Injuries to Deployed U.S. Army Soldiers Involved in HMMWV Crashes,

MILITARY MEDICINE, 177, 8:963, 2012 Injuries to Deployed U.S. Army Soldiers Involved in HMMWV Crashes, 2002 2006 CPT Samuel M. Peik, MC USA*; Keshia M. Pollack, PhD, MPH ; Michelle Canham-Chervak, PhD, MPH ; Keith G. Hauret, MSPH, MPT ; Susan P. Baker, MPH ABSTRACT Highly mobile multipurpose wheeled vehicle (HMMWV or Humvee) crashes present an important issue for the U.S. military. The aim of this study was to provide a descriptive analysis of occupants of military motor vehicle (MMV) crashes involving HMMWVs that occurred among deployed U.S. Army Soldiers. Crash-related data were collected from the U.S. Army Combat Readiness/Safety Center on MMV crashes among active duty Army personnel between 1999 and 2006. Records for 964 occupants with injuries from HMMWV crashes were analyzed, which represented 52% of the total occupants of MMV crashes. A significant association was observed between injury and engagement in combat, odds ratio 1.49 (1.03, 2.16). The risk of injury was greatest for gunners, odds ratio 2.37 (1.43, 3.92), and injury cost related to the crash was significantly related to prior deployment status (p < 0.001) and role of Soldier in the vehicle (Operator p = 0.005, Gunner p = 0.003). There was also a decrease over time in the number of crashes resulting in injury (p < 0.001). These data support the development of interventions that address the specific risks detailed, including the use of combat simulation training, increased protection for vulnerable positions, and enforcement of safety regulations. INTRODUCTION Injuries in the U.S. military represent an enormous burden to the force. They are the largest health issue the military faces, both in peacetime and during combat operations. 1 4 Historically, injuries have been the leading cause of death, disability, and medical encounters across all U.S. military services. 5,6 Estimates also show that approximately 25 million days of limited duty are as a result of injuries. 7 Motor vehicle crashes (both privately-owned vehicles [POVs] and military motor vehicles [MMVs]) account for nearly onethird of U.S. military fatalities annually, and also are among the top five causes of hospitalization. 8 Injuries have a tremendous impact on the combat effectiveness and health of military personnel and can adversely affect force readiness. 1,3,5 As noted by Krull et al, 9 key differences exist between operating POVs and MMVs, and although a number of studies have been performed on POVs, their findings may not be generalizable to combat vehicles. New tactical situations present uncertain, difficult, stressful circumstances, and operators wear heavy gear, which can limit physical mobility, hearing, vision, and reactions. The military population also often engages in risky driving behaviors because of the predominance of young males who have less experience driving vehicles. 9 For these reasons, experiences with MMVs differ from those of other vehicles and may present unique challenges. *Division of Preventive Medicine, Walter Reed Army Institute of Research, 503 Robert Grant Avenue, Silver Spring, MD 20910. Department of Health Policy and Management and Johns Hopkins Center for Injury Research and Policy, Johns Hopkins Bloomberg School of Public Health, 624 North Broadway, Baltimore, MD 21205. Injury Prevention Program, U.S. Army Public Health Command, E5158 Blackhawk Road, Aberdeen Proving Ground, MD 21010. The views expressed in this publication are those of the authors and should not be construed to represent the positions of the Centers for Disease Control and Prevention, the Department of the Army, or the Department of Defense. Few studies exist on injuries resulting from MMVs. 10 A recently completed study by Rossen et al, using the same data set as this study, focused on a broader descriptive analysis and detailed costs and trends over time. 11 The study reported here focused on describing MMV crashes, specifically those involving highly mobile multipurpose wheeled vehicles (HMMWVs), also known as Humvees, among deployed U.S. Army Soldiers from 2002 to 2006. The purpose of this study was to identify risk factors for injury from crashes involving HMMWVs in a deployed environment. METHODS Data Collection and Variables Data on U.S. Army motor vehicle (AMV) crashes from 1999 to 2006 were collected from the Army Safety Management Information System (ASMIS), which is maintained by the U.S. Army Combat Readiness/Safety Center. Information was collected in accordance with Department of the Army Pamphlet (DA PAM) 385-40: Army Accident Investigations and Reporting, which includes DA Form 285-AB-R: U.S. Army Abbreviated Ground Accident Report (AGAR) and DA Form 285: U.S. Army Accident Report. Unit safety officers were responsible for collecting information and completing these forms, which were subsequently used to populate ASMIS data fields. These records provided extensive data including but not limited to date, place, and circumstances of the incident; description of vehicles involved; description of injuries sustained; details of costs related to sustained injuries; and demographic data of Soldier occupants. Vehicle categories were defined as outlined in DA PAM 385-40. AMVs are defined as Army-owned or leased wheeled vehicles typically used to transport cargo or personnel. Army combat vehicles (ACVs) are defined as tanks, self-propelled MILITARY MEDICINE, Vol. 177, August 2012 963

weapons, tracked armored personnel carriers, amphibious vehicles ashore, and similar equipment. Other Army vehicles are defined as those that involve operation of an Army vehicle other than those already listed. POVs are defined as motor vehicles that are privately owned and operated by an individual, including on official business. Occupant role in the vehicle was determined by examining the qualitative account of the incident and assigning each occupant to one of five categories: supervisor, passenger, operator, gunner, or bystander struck by a vehicle. Supervisor was defined as an individual having any kind of leadership role in the vehicle team; passenger was defined as an individual present in the vehicle but with no leadership or operating role; operator was defined as being behind the wheel of the vehicle; gunner was defined as being either in a gun position on top of the vehicle or designated as a gunner inside the vehicle; and bystander was defined as not inside the vehicle. Injuries to bystanders were rare, with mixed descriptions and results. These bystander cases were included in the demographic analysis but not in the multivariate models. Costs were determined in accordance with Department of Defense Directive 6055.7: Accident Investigation, Reporting, and Record Keeping, as well as DA PAM 385-40: Army Accident Investigations and Reporting. In general these documents outline a set amount of costs based on the severity of injury but are also based on length of injury-related recovery. Also, as detailed in DA PAM 385-40, combat is defined as an incident in a theater of hostile fire or enemy action, but not as the result of such fire/action. This includes direct preparation for combat, actual combat, or redeployment from a combat theater immediately following combat. For example, primary combat injuries such as directly being hit by enemy fire are not included, but injuries that result from crashes are, even if the crash occurs in a combat setting. Study Sample Of the total documented cases (n = 14,109), which included both deployed and nondeployed crashes and both POV and MMV, 2,640 were excluded because of absence of personal identifiers (i.e., Social Security Number or name) to link to demographic and deployment-related data obtained from the Defense Medical Surveillance System. This resulted in a total of 11,469 records, which were assigned unique identification numbers and included in the data set for analysis. These records were further narrowed to focus on the subpopulation of active duty U.S. Army, Army Reserve, and National Guard Soldiers involved in military vehicle crashes in the countries of Iraq, Kuwait, and Afghanistan, occurring during Operation Iraqi Freedom and Operation Enduring Freedom. This narrowing resulted in identification of 2,074 records. For further analyses, the population was restricted to include only the occupants of HMMWV crashes. These vehicles were chosen because they were the most common tactical vehicle type involved in crashes reported during this time frame and provided the best subset for analysis. A total of 964 occupant records were identified. These records were used for descriptive analysis, however, bystanders (n = 36) and occupants with missing variables (n = 78) were not used in the multivariate analyses, resulting in 850 records for these models. Each record represented a Soldier, and injury was defined as anything requiring medical attention (including first aid, restricted duty, lost workday, disabled, or death). Crashes meeting these criteria were present only beginning in 2002, coinciding with the beginning of military operations in Iraq and Afghanistan, hence our examination of the time period of 2002 2006 for this analysis. Analyses Descriptive data were analyzed for vehicle type and Soldier demographics. Descriptive data on severity of injury, number of reported operator mistakes (errors) related to the crash, and types of mistakes were examined and tested for significance with c 2 tests. Data on mistakes were available for 488 occupants; mistakes were categorized by examining the qualitative text account and assignment to one of thirteen error categories. Some records listed multiple mistakes; however, only the mistake listed in the first position on the form was included since it was likely to be the primary and most important reason for the crash. A multivariate logistic regression analysis was used to determine if combat, deployment experience (measured by number of deployments), or role within the vehicle had any effect on the outcome: odds of any injury (fatal or nonfatal) from the crash. Reference levels for experience and role in the vehicle were no prior deployment and supervisor, respectively. Analysis of variance (ANOVA) was performed to explore associations between direct costs (log-transformed) and selected crash circumstances and occupant characteristics. Rank, gender, and age were included in the multivariate models as potential confounding variables. All data entry and subsequent analyses were performed using STATA version 11. All significant results are reported with a threshold of p = 0.05 and 95% confidence intervals (95% CIs) are displayed for ORs. This study was reviewed and approved by the Johns Hopkins Bloomberg School of Public Health Institutional Review Board. RESULTS Descriptive A total of 2,074 U.S. Army Soldiers were involved in motor vehicle crashes in Iraq, Kuwait, and Afghanistan from 2002 to 2006 as reported by the Army Combat Readiness/Safety Center. Over three-fourths (76%) of these vehicles were AMVs (Table I). Table II shows the breakdown by specific vehicle type for all occupants of the vehicle, as well as how many crashes resulted in one or more injuries by vehicle type. Although twenty different types of vehicles are listed, nearly half of the occupants were in HMMWV crashes (47%, n = 964) and nearly half of all occupants had an injury (49%, n = 590). 964 MILITARY MEDICINE, Vol. 177, August 2012

TABLE I. Categories of Vehicles in Which Soldier Occupants Were Involved in Crashes, Iraq and Afghanistan, 2002 2006 (n = 2,074) Category No. of Occupants (%) Army Motor 1,577 (76.0) Army Combat 379 (18.3) Other Army 59 (2.8) POV 14 (0.7) Motorcycle 10 (0.5) Unknown 35 (1.7) Total 2,074 (100.0) The demographic characteristics of the Army occupants (n = 964) in crashes involving HMMWVs are detailed in Table III. Of note, nearly all persons were involved in vehicle crashes between 2003 and 2006, and the majority of occupants were white (57%), male (93%), and between ages 21 and 25 (40%). Enlisted personnel (93%) and those with at least one prior deployment (88%) also predominated. No injury was reported for 39% of occupants; on the other hand, 10% of the occupants had fatal injuries. Most occupants were classified as being in combat at the time of the crash (81%). Operator was the most common occupant role for TABLE II. Specific Types of Vehicles in Which Soldier Occupants Were Involved in Crashes, Iraq and Afghanistan, 2002 2006 (n = 2,074) No. of Type No. of Occupants (%) Occupants with Injury (%) HMMWV (Tactical) 964 (46.5) 590 (48.5) Family of Medium 138 (6.7) 99 (8.1) Tactical Vehicles Other Fighting Vehicle 108 (5.2) 48 (3.9) Tank-M1 99 (4.8) 36 (3.0) 5 Ton Truck (Tactical) 97 (4.7) 54 (4.4) Heavy Expanded Mobility 92 (4.4) 48 (3.9) Tactical Truck Stryker Series 76 (3.7) 46 (3.8) Over 10 Ton Truck 71 (3.4) 31 (2.5) (Tactical) Tactical Trailers 34 (1.6) 11 (0.9) Carrier-M113 27 (1.3) 18 (1.5) Special Purpose 26 (1.3) 19 (1.6) Armored Vehicle CJ5/6/7 (Commercial Jeep) 23 (1.1) 16 (1.3) Other Tactical 22 (1.1) 9 (0.7) Carrier-Other 17 (0.8) 14 (1.2) Over 2 Ton Truck 16 (0.8) 14 (1.2) (Commercial) Vehicle Track Recovery 15 (0.7) 7 (0.6) Heavy Transport Equipment 15 (0.7) 5 (0.4) (Tactical) 2.5 Ton Truck (Tactical) 12 (0.6) 7 (0.6) Sedan/Station Wagon 11 (0.5) 5 (0.4) (Commercial) SP Guns/Howitzers 7 (0.4) 1 (0.1) Unknown 204 (9.8) 138 (11.3) Total 2,074 (100.0) 1,216 (100.0) TABLE III. Demographics of Occupants, HMMWV Crashes, Iraq and Afghanistan, 2002 2006 (n = 964) Category No. of Occupants (%) Year 2002 5 (0.5) 2003 168 (17.4) 2004 213 (22.1) 2005 370 (38.4) 2006 208 (21.6) Race White 546 (56.6) Black 139 (14.4) Hispanic 107 (11.1) Asian 36 (3.7) Am-Indian 15 (1.6) Unknown 121 (12.6) Gender Male 899 (93.2) Female 49 (5.1) Unknown 16 (1.7) Rank Enlisted E1-3 199 (20.6) Enlisted E4-5 570 (59.1) Enlisted E6-9 128 (13.3) Warrant W1-3 10 (1.0) Officer O1-5 57 (5.9) Role Supervisor 170 (17.6) Operator 465 (48.2) Passenger 145 (15.0) Gunner 148 (15.3) Bystander 36 (3.7) Age 18 20 141 (14.6) 21 25 389 (40.4) 26 30 161 (16.7) 31 35 83 (8.6) 36 40 68 (7.1) 41 45 43 (4.5) 45+ 12 (1.2) Unknown 67 (7.0) Mean 26.5 (SD 7.0) Range (18 59) Prior Deployments 0 117 (12.1) 1 463 (48.0) 2 274 (28.4) 3+ 110 (11.4) Mean 1.43 (SD 0.96) Range (0 6) Severity of Injury No Injury 374 (38.8) First Aid 127 (13.2) Restricted Work 42 (4.4) Lost Workday 306 (31.7) Disability 16 (1.6) Fatal 99 (10.3) Combat Status Combat 799 (80.8) No Combat 185 (19.2) MILITARY MEDICINE, Vol. 177, August 2012 965

Soldiers involved in HMMWV crashes (48%), although supervisor, passenger, and gunner represented at least 15% each (Table III). Injury Logistic regression analysis of injury (Table IV) showed that occupants of vehicles in crashes that were combat-related were associated with statistically significant higher odds of injury (odds ratio [OR], 1.49; 95% CI, 1.03 2.16) compared to occupants of crashes that did not occur during combat situations. The role of gunner was also shown to have significantly higher odds of injury than supervisors involved in a crash (OR, 2.37; 95% CI, 1.43 3.92). No other associations were statistically significant. Costs ANOVA of the log-transformed total cost of injury is shown in Table V. The combined model was statistically significant (F = 10.1, p < 0.001), and the combination of variables explained approximately 34% of the variance in injury costs (R 2 = 0.34). Personnel with no prior deployment had significantly higher costs of injury compared to personnel with prior deployments ( p < 0.001). Occupant role also had significant associations; injured operators (F = 8.02, p = 0.005) and gunners (F = 8.78, p = 0.003) had statistically significant higher injury costs. Interactions between variables were tested but were not significant, and are not shown. Trends Figure 1 shows the percentage of occupants of HMMWV crashes that were fatally and nonfatally injured. The proportion of occupants that were not injured increased from 2004 to 2006, compared to 2002 2003 ( p < 0.001). Although the percentage of fatalities decreased slightly after 2003, this was not statistically significant ( p = 0.32). TABLE IV. Multivariate Logistic Regression Analysis of Injury on Combat, Deployment, and Occupant Role (n = 850) Combined Model OR (95% CI) p-value SE No Combat 1.00 (Ref) Combat 1.49 (1.03, 2.16) 0.036* 0.28 Prior Deployment 0 1.00 (Ref) 1 1.12 (0.71, 1.76) 0.633 0.26 2 1.19 (0.73, 1.92) 0.489 0.29 3+ 1.65 (0.91, 2.96) 0.097 0.49 Role Supervisor 1.00 (Ref) Operator 1.08 (0.65, 1.80) 0.761 0.28 Passenger 0.87 (0.58, 1.31) 0.506 0.18 Gunner 2.37 (1.43, 3.92) 0.001* 0.61 Controlled for age, rank, and gender; excludes bystanders and occupants with missing variables. *Statistically significant p-values. TABLE V. ANOVA of Log Injury Cost on Combat, Deployment, and Occupant Role (n = 850) F statistic p-value Combined Model 10.1 <0.001* Combat 0.48 0.489 Prior Deployment 0 68.1 <0.001* 1 0.50 0.480 2 0.34 0.563 3+ - - Role Supervisor 1.34 0.248 Operator 8.02 0.005* Passenger 3.44 0.064 Gunner 8.78 0.003* Controlled for age, rank, and gender; excludes bystanders and occupants with missing variables. *Statistically significant p-values. Mistakes Failure to be attentive or follow procedures, failure to follow traffic laws, improper vehicle operation, and nonuse of safety equipment were among the most commonly recorded human factors contributing to HMMWV crashes (Table VI). Figure 2 shows the number and percentage of incidents with reported occupant mistakes per year. Although there was significant difference over the years in the percent of occupants with mistakes (p = 0.017), the trend is not in a consistent direction. DISCUSSION This analysis was the first published study, to our knowledge, to explore injuries to occupants of HMMWVs in a deployed setting using detailed safety report data. These results support the conclusions that HMMWV crashes were a significant problem among deployed forces from 2002 to 2006, and that involvement in combat and occupant role influenced the likelihood of injury following a crash. Occupant roles of gunner and operator were also associated with higher injury costs. As noted earlier, evidence on MMV crashes is limited in both number and quality of studies. Existing evidence identified by a 2010 systematic review detailed only 13 studies, with most coming from case reports or series and only one intervention study. 10 The review also pointed out a lack of epidemiologic studies assessing injury rates, type, severity, and risk factors, as well as interventions. Although the present study can only propose potential interventions, it may serve as a useful assessment of injury costs and risk factors and provide evidence to guide further prevention efforts and intervention research. A broader descriptive analysis of safety report data on motor vehicle crashes among Army personnel by Rossen et al showed a greater proportion of crashes occurred among younger males of enlisted rank, which is consistent with the results of this study. 11 Therefore, although conclusions in 966 MILITARY MEDICINE, Vol. 177, August 2012

FIGURE 1. Fatal and nonfatal injuries as percentage of occupants by year (n = 964). Chi-square for trend, p = 0.323 (Fatality), <0.001 (Injury and No injury). this article should be limited to military HMMWVs in a deployed setting, some of the findings may generalize to other vehicle types, especially those similar to HMMWVs. Soldiers in this study mirror common demographics in the U.S. Army (enlisted, under age 26, and/or on a second deployment), which may provide particular subgroups to focus interventions. Training and experience are widely regarded as integral to performance in a military environment, and the findings of TABLE VI. Primary Mistake Made by Type, Occupants of HMMWVs, 2002 2006 (n = 488) Mistake Type Mistakes Percentage of Total Mistakes Failed to Stay Alert or Attentive 87 17.8 Failed to Take Appropriate Precautions 62 12.7 or Follow Procedures Failed to Use Required Safety Equipment 58 11.9 Failed to Ensure Adequate Clearance 51 10.5 or Space Excessive Speed 42 8.6 Other Traffic Violations 42 8.6 (Improper Turns, Lane, Right-of-Way) Improper Vehicle Operation 33 6.8 (Brakes, Steering, etc.) Inadequate Preparation, 28 5.7 Planning, or Inspection Inadequate Crew Coordination 26 5.3 or Communication Improper Use of Equipment 21 4.3 (Other Than Vehicle) Operating While Fatigued 16 3.3 or Under Influence Improper Use of Ground Guide 11 2.3 Insufficient Evidence to Determine 11 2.3 Total 488 100 this study support this. Odds of being injured were higher when the crash occurred in combat, which suggests that in a high-stress situation with other priorities, the Soldier may be distracted or less likely to take self-protective measures or follow safety regulations. Thus, training in combat-like situations for all drivers could be important. However, since combat was not associated with increased cost, if an injury does occur, the estimated cost is similar regardless of the involvement of combat. This could also be explained by the cost calculations used in the relevant military documents, which have several limitations, including the fact they do not include air evacuation and other transportation costs that are relevant in theater. Military experience was defined by prior deployments, and although this is an imperfect proxy because of the absence of actual training information, it appears to have a highly protective effect on the Soldier for resulting injury cost. This indicates that those Soldiers who have seen and experienced more situations in theater are better prepared to protect themselves or follow protective procedures, and as a result suffer less severe (and less costly) injuries. In fact, further analysis showed that a large proportion of the fatalities (92%, n = 91) occurred in Soldiers on their first deployment. Since fatalities have a much higher cost associated with them, this helps explain the very large protective effect of prior deployments. It is also important to note that the likelihood of any injury actually appears to increase with prior deployments, although this was not statistically significant and does not take injury severity into account. One of the more important results from this study is the effect of occupant role on injury, which has until now been supported only by anecdotal evidence. Study results strongly suggest that gunners have higher odds of injury and also incur greater injury cost. This is not a particularly surprising result, MILITARY MEDICINE, Vol. 177, August 2012 967

FIGURE 2. Number and percentage of crashes with mistakes, HMMWVs by Year (n = 488). Chi-square for trend, p = 0.017. c 2 for trend, p = 0.017. as gunners are often in more exposed positions, especially when they are on top of the vehicle. Rollover crashes in particular would leave gunners, especially, susceptible to injury, and seat belt use or other protective devices could influence results. Although this information was not examined, it would be of interest for future research. Operators also have higher costs, which may be explained by confinement in the vehicle and having an additional avenue for injury (being behind the wheel). In contrast, passengers and supervisors may be in a more protected position or may be able to take actions to protect themselves in the event of a crash. Analysis of the number and type of mistakes involved is also of interest to better understand the root cause of crashes. Although the data do not show a clearly significant trend, it appears the proportion of accidents with mistakes peaked in 2004 and subsequently decreased in 2005 2006 (2002 had the highest proportion but only had five records). The proportion of injured occupants decreased over time, but proportion of fatalities did not have a significant trend, indicating an increase in the proportion of fatalities among all injuries. Further analysis of injury severity (performed but not shown) showed no compelling trend in either direction. The decrease in injuries from 2003 to 2006 may reflect increased experience or improved training as the force adapted to the deployed environment. It is also important to note that the majority of mistakes are still preventable by following procedures, obeying traffic laws, and using equipment appropriately. This study had several limitations. Data are limited to those crashes reported to the Army Combat Readiness/Safety Center. Although Army safety regulations require reporting of all military vehicle crashes resulting in property damage and/or injury, it is possible that some crashes were not captured. As with many other administrative data systems, missing data was a problem. In this study, data was missing for key potential confounders such as operator license status (23%), seat belt status of occupants (29%), date of last operator training (49%), and time in unit (79%), thus, they could not be controlled for in the analysis. However, we were able to control for other known confounders such as gender, rank, and age. Other important factors, such as the level of danger or intensity of the combat situations would have likely impacted the results but were not available in the database. The model of injury cost was likely influenced by several factors that were not included in this model, since it only explained one-third of the variance in cost. As mentioned earlier, the cost calculations were known to be limited as they do not take into account transportation or evacuation from the crash site. Type of crash was included in the data but was inconsistently recorded, and no clear avenue was identified to categorize these into rollover vs. other crashes, which would be of interest. Occupant role and mistakes were categorized from examination of qualitative entries. It was presumed that the most important mistake was listed first; however, we were unable to confirm this was the case for all crashes. RECOMMENDATIONS Despite these data limitations, some recommendations for policy and research can be drawn from this study. One potential intervention would be increased combat-related vehicle training that closely mimics in-theater operations, including simulation and live training exercises, before deployment. This could decrease the number of injuries sustained while deployed and also the number of crashes because of preventable mistakes. Training could also focus on following established procedures, following traffic guidelines, and using equipment properly. To inform training needs, future epidemiologic analyses should explore the data in further detail, including analyses of the types of mistakes made in combat and noncombat situations and differences in the types of mistakes made by injury severity. In addition, targeted external protection could be provided to gunners and operators of vehicles, who are particularly vulnerable to crash-related injury. Enforcement of seat belt use and application of additional injury prevention devices such as air bags should also be 968 MILITARY MEDICINE, Vol. 177, August 2012

considered based on prior evidence. 12,13 Although these two measures were not directly studied because of data limitations, the nature of the occupant roles at increased risk may support their use. Finally, injury reporting should be improved in the U.S. military because quality data is needed to direct changes in policy. Many of the limitations of this study resulted from inadequate data, and more useful and accurate information could lead to improved study design, which can better answer relevant questions. Further epidemiologic study on the topic of MMV crashes in deployed settings is warranted. Although traditional HMMWVs are being used less frequently (in favor of more explosion-resistant vehicles), they are nonetheless still an integral part of the military mission. Examining if injury patterns are impacted by the use of the new vehicle types would be useful to explore. Further analysis on this population could also provide additional valuable information on rollovers and protective equipment. CONCLUSIONS Injuries, and in particular vehicle crashes, remain a great threat to U.S. military force readiness and are costly from both a resource and manpower perspective. This study found that combat is associated with greater chance of injury, and gunners, operators, and first-time deployers have higher injury costs. Although not all crashes and injuries are preventable in the deployed setting and combat environment, significant measures could be taken to prevent or mitigate the effects of HMMWV crashes. These include training, protective devices, and enforcement of safety regulations. ACKNOWLEDGMENTS The authors thank Lauren M. Rossen, MS, from the Johns Hopkins Bloomberg School of Public Health for her assistance with data set preparation. This publication was supported in part by Grant Number 5R49CE001507 from the Centers for Disease Control and Prevention to the Johns Hopkins Center for Injury Research and Policy. The research was also supported by the Defense Safety Oversight Council through Concurrent Technologies Corporation in June 2007. REFERENCES 1. Hauret KG, Taylor BJ, Clemmons NS, Block SR, Jones BH: Frequency and causes of nonbattle injuries air evacuated from Operations Iraqi Freedom and Enduring Freedom, U.S. Amy, 2001 2006. Am J Prev Med 2010; 38(1 Suppl): S94 107. 2. U.S. Army Medical Surveillance Activity. Frequencies and characteristics of medical evacuations of Soldiers by air (with emphasis on nonbattle injuries). Operations Enduring Freedom/Iraqi Freedom (OEF/ OIF). January-November 2003. MSMR 2004; 10(3): 8 12. 3. Jones BH, Canham-Chervak M, Canada S, Mitchener TA, Moore S: Medical surveillance of injuries in the U.S. Military: descriptive epidemiology and recommendations for improvement. Am J Prev Med 2010; 38(1 Suppl): S42 60. 4. Jones BH, Perrotta DM, Canham-Chervak ML, Nee MA, Brundage JF: Injuries in the military: a review and commentary focused on prevention. Am J Prev Med 2000; 18(3 Suppl): 71 84. 5. Writer JV, DeFraites RF, Keep LW: Non-battle injury casualties during the Persian Gulf War and other deployments. Am J Prev Med 2000; 18(3 Suppl): 64 70. 6. U.S. Army Medical Surveillance Activity. Estimates of absolute and relative health care burdens attributable to various illnesses and injuries, U.S. Armed Forces 2005. MSMR 2006; 12(3): 2 23. 7. U.S. Army Medical Surveillance Activity. Absolute and relative morbidity burdens attributable to various illnesses and injuries, U.S. Armed Forces 2009. MSMR 2010; 17(4): 2 21. 8. Ruscio B, Smith J, Amoroso P, et al: DoD Military Injury Prevention Priorities Working Group: Leading Injuries, Causes, and Mitigation Recommendations. Washington: Office of the Assistant Secretary of Defense for Health Affairs, Clinical and Program Policy, 2006. Available at www.stormingmedia.us/75/7528/a752854.html; accessed March 21, 2012. 9. Krull AR, Jones BH, Dellinger AM, Yore MM, Amoroso PJ: Motor vehicle fatalities among men in the U.S. Army from 1980 to 1997. Mil Med 2004; 169(11): 926 31. 10. Krahl PL, Jankosky CJ, Thomas RJ, Hooper TI: Systematic review of military motor vehicle crash-related injuries. Am J Prev Med 2010; 38(1 Suppl): S189 96. 11. Rossen LM, Pollack KM, Canham-Chervak M, Canada S, Baker SP: Motor vehicle crashes among active duty U.S. Army personnel, 1999 to 2006. Mil Med 2011; 176(9): 1019 26. 12. Richter M, Pape HC, Ote D, Krettek C: Improvements in passive car safety led to decreased injury severity a comparison between the 1970s and 1990s. Injury 2005; 36: 484 8. 13. U.S. Department of Transportation National Highway Traffic Safety Administration. Traffic Safety Stats 2009 Annual Report. Available at http://www.nhtsa.gov; accessed March 21, 2012. MILITARY MEDICINE, Vol. 177, August 2012 969