Sports safety matting diminishes cardiopulmonary resuscitation quality and increases rescuer perceived exertion

Objectives Compliant surfaces beneath a casualty diminish the quality of cardiopulmonary resuscitation (CPR) in clinical environments. To examine this issue in a sporting environment, we assessed chest compression quality and rescuer exertion upon compliant sports safety matting. Methods Twenty-seven advanced life support providers volunteered (13 male/14 female; mass = 79.0 ± 12.5 kg; stature = 1.77 ± 0.09 m). Participants performed 5 × 2 min, randomized bouts of continuous chest compressions on a mannequin, upon five surfaces: solid floor; low-compliance matting; low-compliance matting with a backboard; high-compliance matting; high-compliance matting with a backboard. Measures included chest compression depth and rate, percentage of adequate compressions, and rescuer heart rate and perceived exertion. Results Chest compression depth and rate were significantly lower upon high-compliance matting relative to other surfaces (p<0.05). The percentage of adequate compressions (depth ≥50 mm) was lowest upon high-compliance matting (40 ± 39%) versus low-compliance matting (60 ± 36%) and low-compliance matting with a backboard (59 ± 39%). Perceived exertion was significantly greater upon high-compliance matting versus floor, low-compliance matting, and low-compliance matting with a backboard (p<0.05). Conclusion Providers of CPR should be alerted to the detrimental effects of compliant safety matting in a sporting environment and prepare to alter the targeted compression depth and rescuer rotation intervals accordingly.


Introduction
Global survival rates following cardiac arrest are low (2.8-47.4%) [1] and influenced by factors including the quality of cardiopulmonary resuscitation (CPR) and automated defibrillator (AED) access [2]. Although cardiac arrest during exercise is rare, with an estimated incidence of 4.5 per million per year [3,4], sport and recreation venues have been identified as higherrisk environments [5]. In such venues, however, prompt bystander-delivered CPR and AED deployment are more likely and, therefore, confer a better outcome [6,7].
Chest compression depth and rate, two primary indices of CPR quality, partly determine outcome in cardiac arrest. Importantly, chest compressions should be performed upon a firm surface, and achieve a depth of 50-60 mm at a rate of 100-120 compressions per minute [2]. Achieving these targets may be influenced by many factors in a sports setting, including the surface present beneath a casualty, a factor which has been found to attenuate chest compression quality in clinical settings [8][9][10].
Sporting facilities are a non-clinical environment where compliant undersurfaces are frequently encountered (e.g. gymnastics facilities and martial arts gymnasiums). Indeed, some venues such as indoor climbing gymnasiums may feature very large expanses of floor area confluently overlaid with high-compliance matting. In these environments, considerable and/or unacceptable delay may be incurred while relocating a casualty to a firm surface upon which to perform chest compressions. Despite safety matting being commonplace in the sports and leisure industry, the influence of such surfaces on CPR quality has, to our knowledge, not yet been investigated.
To facilitate the development of future guidance on CPR delivery in sporting facilities which feature compliant matting surfaces, the aims of this study were: i) to ascertain the effect of compliant sports safety matting upon markers of chest compression quality and rescuer exertion, and ii) to investigate whether the use of a CPR backboard attenuated any detrimental effects of a more compliant surface.

Participants
Thirty-two advanced life support (ALS) providers (13 male/19 female; age: 30.1 ± 6.8 y; mass: 75.7 ± 14.1 kg; height: 1.75 ± 0.10 m; mean ± SD) were recruited from local paramedic (n = 10 participants), critical care (n = 6) and emergency medicine (n = 16) departments. Using previous work by Perkins et al. [8] and Tweed et al. [9] a power calculation was performed a-priori (G � Power, Dusseldorf, Germany; α: 0.05, β: 0.8, two-tailed), anticipating a medium effect size for the primary dependent variables (partial eta squared; η 2 p = 0.06). Employing a Repeated-Measures Analysis of Variance (ANOVA), n = 26 was estimated as sufficient. Leeds Beckett University ethical review board granted approval, and participants provided written, informed consent. All participants were free from known pre-existing cardiorespiratory illness (medical questionnaire) and medications known to affect the cardiovascular response to exercise. Before testing, participants were asked to abstain from alcohol and caffeine for 24h.

Experimental protocol
The effects of matting compliance on compression quality were assessed using a randomized, single-blind, crossover design at a commercial indoor climbing gymnasium. Participants delivered chest compressions on a training mannequin beneath which were matting surfaces of varying compliance. Measures of chest compression quality and rescuer exertion were recorded throughout.

Procedures
Following 10-min of quiet sitting, baseline telemetered heart rate (HR; Polar T31, Polar Electro, Finland) and rating of perceived exertion (RPE; CR-100 scale; [11]) were recorded. Prior to testing, a single 20 s bout of compressions was performed on the resuscitation mannequin with visual feedback from the software in an attempt to ensure consistent performance among participants. Following a brief rest, participants then performed 5 × 2-min bouts of continuous, compression-only CPR upon various surfaces (Table 1) in a randomized order, each separated by a 5-min seated rest period (Fig 1). Participants were instructed to adhere to the 2015 ERC resuscitation guidelines (Compression rate: 100-120 min -1 ; depth: 50-60 mm; [2]).

Measures
Chest compressions were performed on a commercially available resuscitation training mannequin (Resusci-Anne QCPR; Laerdal Medical, Norway), incorporating an electronicallymetered internal depth gauge and recoil spring. To achieve a torso mass equivalent to a 75 kg male phantom (32.6 kg; [12]), lead sachets were secured within the thorax compartment of the mannequin.
Compression data. Data on each compression were recorded automatically by the mannequin's proprietary software (SkillReporter PC, V 3.2.0.1; Laerdal Medical, Norway) for exactly 2-min, initiated at the first compression. Maximum and minimum compression depth (mm) was recorded for each compression (compression depth and depth at end of release). Correspondingly, the percentage of compressions attaining the target depth threshold (�50 mm) were recorded. Mean compression rate (min -1 ) was calculated from the inter- Table 1. Descriptions of surfaces beneath the mannequin during each compression bout.

Surface Abbreviation Description
Floor Floor Solid concrete floor.

LCBB
As for LC but with a rigid backboard (800 mm x 500 mm x 18 mm rigid board) inserted between the manneqin and matting.

HCBB
As for HC but with the rigid backboard.
All surfaces except for the Floor comprised three 100 mm layers of foam laid to a total depth of 300 mm, as is standard in venues similar to that where testing was undertaken. All surfaces were covered with an opaque sheet to remove visual cues for participants. N = Newtons.
https://doi.org/10.1371/journal.pone.0254800.t001 compression time interval. Mean compression data were reported as that averaged over the entire 2-min bout, unless specified. Rescuer exertion. Participant HR was recorded during the 60 seconds prior to each trial (HR Pre-Task ) and in the final 20 seconds of compressions (HR Peak ). Perceived exertion was rated immediately before each compression bout (RPE Pre-Task ) and immediately following the 2-min epoch (RPE Peak ).

Statistical analyses
Raw participant data were exported from the mannequin software to Microsoft Excel (Microsoft Corporation, Washington, USA) and analyzed using SPSS Version 26 (IBM, Chicago, USA). Normality of distribution was assessed using the Shapiro-Wilk test. Compression variables (compression rate, depth and percentage of adequate compressions) were compared among surfaces using a One-Way Repeated-Measures Analysis of Variance (ANOVA). Participant heart rate (HR Pre-Task vs. HR Peak ) and RPE (RPE Pre-Task vs. RPE Peak ) were analyzed using a Two-Way Repeated-Measures ANOVA. The Greenhouse-Geisser correction was applied where the assumption of sphericity was violated and all post-hoc analyses were performed with Bonferroni adjustment. Effect size was estimated using the partial eta-squared (η 2 p ) method and categorized as small = 0.01, medium = 0.06, or large = 0.14 [13]. Data are presented as mean ± SD with critical alpha-level applied as 0.05, and confidence interval of 95%.

Participants
From the 32 participants recruited, five subjects (all female) were excluded from the analysis owing to a failure to deliver chest compressions of the pre-determined cutoff depth (>40 mm; 80% of minimum target depth) on solid floor (mean compression depth: 34.4 ± 3.4 mm). When compared to the remaining female participants (n = 14), those excluded were of a similar age (28.2 ± 3.  Compression data Compression depth. Compression metrics among all surfaces are shown in Table 2 Table 2 and Fig 2). Post-hoc testing revealed that the HC and HCBB surfaces resulted in significantly lower compression rates relative to the Floor (p<0.001 and = 0.005, respectively), LC (p = 0.002 and = 0.007, respectively) and LCBB surfaces (p = 0.006 and = 0.005, respectively). No difference was observed between the HC and HCBB surfaces (p = 1.000), or between the Floor, LC and LCBB surfaces (all p = 1.000).

Rescuer exertion
Rating of perceived exertion. Data for RPE following post-hoc analyses are shown in Table 3 Post-hoc analyses revealed that the HC and HCBB surfaces elicited a greater RPE Peak than the Floor (p = 0.002 and = 0.002, respectively) and LC surfaces (p<0.001 and = 0.010, respectively), with the HC surface also significantly different to the LCBB surface (p = 0.001). There was no difference between the HC and HCBB (p = 1.000) surfaces or between the HCBB and LCBB surfaces (p = 0.101).
Heart rate. Data for HR following post-hoc analyses are shown in Table 3

Discussion
This study examined the effect of compliant safety matting on chest compression quality in a sporting environment. Our findings show that high-compliance safety matting found in sporting environments significantly reduced chest-compression quality (compression depth and rate), and increased rescuer perceived exertion. Both were partially improved with use of a backboard.
High-compliance safety matting, frequently encountered in sporting venues, significantly reduced mean compression depth by~3 mm, and attenuated the percentage of adequate compressions by up to 20% when compared to the Floor, LC and LCBB surfaces. The effect of the high-compliance surface was similar to that attributed to air-and foam-filled mattresses in a clinical environment [8,9,14]. Although the absolute decrement in chest compression depth with the HC surface may be relatively small, this difference in depth is comparable to that differentiating survival from non-survival during cardiac arrest [15]. It is also similar in magnitude to that conferring improved survival rates to hospital admission [16] and hospital discharge [17], suggesting this decrement might still be clinically meaningful.
Compression rate was also reduced upon both the HC and HCBB surfaces relative to the Floor, LC and LCBB surfaces. This is again consistent with hospital studies which show reduced compression rates on higher-compliance surfaces [18]. Compression depth and rate often show an inverse association [19][20][21], where the reduction in rate is due to the greater vertical hand-travel required to attain deeper compression depths [10,22], as may be the case on more compliant sports matting.
Notably, no differences were apparent between the Floor, LC and LCBB surfaces for compression depth or rate, suggesting effective CPR may be delivered at least upon certain lowercompliance matting types found in some sports venues. As the transport of a collapsed casualty to a firmer surface is likely to introduce an unacceptable delay to treatment [23], and in the absence of specific knowledge of the effects of all safety matting varieties, it seems prudent that rescuers performing CPR should anticipate a negative effect from sports matting, and aim to compensate appropriately.
Introducing a backboard partially attenuated the loss of compression quality on the HC surface, resulting in a 9% increase in the percentage of compressions achieving an adequate depth, and an increase in absolute mean compression depth of~1.6 mm. Although these measures did not attain statistical significance, research within hospital environments has shown reductions in mattress displacement, and associated rescuer workload, when backboards were used upon compliant hospital beds [10,22]. No difference in compression depth or the percentage of adequate compressions occurred when a backboard was used upon the low-compliance surface, and similarly no effects were observed on any surface for compression rate. Notwithstanding the above, the use of a backboard upon complaint sports surfaces needs further consideration, and where high-compliance undersurfaces are present, we would advocate backboard use if placement does not delay CPR.
Rescuer exertion is an important factor in the delivery of high-quality CPR. Relative to pretask values, heart rate and perceived exertion increased during all 2-min bouts of chest compressions (Table 3), with the HC and HCBB surfaces eliciting values for perceived exertion that were significantly higher than the remaining firmer surfaces. Previous work has described a rating of~35 on the CR-100 scale to correspond with the anaerobic threshold in active young adults [24]. Importantly, the HC and HCBB surfaces, elicited levels of effort closer to the

PLOS ONE
"Heavy" threshold (rating of 50 on the CR-100 scale; [11]). This higher exertion might be attributed to greater mechanical work associated with increased total vertical hand movement, as suggested by Noordergraaf et al. [10], and the elevated muscular effort required to stabilize the lower body on a compliant surface [9]. Our findings suggest that chest compressions upon high-compliance sports matting, may be associated with considerable physiological strain, limiting the duration for which effective CPR can be sustained, even by experienced providers. Such a scenario may, therefore, require more frequent rescuer changeovers during CPR. Finally, five participants were excluded from the analysis owing to a failure to attain the pre-defined mean compression depth on a solid floor. All were female and of a significantly lower body mass, height, and BMI compared to the remaining female participants (n = 14). Previous work has shown a positive relationship between body mass and stature in CPR efficacy [25]. What remains unclear is whether a more complaint surface may exacerbate this interaction between anthropometry and chest compression quality and warrants future investigation. It may be worth noting that inclusion of these participants in the analysis on an intention-to-treat basis did not alter our conclusions.
Sporting environments are considered higher-risk locations for cardiac arrest [5]. Yet the probability of early recognition of such an event, instigation of high-quality, bystander-delivered CPR, and greater defibrillator access all likely contribute to the superior outcomes observed following cardiac arrest in sporting venues [4,6,7]. Our findings highlight additional factors that may be modified or prepared for in order to improve the quality of CPR delivery in such environments. Although further research is required, particularly examining a broader range of matting compliance levels, future guidance should consider the influence of compliant safety matting upon chest compression depth and rescuer fatigue. Where appropriate, these effects should be acknowledged by medical personnel and also addressed during site-specific CPR training courses for sports venue staff.

Limitations
Firstly, we tested safety matting comprised of only two compliance levels and thus are unlikely to represent the entire range of sports protection surfaces. Nevertheless, the two matting types assessed were felt to be representative of those commonly encountered in most venues by the research team, participants and gymnasium staff. Moreover, the use of backboards gave our study an additional level of insight as to the effects of surface compliance on chest compression quality. Secondly, while every effort was made to blind our participants to the surfaces encountered, it is probable that most were able to detect some difference between at least matting, if not backboard conditions. In any event, if our participants were not effectively blinded, it may be that the magnitude of the decline in compression quality was underestimated.
Finally, we recruited a relatively homogenous group of highly-trained and experienced participants as an attempt to control for influences outside of the matting characteristics alone. The addition of data relating to trained bystanders or gymnasium staff may have been beneficial, particularly as these groups are likely to deliver the initial response to a collapsed patient while awaiting trained medical assistance.

Conclusions
We report a detrimental influence of high-compliance safety matting upon chest compression quality and rescuer exertion during simulated CPR in a sporting environment. This information may be used to guide improved CPR quality upon similar surfaces and may be particularly relevant within venues featuring large areas of compliant matting. These effects should be highlighted to first responders, medical staff and those working within sporting environments. Future research should aim to evaluate the clinical relevance of the decrement imposed by high-compliance matting and the influences of a broader range of surfaces. Moreover, the influence of a CPR backboard and the interactions among surface-type, rescuer training and experience level warrant further investigation.
Supporting information S1 File. Compression decay and degree of chest leaning. Additional results relating to magnitude of compression decay and chest leaning present upon each surface. (DOCX) S1