Documentation:FIB book/Biomechanics of Cardiopulmonary Resuscitation-Related Injuries

From UBC Wiki

Introduction & Background

Sudden cardiac arrest claims the lives of 300,000 people annually in the United States[1]. Receiving cardiopulmonary resuscitation (CPR) greatly increases the chance of survival, regardless of a person's skill set in CPR[2]. Cardiac arrest is when a person's heart is no longer beating, and thus the heart is incapable of pumping blood to the brain and lungs[3]. Without assistance, death happens in minutes[3]. When someone performs CPR, they are compressing and relaxing a person’s chest in intervals, resembling how the heart pumps blood throughout the body[3]. When performing CPR, it requires at least 2 inches of depth to be effective [4].

With that said, fractures in ribs and sternum as a result of chest compressions are a common complication from CPR [5]. Notably, broken ribs tend to only cause complications if they are broken in many different places and are dislocated [5]. If there are many fractures on each side of the thoracic cage, there is risk of flail chest[5]. Flail chest is a result of traumatic injury where the chest wall moves independently and significantly impedes respiratory physiology. It is likely to occur when more than 3 ribs are broken[6]. Loose fragments of rib and sternum can cause soft tissue damage, such as injury to the heart and lungs, like hemothorax and pneumothorax[5]. A summary of CPR-associated injuries are outlined in Figure 1. In 2015, 64.5% of a 31-person study group (with a mean age of 70.2 years) suffered rib fractures resulting from cardiopulmonary resuscitation[1]. All were resuscitated following the 2015 European Resuscitation Council (ERC) Guidelines, which recommend a compression depth between 2 to 2.5 inches for the average adult[7],[8].

Figure 1: CPR Associated Injuries[9]

The survival rate of patients requiring CPR greatly relies on the quality of CPR in the first few minutes, such as compression point location and compression depth[8]. A study of 859 out-of-hospital cardiac arrests (OHCA) were split into two groups: those who died in CPR and those who were successfully resuscitated[7]. In the 859-person study, 628 had died and 231 survived. An injury associated with CPR occurred in 13% of alive patients and 87% of the deceased[7]. Those who died were more likely to have thoracic injuries, like fractures to the ribs and sternum[7].

Factors Affecting the Biomechanics of the Thorax and CPR Injury Mechanisms

Biomechanics of CPR

CPR can load the thorax to a maximum chest deflection, comparative to those seen in frontal crash environments[10]. Maximum chest deflection is the maximum distance the chest is compressed. The force applied during chest compressions is transmitted to the sternal hinge which aids in the creation of intrathoracic pressure variations during the compression-decompression phase of CPR[10]. The sternal hinge is fixed at its superior end by flexible sternoclavicular joints and assists the ventral movement of the sternum[10]. During compression of the thoracic cage, the cartilage-rib system and the intrathoracic viscera reduce the impact force to aid recoil of the sternum to its original position[10]. For reference, an image of the human thorax can be seen in Figure 2.

Figure 2: Labeled diagram of Human Thorax[11]

Effect of Age on Thoracic Cage Injuries from CPR

Studies have found that older individuals are at higher risk of thoracic cage injuries, compared to younger individuals[12]. Age is directly related to a person's stiffness of their thoracic rib cage[13]. Notably, CPR causes the sternum to deflect towards the spine of a patient to induce blood flow, since chest compressions are applied directly over the sternum[13]. The sternum in an adult is made of six bones (the manubrium, sternebrae 1 to 4, and the xiphoid process) joined together[12],[14]. This can be seen above in Figure 2. The xiphoid process forms after a person is 3-6 years old, and the sternebrae start fusing together around 4 years old and will continue to fuse, until a person reaches 20 years of age, increasing stiffness[14]. Additionally, the costal cartilage that connects the ribs to the sternum calcifies with age, which affects the stiffness and flexibility of the thoracic cage[14]. The modulus of elasticity of the rib cage thus changes over time, due to structural changes of the thorax and the materials within it, including bone, cartilage, lung tissue and smooth muscle [14],[15]. It is largely unknown how the properties of these materials change over time due to aging, and the relative changes compared to other materials[15]. This makes it difficult to apply the scaling method to data used for building ATD’s with biofidelic thoraxes and to identify a single effective modulus of elasticity that can be used to study thoracic cage injury in different age groups[15].

Maltese et al. (2008) found that the elastic stiffness of the thoracic cage increases from approximately 8 years to 40 years old, and then decreases after age 40[13]. Both the viscous force and the elastic force increased with age in chest models used during testing, exhibiting the progressive stiffening of the chest[13]. These conclusions were drawn after removing the mattress deflection from the chest deflection signals, by constructing a spring and mass model of the patient-mattress system[13].

The biological changes occurring over time that impact thorax stiffness has led to the observation that children rarely suffer from rib fractures (0-2% probability), whereas adults are at an increased risk of rib fracture (13-97% probability)[13]. However, children are still at high risk of soft tissue or pulmonary injuries occurring, due to the force applied during CPR[13]. This was evident when Ouyang et al. (2006) conducted thoracic impact testing using pediatric cadavers, where 6 of the 9 specimens developed pneumothoraxes[14],[16]. Whereas when Kroell et al. (1974) used a similar procedure for thoracic impact testing on adult cadavers, 18 of 22 cadavers developed rib fractures[15],[17]. This leads to the conclusion that older individuals are at increased risk of skeletal thoracic cage injuries from CPR.

Effect of Environmental Factors on Thoracic Cage Injuries from CPR

There are many situations involving cardiac arrest which result in modified thoracic cage biomechanics. The physical environment in which a patient experiences a cardiac arrest, as well as when and by whom CPR is administered, are all variables that affect patient outcomes.

Physical Environment

Patients who have undergone cardiac arrest while afflicted with hypothermia or asphyxia may have modified thoracic cage biomechanics or CPR administration [16],[18]. Hypothermia occurs when the core temperature of the body falls below 35°C[18]. Since hypothermia lowers oxygen consumption at the cellular level, the heart and brain can tolerate longer cardiac arrest times at moderately lower temperatures with greater chances of complete neurological recovery than at the physiological body temperature[18]. Biomechanically, hypothermic individuals have greater thoracic stiffness, so chest compressions require a higher application force[18]. The higher the application force, the more likely that a CPR-induced injury will occur[18]. Quantitative injury thresholds relevant to this situation are described in section 3.

When and How CPR is Delivered

There are many other factors that can significantly affect CPR patient outcomes, such as the time of day at which a patient receives CPR, if a patient already has thoracic chest injuries, whether the rescuer is a human or an automated machine, and if the patient experiences a cardiac arrest out-of-hospital versus in-hospital.

Patients receiving CPR at medical centers at night receive lower-quality CPR, characterized by a higher rate of occurrence of chest injuries and fewer chest compressions that occur at an adequate depth [17],[19]. Chest compressions at depths greater than 6 cm are more likely to increase chest injury occurrence and severity[19]. In a study involving 1254 patients, it was found that the night-time patients had a shorter CPR duration (23.6 minutes versus 27.8 minutes), a higher incidence of chest injuries (67% versus 40%), and a lower rate of spontaneous circulation return (26.5% versus 38.4%) in comparison to the day-time patients[19]. No significant differences in survival were observed[19]. This higher prevalence of injury at night could be due to the difficulty in maintaining high-quality CPR at night when medical staff are tired, less motivated, and limited in number[19].

Regardless of the time of day, a patient’s thoracic and pulmonary compliance at the time of CPR can change the effectiveness of CPR[19]. Pulmonary compliance is the measure of the expansion of the lung and is calculated by dividing the change in lung volume (liters) by the change in pleural pressure (cm H20)[20]. Thoracic compliance failure, due to bone fractures or pneumothorax, can result in ineffective chest compressions[19]. Therefore, if a patient already has thoracic chest injuries, like bone fractures of the ribs and sternum, then the CPR they receive will have reduced effectiveness.

Compared to CPR delivery by experienced medical professionals, CPR delivered via mechanical chest compression devices (CMMD) results in poorer patient outcomes[19]. One common CMMD is the LUCAS 2 seen in Figure 3, which was designed to address the challenge faced by humans of achieving consistent compression depths. However, automated compression devices cause more severe and more frequent sternal fractures, including displaced-type fractures[9]. Additionally, only in patients who received mechanical CPR were visceral injuries observed, including contusions of the heart, skin, fat, or muscle[21]. The difference in injury occurrence between these two groups could be due to the longer CPR duration times associated with mechanical CPR since they are commonly used in ambulances when transporting patients long distances[21]. There is a strong association between the duration of CPR and the occurrence of injury[21].

Figure 3: Typical set-up of the LUCAS 2 mechanical chest compression device on a CPR training dummy[22]

Patients with sudden cardiac arrest out-of-hospital have a 40% lower survival rate, compared to in-hospital sudden cardiac arrests[23]. This could be due to a few factors: 1) the time to defibrillation is higher when patients arrest out-of-hospital, 2) the time to start CPR is higher in out-of-hospital settings, and 3) low-quality or nonexistent bystander-initiated CPR[23]. The longer it takes to defibrillate and start CPR, the more cerebral damage that occurs and the lower the patient’s chances of survival[23]. Poor quality CPR can cause severe skeletal injuries that can further decrease survival rates out-of-hospital[23].

Biomechanics Research

Porcine animal models have been traditionally used for chest compression injury analysis/testing, but this data is not fully transferable to human situations/anatomy[24]. To overcome this limitation, postmortem human surrogates (PMHS), finite element (FE) simulations, and anthropometric test dummies (ATDs) are commonly used[24]. These models have all been used to reconstruct the mechanics of CPR to determine the forces needed to achieve required compression depths and further inform resuscitation guidelines[24].

PMHS Models for CPR

Confidence in PMHS response as an accurate representation of living human response, particularly in the thorax, is vital in the development of CPR guidelines and protocols[25]. Arbogast et al. (2006) analyze the differences between the thoracic response of adult PMHS and live human adult chests[25]. To test living human subjects, the research team created a load cell and accelerometer sensor package that was integrated with a clinical monitor defibrillator to track chest compression and applied force during CPR[25]. The sensor was located between the sternum of the patient and the hands of the person administering CPR[25]. The accelerometer signal generated was processed to yield deflection data for a living subject during chest compressions[25]. CPR for this study was administered by paramedics on 91 patients following CPR guidelines which recommend that chest compressions be applied so that the sternal displacement is 3.8-5.1 cm at a rate of 100 compressions per minute[25]. This method was validated against data generated from chest compressions delivered using the force deflection sensor in a standard CPR manikin, which was instrumented with a linear potentiometer to measure compression[25]. The potentiometer output was compared to the displacement calculated from the acceleration signal[25].

Once the method was validated for live patients, it was further compared to PMHS thoracic response tests for chest deflection[25]. To do this, PMHS were set in the supine position, and a hydraulic cylinder arrangement was used with materials testing machine that administers controlled chest deflections to the PMHS[25]. The load was applied using a 15.2 cm diameter steel hub with a load cell mounted onto a sliding track[25]. The hub was located at the intersection of the midsagittal plane and the fourth intercostal space which was representative of the location of CPR compression[25]. Chest deflection was measured anteriorly with a string potentiometer that adhered to the hub[25]. The test setup involves a 15.2 cm-diameter steel hub mounted on a sliding track with a load cell placed posteriorly to the PMHS [25]. Force in this system was measured in two ways: load cells positioned on the hub to measure anterior loads and load cells positioned posteriorly to determine the reaction forces generated by the deforming thorax[25]. PMHS models were loaded to a depth of 15-20% of the undeformed chest depth[25].

Force-displacement data was captured for 91 adult patients (61 men and 30 women with mean age of 70 years) who were in cardiac arrest[21]. These live volunteers had a maximum applied force of 297 N and a mean maximum compression depth of 4.16 cm[25]. For the PMHS data, the mean maximum applied force was 474 N, and the mean maximum compression depth was 3.1 cm[25]. The differences in this data can be explained by the discrepancies in the area of load application and thoracic structures engaged[25]. The clinical data collected also did not record the presence of sternal or rib fractures that may develop during CPR[25]. Studies on thoracic cage fractures as a result of chest compressions in PMHS models have not been done[25]. Arbogast et al. (2006) show a mechanism to create fundamental biomechanical data on the mechanical response of the adult thorax during CPR. This model could also be used in conjunction with a study by Suazo et al. (2022) that associates higher survival rates with a compression depth between 4.03 and 5.53 cm, and a survival peak at 4.56 cm[8]. This data is comparable to the findings on live patients with a mean maximum compression depth of 4.16 cm[8]. Statistical comparison between both the live patient data and PMHS model data shows that compressions in CPR in live subjects generate less force at equivalent deflection than PMHS[25].

Arbogast et al. (2006) were able to present fundamental biomechanical data on the mechanical response of the adult thorax to increase the understanding of differences between PMHS and living humans. This study, however, was not without its limitations. First, the CPR model of human patients did not account for the compliance of the surface the subject was laying on, and thus true chest deflection may be overestimated, and stiffness of the thorax underestimated[25]. Another limitation of this study was that clinical data collected on live patients did not record the presence of sternal or rib fractures that may develop[25]. However, the most significant limitation is the difference in contact surface area between the hub on the PMHS chest and the CPR force deflection sensor. The hub on the PMHS chest was larger than the CPR force deflection sensor contact area, providing a smaller stress for a given force in the PMHS data[25]. This could also impact the stiffness of the thorax, as stiffness variation correlated to variation in the area of load application[25]. Finally, load application between the clinical data and PMHS data engages different anatomical structures[25]. This effect is not quantified in this study but has the potential to be using FE models.

FE Models for CPR

CPR studies using PMHS models are known to be challenging and as described above, focus on compression depth and force applied during the procedure rather than the risk of injury due to these applied forces. To overcome these challenges, FE simulations can be used as an alternative to determining indicators of CPR performance and rib fracture risk which correlates to the displacement of the rib cage and maximum rib stress level[8]. An FE model published by Suazo et al. (2022) leverages computerized tomography images available in the BodyParts3D database to model the sternum, 7 pairs of costal cartilage, 10 pairs of osseous ribs, and 9 sections of intercostal muscle[8]. Three-dimensional triangular surface meshes for each individual element in the model were downloaded from the database to be refined and made consistent by removing undesirable intercepts between contiguous elements[8]. The consistent surface meshes were used to build 3D volume meshes that were compounded into a global computational mesh[8]. This model was used to investigate the effects of compression locations on the chest, which was modeled in 5 regions as a surface patch with a 10 cm2 area[8]. The rib cage model assumes the spine to be fixed, and bone is modeled as a homogenous material[8]. Testing against previous studies, as well as trial and error, were used to determine the characteristic parameters of the elasto–plastic bone model[8]. The irreversible plastic behavior was in the region for applied forces larger than 60 N and the quantitative assessment of fracture risk was determined based on a referenced level of Von Mises tension[8]. Von Mises values give information on whether a material will yield or fracture[26].

The model was used for 9 compression forces between 200 N and 600 N in increments of 50 N, which were applied to each of the 5 compression locations specified by the study researchers[8]. The 5 regions of interest, as seen in Figure 4, were the sternum (P1 and P2 area) and the 4th, 5th and 6th left ribs (P3, P4, and P5 respectively)[8]. The relationship between compression depth and force reveals that the stiffness of the rib cage tends to increase with increasing compression force[8]. The FE model revealed that standalone ribs when compressed become increasingly softer, whereas the whole rib cage tends to become stiffer with increasing applied force[8]. Figure 5 provides a quantitative assessment of the relationship between compression depth and force[8]. This reveals that less force is required to reach a certain compression depth in the sternum P2 area than to achieve the same depth in the P1 area regardless of the compression location[8]. Therefore, by caudally shifting the compression location to 16% of the total length of the sternum, this results in more effective CPR, in terms of the compression depth achieved and force needed[8].

Figure 4: The approximate location of each of the five compression areas considered [8]
Figure 5: Maximum compression depth obtained in the compression area (A) P1, (B) P2, (C) P3, (D) P4, and (E) P5 as a function of the amplitude of the force applied during the CPR [8]

The maximum stress values can be used to determine the risk of injury during chest compressions. The model was able to predict the highest percentage of fractures in the 2nd-6th ribs as seen in Figure 6 which resembles findings in a previous study of 1480 patients that sustained skeletal chest injuries[8]. Region P3 was also found to have the poorest performance, in terms of compression efficiency and potential risk of injury[8]. High force application is required to reach the recommended compression depth at P3, which results in a high risk of fracture to the 2nd and 4th ribs, as well as costo-sternal separation of the 1st left rib[8]. P3 was indicated as a region to avoid CPR, because of the high-risk factors of injury and inefficiency.   

Figure 6: Von Mises stress distributions (σv) on the rib cage when a 600 N force was applied at the compression region: (A) P1, (B) P2, (C) P3, (D) P4, and (E) P5. These measures were used to determine rib fracture [8]

FE models offer the ability to apply different conditions to a unique geometry, which are not possible measurements in real patients, as CPR is performed using only one compression location[8]. The FE model also has several inherent limitations. For one, the subject being modeled is a young male with a normal body mass index, which contrasts to the common recipients of CPR that tend to be old and obese individuals[8]. As discussed previously, age and body mass impact the aspect ratio of the rib cage. The osseous rib was also modeled as a single material, when in reality it is composed of cortical and trabecular bone[8]. Another limitation is the FE model was considered to be rigid, due to the omission of sternochondral joints[8]. This results in the inability to assess the risk of separation of the costal cartilage from the sternum during CPR maneuvers, using the FE model[8]. Evidently, the FE model is limited by the use of a single geometry model and the absence of anatomical structures but is able to provide a tool to assess the risk of injury due to alternative compression locations for CPR.

ATDs for CPR

The study of CPR using ATDs is limited in the available literature. ATDs are mostly used to study motor vehicle accidents at high inertia and impact velocities and have not been frequently used to explore CPR-related injuries.

Biofidelity of the thorax in ATDs must be optimized for accurate modeling of CPR-induced injury criteria. It is vital that the rib cage model accurately imitates rib deflection and torsion under an acceptable force range, with models undergoing peak loads as high as 10 kN in certain crash tests, and only 500 N in lower load cases such as CPR testing. In fact, rib fractures mainly occur from bending, opposed to torsion[27]. The combined deflection criterion (DC) “uses chest displacements at four locations to compute overall and differential deflections resulting in bending strains”[27]. Whereas the number of fractured ribs (NFR) criterion “uses locally measured strains at individual ribs to identify those ribs for which the bending strains at any location exceeded a critical value”[27]. Both criteria were implemented in the design of the modern THOR ATD design, which specializes in the biofidelity of the thorax and shoulder.

Manikins used for commercial CPR training present similar biofidelity challenges to ATDs in their design. Low accuracy in chest stiffness and material properties of available training manikins can lead to substantial deficiencies in the rescuers' learning[28]. To improve training manikin design, one study collected data from 59 cardiac-arrest patients in ambulances using defibrillators outfitted with an accelerometer and a force sensor to measure chest deflection against force[29]. As a result, existing manikin designs were modified to have variable chest stiffness to model different population size bands[29]. Collected data also allowed for the improved manikins to “mimic the viscous damping properties” of patient chests through more accurate spring damping.

Commercial training manikins present low levels of biofidelity, in comparison to ATDs[29]. As a crash test dummy, THOR is frequently used in high-impact and high load testing scenarios[15]. However, using THOR at lower forces with high load cycles could provide valuable information on CPR-induced injury mechanisms[15]. A complementary study analyzed the incorporation of data from CPR cases into the chest response to ATDs[15].

Figure 7: Force-deflection corridors for a) 6 year old children, b) to year old children, c) 12 year old children versus 5th percentile adult females and d) 50th percentile 20 year old adult males versus 50th percentile 63 year old men [15]

CPR data was gathered from two separate studies to understand the biofidelity of ATDs for assessing thoracic injury criteria. Together both studies redefined the biofidelity corridors and the design of ATD chests, as data from these studies were used to improve limited existing data from more injury-prone experimental tests[15]. Figure 7 shows that the maximum load for a 6-year-old was 7% less than the preceding ATD specifications, whereas the maximum displacement was 8% higher[15]. Additionally, the corridors for 12-year-olds were strikingly similar to that of the 5th percentile adult female, proving that existing female ATDs could be used for CPR related studies in children around the age of 12, provided similar loading cases are adopted[15]. Notably, the corridors for 20-year-old and 63-year-old adult men significantly differ, as seen in Figure 7d[15]. This challenges the previous assumption in ATD design that no change in thoracic stiffness occurs throughout adulthood[15].

Hence, many modifications to the current ATD design in relation to thoracic modeling may be proposed. Maltese et al. suggest incorporating tunable thoracic stiffness in future ATDs, similar to that implemented in existing training manikins, to represent multiple age ranges[13],[15]. It is also suggested to consider changes in spinal flexibility with age, to more accurately model pediatric case scenarios. Finally, biofidelity corridors for the thorax model can be improved further by expanding the sample size of CPR subject data with referenced results being preliminary[15].

Comparison of Methods

The previously analyzed methods each hold clear benefits and drawbacks. It is apparent from the literature in the field of ATD application for CPR research that this is still a relatively new field of study. Hence, ATDs need further development in order to substantially improve CPR-related research. While the PMHS testing results discussed earlier did not exactly compare to that of the live human experiments, presumably due to discrepancies in contact surface area and structural aspects engaged in the experiment setup, PMHS testing is a valid strategy to optimize ATD models for research and training in the field. Arbogast et al. (2006) indicates that PMHS subjects should not show significant differences in chest deflection with live subjects under identical test set-ups, suggesting that muscle activation in the thorax does not need to be accounted for in CPR models[25].

FE modeling is useful for analyzing the thoracic structure under CPR loading to determine the risk of injury and indicators of CPR performance. Using previous data using human bodies, modern FE models are able to accurately predict the risk of thoracic injury depending on the magnitude and location of load application on the sternum. While this is an effective tool for research in the area, it may be a less effective training tool for practitioners to optimize CPR administration. Contrarily, ATDs and manikins significantly bolster learning of proper CPR techniques, but PMHS and FE models help determine their model parameters. Consulting doctors is also a proven method to improve the biofidelity of such models, evaluating the accuracy of such models with doctors with invaluable experience in the field[29]. Ultimately, all methods described should be adopted in conjunction to optimize models for CPR training.

Discussion

Limitations

The primary limitation in this field is that public studies around biofidelity testing of the thorax for low-impact loadings, like CPR, are highly limited, especially for broad demographics. Research in the field should be heavily extended to model CPR procedures for training and biofidelity testing more accurately. For example, more extensive testing may be performed by evaluating thoracic stiffness by measuring load at a broader range of chest displacement levels in patient testing and testing a larger sample of male and female patients with a substantial age distribution.

PMHS Limitations

One general limitation of PMHS-specific CPR testing is the difficulty in acquiring subjects for testing that are representative of a wide demographic. Furthermore, PMHS testing for CPR is limited to testing chest deflection against force, rather than the risk of injury due to extenuating circumstances in using postmortem subjects. Overall, this makes PMHS testing challenging, especially when studying more varied demographics such as infants and children, which are proven to show substantially different biomechanical responses to CPR loading.

FE Modeling Limitations

The key disadvantage of FE modeling is the difficulty in assessing the biofidelity of the model’s response to different loadings. Physical models are often tested with practiced doctors to evaluate their accuracy, purely by asking doctors how closely the model resembles real patients in CPR. While FE models allow for much more efficient research in the field of biomechanics, it is more difficult to validate these models and use them in the context of training.

ATD Limitations

ATD design in the field of CPR modeling is also limited by the lack of consideration for structural fatigue in the thorax. Since CPR requires high cycle quantities, this may affect the material properties of bone and tissue on a microscopic level, which is disregarded in experiments found in available literature on the subject.

Future Research

There is currently little consistency within CPR-related rib and sternal fracture research, which creates limited data sets that can be compared[30]. Therefore, consistent study protocols are needed to enhance interstudy comparability and improve CPR maneuvers. This would result in a better understanding of CPR-related compression injuries to the thorax. To collect this information, additional future research using PMHS models, FE models, and ATDs, is needed. PMHS models are able to present fundamental biomechanical data on the mechanical response of the adult thorax. Further testing could be done to develop a data set that can be used to inform injury risk and PMHS should be used for this purpose, as injuries cannot be modeled using volunteers or ATDs. FE models also offer the ability to predict injury, and thus future models should be created for a larger demographic. The inherent nature of FE models can be used to apply different mechanical properties to unique geometries and make measurements that are not capable on real patients. Finally, in terms of ATDs, modifications to the current ATD design in relation to thoracic modeling may be proposed. Biofidelity corridors for the thorax model can be improved further by expanding the sample size of CPR subject data, while considering any previously referenced results as preliminary. The increase of data collection through human volunteers and PMHS experimentation will foster the further development of accurate FE models and biofidelic ATDs for CPR injury prevention research.

References

  1. 1.0 1.1 J. Kelley, P. B. Richman, G. A. Ewy, L. Clark, B. Bulloch, and B. J. Bobrow, “Eighth grade students become proficient at CPR and use of an AED following a condensed training programme,” Resuscitation, vol. 71, no. 2, pp. 229–236, Nov. 2006, doi: 10.1016/j.resuscitation.2006.03.015.
  2. “Cardiopulmonary resuscitation (CPR) complications encountered in forensic autopsy cases | SpringerLink.” https://link.springer.com/article/10.1186/s12873-019-0234-5 (accessed Nov. 09, 2022).
  3. 3.0 3.1 3.2 CDC, “Three Things You May Not Know About CPR | cdc.gov,” Centers for Disease Control and Prevention, Apr. 26, 2022. https://www.cdc.gov/heartdisease/cpr.htm (accessed Dec. 04, 2022).
  4. “How To Perform CPR,” Cleveland Clinic. https://my.clevelandclinic.org/health/articles/17680-cardiopulmonary-resuscitation-cpr (accessed Dec. 04, 2022).
  5. 5.0 5.1 5.2 5.3 W. Lederer, D. Mair, W. Rabl, and M. Baubin, “Frequency of rib and sternum fractures associated with out-of-hospital cardiopulmonary resuscitation is underestimated by conventional chest X-ray,” Resuscitation, vol. 60, no. 2, pp. 157–162, Feb. 2004, doi: 10.1016/j.resuscitation.2003.10.003.
  6. T. B. Perera and K. C. King, “Flail Chest,” in StatPearls, Treasure Island (FL): StatPearls Publishing, 2022. Accessed: Dec. 02, 2022. [Online]. Available: http://www.ncbi.nlm.nih.gov/books/NBK534090/
  7. 7.0 7.1 7.2 7.3 J. Karasek et al., “CPR-related injuries after non-traumatic out-of-hospital cardiac arrest: Survivors versus non-survivors,” Resuscitation, vol. 1, pp. 90–95, 2022, doi: 10.1016/j.resuscitation.2021.12.036.
  8. 8.00 8.01 8.02 8.03 8.04 8.05 8.06 8.07 8.08 8.09 8.10 8.11 8.12 8.13 8.14 8.15 8.16 8.17 8.18 8.19 8.20 8.21 8.22 8.23 8.24 8.25 8.26 8.27 8.28 8.29 M. Suazo, J. Herrero, G. Fortuny, D. Puigjaner, and J. M. Lopez, “Biomechanical response of human rib cage to cardiopulmonary resuscitation maneuvers: Effects of the compression location,” Int. J. Numer. Methods Biomed. Eng., vol. 38, no. 4, p. e3585, 2022, doi: 10.1002/cnm.3585.
  9. 9.0 9.1 A. Karatasakis et al., “Prevalence and Patterns of Resuscitation-Associated Injury Detected by Head-to-Pelvis Computed Tomography After Successful Out-of-Hospital Cardiac Arrest Resuscitation,” J. Am. Heart Assoc., vol. 11, no. 3, p. e023949, 2022, doi: 10.1161/JAHA.121.023949.
  10. 10.0 10.1 10.2 10.3 “Abstract 131: Chest Molding in Extended Cardiopulmonary Resuscitation | Circulation.” https://www.ahajournals.org/doi/10.1161/circ.142.suppl_4.131 (accessed Nov. 09, 2022).
  11. L. M. Biga et al., “7.5 The Thoracic Cage,” Sep. 2019, Accessed: Dec. 03, 2022. [Online]. Available: https://open.oregonstate.education/aandp/chapter/7-5-the-thoracic-cage/
  12. 12.0 12.1 “thoracic cavity | Description, Anatomy, & Physiology | Britannica.” https://www.britannica.com/science/thoracic-cavity (accessed Nov. 10, 2022).
  13. 13.0 13.1 13.2 13.3 13.4 13.5 13.6 13.7 M. R. Maltese et al., “Methods for determining pediatric thoracic force-deflection characteristics from cardiopulmonary resuscitation,” Stapp Car Crash J., vol. 52, pp. 83–105, Nov. 2008, doi: 10.4271/2008-22-0004.
  14. 14.0 14.1 14.2 14.3 14.4 K. B. Arbogast et al., “Expert clinical assessment of thorax stiffness of infants and children during chest compressions,” Resuscitation, vol. 80, no. 10, pp. 1187–1191, Oct. 2009, doi: 10.1016/j.resuscitation.2009.07.005.
  15. 15.00 15.01 15.02 15.03 15.04 15.05 15.06 15.07 15.08 15.09 15.10 15.11 15.12 15.13 M. R. Maltese et al., “Incorporation of CPR Data into ATD Chest Impact Response Requirements,” Ann. Adv. Automot. Med. Assoc. Adv. Automot. Med. Annu. Sci. Conf., vol. 54, pp. 79–88, 2010.
  16. 16.0 16.1 J. Ouyang, W. Zhao, Y. Xu, W. Chen, and S. Zhong, “Thoracic impact testing of pediatric cadaveric subjects,” J. Trauma, vol. 61, no. 6, pp. 1492–1500, Dec. 2006, doi: 10.1097/01.ta.0000233711.07823.40.
  17. 17.0 17.1 C. K. Kroell, D. C. Schneider, and A. M. Nahum, “Impact Tolerance and Response of the Human Thorax II,” SAE International, Warrendale, PA, SAE Technical Paper 741187, Feb. 1974. doi: 10.4271/741187.
  18. 18.0 18.1 18.2 18.3 18.4 D. Cimpoesu et al., “Cardiac Arrest in Special Circumstances-Recent Advances in Resuscitation. [Review],” Am. J. Ther., vol. 26, no. 2, pp. e276–e283, Apr. 2019, doi: 10.1097/MJT.0000000000000927.
  19. 19.0 19.1 19.2 19.3 19.4 19.5 19.6 19.7 W. Takayama, A. Endo, H. Koguchi, K. Murata, and Y. Otomo, “Differences in durations, adverse events, and outcomes of in-hospital cardiopulmonary resuscitation between day-time and night-time: An observational cohort study,” Resuscitation, vol. 1, pp. 14–20, 2019, doi: 10.1016/j.resuscitation.2019.01.023.
  20. Z. Edwards and P. Annamaraju, “Physiology, Lung Compliance,” in StatPearls, Treasure Island (FL): StatPearls Publishing, 2022. Accessed: Oct. 27, 2022. [Online]. Available: http://www.ncbi.nlm.nih.gov/books/NBK554517/
  21. 21.0 21.1 21.2 21.3 L. Milling, B. S. Astrup, and S. Mikkelsen, “Prehospital cardiopulmonary resuscitation with manual or mechanical chest compression: A study of compression-induced injuries,” Acta Anaesthesiol. Scand., vol. 63, no. 6, pp. 789–795, 2019, doi: 10.1111/aas.13347.
  22. J. Godbout, “Mechanical CPR: Past, Current, and Future,” EMOttawa Blog, Mar. 14, 2019. https://emottawablog.com/2019/03/mechanical-cpr-past-current-and-future/ (accessed Dec. 04, 2022).
  23. 23.0 23.1 23.2 23.3 J. Herlitz et al., “A comparison between patients suffering in-hospital and out-of-hospital cardiac arrest in terms of treatment and outcome,” J. Intern. Med., vol. 248, no. 1, pp. 53–60, 2000, doi: 10.1046/j.1365-2796.2000.00702.x.
  24. 24.0 24.1 24.2 “A new physiological model for studying the effect of chest compression and ventilation during cardiopulmonary resuscitation: The Thiel cadaver - ScienceDirect.” https://www.sciencedirect.com/science/article/pii/S0300957218300121 (accessed Nov. 09, 2022).
  25. 25.00 25.01 25.02 25.03 25.04 25.05 25.06 25.07 25.08 25.09 25.10 25.11 25.12 25.13 25.14 25.15 25.16 25.17 25.18 25.19 25.20 25.21 25.22 25.23 25.24 25.25 25.26 25.27 K. B. Arbogast, M. R. Maltese, V. M. Nadkarni, P. A. Steen, and J. B. Nysaether, “Anterior-posterior thoracic force-deflection characteristics measured during cardiopulmonary resuscitation: comparison to post-mortem human subject data,” Stapp Car Crash J., vol. 50, pp. 131–145, Nov. 2006, doi: 10.4271/2006-22-0006.
  26. “von Mises yield criterion,” Wikipedia. Nov. 16, 2022. Accessed: Dec. 04, 2022. [Online]. Available: https://en.wikipedia.org/w/index.php?title=Von_Mises_yield_criterion&oldid=1122150735
  27. 27.0 27.1 27.2 J. Carroll, “An advanced thorax-shoulder design for the THOR dummy,” Jan. 2013, Accessed: Nov. 10, 2022. [Online]. Available: https://www.academia.edu/68860797/An_advanced_thorax_shoulder_design_for_the_THOR_dummy
  28. “BODDEE BUDDEE: Evaluation of Different Foams and Thermoplastics to Develop a Biofidelic Manikin for Cardiopulmonary Resuscitation - PubMed.” https://pubmed.ncbi.nlm.nih.gov/31679238/ (accessed Dec. 02, 2022).
  29. 29.0 29.1 29.2 29.3 “Manikins With Human-Like Chest Properties—A New Tool for Chest Compression Research | IEEE Journals & Magazine | IEEE Xplore.” https://ieeexplore.ieee.org/document/4666711 (accessed Dec. 02, 2022).
  30. K. L. Wininger, “Chest compressions: biomechanics and injury,” Radiol. Technol., vol. 78, no. 4, pp. 269–274, 2007.