|Home | About | Journals | Submit | Contact Us | Français|
Determine the impact of backboard placement, torso weight and bed compression on chest compression (CC) depth feedback in simulated cardiac arrest patients.
Epochs of 50 high quality CCs with real-time feedback of sternum-to-spine compression depth were provided by a blinded BLS/ACLS/PALS certified provider on manikins of two torso weights (25 vs. 50 kg), using three bed surfaces (stretcher, Stryker hospital bed with Impression mattress, soft Total Care ICU bed), with/without a backboard (BB). Two BB sizes were tested (small: 60 cm × 50 cm; large: 89 cm × 50 cm) in vertical vs. horizontal orientation. Mattress displacement was measured using an accelerometer placed internally on the spine plate of the manikin. Mattress displacement of ≥5 mm was prospectively defined as the minimal clinically important difference.
During CPR (CC depth: 51.8 ± 2.8 mm), BB use significantly reduced mattress displacement only for soft ICU beds. Mattress displacement was reduced (vs. no BB) for 25 kg torso weight: small BB12.3 mm (95%CI 11.9–12.6), horizontally oriented large BB 11.2 mm (95%CI 10.8–11.7), and vertically oriented large BB 12.2 mm (95%CI 11.8–12.6), and for 50 kg torso weight: small BB 7.4 mm (95%CI 7.1–7.8), horizontally oriented large BB 7.9 mm (95%CI 7.6–8.3), and vertically oriented large BB 6.2 mm (95%CI 5.8–6.5; all p < 0.001). BB size and orientation did not significantly affect mattress displacement. Lighter torso weight was associated with larger displacement in soft ICU beds without BB (difference: 6.9 mm, p < 0.001).
BB is important for CPR when performed on soft surfaces, such as ICU beds, especially when torso weight is light. BB may not be needed on stretchers, relatively firm hospital beds, or for patients with heavy torso weights.
Best estimates suggest pediatric cardiac arrest occurs in about 16,000 children each year, with about half of cardiac arrests inhospital.1,2 High quality CPR has repeatedly been associated with improved patient outcomes.3–5
Chest compressions (CCs) provided for children during inhospital cardiac arrest are typically performed on soft surfaces such as ICU beds, and less frequently on firm hospital beds or firm transport stretchers.6,7 Traditionally, clinicians have placed a backboard under the patient to minimize the mattress displacement to provide more effective CC and decrease work of CC delivery. However, recently published evidence-based resuscitation guidelines recommend an optional use of backboards, due to a lack of evidence to support or refute backboard effectiveness to improve quality of CPR.8,9
Our previous study using forensic reconstruction techniques demonstrated substantial movement (displacement) in supporting systems (mattress and bed) during actual CCs. 7 However, very few studies evaluate the effectiveness of the backboard directly, and the results are inconclusive.10–17 An important reason the prior results are inconclusive is that the quality of CC is often not optimized in these experiments.11
CPR feedback and coaching devices to guide CC providers have become more available over the last a few years.18 However, devices that coach based upon sternal movement may overestimate the patient’s sternum-to-spine CC depth due to the mattress displacement, and potentially result in shallower CCs and worse patient outcomes.14 Smarter CPR feedback devices are on the horizon to correct those errors and minimize this overestimation by compensating for mattress displacement or accurately measuring actual sternum-to-spine CC depth. Therefore, we attempted to replicate high quality CPR with accurate 2010 guideline compliant sternum-to-spine depth feedback to evaluate the effectiveness of backboard use on a variety of hospital bed surfaces.
Our objective for this study was to evaluate the impact of backboard placement, torso weight, and bed/mattress displacement on anterior sternal movement based CC depth in simulated in-hospital cardiac arrest patients. We hypothesized that: (1) the backboard use would reduce displacement of the supporting system ≥ 5 mm (i.e., firm stretcher, hospital bed with firm mattress, soft ICU bed) and (2) the effectiveness of a backboard to reduce mattress displacement would be affected by bed surface stiffness, torso weight, and backboard orientation (horizontal vs. vertical).
The study protocol was approved by the institutional review board at The Children’s Hospital of Philadelphia.
Simulated CPR settings were generated controlling for (1) Manikin characteristics, (2) Torso weight, (3) bed surface stiffness, (4) backboard size, and (5) backboard orientation.
A Voice Advisory Manikin with Skill Reporter System (Laerdal, Wappinger Falls, NY, USA) was used as a testing manikin. The internal displacement sensor reported the manikin anterior sternum-to-spine compression depth for every CC. A Heart-start 4000 (Philips Medical Systems, Andover, MA, USA) monitor/defibrillator equipped with a force and displacement sensor (FDS) placed on top of the manikin’s sternum was used to collect total compression depth. The monitor/defibrillator records data on CC quality (CC rate/min, depth (mm), force (kg). Two different weights, cut to the shape of the manikin torso were placed underneath the manikin on the backboard to adjust the torso weight to 25 kg and 50 kg, representing a child and adult torso, respectively. Our previous studies have shown that the stiffness of the child and elderly adult chests during CPR are similar,19 thus justifying the use of the same manikin stiffness (i.e., spring stiffness) to represent elderly adult and child, and change only the manikin torso mass.
Three different bed surfaces were used to evaluate the effect of backboard placement on displacement of the mattress during CCs: (1) firm stretcher: Steris Hausted Horizon (thickness 6 cm; Steris Corp, Mentor, OH); (2) firm hospital bed: Stryker bed with Impression mattress (thickness: 13 cm; Stryker Medical, Portage, MI, USA), and (3) standard soft ICU bed: Total Care Bed with a therapy mattress (thickness: 14 cm; Hill Rom Corp, Batesville, IN, USA). Two different size backboards were assessed: (1) small backboard (Small backboard: 59 cm × 50.5 cm × 1 cm, weight 3.5 kg) which is currently used in our hospital and (2) a large backboard (Large backboard: 88.7 cm × 50.5 cm × 1 cm, weight 5 kg).
A single American Heart Association Basic Life Support, Advanced Cardiac Life Support, and Pediatric Advanced Life Support certified provider experienced in adult and pediatric resuscitation performed CCs. Real-time sternum-to-spine CC depth feedback from the Voice Advisory Manikin with Skill Reporter System was used to guide high quality 2010 guideline compliant CC. Chest compressions were targeted using real-time feedback to achieve 50 mm depth with a rate of 100/min and complete release between CCs. The provider was blinded to the presence, size and orientation of the backboard. Specifically, each backboard condition was prepared by separate investigators without the provider present, and the bed surface was covered by white sheets. The CC provider focused on the real-time feedback displayed on the computer screen that was at eye height level across the bed. Fifty consecutive CCs were performed in each setting. A step stool was utilized as needed to achieve real-time feedback-guided CC goals.
To evaluate the effect of backboard presence, size and orientation, a total of 24 conditions were evaluated based on the possible combinations of two different torso weights (25 and 50 kg), three bed types (firm stretcher, firm hospital bed, soft ICU bed), and four different backboard conditions (no backboard, small backboard, large backboard in horizontal position, large backboard in vertical position) [Fig. 1].
Mattress displacement was directly measured by an accelerometer placed on the spine plate inside the manikin torso. This method was previously published to measure the mattress displacement during simulated CCs.7 Our measurement of mattress displacement was performed by accelerometer-based measures with an error range of 3 mm.7,19–22 The mattress displacement for each CC was calculated by double integration from the accelerometer measurement. The CC depth measures were recorded and stored in the Heart-start 4000 defibrillator and Voice Advisory Manikin (VAM) with Skill Reporter System.
Our primary outcome measures are the difference in mattress displacement during CCs with or without backboards. The mattress displacement and chest compression data were extracted to an Excel spreadsheet (Microsoft Corp, Redmond, WA, USA). Summary data were reported as mean ± standard deviation (sd). An independent t-test was used for univariate analysis. The effect of backboard was reported as the reduction in mattress displacement. We prospectively defined ≥5 mm reduction in mattress displacement as a minimally clinically important difference from the existing literature.4,23 An estimated mean with 95% confidence interval (CI) was reported for the difference between testing conditions. A two-sided test with alpha = 0.05. STATA 11.0 (Stata Corporation, College Station, TX, USA) was used throughout.
A total of 24 conditions with 1191 CCs were analyzed. High quality CCs guided by real-time sternum-to-spine feedback were delivered with mean depth 51.8 ± 2.8 mm and mean force 42.8 ± 4.9 kg. Table 1 demonstrates the delivered CC force and sternum-to-spine depth as well as the measured mattress displacement under various conditions.
Backboard use reduced the mattress displacement significantly (i.e. ≥5 mm) only in soft ICU beds for 25 kg and 50 kg torso weight and in the 25 kg patient on a firm stretcher (Table 2). Reduction in mattress displacement for the 25 kg torso weight on the soft ICU bed was 12.3 mm (95% CI 11.9–12.6 mm) with a small backboard, 11.2 mm (95% CI 10.8–11.7 mm) with a horizontally placed large backboard, and 12.2 mm (95% CI 11.8–12.6 mm) with a vertically placed large backboard (p < 0.001 for all comparisons). Reduction in mattress displacement for the 50 kg torso weight on the soft ICU bed was: small backboard 7.4 mm (95% CI 7.1–7.8 mm, p < 0.001), horizontally placed large backboard 7.9 mm (95% CI 7.6–8.3 mm, p < 0.001) and vertically placed large backboard 6.2 mm (95% CI 5.8–6.5 mm, p < 0.001). Reduction in mattress displacement for the 25 kg torso weight on the firm stretcher was 10.9 mm (95% CI 10.6–11.2 mm, p < 0.001) with the backboard placed vertically.
The size and orientation of backboard did not significantly affect mattress displacement except in the condition utilizing a vertically placed large backboard on a firm stretcher with 25 kg torso weight (difference between small vs. vertically placed backboard, 6.2 mm [95% CI 6.0–6.4 mm, p < 0.001], between horizontally placed large backboard vs. vertically placed large backboard, 8.2 mm [95% CI 7.8–8.4 mm, p < 0.001]). The difference was not significant for the 50 kg torso weight (difference between small vs. vertically placed large backboard, 1.6 mm [95% CI 1.5–1.8 mm, p > 0.99], between horizontally placed large backboard vs. vertically placed large backboard, 0.2 mm [95% CI 0.0–0.4 mm, p > 0.99]).
Overall, lighter torso weight (25 kg) was associated with larger mattress displacement in all bed/backboard conditions. This, however, became significant only in CCs delivered on a soft ICU bed without backboard (difference 6.9 mm, 95% CI: 6.4–7.3 mm, p < 0.001).
Use of a backboard during in-hospital CPR is traditionally recommended to improve the quality of CC depth. 24 However, few studies have evaluated the effect of the backboard size and orientation, and the results are conflicting, which led to an inconclusive statement in the current resuscitation consensus and guidelines.8,9 In this study, we evaluated the effectiveness of various backboards to decrease the mattress displacement under realistic and varying realistic clinical conditions. Real-time sternum-to-spine depth feedback to the CPR provider was used to maintain high quality CC, regardless of bed surface. In order to avoid “statistically significant but not clinically important” results, we a priori defined the minimal clinically important difference in mattress displacement as ≥5 mm, based on previous clinical investigations (i.e., how much depth loss in the mattress would affect patient outcome).4,23
Our study results were somewhat surprising. Use of backboards did not significantly reduce the mattress displacement during CPR on firm stretchers or firm standard hospital beds, except on a stretcher with a lightly weighted torso when the large backboard was oriented vertically. On soft ICU beds, however, the backboard induced reduction in mattress displacement was >5 mm in all conditions, and ≥10 mm when the torso weight was light (25 kg). This suggests that backboards should be used for CPR events in pediatric and adult ICUs where the majority of patients are in soft ICU beds. The finding is intuitive if we acknowledge that lighter torso weights sink less into soft bed mattresses and thus have more potential for mattress displacement during CCs. Heavier torso weights cause the torso to sink into the mattress even when the CC force is released completely (i.e. no leaning), therefore leaving less potential for mattress displacement during chest compression.
Compared to our previously reported clinical study in older children on similar mattress surfaces and backboard positions, our study demonstrated a larger bed surface displacement.7 This difference is likely due to deeper sternum-to-spine CC depth targets with new 2010 International Liaison Committee on Resuscitation (ILCOR) guidelines (increasing from ≥38 mm to >50 mm), and delivery of more forceful CC.
Boe and Babbs previously evaluated the effect of the bed surface and backboard use applying a sophisticated mathematical model.25 In their study, the force of compression was held constant at 400 Newton (40.8 kg), described as a constant peak force technique. Interestingly, our findings of the 20% increase in total compression when a backboard is not placed are similar to the 15% decrease in sternum-to-spine compression distance that was reported by Boe and Babbs on a typical stiff hospital mattress without a backboard.
Our study results are different from other published studies, and fill in gaps that will inform future guidelines (see Table 3). One previous study with a similar design (targeting an internal sternum-to-spine compression depth) demonstrated a much larger effect of backboard presence to reduce CPR provider’s hand movement (i.e., to reduce mattress displacement).15 Reduction of mattress displacement at 50 mm CC depth target was approximately 45 mm. This study, however, did not add weight to the torso when CCs were provided without the backboard, which explains this large mattress displacement. Furthermore, torso weight was added only when CC were performed on the backboard. This probably led to an overestimation of the backboard effectiveness.
Perkins et al. reported the effectiveness of a backboard to reduce soft and hard hospital mattress displacement, using a model with CC feedback based on sternum-to-spine compression depth, similar to what we used in our study.14 While mattress displacement accounted for up to 40% of overall sternum movement, their narrow backboard (45.7 cm × 182.6 cm × 3.7 cm) reduced this by 4.7% (95% CI 1.4–8.1%, p < 0.007) and their wide backboard (63.5 cm × 150.9 cm × 0.4 cm) reduced this by 6.6% (95% CI: 3.4–10.0%, p < 0.0001). They concluded that backboard use reduces the mattress compression and therefore the amount of work required from the CPR provider.
They also studied the effect of backboard use on CC depth with a similar VAM on their standard soft hospital bed with foam mattress on top.11 Twenty second-year medical student basic life support instructors were randomized to perform CC guided by a sternum-to-spine depth Voice Advisory Manikin with or without a backboard. Overall CC depth did not meet the current ILCOR guidelines, and they did not find differences between CC with and without backboard (29 ± 7 mm vs. 30 ± 10 mm, p = N.S.). Shallow CC depth performed in this study likely precludes conclusions regarding the effect of the backboard when high quality CPR guided to a 2010 guideline CC depth of >50 mm is provided.
Anderson et al. studied the effect of backboard presence on CC depth using adult manikins on a firm hospital bed without adjustment for a torso weight.10 Twenty-three members of their hospital cardiac arrest team were randomized to perform CPR with or without a small backboard (44 cm × 58 cm) similar to our currently studied small backboard. Subjects were trained and updated on ILCOR 2005 guidelines. They reported the use of backboard improved internal sternum-to-spine measured CC depth from 43 mm to 48 mm (p < 0.0001). This study was conducted without a feedback system and subjects were not blinded to the presence of backboard. A similar result was reported when CC were performed on a CPR manikin on a firm operating room bed with a pressure reduction mattress added on top.17 Our study was different from those studies, quantifying mattress displacement during CCs coached to the deeper 2010 Guideline depth (>50 mm). This suggests that use of an externally placed sternal accelerometer device could potentially underestimate actual sternum-to-spine CC depth and potentially misguide rescuers to achieve shallower CC depth.7,14
Our study did not show completely consistent results regarding the impact of size and orientation of the backboard on mattress displacement. Comparison to a recent prior study16 is difficult, since their study used a target cylinder movement (a net sternal movement) of 50 mm, not the actual sternum-to-spine CC depth of >50 mm, resulting in variable compression force delivery. They reported more sternum-to-spine displacement (i.e. less mattress displacement) when a larger (86 cm × 50 cm) vs. smaller (56 cm × 43 cm) backboard was used on a soft mattress (44 mm vs. 30 mm, p < 0.001). As in our study, this difference was not observed on a harder mattress. They also reported an inconsistent result regarding the effect of horizontal vs. vertical backboard orientation, similar to what we observed in our current study.
Our results should be interpreted with caution. We acknowledge that our study manikin chest may have different stiffness properties compared to actual humans. This might have led to more forceful CC to achieve the target depth. However, based on recent chest stiffness analyses during real CPR in humans, 43 kg of compression force applied in this study is a reasonable range to achieve a target 50 mm depth.20,21,26 We also acknowledge that the horizontally placed large backboard on a stretcher might have less contact to the bed than the surface of the backboard due to the narrow width of the stretcher. In addition, the CC provider could occasionally ‘guess’ the backboard condition in our experiment, especially for a horizontally placed large backboard. However, as CCs were tightly guided by the sternum-to-spine feedback system and the provided CCs consistently met the preset target, this should not have affected our outcome measures. Despite these limitations, our study provides quantitative data about the impact of backboard placement, torso weight, and bed compression on anterior sternal movement based feedback for CC depth in simulated in-hospital cardiac arrest patients.
Backboards should be used for CPR when performed on soft surfaces, such as ICU beds. Backboards may not be needed for CPR on firm stretchers, firm hospital beds, or for patients with heavy torso weights. We evaluated the effect of the backboard presence, size and the orientation in a quantitative manner under various conditions that mimic common clinical situations. We prospectively defined an evidence-based minimally clinically important difference. On soft ICU beds, backboard use was associated with a decrease in mattress displacement as compared to no backboard use. Backboard size and orientation (horizontal vs. vertical) did not have a consistent effect on mattress displacement. Lighter torso weights were associated with larger mattress displacements on soft ICU beds. Future studies should measure the reduction of work and fatigue of CPR providers when backboards are used on soft surfaces, and adequately deep sternum-to-spine high quality CCs are provided.
Laerdal Foundation for Acute Care Medicine and Endowed Chair of Critical Care Medicine at the Children’s Hospital of Philadelphia.
Funding provided by the Endowed Chair of Critical Care Medicine at the Children’s Hospital of Philadelphia and Laerdal Foundation for Acute Care Medicine. We thank Stephanie Tuttle, MBA for her administrative support, Larissa Hutchins APN, RN for arranging beds for testing. Blue Bell Bio-medical, Van Wert, OH supplied the large backboard.
A Spanish translated version of the summary of this article appears as Appendix in the final online version at doi:10.1016/j.resuscitation.2012.01.016.
Conflicts of interest
The authors acknowledge the following potential conflicts of interest. Akira Nishisaki, Dana Niles, Robert Sutton, Matthew R. Maltese and Vinay Nadkarni receive unrestricted research grant support from the Laerdal Foundation for Acute Care Medicine. Robert Sutton is supported through a career development award from the Eunice Kennedy Shriver National Institute of Child Health & Human Development (K23HD062629).