Abstract
Background Conductive energy devices (CEDs) have been temporally associated with morbidity and mortality in police work, but the frequency of use and of complications is not certain.
Methods This is a literature review using PubMed and Google Scholar searches to identify population-based CED studies, studies reporting morbidity and mortality with CEDs, and studies in healthy volunteers.
Results Recent studies indicate that CEDs are used approximately 83 to 338 times per million population per year in the United States. The subjects have a typical profile, including young men with belligerent or bizarre behavior who often have a psychiatric disorder or are intoxicated with drugs. The mortality estimates range from 0.0% to 1.4% of subjects controlled with CEDs. Limited information from autopsy studies indicates that death is frequently associated with confounding factors, especially intoxication with illicit drugs.
Conclusions Conductive energy devices are used frequently during police work and are associated with a low but definite mortality rate. The use of CEDs and the management of at-risk subjects need more study.
Conductive energy devices (CEDs) were introduced into police work as nonlethal weapons in the use of force continuum designed to control belligerent and dangerous subjects and to avert the use of lethal weapons.1The use of CEDs has become controversial and has been temporally associated with multiple deaths.2,3These deaths are frequently attributed to drug intoxication, natural causes, or excited delirium.4However, no single law enforcement jurisdiction has significant experience with in-custody deaths, and the analysis of the potential contributions of CEDs to in-custody death is difficult. More information on CED use and temporally associated morbidity and mortality has recently become available in the medical literature, and we have reviewed recent population-based studies that provide reasonable estimates of the frequency of use and outcomes.
MATERIALS AND METHODS
Searches
We did PubMed searches in 2009 and 2010 using the MeSH terms electric injury/etiology, electric injury/physiopathology, electric injury/complications, electroshock, immobilization, and death and the text words taser and stun gun. We reviewed the reference list of the articles identified in the PubMed search and used the related articles algorithm on the PubMed Web site. We did Google and Google Scholar searches with the phrase “taser-associated deaths” to retrieve articles in the medical literature and reports and position papers from governmental and nongovernmental agencies. We also cross-checked references listed in articles identified through the Google searches. We focused this search on population-based studies, studies reporting morbidity and mortality, and studies in healthy volunteers. The term conducted energy weapon injuries was introduced as a MeSH term in 2011. Stun gun injuries (and variations) and taser injuries (and variations) will map to this term. Only 2 articles were retrieved at present when using this MeSH term (February 7, 2011).
Amnesty International (AI) maintains a database on the use of CEDs in police fieldwork (www.amnestyusa.org).2,3The authors abstracted information from the narrative sections of these reports to summarize the use of CEDs, including the typical profile of subjects involved in police action and outcomes. We collected information on the subject’s age, sex, race, comorbidity, drug use, alcohol, restraint use, number of electroshocks, time to death, and causes of death. This information was not consistently present in each report; we did not verify the reports and did not review autopsy reports.
Terminology
Conductive energy devices are also referred to as conductive energy weapons, conducted energy devices (weapons), stun guns, and tasers. The TASER X26 model fires 2 darts (probe mode) (Taser International, Inc., Scottsdale, AZ) that allow the delivery of short bursts (usually 5 seconds) of high-voltage current (50,000 V). These currents cause neuromuscular incapacitation and pain. This device can also be used in the drive stun mode in which the barrel is held against the subject and the electrical current causes pain.
Excited delirium is a clinical syndrome observed in fieldwork, emergency departments (EDs), and psychiatric hospitals characterized by extreme mental and physical excitement with extreme agitation, hyperthermia, euphoria, hostility, exceptional strength, and endurance without fatigue. This clinical presentation does not have a formal definition in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, but has important implications in understanding the pathogenesis of in-custody deaths. The National Association of Medical Examiners uses this term for reporting deaths, and the American College of Emergency Physicians published a white paper report on this syndrome in 2009.5
Use of force continuum is a phrase that describes the options available to police officers in the management of subjects during routine police work. Eastman et al.6included a table in their article, which describes these options for the Dallas police force. Nonlethal methods are designed to control the subject with a low expectation of death.
Numerical Analysis
We used information from individual articles and reports to calculate the number of CED uses per 1 million population per year.
RESULTS
Frequency of Use
Three recent population-based studies have provided important information on CED use in the United States (Table 1). Eastman et al.6reported a prospective study from the Dallas metropolitan area. These police officers used the TASER X26 model on 426 subjects during a 15-month period (November 2004 to January 2006), resulting in a deployment rate of 273 uses per million population per year. In this study, all activations required administrative review, and all subjects controlled with a CED had a medical review. Bozeman et al.7prospectively studied the use of TASER X26 model in 6 jurisdictions for the US Department of Justice. The police in these communities used CEDs 1201 times for a mean deployment rate of 170 per million population per year. However, there were significant differences in the use rates in these jurisdictions; these ranged from 83 to 338 per million population per year. Strote et al.8reported results from a 6-year retrospective study in the Seattle police department. Conductive energy devices were used on 1101 subjects; this resulted in a use rate of 316 per million population per year. These 3 studies demonstrate that the deployment rate ranges from 83 to 338 per million population per year and that there is a significant variability among jurisdictions. These studies describe police policies that include training and retraining with CED use (and the use of force continuum) and administrative reporting of CED use; these precautions likely influence outcomes in these jurisdictions. Use patterns in other jurisdictions are not available.
Circumstances Requiring CED Use
The medical literature and other reports describe the typical circumstances requiring CED use. Vilke et al.9reviewed in-custody deaths associated with CEDs using lists obtained from AI and the Arizona Republic newspaper. They were able to obtain information on 77 decedents (a 65% response rate) and described the demographic characteristics of these decedents and the tactical considerations underlying the use of CEDs during these police actions. Most decedents were men (96%), white (46%), and between the ages of 21 and 50 years. Tactical considerations included an armed subject (26%), fleeing (30%), noncompliance (75%), and physical aggression (87%). Ninety-nine percent of the CED uses occurred after other force methods failed. Conductive energy devices can be used in a probe mode in which barbed darts are fired into the subject or in the drive stun mode in which the CED is placed in contact with the subject’s skin. The latter mode largely causes pain and was associated with approximately 25% of the deaths that occurred after the use of a CED in this study.
Ho et al.10reviewed the database maintained by Taser International, Inc, (Scottsdale, AZ) to determine whether CEDs were used in subjects who were apparently with mental illness. Approximately 25% of the episodes (10,608 reports total) requiring the use of CEDs occurred in subjects who were potentially with mental illness based on available descriptive information about their behavior in narrative sections of the reports submitted to Taser International, Inc. In 45% of these episodes, the law enforcement agencies indicated that lethal force might have been justified; in 55% of the episodes, lethal force was not justified. Ho et al. concluded that CEDs can prevent an escalation in the force continuum in subjects with mental illness and do not seem to be associated with the deaths of these subjects. This information is potentially important because subjects with psychiatric disorders are at increased risk for the development of excited delirium, which is a potential factor in in-custody deaths. However, this information is based on voluntary reporting, which is potentially selective and biased, and the degree of psychiatric disturbance in these subjects is uncertain. Therefore, any conclusion about the safety of CEDs in subjects with mental illness seems problematic.11
Conductive Energy Device Morbidity
The study of Vilke et al. describes the typical characteristics of subjects involved in police confrontations resulting in an arrest and highlights the safety considerations for both subjects and law enforcement personnel. Legal interventions during fieldwork can cause injuries to both subjects and police officers. Most nonfatal injuries involve physical contact with police. Data collected from the Centers for Disease Control and Preventions National Electronic Injury Surveillance System—Firearm Injury Surveillance Study (years 2005 to 2008) demonstrate that CED use is associated with 11.2% of legal intervention–related injuries reported in EDs.12These injuries usually involved the upper and lower trunk and included puncture wounds, contusions/abrasions, foreign bodies, and lacerations. Most suspects (93.6%) with these injuries were treated and released from the ED. The 3 population studies discussed here (Table 1) reported significant variability in injuries, which ranged from 0% (Eastman et al.,6only first aid needed) to 26.8% (Strote et al.,8295 ED visits and 72 hospital admissions). These differences may reflect police policies or ED policies or both. The study of Strote et al.8also documented that 5% of subjects brought to the ED had rhabdomyolysis, acidemia, elevated troponin level, or their combinations.8Regardless of the exact cause of these abnormalities, some subjects were at risk for clinical deterioration.
The medical literature includes multiple case reports and short reviews of morbidity and unique events associated with CED use.13–17Mangus et al.14reported a case series with 4 patients who sustained significant head injuries after falls secondary to neuromuscular incapacitation after the use of CEDs to control subjects in the field. These injuries included subarachnoid/epidural hemorrhage, concussion, seizures, and forehead hematoma. Naunheim et al.16recently reported a case in which a 17-year-old adolescent boy became intoxicated and violent at home. He was shocked once and collapsed. Paramedics documented ventricular fibrillation within 4 minutes of collapse. Feeney et al.17recently reported a 37-year-old man without a psychiatric history, illicit drug intoxication, or prolonged physical exertion who developed excited delirium after a 3-block chase and 4 CED shocks.
In general, injuries reported with CED use are minor and expected (eg, lacerations). However, some subjects have more serious injuries secondary to falls; metabolic derangement related to exertion, heat, and drugs; and, rarely, cardiac rhythm disturbances. The 2 case reports summarized here have important implications. The case with ventricular fibrillation suggests that CED currents occasionally capture the heart rhythm. The second case raises the possibility that high-voltage electrical currents can induce or increase excited delirium. These are important implications because medical examiners often conclude that arrest-related deaths are explained by excited delirium and that CED use is coincidental.
Conductive Energy Device–Associated Mortality
Ordog et al.15reported information collected on 218 patients brought to EDs after CED deployment for violent or criminal behavior. These subjects were compared with 22 similar subjects shot by police with 0.38 handguns. The mortality in subjects controlled with CEDs was 1.4% (3/218; 95% confidence interval [CI], 0%–2.9%); the mortality rate of those shot with handguns was 50% (11/22; 95% CI, 29%–71%). The long-term morbidity rate was 0% in subjects controlled with CEDs. Short-term complications included contusions, abrasions, lacerations, mild rhabdomyolysis, and testicular torsion. Of the subjects controlled with CEDs, 48% required hospitalizations, usually for preexisting injury, drug toxicity, or psychiatric problems. The high mortality rate (3 deaths in subjects managed with CEDs) might be explained by the use of CEDs on subjects with phencyclidine intoxication. All 3 subjects developed cardiac arrest and presented to medical care in asystole. This report concluded that CEDs were relatively safe, especially when compared with shooting with conventional weapons that carry a high expected mortality rate. However, this study was published in 1987, and the police used earlier models of CEDs. Therefore, this information may not be applicable to current use of force continuum policies and newer CED models. In addition, the mortality secondary to lethal weapons could be different.
The 3 most recent studies may provide a better estimate of morbidity and mortality (Table 1). Eastman et al.6reported 1death (0.2%) in the prospective study in Dallas, and this was attributed to toxic hyperthermia related to exertion and cocaine. Eastman et al.6also concluded that 5.4% of CED use prevented the use of lethal force. The study of Bozeman et al.7in 6 jurisdictions reported 3 major injuries (2 intracranial bleeds and 1 rhabdomyolysis case) and 2 unexpected deaths (0.17%) in-custody. These 2 individuals had participated in prolonged struggle, had evidence of drug toxicity (1 with cocaine and 1 with olanzapine), and likely had excited delirium. The medical examiner maintained that the electrical shocks did not contribute to these 2 outcomes. There were no deaths in subjects managed with CEDs in the Seattle study.8Eastman et al. and Strote et al. suggested that we need more studies on subjects who are at risk for in-custody death and suggested that subjects managed with CEDs might identify a subgroup at risk.
Four subjects in the Seattle study had excited delirium, and this information provides an estimate of the frequency of excited delirium in subjects who required CED use during arrests.8Our calculation indicates that 0.61 cases per 1000 subjects managed with CEDs per year likely have excited delirium. Stratton et al.18reported that approximately 36 subjects per year (from a mean of 196,261 emergency medical service [EMS] calls per year) had the clinical features of excited delirium and had been restrained by the police in Los Angeles County, California. This produces an incidence of 0.18 cases per 1000 EMS calls per year. This is a minimal estimate because some subjects are taken directly to a medical facility without calling EMS. The mortality rate in these subjects in LA County, California, with ED was 10.95% (23/210; 95% CI, 6.7%–15.2%). These field studies demonstrate that mortality is low in confrontations requiring nonlethal force and that excited delirium is relatively uncommon. However, the risk for mortality in subjects with excited delirium is high in some situations.
Lee et al.19collected information from 126 police and sheriff departments in California using mail surveys to determine when electronic weapons were introduced into police work and the number of in-custody deaths (without lethal force), the number of lethal force deaths, and the number of officer injuries during police work. Most of the information came from 50 police and sheriff departments that had adequate data on deaths 1 year or longer before deployment and 1 year or longer after deployment. This study demonstrated that the number of in-custody sudden deaths per 100,000 arrests increased from 0.93 per 100,000 during the baseline period to 5.96 per 100,000 in the first full deployment year of electronic weapon use. In addition, the number of lethal force deaths increased during the first full year of electronic weapon use. These numbers subsequently fell toward the baseline rate during years 2 to 5 after the introduction of electronic weapons. Therefore, this study suggests that the introduction of CEDs actually increased the number of in-custody deaths and lethal force deaths without a reduction in the number of serious officer injuries. With training and experience, these numbers fell toward the preintroduction baseline. This study also indicates that the number of deaths occurring in the absence of lethal force had the same order of magnitude as the number with lethal force (1.44–9.1 per 100,000 arrests). These reports did not directly implicate CEDs as the cause of death, and the exact frequency of use in these arrest-related deaths is unknown. Conductive energy devices were introduced in these cities between 1985 and 2007. Consequently, secular trends in drug use or violent crime could not explain these results unless similar trends occurred in these cities at quite different times. Finally, this study did not include the large metropolitan areas outside California, which declined to participate. Therefore, this information may not reflect the use pattern in some major cities accurately.
Ho et al.20also collected information using an open-source Web-based method for 12 months. Search terms included excited delirium, prehospital death, custody, dies, and death, among others. However, the terms taser, conducted energy device, stun guns, and others, were not used in these searches. Approximately 30% of the arrest-related deaths occurred after the use of a CED. None of these deaths occurred instantaneously. This study almost certainly underestimates the number of arrest-related deaths and minimizes the potential association with CEDs.20These last 2 studies demonstrate that the analysis of arrest-related deaths associated with nonlethal force is very complicated and that many factors potentially limit analysis. Uniform and mandatory reporting by police departments would significantly improve the quality of information.
Autopsy Information
Amnesty International collected information on 152 deaths associated with CED use in North America.2,3We have tabulated the information from these databases and summarized it in Table 2. Most subjects were men with a mean age of 35 years. Of the subjects with toxicology studies, 60% had at least 1 drug detected. Moreover, 86% required control with additional restraints. The mean number of CED strikes was 3.3 with a range of 1 to 20. The time to death was within 0 to 24 hours in 72.5% of the subjects. The cause of death was attributed to cardiac disease, drug use, and agitated delirium. These reports suggested that CED use contributed to death in 14.4% of cases. Autopsies were done in 21.7% of cases; we summarized this information in Table 3 for comparison with the study of Kornblum and Reddy21(with 16 subjects).
This autopsy information again emphasizes the complexity of these death investigations and the difficulty in establishing causal relationships. Both internal subject factors (comorbidity and drug toxicity) and external factors (environment, exertion, and use of force) influence outcomes. The interaction among these factors (eg, additive, synergistic, conditional) is unknown. Finally, some in-custody deaths occur without the use of CEDs, and all these situations need more investigation.
Pathophysiologic Effects
Recent studies have evaluated cardiac, respiratory, and metabolic responses to CED applications in volunteers. These studies usually involve healthy volunteers who meet both inclusion and exclusion criteria; the CED darts are applied to the back in a controlled environment, with subjects supported so that they do not fall. These studies have demonstrated that CEDs do not cause significant arrhythmias, changes in blood pressure, or increases in cardiac biomarkers in resting subjects.23,24However, Nanthakumar et al.25and Walter et al.26have reported that CED discharges with the vector across the heart in anesthetized pigs stimulated the heart and increased the heart rate in 80% to 100% of animals. In addition, under conditions of simulated adrenergic stress with epinephrine infusions, CED discharges stimulated the heart in 13 of 16 trials and caused 1 episode of ventricular tachycardia and 1 episode of ventricular fibrillation.25These animal studies demonstrate that the orientation of the electrical discharge in relationship to the cardiac axis strongly influences the frequency of cardiac stimulation and the development of ventricular arrhythmia. Hence, studies in human volunteers may not reproduce CED use in the field during periods of stress, especially when dart application is remote from the heart.
The 5- and 30-second CED exposures do increase the respiratory rate and minute ventilation, do increase lactate levels, and do decrease the pH.23,27These physiological changes are not associated with immediate or near-term clinical consequences in these volunteers. Conductive energy device exposure after prolonged physical activity (push-ups and treadmill running until exhaustion) had the same effect as additional exertion.28These subjects had pH values as low as 7.11. Conductive energy device application after exercise to 85% of predicted heart rate increased the heart rate for at least 60 minutes in healthy volunteers.29A 5-second CED exposure increases cortisol secretion for 60 minutes.30Exposure to CED increases lactate levels in healthy volunteers but not to the same degree as a 150-m sprint with a wall hurdle or 45 seconds of striking a heavy bag.31Pasquier et al.32recently reviewed this literature and concluded that most of these studies had industry support or links and had poor methodological scores. Consequently, there is a potential for bias in the results and reporting.
In aggregate, these studies demonstrate that the electrical current delivered by CEDs can cause cardiac, respiratory, and metabolic stress, including lactic acidemia. They have an additive effect when added to exertion. These effects are usually not important in healthy volunteers but may be very important during police work when the situation likely involves young men, belligerent behavior, prolonged physical struggles, and/or drug toxicity.33Effects on the respiratory system could cause dyspnea and severe anxiety in subjects with respiratory impairment (such as obese subjects). The prolonged effect on heart rate could increase myocardial oxygen demand and potentially contribute to cardiac ischemia and cardiac failure in some subjects. This might explain the deterioration observed after a period of quiet when the physical struggle stops.
DISCUSSION
Our review demonstrates that CEDs are used frequently in police work; we calculated a mean deployment rate of approximately 250 per million population per year using the 3 population-based studies.6–8Subjects requiring control with these devices have common characteristics, including young men, often minorities, who display violent or bizarre behavior and who are often intoxicated with cocaine, phencyclidine, and other drugs.8,9The frequency of excited delirium in the Seattle study is approximately 1 case per 1000 per year in subjects managed with CEDs and brought to medical attention. The mortality rate temporally associated with CED use ranged from 0.0% to 1.4%; the mean mortality was 0.2% (6/2946; 95% CI, 0.04%–0.36%).6–8,15Based on a US population of approximately 300 million, the deployment rate would average 76,000 times per year, with a mortality rate ranging from 30 (lowest) to 275 (highest) per year. Therefore, the number of deaths in any given municipality would be very small. However, these numbers have the same order of magnitude as estimates of justifiable homicide cases by police reported by Loftin et al.34and fatal police shootings reported in Harris County, Texas.35These estimates range from 376 (Loftin et al.34) to 1011 per year (our extrapolation from Strote et al.22).35These CED-associated deaths are usually attributed to drug toxicity, natural causes, and excited delirium, and the contribution of CEDs is difficult to determine.4Statistics presented by police departments generally demonstrate that these devices are used in appropriate situations in an effort to avoid the use of lethal handguns. However, the study from California demonstrated that the number of in-custody deaths and lethal force deaths actually increased during the first year of CED use.19In addition, some reports in newspapers suggest that the devices are used in situations that do not represent any immediate risk to the police or bystanders and, occasionally, are used on people with medical emergencies. The widespread use of these devices at times has generated controversy and concerns from advocacy groups. However, the discussion of police action, CED use, and in-custody deaths in the media, medical literature, and/or official records cannot convey the complexity of decision making in the field when confronting dangerous, possibly armed, individuals who require control for their safety and for public safety.
We have suggested that these devices can cause an acute stress cardiomyopathy in certain individuals and that this contributes to their deaths through acute cardiac decompensation and not necessarily through cardiac arrest.36–38Data from studies on healthy volunteers demonstrate that neuromuscular incapacitation and pain cause acute cardiorespiratory distress and cortisol secretion. Some subjects with or without comorbidity, heat stress, exertion, and drug toxicity may not tolerate this stress and die in custody. However, other commentators have suggested that CED deployment interrupts excited delirium and prevents progression. These differences cannot be resolved with the current medical or epidemiological information.
LIMITATIONS
Our review has definite limitations. Our main goals were to determine the frequency of complications associated with CED use and to determine the contribution of CEDs to mortality. There is no national database available with comprehensive forensic case review, toxicology, and autopsy to determine the causes of deaths in these subjects. The National Institute of Justice Bureau of Research Statistics tracks arrest-related deaths but does not classify CED-associated deaths into a separate category. Therefore, we used information from the medical literature to make the estimates and to reach tentative conclusions. Our estimates on the complications associated with these weapons clearly depend on reporting accuracy and may be low.39In addition, most of the published information comes from larger police jurisdictions that may provide more training; outcomes with less-trained personnel or in other communities with more drug use may have higher morbidity and mortality. Obviously, the relatively low frequency of excited delirium and in-custody deaths limit the experience of ED physicians and forensic pathologists and reduce the information available in the medical literature. Our effort to determine the contribution of CEDs to mortality is even more difficult because multiple factors likely influence the outcomes in these situations, and our current understanding of the pathogenic events in these cases precludes definitive conclusions. More information on the molecular pathology with molecular signatures in cases with excited delirium, drug toxicity, and profound sympathetic stress might allow the classification of the primary causes of death better.40In addition, we need more comparisons among different methods of nonlethal force and outcomes.41
CONCLUSIONS
In summary, our review suggests that CEDs are used approximately 250 times per million population per year as part of a nonlethal force continuum for the management of dangerous subjects. These subjects sometimes have the clinical features of excited delirium and are at risk for death. Pain and some physiological effects of CEDs, especially falls, may contribute to the morbidity and mortality associated with nonlethal force. We need more information about in-custody deaths to determine whether CEDs represent an additional risk or are used in dangerous situations in which there is a low and unpredictable mortality rate even if lethal weapons are not used. We need registries to collect this information on the treatment of subjects with excited delirium and autopsy results from decedents. We also need more clinical and experimental information on the effects of electrical current and associated pain on the sympathetic nervous system and the cardiac system in subjects under severe stress. Finally, physicians in EDs and intensive care units need to remember that CED use may help identify subjects at risk for death regardless of the underlying mechanism of injury.
ARTICLE SUMMARY
Why is this topic important? Conductive energy devices have been associated with morbidity and mortality in police work, but the frequency of use and of associated complications is not certain.
What does this study attempt to show? We have reviewed recent population-based studies that provide reasonable estimates of the frequency of use of CEDs and the outcomes after their use.
What are the key findings? We found that CEDs are used frequently in police work, and we calculated a mean deployment rate of approximately 250 per million population per year using 3 population-based studies. The overall mortality in these studies was low.
How is patient care impacted? Subjects controlled with CEDs sometimes have the clinical features of excited delirium and are, therefore, at risk for death. Physicians in EDs and intensive care units need to remember that CED use may identify patients at risk for death regardless of the underlying mechanism(s) of injury.