed PMCT data for all cases of the following

compression-associated injuries: rib fractures (number and location), haemothorax, pneumothorax, sternal fracture, retrosternal haematoma, pericardial haemorrhage, haemoperitoneum, retroperitoneal haemor- rhage, Free air, and vertebral fractures. We included both complete and incomplete rib fractures [7]. We divided the third locations (anterior, lateral, or posterior) according to the distance of the rib frac- ture from the costochondral junction [11]. We defined haemorrhage as fluid collection with an attenuation of >20 HU [12]. Cases of critical haemorrhage as the likely cause of death were excluded from the examined compression-associated injuries (e.g. pericardial haemor- rhage associated with aortic dissection and haemoperitoneum associ- ated with ruptured Abdominal aortic aneurysm). We decided the cause of death from both medical records and PMCT data. Data assessors were not blinded to the intervention. A 64-row, whole-body PMCT (SOMATOM Definition, Siemens, Forchheim, Germany) was performed on all patients, and the CT slice thickness was 4 mm for the head and 5 mm for the thorax and pelvis.

• 1. Study outcome

The primary outcome was the incidence of compression-associated injuries: rib fractures (>=3 and location), haemothorax, pneumothorax, sternal fracture, retrosternal haematoma, pericardial haemorrhage,

haemoperitoneum, retroperitoneal haemorrhage, free air, and vertebral fractures.

• 1. Statistical analysis

We used frequencies and percentages for categorical variables and medians (interquartile range [IQR]) for continuous variables. We used the chi-square test to compare categories and the Mann-Whitney U test to compare continuous measures. A two-sided p value <0.05 was considered statistically significant. According to a previous study, we adjusted for age, sex, bystander CPR performance, and CPR duration [13]. Compression-associated injuries were tested using logistic regres- sion models adjusting for covariates listed previously [13]. All statistical analyses were performed using EZR, a graphical user interface for R (version 4.0.3; the R Foundation for Statistical Computing, Vienna, Austria). Specifically, EZR is a modified version of the R commander (version 2.7-1), and was designed to add statistical functions frequently used in biostatistics [14].

Fig. 2. Study participant flow.

• 1. Ethical statement

The study was approved by the institutional review board of Fukui Prefectural Hospital (No. 22-16), which waived the requirement for informed consent due to the retrospective nature of the study.

1. Results

A total of 245 patients had OHCA between April 2019 and August 2022. We excluded 55 patients who met the exclusion criteria and one patient who were performed emergency department thoracotomy for abdominal aortic aneurysm rupture, a total of 189 patients (77.1%) were included in the study. We categorized the patients into two groups: mechanical (using CLOVER3000; n = 80) and manual CPR (n = 109) groups (Fig. 2). Before April 2021, Manual CPR was per- formed in all 94 patients. After Fukui City EMS started to use CLO- VER3000, 80/95 (84.2%) patients were used CLOVER3000 and 15/95 (15.8%) patients were performed manual CPR in the study period.

Table 1 shows the patient characteristics. The median ages were

83 years (IQR: 72.75-90 years) in the CLOVER3000 group and

Table 1

Patient characteristics.

78 years (IQR: 70-85 years) in the manual CPR group. The median dura- tions of CPR were 32 min (IQR: 27-36 min) in the CLOVER3000 group and 31 min (IQR: 25-40 min) in the manual CPR group.

Table 2 shows the compression-associated injuries. The overall incidence of compression-associated injuries was similar between the two groups (92.5% vs 94.54%; adjusted odds ratio (AOR), 0.62 [95% con- fidence intervals (CI), 0.06-1.44]; p = 0.45). The most common injury was anterolateral rib fracture for both groups. Likewise, its incidence was similar between groups (88.7% vs 88.9%; AOR, 1.03 [95% CI: 0.38-2.78]; p = 0.95). The median number of rib fractures was

also similar between the groups (6 vs 7; AOR, -0.67 [95% CI -1.71 to 0.37]; p = 0.25). The second most common injury was sternal fracture for both groups (53.1% vs 56.7%; AOR, 0.68 [95% CI, 0.36-1.30]; p =

0.24). The incidence rates of other associated injuries were not statisti- cally different between the two groups.

1. Discussion

In summary, we observed a similar overall incidence of compression-associated injuries between the CLOVER3000 and manual CPR groups. The most common injury was anterolateral rib fracture, followed by sternal fracture for the both groups. To the best of our knowledge, this is the first study to characterize compression- associated injuries between CLOVER3000 and manual CPR.

CLOVER3000

group,

n = 80

Manual CPR group,

n = 109

p value

In our study, the overall incidence of compression-associated injuries was similar between the groups (92.5% vs. 94.54%, p = 0.45). In a retrospective study by PMCT, the LUCAS and manual CPR groups

Age (years, median, IQR) 83 (72.75-90) 78 (70-85) 0.054

Male, n (%) 49 (61.2) 59 (54.1) 0.4

Witnessed cardiac arrest, n (%) 21 (26.2) 32 (29.3) 0.76

Bystander CPR, n (%) 44 (55) 55 (50.4) 0.63

Initial rhythms, n (%) 0.67

Asystole 63 (78.7) 81 (74.3)

Pulseless electrical activity 15 (18.7) 23 (21.1)

Ventricular fibrillation or pulseless 2 (2.5) 5 (4.5) ventricular tachycardia

CPR duration (min), median, IQR 32 (27-36) 31 (25-40) 0.53

Airway management, n (%) 0.39

mask ventilation 29 (36.2) 34 (31.1)

supraglottic airway device 37 (46.2) 61 (55.9)

endotracheal tube 14 (17.5) 14 (12.8)

Suspected cause of death, n (%) 0.67

Heart disease 12 (15) 18 (16.5)

thoracic aortic dissection 6 (7.5) 14 (12.8)

Others 26 (32.5) 32 (29.3)

Uncertain 36 (45) 45 (41.2) Abbreviations; CPR: cardiopulmonary resuscitation, IQR: interquartile range.

had a similar overall incidence of compression-associated injuries (100% vs. 95%, p = 0.455) [15]. In contrast, in a systematic review and meta-analysis, by autopsy, the LUCAS group had a significantly higher overall incidence of compression-associated injuries than the manual CPR group (87.6% vs. 68.6%; OR, 1.29 [95% CI, 1.19-1.41]; I2 = 21.83%)

[7,16-19]. These differences may be associated with the chosen exami- nation method (i.e. PMCT or autopsy). Another retrospective study re- ported that compared with autopsy, PMCT detected more cases of incomplete rib fractures but fewer ribs (ribs 1-3 and 7-12) and fracture locations (posterolateral and paravertebral) [20]. Therefore, our find- ings suggested that the CLOVER3000 may be as safe as manual CPR.

In our study, the most common injury was anterolateral rib fracture for both groups (88.7% vs. 88.9%, p = 0.95). The incidence of multiple rib fractures (>= 3) was also similar between groups (77.5% vs. 84.4%, p = 0.24). In a systematic review and meta-analysis, the LUCAS group had an increased overall risk of rib fractures (OR, 1.23 [95%

CI, 1.12-1.35]; I2 = 22.35%) and multiple rib fractures (OR, 1.45 [95%

Table 2 Compression-associated injuries.
	CLOVER3000 group,	Manual CPR group,	Unadjusted OR	p value	Adjusted OR (95%CI)	p value
	n = 80	n = 109	(95% CI)
Overall compression-associated injuries, n (%)	74 (92.5)	103 (94.4)	0.78 (0.22 to 2.32)	0.58	0.62 (0.06 to 1.44)	0.45
Number of rib fractures, median (IQR)	6 (3-9)	7 (4-10)	-0.50 (-1.60 to 0.58)	0.36	-0.67 (-1.71 to 0.37)	0.25
Rib fractures ?3, n (%)	62 (77.5)	92 (84.4)	0.63 (0.30 to 1.33)	0.23	0.53 (0.24 to 1.18)	0.12
Anterolateral rib fracture, n (%)	71 (88.7)	97 (88.9)	1.11 (0.43 to 2.86)	0.82	1.03 (0.38 to 2.78)	0.95
Posterior rib fracture, n (%)	3 (3.7)	9 (8)	0.43 (0.11 to 1.65)	0.22	0.48 (0.12 to 1.93)	0.3
Haemothorax, n (%)	1 (1.2)	5 (4.5)	0.26 (0.03 to 2.30)	0.22	0.45 (0.042 to 3.84)	0.43
Pneumothorax, n (%)	4 (5)	8 (7.3)	0.66 (0.19 to 2.29)	0.51	0.68 (0.19 to 2.41)	0.55
Sternal fracture, n (%)	42 (53.1)	59 (56.7)	0.86 (0.48 to 1.56)	0.63	0.68 (0.36 to 1.30)	0.24
Retro sternal hematoma, n (%)	19 (23.7)	22 (20.1)	1.23 (0.61 to 2.47)	0.55	1.01 (0.48 to 2.10)	0.98
Pericardial haemorrhage, n (%)	4 (5)	6 (5.5)	0.90 (0.24 to 3.31)	0.87	0.82 (0.21 to 3.18)	0.77
Haemoperitoneum, n (%)	3 (3.7)	2 (1.8)	2.08 (0.34 to 12.80)	0.42	1.58 (0.22 to 11.1)	0.64
Free air, n (%)	0 (0)	1 (1.2)	n/a	n/a	n/a	n/a
Retroperitoneal haemorrhage, n (%)	0 (0)	0 (0)	n/a	n/a	n/a	n/a
Vertebral fracture, n (%)	4 (5)	1 (1.2)	5.68 (0.62 to 51.9)	0.12	4.71 (0.46 to 48.1)	0.19

Abbreviations; OR: Odds Ratio, CI: Confidence Intervals, CPR: cardiopulmonary resuscitation, IQR: interquartile range, n/a: not applicable. Compression-associated injuries were tested using logistic regression models, adjusting for age, sex, bystander CPR, and CPR duration.

CI, 1.13 to 1.87]; I2 = 62.36%) than the manual CPR group [7]. Six of seven studies included in the meta-analysis of rib fractures were evalu- ated by autopsy, and in one study by PMCT, the LUCAS and manual CPR groups had a similar overall risk of rib fractures (OR, 1.29 [95% CI, 1.19-1.41]) [15]. All studies included in the meta-analysis of multiple rib fractures were evaluated by autopsy. Thus, CLOVER3000 may be as safe as manual CPR in terms of rib fracture risk.

In our study, the incidences of other injuries were similar in both groups. In a systematic review and meta-analysis, the LUCAS group had an increased overall risk of sternal fracture, and haemoperitoneum than the manual CPR group [3]. Only one of eight studies included in the meta-analysis of Sternal fractures were evaluated by PMCT, the LUCAS and manual CPR groups had a similar overall risk of rib fractures (OR, 1.18 [95% CI, 0.76-1.83]) [15]. All studies included in the meta- analysis of haemoperitoneum were evaluated by autopsy. Thus, CLOVER3000 may be as safe as manual CPR in terms of sternal fracture, and haemoperitoneum risks.

Our findings suggested that CLOVER3000 might be as safe as manual CPR in terms of compression-associated injuries in non-survivor patients of OHCA. Further larger studies are needed on the safety of CLOVER3000 compared to manual CPR and other mechanical CPR devices. Future studies should focus on compression-associated injuries of CLOVER3000 compared to manual CPR and other mechanical CPR devices in survivor patients of OHCA.

1. Limitations

This study had several limitations. First, this was a retrospective study of a single center in Japan and was based on a small sample size. Second, autopsy was not conducted in this study, which is the gold standard for detecting compression-associated injuries. However, PMCT, which was included in the study, has been suggested as an alternative in recent years [20]. Third, PMCT data assessors were not ra- diologists. Therefore, PMCT data might be misinterpreted. Fourth, our study was for non-survivor patients with CPR duration longer than 30 min. Therefore, compression-associated injuries in survivor patients with shorter CPR duration are unknown. Finally, all patients were Japanese; thus, the results may differ in other groups.

1. Conclusion

We observed a similar overall incidence of compression-associated injuries between the CLOVER3000 and manual CPR groups on small sample size. The most common injury was anterolateral rib fracture, followed by sternal fracture for both groups. Our findings suggested that CLOVER3000 might be as safe as manual CPR in terms of compression-associated injuries in non-survivor patients of OHCA.

Funding

None.

CRediT authorship contribution statement

Minoru Hayashi: Writing - review & editing, Writing - original draft, Visualization, Validation, Software, Resources, Project administra- tion, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Shinsuke Tanizaki: Writing - review & editing, Supervision. Naru Nishida: Writing - review & editing. Haruki Mizuno: Writing - review & editing. Kenichi Kano: Writing - review & editing. Jyunya Tanaka: Writing - review & editing. Hiroyuki Azuma: Writing - review & editing. Makoto Sera: Writing - review & editing. Hideya Nagai: Writing - review & editing. Shigenobu Maeda: Writing - review & editing.

Declaration of Competing Interest

None.

Acknowledgement

We thank Editage for the English language editing.

References

1. Panchal AR, Bartos JA, Cabanas JG, et al. Part 3: adult basic and advanced life support: 2020 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation. 2020;142:S366-468. https://doi.org/10. 1161/CIR.0000000000000916.
2. Krarup NH, Terkelsen CJ, Johnsen SP, et al. Quality of cardiopulmonary resuscitation in out-of-hospital cardiac arrest is hampered by interruptions in chest compressions-a nationwide prospective feasibility study. Resuscitation. 2011;82: 263-9. https://doi.org/10.1016/j.resuscitation.2010.11.003.
3. Wik L, Olsen JA, Persse D, et al. Manual vs. integrated automatic load-distributing band CPR with equal survival after out of hospital cardiac arrest. The randomized CIRC trial. Resuscitation. 2014;85:741-8. https://doi.org/10.1016/j.resuscitation. 2014.03.005.
4. Rubertsson S, Lindgren E, Smekal D, et al. Mechanical chest compressions and simultaneous defibrillation vs conventional cardiopulmonary resuscitation in out- of- hospital cardiac arrest: the LINC randomized trial. JAMA. 2014;311:53-61. https://doi.org/10.1001/jama.2013.282538.
5. Perkins GD, Lall R, Quinn T, et al. Mechanical versus manual chest compression for out-of-hospital cardiac arrest (PARAMEDIC): a pragmatic, cluster randomised con- trolled trial. Lancet. 2015;385:947-55. https://doi.org/10.1016/S0140-6736(14) 61886-9.
6. Soar J, Bottiger BW, Carli P, et al. European resuscitation council guidelines 2021: adult advanced life support. Resuscitation. 2021;161:115-51. https://doi.org/10. 1016/j.resuscitation.2021.02.010.
7. Gao Y, Sun T, Yuan D, et al. Safety of mechanical and manual chest compressions in cardiac arrest patients: a systematic review and meta-analysis. Resuscitation. 2021; 169:124-35. https://doi.org/10.1016/j.resuscitation.2021.10.028.
8. Product information. CLOVER 3000: KOHKEN Medical, Tokyo, Japan. (Accessed 23 March 2023). https://www.kohkenmed.co.jp/products/kom330/; 2023.
9. Tomlinson AE, Nysaether J, Kramer-Johansen J, Steen PA, Dorph E. Compression force-depth relationship during out-of-hospital cardiopulmonary resuscitation. Resuscitation. 2007;72:364-70. https://doi.org/10.1016/j.resuscitation.2006.07.017.
10. Product information. LUCAS: Jolife AB, Sweden. (Accessed 10 November 2022). https://www.lucas-cpr.com/product_specifications/; 2023.
11. Yang K, Lynch M, O’Donnell C. “Buckle” rib fracture: an artifact following cardio- pul- monary resuscitation detected on postmortem CT. Leg Med. 2011;13:233-9. https:// doi.org/10.1016/j.legalmed.2011.05.004.
12. Koga Y, Fujita M, Yagi T, et al. Effects of Mechanical chest compression device with a load-distributing band on post-resuscitation injuries identified by post-mortem computed tomography. Resuscitation. 2015;96:226-31. https://doi.org/10.1016/j. resuscitation.2015.08.013.
13. Ram P, Menezes RG, Sirinvaravong N, et al. Breaking your heart-a review on CPR- related injuries. Am J Emerg Med. 2018;36:838-42. https://doi.org/10.1016/j.ajem. 2017.12.063.
14. Kanda Y. Investigation of the freely available easy-to-use software “EZR” for medical statistics. Bone Marrow Transplant. 2013;48:452-8. https://doi.org/10.1038/bmt. 2012.244.
15. Baumeister R, Held U, Thali MJ, Flach PM, Ross S. Forensic imaging findings by post- mortem computed tomography after manual versus mechanical chest compression. J Forensic Radiol Imaging. 2015;3:167-73. https://doi.org/10.1016/j.jofri.2015.08.001.
16. Smekal D, Lindgren E, Sandler H, Johansson J, Rubertsson S. CPR-related injuries after manual or Mechanical chest compressions with the LUCASTM device: a multicentre study of victims after unsuccessful resuscitation. Resuscitation. 2014;85:1708-12. https://doi.org/10.1016/j.resuscitation.2014.09.017.
17. Lardi C, Egger C, Larribau R, Niquille M, Mangin P, Fracasso T. Traumatic injuries after mechanical cardiopulmonary resuscitation (LUCAS2): a forensic autopsy study. Int J Leg Med. 2015;129:1035-42. https://doi.org/10.1007/s00414-015-1146-x.
18. Ondruschka B, Baier C, Bayer R, Hammer N, Dressler J, Bernhard M. Chest compression-associated injuries in cardiac arrest patients treated with manual chest compressions versus automated chest compression devices (LUCAS II) - a fo- rensic autopsy-based comparison. Forensic Sci Med Pathol. 2018;14:515-25. https://doi.org/10.1007/s12024-018-0024-5.
19. Smekal D, Johansson J, Huzevka T, Rubertsson S, et al. Resuscitation. 2009;80: 1104-7. https://doi.org/10.1016/j.resuscitation.2009.06.010.
20. Hamanaka K, Nishiyama K, Nakamura M, Takaso M, Hitosugi M. Both autopsy and computed tomography are necessary for accurately detecting rib fractures due to cardiopulmonary resuscitation. Diagnostics (Basel). 2020;10:697. https://doi.org/ 10.3390/diagnostics10090697.