Are CEDs Safe and Effective?
Thousands of law enforcement agencies throughout the United States have adopted conducted energy devices (CEDs) as a safe method to subdue individuals, but are these devices really safe? What policies should agencies adopt to ensure the proper use of this technology? This NIJ Conference Panel discusses the physiological effects of electrical current in the human body caused by CEDs, as well as how this technology can reduce injuries to officers and suspects when appropriate policies and training are followed.
Joseph Cecconi: Good afternoon. This is going to be the “Are CEDs Safe and Effective?” session. I'm Joe Cecconi. I am with NIJ, moderator of this panel here, and just a quick intro here.
At the office, I use a definition for “less lethal” as a device when compared to firearms, reduces the probability of a bad outcome for police, for bystanders, for suspects. I don't use the term “weapon.” I use the word “device.” “Weapon” has the implication of causing harm, death or maiming. There is no intent to do that.
So, notionally, if one were to say, if you were to compare yourself to firearms, this is basically what you'd see. You'd see your injuries over in the light blue on the left side there, very high, and if you look at baton CEDs, physical controls, OC, others, you are going to be seeing a significant, significant reduction in bad outcomes.
So this basically shows that the community, the less-lethal community is doing their job, and they are improving outcomes. Now, this also, you will note down there, of all situations, 1.5 percent require the use of force, so this is a very small number that actually went into force.
So why is that notional? There is a lot of moving parts in trying to do these numbers insofar as there's different devices out there. There's probably 10, 15, 20 different types of kinetic energy launchers and rounds. The physiological outcome is different for each person, and there is no funding been put into psychological effects, how these devices reduce injuries by using … by psychological effects.
The effectiveness also is poorly studied in the community, and the other areas, use of force policies, procedures and training (PP&T), when you look at that, it gets very difficult, too, because we've got 21,000 different law enforcement agencies doing the use of force policies, procedure and training. And when you start looking at this, it gets complicated by the fact there's a triad here between the officer, suspect, and stand by … and bystanders. If you make it too difficult for the officer, he can't react quick enough. That puts bystanders in jeopardy. It makes it safer for the suspects, but if you make it too lenient for the officer, suspects will bear the brunt of it.
So, with that, I ask the question: Are we ready for the next generation, RF less lethal devices?
And with that, I want to introduce Dr. John Hunsaker. What we did at NIJ a while back is convene a panel of medical examiners, cardiologists, emergency room docs, and we met a bunch of times, reviewing in-custody deaths when a Taser was used, and we had come out with an interim report, and we're going to be coming out with a final report here shortly.
And Dr. Hunsaker is one of the co-chairs on the panel. He's the associate chief medical examiner in the Kentucky office of the chief medical examiners, and he's on the editorial board of the American Journal of Forensic Medicine and Pathology and Forensic Science, Medical and Pathology.
John Hunsaker: The way this was phrased in terms of what the purpose of this session was, specifically, asked this. And, again, I'll have to get this in the light. “A panel will discuss the physiologic effects of electrical current in the human body as caused by CEDs” and so forth, and that's, in a sense, a question begging formulation in the sense many folks argue that current is not really in the body unless you count those muscles immediately below the skin but certainly not in the chest cavity.
What I want to do is summarize what this panel did, and as Joe mentioned, the idea is that it will be, hopefully, the final report will be coming out sometime midsummer before the fall.
We started back in 2006, and back in those days, we were calling them EMDs, and now it's morphed over to CEDs, and some people even call them CEWs, which, apparently “W” is a bad word not only in the last administration but in connection with these devices. And our basic question was, “Can CEDs cause death under certain circumstances, contribute to death or have no bearing on death at all?” And that was in the context that back then, there were well over 300 incidents around the countries of death during attempted restraint or “subdual” — is the word of art we use — of various folks by law enforcement and who died after the use of one or more different types of CEDs.
So we looked at those cases and winnowed that down to 22 specific cases from around the country with the help of the International Association of Chiefs of Police and their folks in terms of getting as much information as possible about each case and batting those around over a series of the last meeting — roughly three or four times over the last three years.
I will say that, as you might imagine, there was a lot of differences of opinion concerning a lot of the topics that we went over, including the role of the device in a specific death investigation, whether it had to do reviewing someone else's work product, if you will, a medical examiner out in X state or jurisdiction and then those of us in the panel looking it over.
The panel, specifically the medical panel, was made up of five medical examiner forensic pathologists from around the country, both in large jurisdictions and relatively small jurisdictions, most of whom had had some experience in their work in dealing with such an event. And then as Joe mentioned, there was an electrophysiologist, cardiologist, an emergency room doctor, a forensic toxicologist of note, and we heard during over this time, we believe of as broad a spectrum as we could get, from those particular human rights groups who had been very vocal in claiming that the devices should not be used until it was clear whether they were causing fatalities or not, two representatives of the manufacturers of the devices on more than one occasion, and various of their representatives as well, not only the officers of the corporation but some of the scientists who work for them.
We did hear a lot of information from researchers, the pure researchers looking at animals, and also researchers who had done some limited studies on human beings.
Now, I want to say from the outset that the stance or the position of the panel was to be neutral or not affected by any of the interests from one side or the other. We did hear folks who have some association with the manufacturer of the Taser devices, who have done a lot of studies, and we wanted to … we took them on face value without reference to whether they were connected with Taser or not and analyzed their reports on the basis of what the report said and not because of who they worked for or at least were associated with.
After doing that and then coming up with some definitions and glossary terms that applied to the operation, we came out with the interim report a couple years ago, and I will say that the final report will not be substantially different from the interim report for those of you who may have seen that.
What I would like to do is summarize the main points or conclusions of this panel that will show up in the final report.
Number one, as of this year, based upon the science and, to a certain extent, the art of this type of situation, the panel is of the view that there is no conclusive medical evidence within the research as it stands now that there is a high or, another term, significant risk of serious injury or death to humans from the direct or indirect cardiovascular or metabolic effects of short-term CED exposure in healthy, normal, nonstressed and non-intoxicated persons. A little bit more on that later.
Number two, law enforcement need not refrain from using CEDs to place uncooperative or combative subjects in custody, provided the devices are used in accordance with accepted national guidelines and appropriate use-of-force policy.
I realize there are a lot of words in there, some of which are qualitative or subjective. You know, you might put your own definition as what is accepted national guidelines or appropriate use-of-force policy, and that may be covered to a certain extent by some of the successors to the podium here today. It is covered to a certain degree with more specificity in the report itself.
Number three, all deaths following deployment of a CED should be subject to a complete medical/legal investigation, which includes a complete autopsy by a forensic pathologist in conjunction with a medically objective investigation independent of law enforcement involvement.
Number four, unlike secondary injury due to such events as falling as a result of the neuromuscular incapacitation or other types of traumatic injury, human death due directly to the primary — due directly or primarily, excuse me — to the electrical effects of CED application has not been — one of those words again — conclusively demonstrated.
We did review a lot of papers in which the assertion was made, and we certainly went over a lot of anecdotes. As a sidebar, I will say that or something that we had heard from various representatives that it was physiologically not possible for a CED, appropriately used, to cause an individual to experience ventricular fibrillation or the quivering heart because, in order to do that, you have to invoke the model of electrocution meaning that the electrons flow not above and through the skin or in the muscles but into the heart itself and then disrupt the internal functioning, the electrophysiological functioning of the heart, if you will.
There are some individuals in the field of electrophysiology who have stated initially that that was the case, but more recently, in particular, a fairly well-known electrophysiologist on the West Coast has re examined some cases and reversed his opinion at least in one instance. So that is still a matter of ongoing debate like a lot of the topics that we covered.
Number … next, I think it's five. In general, the stress of CED deployment should be considered to be a stress of a magnitude that is comparable to other components of subdual. In brief — and this was directed more toward one of the … one portion of the audience that this report is geared toward and that is the medical examiners who are faced with doing the investigations along with all sorts of other specialists and certifying the death certificate. And so that gives them the leeway or at least our suggestion is to give them the leeway that if they, looking at the totality of the circumstances and realizing the limitations of the science, of assigning in some fashion or another, either as a cause, causally, directly causal, or as a contributory factor, a role as part of the subdual process of the CED without having to say “aha, the person was electrocuted by the CED.” In other words, during the process of most subduals, there are all sorts of ways in which the person who is being or attempted to be subdued is under stress: chemical, physiological and so forth, and it was deemed at this point in the state of the science by the panel that it was not possible to sort out the CED-related stress, however little or great, from other types of stressors that may occur during the process.
And lastly, of the main points of the report, caution is urged in using multiple or prolonged activations of a CED as a means to accomplish subdual. One of the sections in the report attempts to define what “prolonged” means, and I will say, generally speaking, we have, at this point, reached a conclusion that any type of exposure, whether it would be continuous or segmented into five-second bursts, should not be greater than 15 seconds under most circumstances.
Now, there are a lot of other considerations that the panel has worked through and some of which may be outside the scope in terms of certain policy implications, but our task was to come up with the best conclusions based on the state of the science, and I believe that it's fair to say that the conclusions in this report would not be the type of conclusions that the manufacturer of the CEDs would want and, on the other hand, would not necessarily meet the criteria for certain groups who want to abolish them all together or stop their use until it's proved absolutely that they cannot, quotes, “kill anyone.”
So this, I hope, gives you an idea, and I am sorry I didn't have something to project on the screen to make those points perhaps a little more clear, but the report should be coming out, as I say, hopefully, before you go back to school this fall.
So there will be an opportunity to discuss this, I believe, at the end of the session as I step aside for the next presenter.
Scott Hammack: Joe, thank you for the introduction and for inviting me to participate in today's panel and at the conference. As Joe mentioned, I'm an attorney who served on the Maryland Attorney General's Task Force on Electronic Weapons. Fortunately, I was not tasked with coming up with all of the best practices on my own. It was a collaborative effort, and I'd like to spend my time today to share the — to talk a little bit about what the task force is, what its goals were, and also to highlight some of the key themes and the key recommendations that were in the report that the task force issued at the end of last year.
So the task force was created in October of 2008 when Maryland Attorney General Doug Gansler created the group and tasked it with the mission to come up with a set of best practices for the use of — we're calling it “ECWs,” electronic control weapons. Obviously, “CEDs” is another option. Many people use the brand name, Tasers. I think we all know what we're talking about, though. But the mission was to create a set of best practices, and what was unique about the approach that this task force had was that it attempted to identify all of the various stakeholders out there and bring them to the table to try to find common ground and to create a set of best practices that addressed everyone's needs and concerns.
So, while there's a number of best practice guides out there, certainly PERF and the IACP and the ACLU has some and Amnesty International, many of those have kind of a single perspective at the end of the day where they're a single interest that they're focused on, and this task force was designed to break down those interests and to really create a common ground and find policies that would work for everyone, which, of course, means that nobody is fully happy with all of them, but that's usually a good sign that good compromises were made.
The membership of the task force reflected this goal and had a set of broad and diverse policy or broad and diverse groups of stakeholders. Among the 15 members, there are a number of law enforcement officials from both large and small jurisdictions and also from federal agency. There were members of the medical community, representatives of civil rights organizations, including the ACLU and NAACP, and also representatives of the legal community. There was a prosecutor, attorneys who have worked both on the plaintiff and the defense side of excessive force cases, and members of legal academia.
Over the 14 months that the task force met regularly, it conducted … spent an extensive amount of time in the fact-gathering phase. In fact, I think about the first 10 months of the task force, the task force's efforts were focused on getting up to speed as to what was already out there, and we reviewed scores of reports and studies and articles.
We had presentations sent to the task force by law enforcement agencies, by Taser International, by the Maryland Medical Examiner's Office. Public hearings were held where a number of people were invited and we had law enforcement officials, civil rights advocates, mental health advocates, Taser International and also a variety of elected officials show up to share their views on some of the risks and concerns that they see with CED usage.
And, finally, we also conducted two surveys of Maryland jurisdictions. One was a survey of 32 Maryland jurisdictions for what their policies were for use of force and training and reporting and so forth, for ECW use, and another policy that was focused on the use-of-force data and demographics thereof.
So, after all the data was collected and the reports had been reviewed and we ended the information gathering phase, the task force turned to trying to come up with a set of recommendations, and as you might expect, this process was not the most simple in the world and was often very contentious, but at the end of the day, the task force was able to agree to 60 best practice recommendations, and what was, I think, most important about this was every one of those recommendations was unanimously endorsed by every member of the task force.
The areas that the recommendations were in, were in planning and implementation, training, use of force, medical care, reporting and investigation, monitoring and data collection, and the task force also identified some suggested areas of research.
Now, I'll talk about a few of the recommendations later on. Obviously, there are 60 recommendations, and I think I have 20 minutes, so that means we're not going to talk about all of them. I do have copies of the recommendations here. I didn't bring the full report, but if anyone is interested in seeing all of the recommendations, I'll have some copies up here at the end of the program. And at any point, the report is available online. The website is at the bottom of each of the slides.
But before we talk about the recommendations, I want to talk about a few themes that emerged that really, I think, serve to form the foundation of the recommendations and the report.
The first of these themes is that ECWs are a valuable law enforcement tool when used properly. The fact that this emerged as a key theme was somewhat surprising given where the task force started off and where I think a number of the members of the task force came to it thinking that Tasers were not good devices, that they were things they wanted to see moratoriums or completely not used, but once everyone had a chance to understand what the devices do, how effective they can be when used properly, there was essentially no discussion about endorsing a moratorium or suggesting that ECWs not be used but rather trying to figure out how to use them in a way that allows officers to maximize their benefits while minimizing their risks.
The need to minimize the risks, I think, is really the key thing and what emerges the second theme, which is that the risks associated with ECW use are not really particularly well understood, and there's a couple of aspects to this.
First, there's obviously an ongoing debate over the exact extent of the risk of ECW use, and that's above my pay grade to figure out. Hopefully, some of the folks with the doctorates over here can be helping us with that. But what there isn't much debate about is that ECWs can and have caused death, perhaps not conclusively with the electrical currents, but certainly secondary risks and resulting impacts from what happens after someone has been shocked.
So, given that, there seems to be a disconnect among a number of officers and agencies where, instead of viewing the ECWs as a less-lethal device, they're viewed as almost completely nonlethal, and we saw a number of examples of this disconnect.
One of the ones that sticks out in my head is there was a Take Your Child to Work Day at a prison down in Florida, and the correctional officials there thought that it would be a good idea as one of the activities for the day to discharge their ECWs on their own children.
Another one here, actually from here in Virginia, there was a woman, a mentally disabled woman, who apparently liked to listen to loud music while she Hula-hooped, and when an officer came to ask her to turn her radio down she refused, and she was shocked with an ECW.
There are numerous stories like this that all tell us that there is a disconnect, that people do not fully understand the risks, even if they are minimal risks, of what these devices can do.
A second reason why some of the risks are misunderstood is that the physiological effects simply have not fully been studied. You know, as Dr. Hunsaker noted, it is very difficult to replicate actual field conditions in a controlled manner for scientific research. You can't capture all the stressors. You know, if someone's under the influence of drugs, if somebody has some sort of pre existing condition, these are very difficult things to capture in scientific research.
So, for that reason, effects are not known, particularly with respect to certain vulnerable populations, such as those in medical or mental health crises, people that are under the influence of drugs and alcohol, pregnant women, children, the elderly, and so on.
Shifting to the other end of the spectrum, there is — in addition to the people who view ECWs as nonlethal devices, I think there is a significant population of people who view the ECWs as — I don't know — perhaps a more lethal device. That, we've seen with Amnesty and with other groups, an apparent non-appreciation for what the ECWs can do and fears that they are going to cause just huge amounts of death or injury. And part of that we have attributed to the fact that a lot of the early research out there, the researchers who did the work have either funding or some sort of affiliation with ECW manufacturers. It's not particularly surprising. This happens all the time in the pharmaceutical industry. The people who have the interest in funding studies are obviously going to be the manufacturers.
So, whether or not there is actually any linkage there, what happens is it creates an appearance of a conflict of interest and it undermines the credibility of some of the research and allows people to push aside perhaps valid scientific studies by simply saying, “Oh, look at the funding of it. This can't be potentially … this can't be unbiased.”
The third key theme that emerged was that many agencies' policies do not provide sufficient guidance, tell our officers to manage the risks associated with ECW usage. This became apparent when we did our survey of agencies in Maryland and started to take a look at what some of their policies were for when ECW use was allowed.
These were some examples: one, to incapacitate a resistive person; to be used on simply noncompliant individuals; to control the situation; and when circumstances are tense, uncertain and rapidly evolving. These types of vague standards do not provide the type of guidance the officers need to understand when the use of the ECW is justified by the risks it presents. As a result, we see an over reliance on ECWs and an increase of risk of injury or death. This is particularly true when ECWs are turned to in lieu of verbal de-escalation techniques or other nonforce options.
Also, increases the agencies and officers' exposure to liability. Just in the last month or so, I've seen a couple multi-million-dollar settlements. There was a city, Fort Worth, I guess, made the news about two weeks ago offering $2 million to settle a death that occurred after an ECW deployment, and a city in Missouri just recently settled a case for 2.4 million, and more importantly, they had to agree to a moratorium on ECW use following that settlement.
And that brings me to my next point that these type of policies and these types of uses really increases the public resistance to ECWs. It makes it more difficult for agencies to implement new ECW programs or to expand existing ones.
Turn now to some of the recommendations and picking out a few I think were really the key recommendations of the task force, and I think the recommendation that really serves as the foundation for all other recommendations is what the use-of-force policy is going to be, and not surprisingly, this was one of the more contentious ones that worked out or to work out, and ultimately, we reached this formulation, which everyone signed off on as being the appropriate one. And that is that ECW should be allowed only when an individual's actions pose an imminent threat of physical harm to themselves or others.
Now, how this differs from a number of ECW policies that we saw is it does not allow for ECW use merely to gain the compliance of an individual who is otherwise not threatening. So, in some of the cases that I described earlier, that's the type of use we saw, and most of the cases that really sparked the public outrage about ECW usage are in these types of situations where there is no apparent risk of injury, and it's just an officer who's frustrated with a noncompliant individual and is using this new device to gain that compliance.
The task force also recommended that the ECW use of force be included in a broad use-of-force policy that includes all other uses of force as well as de-escalation techniques and other nonforce options, so that officers will have basically a menu of options at their disposal and can know which options are appropriate for whatever situation that's confronting them and make it easier for them to transition to other options if their option of first choice is not working.
There were a couple of circumstances where the task force found that the standard use-of-force policy wasn't quite sufficient and where heightened awareness and justification is needed.
One of these are in situational examples where there's a heightened risk of death based on the surrounding circumstances, and in these instances, because the risk of death is elevated, the task force recommended that ECWs be allowed only when deadly force otherwise would be allowed. And these types of situations, persons in elevated positions who are at risk of falling, persons with heart disease or pacemakers — I'm sorry, I am reading off my wrong notes here — persons operating vehicles and machinery, persons fleeing on foot — and the fleeing on foot, there are two concerns. One is the risk of falling, if someone is running quickly can fall and hit their head. There's also a concern if the person is of small stature and they have a dart in their chest that potentially could pierce their chest and make contact with the lungs or the heart; persons restrained in handcuffs, persons in danger of drowning, and in or around combustible vapors or liquid such as OC spray.
The other area where additional restraint is required is again certain heightened risk populations. Here, these populations are listed there, again, persons of small stature, persons with heart disease, elderly persons, pregnant women, individuals in mental or medical crisis, and persons under the influence of drugs or alcohol. And in these situations, the law enforcement officials need to be equipped with the latest knowledge about the uncertainties and the potential risks to these individuals, and if they are confronted with a situation where they need to choose whether to use an ECW, they need to be able to articulate justifications to why applying that increased risk was justified in that circumstance.
The issue of the use of ECWs against individuals in mental health crisis is certainly one of much debate right now. There is two very different approaches. One side suggests that the proper thing to do here is to use the ECW quickly, so that you can get the individual the medical care they need as quickly as possible. The other, the competing side is that fear that using the ECW will cause or exacerbate the symptoms that the person is experiencing and could potentially increase the risk of death.
So, given the uncertainty about this, the task force felt that the simple display of the symptoms of a mental health crisis should not in and of itself warrant use of an ECW. Rather, only when there are no other alternative restraint methods available and if the individual is posing a threat to himself or herself or someone else should an ECW be used. Otherwise, the officer should be using the training in the de-escalation techniques and other mental health crisis intervention techniques that they've learned to try to find alternative ways of containing the situation.
Moving on to training, the general theme of the task force's recommendations on training are that officers need to be equipped, not just with how to use the ECW but also when to use the ECW, and when we did our survey of Maryland agencies, we found that the trainings were often very, very heavy on the mechanics of actually using the device but very light on when to use the device and the use-of-force-type considerations.
So, with that in mind, the recommendations say that the training should make clear where the ECW falls within the agency's use-of-force policy, should provide training on de-escalation techniques and other mental health crisis intervention techniques, should make sure the officers are educated about which populations or situations could create a higher risk of death or injury, and should include a scenario-based training that will help the officers make the types of decisions that they're going to be confronted with in the field.
In addition, the training program should have a certification component with an annual recertification, and the agencies should also not rely exclusively on materials provided by ECW manufacturers. Our review of those materials found that, obviously, they didn't have any of the agency-specific-type information about the agency's policies or use of force, where it fell in the spectrum, and also, it appeared to understate the risks or at least the uncertainties associated with using the devices.
Another area that the task force made some recommendations was on data collection. They identified about 20 or so specific categories of data that all agencies should capture and track following any sort of ECW use of force.
We also recommended that a state entity be tasked with collecting that data and publishing it on an annual basis, that when an agency is there collecting this data, use that as an opportunity to examine the data, to see if there are any types of changes that need to be made to their training or use-of-force policies.
A couple of medical examiner-specific recommendations in dealing with this concept of excited delirium: First, the task force recommends that medical examiners identify or for any ECW-proximate death, specifically indicate whether the use of the ECW may have or did contribute to the death. And also if the finding is that excited delirium was the cause of death, the task force recommends that the medical examiner actually identify and explain the actual cluster of symptoms that led to that finding, to allow people to go back later when additional information is known and be able to have better information about the death itself and to see if there are other linkages once further study has been done.
The last area I'm going to mention is the proposed research that the task force thought needed to be done. As I mentioned earlier, there isn't real conclusive research on the physiological effects of ECWs, especially on the non-healthy individuals in controlled situations.
So, with that in mind, we recommend additional research for repeated or prolonged applications of ECWs on persons in heightened risk populations, on persons in medical or mental health crisis, and also when they're discharged in certain areas, specifically the neck, head or the chest area.
Then, finally — and I think this should be a good lead-in to the next presentation — is research on whether ECW deployments affect the numbers, the amounts of use of force and also the number of incidents of deadly force.
Again, those are just some of the highlights of the 60 recommendations. Anyone who wants to take a look at them, I'll have them up here at the end or feel free to go onto the website and download the full 100 page report or so.
Thank you, all.
Eugene Paoline: Hello. I'm Gene Paoline from the University of Central Florida, and today, my colleague, Bill Terrill, and I will be presenting the results from our NIJ-funded use-of-force project that examines the impact of conducted energy devices and injuries to suspects and officers.
CEDs have been around in policing since the late 1970s, but over the last decade, their popularity has increased dramatically. In fact, in a national survey, as part of our project, two-thirds of the responding 662 agencies noted that they used some form of a Taser; 13 percent said another form of a CED. So we have roughly 80 percent of the agencies that responded saying that they were using CEDs.
As one of the contemporary and primary weapons that officers are equipped with, CEDs, like chemical sprays before them, have been the center of controversy between the police and the public. The controversy that surrounds CEDs, as we've mentioned on this panel already, really focus on two things: one, their intended purpose of incapacitating individuals, and two, the injuries that result from their use.
From a police perspective, CEDs are perceived as safer injury-wise for officers that may not have to engage closely or on the ground with citizens that are resisting them, while also safer for citizens that don't have to be engaged with the police in longer physical struggles or by other weapons used against them like batons. Citizens, on the other hand, question this latter point regarding the safety of the CEDs. Advocacy groups often cite the serious injuries and the deaths as a result of forced encounters that have included the use of a CED.
Police officials are aware of this controversy. Some officials do not have CEDs in their departments because they don't want to deal with this controversy; in other instances, CEDs have been taken away — “it's just not worth it; we're not going to use it any longer.”
Other departments and one of our agencies that we have in our research project only reserves CED use for tactical groups, so general patrol officers are not using them, and other agencies have responded by putting the CED at the high end of the force continuum right before deadly force, so police officials are aware of this.
Unfortunately, given that CED research is really in its infancy, we lack consistent findings regarding the impact of CEDs on injuries, and we could use this information, especially in the field, to make some more well-informed decisions, and even with the research that has been done on CEDs, there are some pretty well-documented shortcomings, some of which have already been mentioned.
First and foremost, a noted criticism is that a good deal of the research is being conducted by advocacy groups, your pro-CED folks saying, “Yes, it's safe,” the pro-citizen groups saying, “No, don't use it.” Both groups have been criticized for their research designs, the data that comes from these studies, as well as their inability to act as independent researchers, unbiased in that sense.
A second shortcoming is this focus focused heavily on serious injuries, things like broken bones and deaths. These instances are rare. For us as social scientists, it's analogous to what we do in studying the everyday use of force versus what other people think or wish that we would, excessive force. The excessive force part is certainly the sexy research topic everybody gets geared up, but it's not the modal event.
Other studies have been descriptive in nature, and there's nothing wrong with that. It's good for building hypotheses, but it does limit our ability to draw causal inferences, research in this manner. Small sample sizes are another concern. The fact that we are dealing with generalizability issues — sometimes the research is based on select members of the department having the CEDs. Sometimes it's based on select interactions with certain types of groups. This limits our generalizability from a social science perspective.
Finally and what we regard as probably extremely important is the prior research has failed to adequately control for theoretically relevant factors, and what you're trying to do here is isolate that independent effect of the CED and the injury. This would include looking at CED-only cases versus instances where no CED was used or how is the CED with hands on different from CED and other weapons.
Studies should also look at citizen resistance as a control variable. This can impact injuries to both citizens as well as officers.
The study presented here today attempts to deal with these past research criticisms by empirically examining CEDs and their impact on both suspect injuries as well as officer injuries. In doing so, we utilized data collected as part of our multi agency, multi-method use-of-force study. As Bill will mention shortly, we analyzed over 14,000 use-of-force incidents, 2,600 of which involved a CED.
So, in our analysis, we're interested in independently assessing whether or not there was an injury. We're not wrapped up in the serious/nonserious dichotomy at this point. For us, the real research issue is injury or not.
Before getting to those results, I want to quickly give an overview of our project, so you can assess where this data is coming from.
This study was collected, was a collaborative effort between Michigan State University and the University of Central Florida, and we had two specific things in mind when we were conducting this project.
One, we wanted to assess the types of use-of-force policies currently in existence, and two, what we'll deal with here is examine the relationship between divergent use-of-force policies and a variety of outcomes. These outcomes include, obviously, the use of force. The project was concerned with types and severities of force.
As a result of these force encounters, suspect and officer injuries, these last three outcomes are not relevant for today's presentation, but we were also concerned with citizen complaints, especially those for use of force as well as discourtesy, lawsuits brought against individual agencies and/or departments, and then we were also interested in officer's perceptions: perceptions of use of force in general, perceptions of their use-of-force policy and perceptions of their overall work climate.
Our data to assess these outcomes, you will see, first, we had a multi-method, multi stage project, and very briefly, we started by administering a mail survey to get a glimpse of and get a read on what types of use-of-force policies are out there. From there, we chose eight departments with varying use-of-force policies, and this variation was either built on their design or the tactical placement, and within those eight agencies, we went in for deeper study.
We collected multiple sources of data. “Use of force” in bold there, that's where this analysis comes from. We'll be able to assess what types of force were used and also which types of injuries occurred to suspects and officers, which was built into the use-of-force reporting system, obviously to deal with those other outcomes, citizen complaint reports, civil litigation records, our patrol officers survey to deal with perceptions and a variety of other official agency data.
Here are our study sites. We have eight medium-to-large police departments: Columbus, Ohio; Charlotte-Mecklenburg; Portland, Oregon; Albuquerque; Colorado Springs; St. Pete; Fort Wayne; and Knoxville. Here, we can assess how these places varied in terms of citizen characteristics and organizational characteristics, and in a simplest sense, they are similar but different. OK.
A couple of quick points here. You'll see “Fort Wayne” is shaded on that column, next to the last column. This analysis today is going to rely on the seven agencies and not the eight, and the reason why we left Fort Wayne out: They were one of the departments that used CEDs only for tactical members. And over the two-year study period, they only deployed their CEDs nine times, so we didn't want to have them in with the rest of the analysis.
The second issue that Bill will talk about, when we get to officer injuries, not suspect but officers, we don't use Albuquerque because we can't. They didn't document officer injuries. They only documented suspect injuries.
William Terrill: OK. I'm the final act today, and I'm very happy to be here because I turned in the final report to Brett Chapman on April 30, and for the last 45 days, I've been stuck in my office running these models you are about to see, so it's great to get a captive audience and you're stuck here until five o'clock for me to go over all my statistical models.
I promise at the end of the run here, I have a nice color-coded summary table, so all of those numbers can be brought home in a pictorial type of fashion. But, if you bear with me, hopefully, this will make sense by the end of it, and it's really not complicated. It's just we wanted to look at it in — you know, we wanted to take empiricism as serious as we could, given the data that we had, and so we wanted to look at it from a variety of angles to make sure that the effects we were seeing were true effects given the statistical controls that we were able to incorporate.
I'm going to start out nice and easy with some frequency breakdowns. As you see here in the red, I highlight the red to kind of fix your eyes on the PowerPoint. I know it can be difficult to find some of the items sometimes in these tables. So real simple here, suspect injuries are running about 3-to-1; 30 percent of our cases for suspects resulted in an injury versus 10 percent or so for officers. And you'll see the N at the bottom there being different. That's, again, Gene had just mentioned, we have seven cities involved in the suspect analysis, and just six in the officer analysis because we lost Albuquerque.
Attendee: Is this all use-of-force or just CED?
Terrill: All use-of-force right now. So we're starting broad, and we're going to narrow it down.
Gene mentioned a little bit about injuries. We debated the issue, and I think at some point, we'll also delve into seriousness of injuries. We wanted to start out, though, with a simple “was there an injury or not documented.”
Serious injuries is a tricky proposition based on how officers document it, and you see up here relatively infrequent, as I have the broken bones, 61 cases across there with broken bones, but more than just saying broken bones is a serious injury, which can be somewhat subjective, if you compare it to lacerations, which often in previous studies got characterized as not serious. But if someone has a gash that required 62 stitches and someone else has a broken pinkie, there might be a debate on what's serious.
So, for us to feel comfortable classifying seriousness of injuries, we're going to have to go back into the narrative accounts and see what we can glean for that.
You'll also notice “other” here, the “other” category, fairly large. That's for two reasons. One, sometimes an officer said there was an injury but didn't give us enough detail or indicate what that injury was, so we couldn't classify it. The other reason was that it didn't fit one of the categories that they were documenting, and so, if the suspect complained of a back spasm or nauseousness or something, a headache from landing on the ground, it didn't fit into one of the categories that they were documenting.
OK. Now we're going to go to — again, we're just starting broad yet, looking at overall force. The reason I put this up here, we split hands-on force and some type of weapon. I know, Joe, you don't like that term, but I'm going to use it because that's the headers up here. But we split it into weapon and weaponless tactics. On this slide, I'm just trying to illustrate that the bulk of the hands-on force is soft, empty-hand control, not a surprise to anyone that covers policing or that studies policing, and you see it at 11 percent. That's the hitting, kicking, punching types of behavior.
When it comes to weapons force, you pretty much have chemical spray, OC, or you have CED. After that, it falls off the chart. The baton is an artifact at this point in terms of our eight cities. Less than 1 percent of the cases involved the baton. That was the old-fashioned PR 24 type or the expandable ASP, less than 1 percent. So, when I'm talking about weapons, I'm generally talking about the CED or chemical spray, although there are still a couple hundred cases of the other types of weapons.
All right. Let's look a little bit now at … hone in on the Taser. This slide kinds of kicks us off, and you see up here with the three columns. What we started with is we have said we wanted to parse out our case and say, “Give us the cases that involve no CED use at all.” That was the bulk of our cases. Then we wanted to say, “Give us the opposite end of that,” cases where only a CED was used, and then you have the middle category.
In the red there, you see right off the get-go, with the bivariate analysis, if a CED is used, we had a significant likelihood, a greater likelihood of injury statistically as indicated down there, 41 to 47 percent versus 28, but we are just at the bivariate level now without any statistical controls, which we'll take a look at.
Before doing that, though, we wanted to go into the blue there. We wanted to go into these 1,826 cases, and say, “How are they breaking out in terms of hands and weapon with CED,” so we'll go into that briefly. And we see when we do this, the bulk of CED uses is with hands only, 1,506 cases, only 82 cases where it was used with another weapon only, although there are 238 where they are using virtually the kitchen sink approach.
If you look at the percentage change here, not much. Even the chi-square is not significant, although you do have a little bit more when you have all types of force being used.
Broke that down even further into if they are using it with hands, what type of hands, soft and hard. Here, you see — I don't think a great surprise — more injuries if it's hard, if they're punching, kicking, slapping, that sort of behavior, more injuries. And if we go into those 82 cases where it was used with another weapon, we don't get a lot out of this slide because you take a 3 by 2 table, 82 cases, your cells get small real fast, and so you've got to be careful with interpreting.
You see the chi-square is not significant, not much difference percentagewise, but, again, the N's are real, real small, only three cases there.
So that moves us into the multivariate. Just looking at some of the descriptions of our net controls, the big one for us was suspect resistance, depicted here in ordinal fashion where we indicated the highest level of resistance, the control for that, the assumption being if an officer is aggressively resisting, the more they resist, the likelihood of their injury is going to go up irrespective of the type of force that's used on them.
We also ran all the models; you're going to see I ran 01s on all of those; where I put them in as dichotomies, same findings. All the physical force comes out more significant. The only one that comes out for injuries less significant is failure to comply. They're not physically resisting in any way. They're just verbally or passively resisting.
Here, you see the various ways we looked at CED and other weapons to make comparisons for injury rates that I'm going to show in the multivariate models. In red, you see what I'm calling or what we're calling the base model, and that's just starting out saying, “Let's compare cases where a CED was used alone, with other force, and compare it to when there's no CED used at all.”
After we do that, then we say, “Let's split that variable out into three categories and see what we have comparing to none.” And then we're going to take these two in red up here and compare them in a series of models to various forms of hands and other weapons.
I'm not going to spend much time on the descriptive stats. Again, 31 percent of the cases, an injury occurred. The mean suspect resistance was defensive physical, which means the officer was physically resisting in a defensive mode. They weren't trying to attack the officer. They were trying to get away from the officer's control.
Highlighted, the base model variables again, 5 percent of the cases. Gene had mentioned there were 2,600 cases of Taser or CED use. In our case, it is Taser because all seven cities use the Taser; 5 percent where they used CED only, 1,826 cases where they used it with other force, and over 11,000 cases in which there was no CED used of the total 1,400, 218.
All right. Model one. I put the controls up top and I put the force emphasis down here, and what we see is confirmed in our multivariate model than what we had in the bivariate model. What this is telling us is that when a CED was used either singularly or in combination with other force, the suspect injury probability goes up, significantly so.
Right there, it tells us 1.53 to 1.74, 53 to 74 percent greater likelihood of injury when a CED is used versus none.
Now let's break this category out a little bit more and see what we get. We broke that other force out, and we compare it to none yet as our reference category. And we see the effects remain from 1.5 to 2.19 on the high end, but the CED effects remain increased in suspect injuries.
Now we collapse that variable back. We know they're related for parsimony purposes, and we're comparing it to any time a weapon was used that wasn't a CED. Effects remain 84 percent to over 100 percent greater likelihood oddswise. It's odds ratio, what we're explaining right here.
We break weapon out and we compared the cases. Remember, chemical spray was the bulk of the weapons cases when a CED wasn't used, so we broke cases out that didn't involve it. Effects remain, and then we did a direct comparison to cases with a CED or with a chemical spray, and the odds ratios go up even more, so 2.3 times more likely, 2.7 times more likely CED versus chemical spray.
We flip-flop here now and we take the weapons into a control, and we compare against any kind of hands-on force. Effects remain 54 to 75 percent greater likelihood.
Now we compare to cases where the bulk of the force is soft hands, so no hard hands at all. The effects remain.
Our final comparison is when we look at hard hands. You see I changed the color from red to blue. Blue is indicating that there is no statistical difference when you use a CED versus hard-hand tactics. So, if the choice is CED versus hard hands, you don't have any statistical difference in terms of where the suspect's injured.
Flip-flop to officer side of the equation now. We'll start out with a simple bivariate. A little bit different picture emerges when we look at officer injuries, and so you see here, 5 percent of the cases result in officer injuries when they only used the CED, but that goes up nearly twice percentagewise when they don't use a CED. So you're having an opposite effect than you had in the suspects with the CED only, but you still have a substantial increase right here when CED is used with other force for officer injuries.
Quickly go into those 1,567, just like we did before. We see that the bulk of the cases, again, when you use a CED, is with hands in this case. Percentagewise, officer injuries are more likely if they use the CED with hands on as opposed to another weapon like chemical spray.
So we broke this out. If anyone's interested in, boy, what combination can I use a CED with versus virtually every other combination, the effects are here one way or another when you look at them.
We broke that into soft and hard. Similar to suspects, if it's hard-hand, greater likelihood of injury. I don't think that's particularly shocking.
Break it into the weapons, and just like before when the 82 cases, now we're down to 80 because we only have six cities, and the N's are real small. I put it up here if anyone is interested, but I wouldn't take too much out of here because of the N being so small in that 3 by 2 table.
I'm not going to spend much time on the descript that's very similar to the officer or very similar to suspect and resistance in these control variables, although we're down to 10 percent because officers just aren't injured as much.
The same thing as before. And now we run through the same sequence of models for officer injuries. We have some similarity to suspects but also a big difference here, and the big difference is this right here. When a CED is used singularly, officer injuries, you see the negative coefficient there? It says officer injuries are significantly less, and we have measures of effectiveness which we haven't got into yet, but that's our next project here, is to look at Taser effectiveness across these eight cities.
The basic assumption we're making here is when an officer only has to use a CED and it works and it incapacitates the suspect, they slap the cuffs on, the officer doesn't become injured. Cases here, where they have to use other force, they use a CED. It didn't work for some reason. One of the probes didn't go in. It missed. They have to go to other types of force, and now the probability of the injury goes up and significantly so, as you see here. So, by itself, less injuries; used with other force, more injuries, compared to none.
Break it out into that other category, as we did with suspects, two of the three significant, one when it's used with weapon not significant. Again, we are comparing to none.
Now let's compare it to hands. The effects remain the same as they did in the base model, less injuries with CED only, more when it's used in combination. Same thing if we compare it to a weapon other than the CED, so impact munitions, for instance, baton usage. Direct comparison with chemical spray, same effect, less and more.
Flip-flop to hands only, same effect, less and more.
Same effect if we're not dealing hard hands, so we're, for the most part, soft hands right there.
We get to the hard-hand comparison. Remember last time with suspects, it was blue; there was no statistical difference. In this particular case, both cases compared to hard-hand tactics result in less injuries. So, in this particular case, if the choice is between hard-hand tactics, the CEDs resulting in less officer injuries.
Here is the little fancy table I was telling you I promised that all those stats would bring home. You have the CED-only with other force, and this is your comparison group, all the none, and we broke it out across all of these categories — I even broke it out in three more models, but I figured I'd really be hitting you over the head — into soft only, hard only, chemical only, not used in combination, and the effects remain the same.
So you have three of the four columns here indicating an increase, officer injury by itself or CED by itself, less officer injuries, two cases where there was a nonsignificant finding.
When used with other force, five of the six show an increase in injuries, one nonsignificant finding.
And finally, what does that all mean from a text perspective? As you see, I break it down into suspect and officers. The message should be fairly clear, even if you are tired at five o'clock at this stage. Without question, using the data from these seven cities, when an officer uses a CED either by itself or in combination with some other type of force, there is an increased probability of injury. When it comes to officer injuries, when CED is used singularly or by itself, there is a decrease in injury, but if they have to use other forms of force with the CED, it increases the injury, just as it did with the suspects.
The two exceptions I put up there are with hard hands, which is either nonsignificant or the CED results in less injuries.
Opinions or points of view expressed in these videos represent those of the speakers and do not necessarily represent the official position or policies of the U.S. Department of Justice. Any commercial products and manufacturers discussed in these videos are presented for informational purposes only and do not constitute product approval or endorsement by the U.S. Department of Justice.