Reviews

Evolution of our knowledge of sudden death due to commotio cordis

Leslie A. Geddes DSc, PhD, FACC*, Rebecca A. Roeder PhD

Department of Biomedical Engineering, Purdue University, West Lafayette, IN 47907-2022, USA

Received 13 December 2003; accepted 15 December 2003

Abstract Commotio cordis refers to circulatory arrest due to a nonpenetrating blow to the chest. First discovered in 1932 in a study using large rabbits, it came to the attention of clinicians who encountered children dying suddenly from a chest blow while engaging in sports activities. This review traces the history of commotio cordis, establishes the conditions necessary for sudden death from a nonpenetrating chest blow, and presents the first ECG record showing that a chest blow landing in the ventricular vulnerable period can produce ventricular fibrillation. The conditions necessary for sustaining ventricular fibrillation and numerous examples of sudden death by commotio cordis are presented.

D 2005

Introduction

There have been many sudden deaths due to a non- penetrating blow to the chest. This cardiopulmonary arrest is now known as commotio cordis. Maron et al [1] listed the following as having produced commotio cordis: baseballs, softballs, hockey pucks, lacrosse balls, cricket balls, soccer balls, hockey sticks, tennis balls (filled with coins for training pitchers), knees, feet, elbows, forearms, shoulders, fists, heads (football helmet), goalposts, play (shadow boxing), parent-to-child discipline, gang rituals, scuffle, plastic- sledding saucers (diameter, 91.4 cm), plastic (hollow) toy bats, snowballs, playground swing carriages, charging pet (collie) dog head blows, hiccup remedy (by a friend), and falls from a playground apparatus (monkey bars).

Dorland’s Illustrated Medical Dictionary [2] does not list commotio cordis. It defines commotio as a disturbance.

* Corresponding author. Tel.: +1 765 494 2997; fax: +1 765 494 1193.

E-mail address: [email protected] (L.A. Geddes).

Other definitions given are violent shaking, contusion, concussion, or the shock that results from violent shaking. Commotio retinae and commotio spinalis are listed. However, it appears that the present-day use of the term commotio cordis is in accordance with these definitions and adequately describes a mechanical stimulus delivered to the heart by a nonpenetrating blow to the chest.

First observations

The fact that a blow to the chest can produce ventricular fibrillation (VF) is due in 1932 to Schlomka and Hinrichs [3], well before the ventricular vulnerable period was discovered by Wiggers and Wegria [4] in 1940. Schlomka and Hinrichs [3] were the first to provide electrocardio- graphic evidence that sudden death can accompany a sudden chest blow. In their studies, they used large anesthetized rabbits (12-23 kg) and delivered blows to the chest over the heart using 2 hammers (50 and 260 g). In their many studies, they recorded limb electrocardiogram (ECG) leads and

0735-6757/$ - see front matter D 2005 doi:10.1016/j.ajem.2003.12.023

Fig. 1 The first ECG record of ventricular fibrillation after a blow to the chest of a hare marked by a plus sign. The ECG (left) shows sinus rhythm. At the point marked by a plus sign, the chest blow was delivered over the heart. Note widening of the R waves that progressed to a ventricular tachycardia in the trachea. Redrawn from Schlomka and Hinrichs [3].

respiration. Both vagus nerves were sectioned and the chest blows were delivered randomly. They encountered brady- cardia, tachycardia, AV block, bundle-branch block, and ST segment shifts. They noted that the larger hammer had more effect; Fig. 1 is from their article and shows the ECG and respiration recordings. At the point marked by a plus sign, the chest blow was delivered. The R waves of the ECG became broad and a tachycardia that progressed to VF (right) developed. They reported that VF was encountered frequently.

In 1934, Schlomka was the first to use the term commotio cordis in an article title [5]. From 1932 to 1934, he had performed a large number of experiments on many large rabbits; from all of his studies [6-8], he concluded the following:

1. A blow at the heart area almost always produced ECG changes, whereas trauma to other areas of the chest did not. Many types of ECG changes were noted, including fatal VF. He considered that elevation of the ST segment was bspecificQ for blunt cardiac injury. However, most nonfatal ECG changes lasted only 1/2 to 2 hours.
2. Arterial pressure fell and venous pressure rose as a result of a trauma. The rise in venous pressure usually lasted 1 to 3 minutes.
3. Heart size, measured roentgenologically, increased after a trauma in 99% of the cases. The degree of dilation was correlated to the severity of the ECG findings and to the changes in arterial and venous pressures.
4. Evidence of permanent heart damage was recorded in 60% of the 86 animals followed up for 80 to 120 days. Aneurysm of the right ventricle was found in 21 animals. Many of the animals died of heart failure after a trauma, showing liver stasis and ascites.

Underlying factors

The factors that underlie commotio cordis can be divided into 2 categories: (1) nature of the blow-producing agent and (2) victim-related factors. Each component of these categories will be analyzed in detail.

Commotio cordis came to the attention of clinicians as the result of a case report in 1978 by Dickman et al [9], who stated that a healthy 7-year-old boy was playing third base in an organized youth league T-ball (softball) game. A 7- year-old peer hit a baseball from the stand and struck him in

the midsternum. The boy picked up the ball, threw it to the first baseman with his usual force, and collapsed a few seconds later. A registered nurse attending the game was unable to feel a pulse or heartbeat and promptly performed cardiopulmonary resuscitation (CPR) for an undetermined time. Resuscitative measures were made continuously while the boy was transported 8 miles by a private automobile to a local hospital where a rhythm-strip electrocardiogram demonstrated VF. Electric cardioversion was successful and sinus rhythm was restored and maintained. After 30 minutes, the patient’s vital signs had stabilized and he was transferred by ambulance to a children’s hospital. During the transfer, tonic Seizure activity, which was controlled by an intravenous diazepam (Valium), developed.

Upon arrival at the hospital, the patient was responsive only to deep-pain stimulation. He demonstrated decorticate posture and minimal pupil response to light. respiratory support was required 36 hours after admission. Despite heroic Resuscitative efforts, he died 10 days after admission.

Maron et al [10] presented an unusual case of commotio cordis. The subject was a 14-year-old boy without prior medical problems who was engaged in a baseball game. While pursuing a batted ball in the air, he collided and was forcibly struck in the mid chest with the right knee of a teammate running toward him. The victim collapsed within

10 seconds. Two bystanders found the victim to be unresponsive, cyanotic, apneic, and pulseless without audi- ble heartbeats; the pupils were dilated and grand mal seizure activity occurred. Within 15 seconds of the accident, a physician bystander, trained in CPR, inflicted a forceful chest bthump.Q After 45 seconds, chest compressions and mouth- to-mouth resuscitation were initiated, as well as a second chest thump. About 90 seconds after the accident, a pulse became palpable and he rapidly stabilized.

A chest wall bruise representing an imprint of the knee blow was obvious to observers. Over the subsequent year, the boy remained stable without apparent psychological or neurological deficit or cardiac symptoms.

There are several unusual aspects to this case reported by Maron et al [10]. For example, the circulatory arrest is unlikely to be because of VF because VF rarely ceases spontaneously in a subject of this size. In addition, a chest thump has not been shown to arrest VF. The AHA stated that bthe precordial thump is not useful for anoxic asystole and cannot be depended upon to convert established ventricular fibrillation.Q However, if ventricular tachycardia were present, a chest blow may terminate it; but a chest

blow during the T wave of ventricular tachycardia could induce VF. In this case, the chest thumps apparently served to mechanically pace the heart. The fact that continued CPR with chest thumps resuscitated the boy indicates that SA arrest, AV block, or ventricular arrest was the reason for the collapse. The SA arrest and AV block are known to accompany a chest blow, as was reported by Schlomka and Hinrichs [3].

Nature of the blow-producing agent

If the energy-producing agent that strikes the chest is a projectile, its mass (m), velocity (v), and stiffness are important factors. The kinetic energy of a projectile is

0.5 mv²; it is this energy that deforms the chest, thereby constituting a mechanical stimulus. Note that the kinetic energy depends on the square of the velocity; therefore, high- velocity projectiles will be more dangerous for the same mass. However, this equation does not tell the complete story. Projectile size, shape, and stiffness are also important factors. If the contact area with the chest is large, the force will be distributed; but if the contact area is small, the force will be highly localized. Shape is important in defining the area of contact. For example, a ball will produce a circular area of blunt force, the extent of which will depend on the diameter and mass of the ball and its velocity and stiffness. For the same velocity and mass, a small ball will produce a higher localized area of impact than a large ball. A hockey puck, which is a short cylinder, will strike the chest on its side and resemble a small-area projectile. A factor that should be recognized is the compliance of the chest at the strike point. Obviously, a thick compliant chest protector will cushion the blow. A compliant projectile will do the same. Link et al [11] have shown that a softer baseball is safer than a standard hard ball. They also showed that the lower the velocity, the safer the ball.

In an earlier swine study, Link et al [12] demonstrated that the optimum site for a chest blow to produce VF during the vulnerable period was directly over the heart; other precordial sites were less effective.

Link et al [11,12] used a swine model to establish the foregoing facts. The study used 40 pigs ranging in weight from 8 to 12 kg and the baseballs were timed to strike the chest 10 to 30 milliseconds before the apex of the T wave of the ECG. Four baseballs were used (one standard, 3 with softer casings) and propelled at 30 and 40 mph. The incidence of VF was measured. The masses of the 4 balls were essentially the same. They found that increasing stiffness and increasing velocity increased the incidence of VF when the baseball struck the chest during the ventricular vulnerable period.

In commenting on their results, Link et al [11] stated that the risk of VF was linearly correlated with the hardness of the baseball. With the softest ball (reduced injury factor [RIF], 1), 10 animals received 26 impacts and VF was produced only 3 times (11%). With the medium-soft ball (RIF, 5), 10 animals received 23 strikes and VF occurred 5

times (22%). With the hardest safety ball (RIF, 10), 21 impacts on 10 animals resulted in 4 episodes of VF (19%). With the standard baseball, 13 impacts on 10 animals resulted in 9 episodes of VF (69%). Differences in the incidence of VF between the standard baseball and each safety ball were statistically significant ( P b .01 for comparison between each safety ball and the standard ball).

Legal implications

It may be surprising to note that commotio cordis has entered the legal system in which homicide was attributed to a blow to the chest that caused death. In the criminal justice system, guilt is established by beyond reasonable doubt. In the civil justice system, guilt is established by a preponderance of the evidence, which is a lesser burden of proof.

In surveying a registry of 124 cases of commotio cordis, Maron et al [13] found 6 cases that ended up in criminal prosecution and subsequent convictions. In each of the 6 cases, a man struck another man or a child in the chest, resulting in the victim’s sudden death. Three of the cases involved children and the blow was for disciplinary reasons. All 6 perpetrators were convicted, 4 for manslaugh- ter, one for first-degree murder, and one for criminal negligence. The following paragraphs summarize typical cases.

Boglioli et al [14] reported a commotio cordis homicide that involved a 28-month-old, 28-lb boy who was struck on the chest or back by the babysitter. The child collapsed immediately and was pronounced dead at the scene and the case was referred to the Medical Examiner. The autopsy was unremarkable.

Two cases of commotio cordis homicide involving a 14-month-old child and a 3-year-old child were reported by Denton and Kalelkar [15]. The first case involved a 14- month-old boy, well developed and well nourished, who was reportedly crying after he was given a bath by his mother. There was no history of prenatal problems, birth trauma, heart defects, or recent illness. The mother placed him, lying on his back, on a firm changing table. She dried him off and stepped out of the room to get a diaper as he continued to cry. The child’s father, in the next room, was irritated by the crying. The father approached the child and struck him with a closed fist once on the chest, on its medial soft side. He described the punch as bhard.Q The child’s eyes rolled upward and he immediately became unresponsive with short gasping breaths. The mother immediately summoned paramedics, who determined that his heart rhythm was VF. Aggressive resuscitative and electroconversion attempts were unsuccessful and the child died in the emergency department. At autopsy, there was no evidence of injury to the skin, subcutaneous tissues, or musculature.

The second case involved a well-developed 3-year-old boy in the care of his mother’s boyfriend. The child was

standing in a bedroom, crying, while the mother’s boyfriend was watching television. The child had recent Cold symptoms and was irritable. There was no history of any other illness. The boyfriend became upset at the crying, went into the adjoining bedroom to tell the child to stop crying, then hit the child on the left side of the chest with his fist. The boyfriend left the bedroom, but when the child continued to cry, the boyfriend returned and struck the child on the mid chest over the sternum. The child fell backward onto his buttocks on the floor, stood up, and then collapsed unresponsive. The boyfriend immediately called paramed- ics, relating the above events. The child was asystolic when paramedics arrived a few minutes later. The aggressive resuscitation attempt was unsuccessful. Two areas of bruising were noted during the autopsy. These bruises were consistent with the boyfriend’s account of the inflicted injuries to the chest.

Confessions were obtained in both cases by investigators. It is interesting to note that the cause of death in the first case was VF whereas that in the second case was asystole, which can occur with a blow to the chest. (Schlomka and Hinrichs [3]).

Hockey-puck accidents

Two cases of commotio cordis involving hockey players wearing protective wear were reported by Kaplan et al [16]. Both involved adolescents who were struck on the chest with a hockey puck. The first case involved a 15-year-old boy, previously in excellent health, who fell on the ice and was hit in the chest by a puck while playing high school hockey. Immediately, he had a generalized tonic seizure that lasted approximately 30 seconds. Following the convulsion, he rose to his feet briefly. Within seconds, he again fell on the ice and had another convulsion lasting 2 to 3 minutes. During the second seizure, he became cyanotic. A pediatrician and a surgeon attended the player during the second episode. Their examination revealed absence of carotid and femoral pulses. Cardiopulmonary resuscitation was immediately initiated. A call was placed to the emergency medical service, and assistance arrived 15 minutes after CPR was begun. When the paramedic teams arrived, the player was intubated. An initial electrocardiogram showed VF. Intensive resuscitative measures were applied. Defibrillation with 200 watt-seconds (J) then with 300 watt-seconds failed to elicit a change in cardiac rhythm. After 25 minutes of resuscitation on the ice, the boy was taken by ambulance to the emergency department of a local hospital. After 20 minutes of CPR in the emergency department, an agonal rhythm was observed on the ECG and CPR was stopped. At autopsy, an oval bruise was present over the left portion of the precordium, but gross and microscopic examination of the heart, lungs, and brain were within normal limits.

The commercially manufactured chest protector worn by the player was of a design that covered both shoulders with

a dense preformed plastic matrix; however, the mid and lower precordium and lateral chest regions were protected by thinner, more conformable 0.7-cm-thick closed-cell plastic foam sandwiched between woven synthetic cloth. The position of the bruise to the chest indicated that the puck had struck the portion of the chest protector composed of the 0.7-cm closed-cell plastic foam.

The second case [16] involved a 15-year-old boy who suffered cardiac arrest after being struck in the left chest by a hockey puck. A pediatrician was present and CPR was initiated within 1 minute of the event. Physical examination was remarkable for cyanosis and absence of pulses and of jugular venous distention. Advanced life support at the scene included orotracheal intubation, CPR, and adminis- tration of intravenous epinephrine (2 mg) and atropine (2 mg). The boy arrived at a hospital’s emergency department 20 minutes after his collapse. An ECG recording initially showed complete heart block with an ECG, atrial rate of 80/min and a ventricular rate of 75/min. However, we do not know the force of the ventricular contractions. Despite heroic resuscitative efforts, the patient did not respond to therapy and his cardiac rhythm rapidly deteri- orated into a wide-complex bradycardia. Resuscitation efforts continued for 25 minutes without success. Autopsy revealed a patterned bruise consistent with the shape of the puck on the left chest and associated fracture of the left fifth rib with contusion of the intervening left lung and left anterolateral pericardial sac. There was no significant injury to the heart and no underlying cardiac disease or other condition.

The chest protector worn by this player covered both the shoulders and chest, incorporating a circular precordial foam pad 30-cm high and 34-cm wide, composed of woven synthetic cloth sandwiching 0.7-cm thickness of closed-cell foam. Centered within this pad was a smaller 23 x 23 cm oversewn pad of similar material creating a thicker central precordial zone of protection. Localization of the puck injury to the area that was covered by the protective garment indicated that the puck had struck predominantly the lateral, thinner region of the chest protector.

These 2 cases reported by Kaplan et al [16] are interesting in that the puck caused VF in the first case and complete atrioventricular block in the second. Both subjects could not be resuscitated. Another interesting fact is the apparent inability of the protective wear to shield the chest adequately against the blow of the hockey puck.

Rubber and plastic bullets

Nonlethal weapons appeared to be an attractive method for controlling riots and political uprisings. Rubber and plastic bullets allow law enforcement personnel to be at a distance from unruly protesters, making it safe for law enforcement personnel. However, there have been incidents in which bodily harm and death have accompanied the use

of such bnonlethalQ weapons, particularly in Northern Ireland.

Rubber bullets

In 1970, the British Army was authorized to use rubber bullets to put down uprisings in Northern Ireland. The manner of use was also specified to minimize injury. The method of use was to fire the bullets so that they would strike the lower limbs; nonetheless, injuries and deaths did occur. The following describes the characteristics of the rubber bullets shown in Table 1.

Rubber bullets are high-velocity projectiles fired from a rifle at 160 mph. Each projectile weighs 5.25 oz (149 gm), about the weight of a baseball. By 1974, 55,688 rounds had been fired in Northern Ireland, causing injuries and 3 deaths. From 1970 to 1972, surgeons at the Royal Victoria Hospital in Belfast, Ireland, reported a survey of 90 patients, aged 7 to 67 years, injured by rubber bullets. Metress and Metress

[17] stated that injuries to the head and neck were frequent, some of which were severe. Skull fractures, ruptured eyeballs, major brain damage, and damage to internal organs were all found.

It is useful to examine the circumstances of the 3 deaths from rubber bullets in view of the severity of the above- described injuries. The first death was that of an 11-year-old boy shot in the head by a rubber bullet at a 5- to 7-yd range. The second was an 18-year-old boy bstruck in the heart Q at point-blank range. The third death involved a boy who was struck in the head by a rubber bullet at a 25- to 30-yd range. The second case is of interest because it may have been a case of commotio cordis because of the high-velocity bullet; however, we will never know so because there was no ECG evidence.

With the foregoing injuries in mind, it is useful to examine the data in Table 1. Note that a 149-g rubber bullet, traveling at 160 mph, struck the chest of the second victim who died. Link et al [11] showed that a 150-g baseball traveling at 30 to 40 mph and striking the chest during the ventricular vulnerable period can cause VF and death (commotio cordis). The sudden death of the second victim might be attributable to this cause.

Plastic bullets

Plastic bullets were introduced as a riot-control weapon by the British Army in Northern Ireland in 1973 and became fully operational by 1975. A plastic bullet is a solid polyvinyl chloride cylinder that measures 4 1/2 x 1 1/2 in

Table 1 Rubber	and	plastic	bulletsa
Item	Rubber			Plastic
Length (in)	5.75			4.5
Diameter (in)	1.5			1.5
Weight (oz)	5.25			4.75
Velocity (mph)	160			200
a Data from Irish News (1981).

and weighs 4 3/4 oz (134 g). The bullets are fired from an antipersonnel gun, a grenade-launching weapon that is 3- feet long. The bullet has an operational range of 36 to 72 yd, allowing for both greater range and greater accuracy. The bullet travels with a velocity of 200 mph. According to Ritchie and Gibbons [18] over 54,000 rounds were fired over a 13-year period. During this time, 4 people were killed by a chest strike and 3 died of VF. The high velocity (200 mph) of the plastic bullet striking the chest indicates possible commotio cordis. The fourth victim died of a hemopneumothorax.

In commenting on their study that involved 80 patients, Ritchie and Gibbons [18] stated that all injuries to the chest caused by plastic bullets should be regarded as potentially life threatening. They recommended that all patients be admitted to a hospital so that cardiac monitoring and enzyme state would be determined and that chest radiography should be repeated after a period of observation to detect hemop- neumothorax and pulmonary contusion, which may present insidiously even in the absence of a fracture. Adequate analgesia should be prescribed to prevent the secondary complications of retention of sputum.

Victim-related factors of commotio cordis

For sudden circulatory arrest to occur after a non- penetrating blow to the chest, 4 conditions must be present:

(1) the blow must strike the chest in the area of the heart with sufficient force, (2) the projectile must have enough mass,

(3) the blow must strike the chest during the ventricular vulnerable period to produce VF, and (4) the ventricles must have a large-enough mass to sustain fibrillation (Critical mass concept). It should be noted that a sudden blow to the chest can cause delayed death without fibrillation; these factors will be discussed in the following sections.

The vulnerable period

The vulnerable period is the time window in the ventricular recovery cycle during which a single stimulus will produce VF. The time window is associated with the T wave of the ECG.

That there is a sensitive time in the cardiac cycle when VF can most easily be precipitated was discovered in 1940 by Wiggers and Wegria [4], who reported its extent using single capacitor and induction-coil shocks delivered to the ven- tricles of dogs. They found that a suprathreshold stimulus was needed to evoke fibrillation. The importance of the ventricular vulnerable period came to the attention of clinicians who implanted fixed-rate cardiac pacemakers. In some patients, AV conduction returned some time after pacemaker implantation and the ventricles experienced competitive pacing; one excitation was conducted through the AV node and the other was a stimulus delivered by the

implanted pacemaker. Sooner or later, a pacemaker pulse was delivered during the vulnerable period. The first to document this fact in human beings were Tavel and Fish [19] in 1964 in a patient who, after pacemaker implantation, regained some degree of AV conduction. A pacemaking-strength stimulus landed in the vulnerable period and produced a ventricular tachycardia that required a countershock. This situation led immediately to the creation of the ventricular-inhibited pacemaker (developed by a team headed by Lemberg et al. [20]), which solved the problem.

Fig. 2 illustrates precipitation of VF by delivery of a single, 10-millisecond stimulus during the ventricular vul- nerable period. Note that the stimulus was delivered during the T wave of the ECG.

The temporal location of the center and the width of the ventricular vulnerable period depends somewhat on the duration of the stimulus pulse. A study by Jones and Geddes

[21] in 1977 revealed this information. In a dog study in which stimulating electrodes were applied directly to the ventricles, they measured threshold peak current required to produce an extrasystole with 5-, 1.0-, and 0.5-millisecond rectangular stimuli, delivered earlier and earlier during ventricular recovery. The results produced strength-interval curves shown in the lower part of Fig. 3A. Then, the stimulus intensity was increased for each pulse duration and the threshold current for VF was determined at varioUS times during the ventricular recovery cycle. The thresholds for precipitating VF during the vulnerable period are shown by the V-shaped curves in the upper part of Fig. 3A. Note that the threshold for fibrillation is much higher than that required for an extrasystole in a normal heart; this is not true with myocardial hypoxia. The extent of the vulnerable period is shown in Fig. 3B, along with the T wave of the ECG. Fig. 3B also illustrates a ventricular transmembrane potential and the vulnerable period is identified on it, extending from about 60% to 90% of the QT interval.

Critical mass concept

For fibrillation to be sustained in cardiac muscle, 3 factors must have the correct relationship: (1) excitation

propagation velocity, (2) refractory period, and (3) adequate path length so that the excitation can be reentrant. This latter factor is now known as the critical mass concept. Because propagation velocity and refractory period are inherent properties of the myocardium, a critical mass is needed to provide an adequate path length for reentrant excitation to be sustained.

The pioneering articles on reentrant excitation are those reported by MacWilliam [22], Mines [23], and Garrey [24]. MacWilliam stated that bspontaneous recov- ery [from fibrillation] may occur readily in the heart of the cat, rabbit, rat, mouse, hedgehog and fowl.Q The studies by Mines [23] used the hearts of several species, and he recognized the importance of the refractory period, heart size, and stimulus frequency using different sized hearts. He described circulating excitation (now called reentrant excitation) and stated, bit [his theory] is capable of further extension to explain the condition of fibrilla- tion, which arises under precisely those circumstances which produce the conditions in heart muscle essential for the excitation.Q

Garrey [24] precipitated VF in the dog heart and then reduced the ventricular size by cutting away slices. The excised portions did not continue to fibrillate. The remain- der of the ventricular mass sustained fibrillation until about 3 quarters had been removed. He thereby proved the need for a critical mass of myocardial tissue to sustain fibrillation. In addition, he showed that shape is also important; long thin strips cut from a fibrillating heart did not continue to fibrillate. However, strips with greater cross-sectional area were able to sustain fibrillation.

In 1967, Louhimo [25] investigated the effect of blunt trauma to the chest on cardiac function. He used 157 rabbits, all weighing less than 2.5 kg. The chest blow was delivered by a cylinder falling in a vertical tube mounted above each animal’s chest so that the weight would always strike the mid or lower sternum. The falling weight was based on a fraction of each rabbit’s weight. In these studies, he encountered AV block, bundle-branch block, and ST segment shifts; however, he encountered VF only once. Recall that the rabbits weighed only 2.5 kg or less, a weight that indicates that the heart was too small to sustain VF.

Fig. 2 Precipitation of ventricular fibrillation by delivering a single 10-millisecond pulse during the T wave of the ECG in an experimental animal.

Fig. 3 Threshold strength-interval curves (A) for extrasystoles using 0.5-, 1.0-, and 5-millisecond pulses delivered earlier and earlier in the ventricular cycle. The V-shaped curves represent thresholds for ventricular fibrillation by vulnerable period stimulation. In (B) is shown the extent of the vulnerable period in relation to the transmembrane ventricular monophasic action potential and in relation to the T wave of the ECG.

In a study by Geddes et al [26] in which the threshold transchest current required for Ventricular defibrillation was being determined over a wide range of body weights up to 450 kg, it was found that rabbits weighing less than about 3.5 kg would not sustain VF. In other words, the ventricles were too small to sustain the reentrant excitation needed to sustain fibrillation. This explains why Schlomka and Hinrichs encountered VF in their study, which used very large rabbits (12-23 kg). Louhimo

[25] encountered VF only once in his animals, which all weighed less than 2.5 kg.

In another defibrillation study, Geddes et al [27] determined the threshold current for the defibrillation of hearts ranging in weight from 60 to 2000 g, with electrodes applied directly to the heart. Hearts weighing less than about 25 g did not sustain VF.

In 1975, Zipes et al [28] performed an interesting experiment analogous to that of Garrey [24]. In their study, Zipes et al rendered parts of the ventricles of dog hearts inexcitable by selective intracoronary injections of potassi- um chloride, a depolarizing agent. They found that the critical mass for sustaining fibrillation was 28% of the total ventricular myocardium. They used dogs weighing from 15 to 30 kg. For a heart weight of 0.6% of the body weight, the corresponding heart weight range is from 90 to 180 g.

Therefore, the critical mass range is 25.2 to 50.4 g, respectively.

Extent of the vulnerable period

From Fig. 3, it is seen that the vulnerable period extends from 60% to 90% of the QT interval. Choosing 0.36 seconds as the QT duration for a heart rate of 70/min, the width of the ventricular vulnerable period is 108 milli- seconds, which is quite a narrow time window; yet the published literature shows that chest blows have occurred in this time window and produced VF.

Pacing with a chest blow

A chest blow can be lifesaving when applied to a patient with cardiac arrest. For example, use of a fist blow to the chest to produce an ectopic heartbeat was reported by Wild and Grover [29], who stated that the ventricles of 3 elderly patients ceased beating because of AV block. In each, chest thumps with the fist at the lower left sternum produced ventricular excitations as shown by ECG. A peripheral pulse accompanied each ventricular excitation. All patients

recovered. Based on these successes, they recommended use of chest thumping in ventricular asystole. The 1973 American Heart Association [30] standards for CPR advocated use of the bprecordial thumpQ for closed-chest pacing only under the following conditions:

1. witnessed cardiac arrest (basic life support)
2. monitored patient (advanced life support)
3. pacing known A-V block (advanced life support)

The standard continued:

The role of the precordial thump in the unmonitored patient or in an unwitnessed cardiac arrest has not been determined yet so specific recommendations cannot be made for those situations. At this time, the precordial thump is not recommended for use on children.

Use of the chest thump to arrest ventricular tachycardia and VF was advocated by Caldwell et al [31]. In their view, a precordial thump should be used bblindly Q in all unmoni- tored patients who develop circulatory arrest. They continue by stating that a precordial thump may be successful early after the onset of VF and during a longer period if the rhythm is rapid ventricular tachycardia or asystole.

The effectiveness of a chest blow during the ventricular venerable period could be affected by chest compliance. For example, a stiff chest would prevent the full force of a blow from reaching the ventricles. Conversely, a compliant chest such as that of a child would favor transmission of the force of the blow to the ventricles, which, if occurring during the vulnerable period, could induce VF.

Conclusion

From the foregoing review of human accidents, animal studies that involved a nonpenetrating chest blow, and ventricular defibrillation studies on animals, several con- clusions can be drawn. For example, if unconsciousness follows immediately from a chest blow, there are 2 possibilities: (1) circulatory arrest from VF due to vulner- able-period stimulation or (2) ventricular arrest due to SA node arrest or a very slow ventricular rate due to AV block. In both cases, immediate CPR is mandatory without ECG evidence. If ventricular arrest is present and confirmed by an ECG, a chest thump can produce an ectopic beat and repeated thumps may be lifesaving. However, if ventricular activity is restored, there is the danger of a chest thump landing in the vulnerable period and inducing VF.

It has been shown conclusively that a nonpenetrating blow to the chest can cause many different types of arrhythmias including SA node arrest, AV block, bundle- branch block, ST segment shifts, bradycardia, and tachy- cardia, which may progress to VF [5].

Finally, sustaining VF requires a critical mass of heart and body weight. From animal studies, hearts weighing less

than about 25 g are not likely to sustain VF. Similarly, body weights less than about 3.5 kg are unlikely to have hearts large enough to sustain VF.

Therefore, all medical personnel that encounter sudden cardiac arrest should be aware of the possibility of commotio cordis being the causative factor after a nonpenetrating chest blow, which can be confirmed by the ECG.

References

1. Maron BJ, Gobman TE, Kyle SB, et al. Clinical profile and spectrum of commotio cordis. JAMA 2002;287(9):1142 - 6.
2. Dorland’s Illustrated Medical Dictionary. Philadelphia (Pa)7 WB Saunders; 1988.
3. Schlomka G, Hinrichs A. Experimentelle Untersuchungen uber den Einfluss stumpfer Brustkorbverletzungen auf das Elektrokardiog- ramm. Z Gesamte Exp Med 1932;31:43 - 61.
4. Wiggers CJ, Wegria R. Ventricular fibrillation due to single localized induction and condenser-discharge shocks applied during the vulnerable phase of ventricular systole. Am J Physiol 1940;128: 500 - 5.
5. Schlomka G. Commotio cordis und ihre Folgen (Die Einwirkung stumpfer Brustwandtraumen auf das Herz). Berlin7 Julius Springer; 1934. p. 1 - 94.
6. Schlomka G, Hinrichs A. Experimentelle Untersuchungen uber den Einfluss stumpfer Brustkorbverletzungen auf das Elektrokardiog- ramm. Z Gesamte Exp Med 1932;85:171 - 96.
7. Schlomka G, Hinrichs A. Experimentelle Untersuchungen uber den Einfluss stumpfer Brustkorbverletzungen auf das Elektrokardiog- ramm. Z Gesamte Exp Med 1932;83:779 - 91.
8. Schlomka G, Hinrichs A. Experimentelle Untersuchungen uber den Einfluss stumpfer Brustkorbverletzungen auf das Elektrokardiog- ramm. Z Gesamte Exp Med 1933;90:301 - 18.
9. Dickman GL, Hassan A, Luckstead EF. Ventricular fibrillation following baseball injury. Physician Sports Med 1978;(July):85 - 6.
10. Maron BJ, Strasburger JF, Kugler JD, et al. Survival following blunt chest impact. Am J Cardiol 1997;79:840 - 1.
11. Link MS, Maron BJ, Wang PJ, et al. Reduced risk of sudden death from chest wall blows (commotio cordis) with safety baseballs. Pediatrics 2002;109:873 - 7.
12. Link MS, Maron BJ, Vander Brink BA, et al. Impact directly over cardiac silhouette is necessary to produce ventricular fibrillation in an Experimental model to produce commotio cordis. J Am Coll Cardiol 2001;37(2):649 - 54.
13. Maron BJ, Mitten MJ, Burnett CG. Criminal consequences of commotio cordis. Am J Cardiol 1997;79:840 - 1.
14. Boglioli LR, Taff ML, Harleman G. Child homicide caused by commotio cordis. Pediatr Cardiol 1998;19(3):436 - 8.
15. Denton JS, Kalelkar MB. Homicidal commotio cordis in two children. J Forensic Sci 2000;45(3):734 - 5.
16. Kaplan JA, Karofsky PS, Volturo GA. Commotio cordis in two amateur ice hockey players despite the use of commercial chest protectors. J Trauma 1993;(Jan):151 - 3.
17. Metress EK, Metress SP. The anatomy of plastic-bullet damage and crowd control. Int J Health Serv 1987;17:333 - 42.
18. Ritchie AJ, Gibbons JR. Life threatening injuries to the chest caused by plastic bullets. BMJ 1990;301:1027.
19. Tavel M, Fish C. Repetitive ventricular arrhythmia resulting from artificial internal pacemaker. Circulation 1964;30:493 - 500.
20. Lemberg L, Castellanos A, Berkovits BV. Pacemaking on demand in A-V block. J Am Med Assoc 1965;191:12 - 4.
21. Jones M, Geddes LA. Strength-duration curves for cardiac pacemak- ing and ventricular fibrillation. Cardiovasc Res Cent Bull 1977;15(4): 101 - 12.
22. MacWilliam JA. Fibrillary contraction of the heart. J Physiol 1887;8:296.
23. Mines GR. On dynamic equilibrium in the heart. J Physiol 1913; 46:349.
24. Garrey WE. The nature of fibrillary contraction of the heart. Its relation to tissue mass and form. Am J Physiol 1914;33:297.
25. Louhimo I. Heart injury after blunt Thoracic trauma. Acta Chir Scand 1967;30:29 - 52.
26. Geddes LA, Tacker WA, Rosborought J, Moore AG, Cabler PS. Electrical dose for ventricular defibrillation of large and small animals using precordial electrodes. J Clin Invest 1974;53(1):310 - 9.
27. Geddes LA, Tacker WA, Rosborough J, Moore AG, Cabler PS, Bailey M, et al. The electrical dose for ventricular defibrillation with

electrodes applied directly to the heart. Thorac Cardiovasc Surg 1974;68(4):593 - 602.

1. Zipes DP, Fischer J, King RM, et al. Termination of ventricular fibrillation in dogs by depolarizing a critical amount of myocardium. Am J Cardiol 1975;36:37 - 44.
2. Wild JB, Grover JD. The fist as an external cardiac pacemaker. Lancet 1970;29(Aug.):436 - 7.
3. American Heart Association/National Research Council. Standards for cardiopulmonary resuscitation and emergency cardiac care. Dallas (Tex)7 American Heart Association; 1973.
4. Caldwell G, Millar G, Quinn E, et al. Simple mechanical methods for cardioversion: defence of the precordial thump and cough version. BMJ 1985;291:627 - 30.