Occupational and Non-Occupational Injuries in the United States Army Focus on Gender Hope M. Tiesman, MSPH, PhD, Corinne L. Peek-Asa, MSPH, PhD, Craig S. Zwerling, PhD, MD, Nancy L. Sprince, MPH, MD, Paul J. Amoroso, MPH, MD Background: The differences in occupational and non-occupational injuries between military men and women have not been documented. This study compares occupational and non-occupational injuries between male and female United States Army soldiers by examining injury hospitalization rates and characteristics. Methods:
The U.S. Army’s Total Army Injury and Health Outcomes Database was searched for hospitalizations with ICD-9-CM codes for injury (800 –959.9) between 1992 and 2002. Injury rates were calculated using yearly U.S. Army population data and compared using rate ratios. Injury characteristics were compared among categories of the Trauma Code (on duty; off duty; scheduled training, schemes, and exercises), stratified by gender.
Results:
Included in this analysis were 792 women for an injury hospitalization rate of 11.0 per 1000 individuals (95% confidence interval [CI]⫽8.5–13.5) and 4879 men for a rate of 15.5 per 1000 individuals (95% CI⫽14.0 –16.9). While women had significantly more injuries during scheduled training, schemes, and exercises than men (p⬍0.0001), there were few differences in the cause of those injuries. Women had longer average hospital stays compared to men due to these injuries (9.3 days vs 7.4 days, p⫽0.002), although these injuries were not more severe (average Injury Severity Score⫽3.5 for men vs average ISS for women⫽3.5, p⫽0.79). There was no difference between the genders in the percent of injuries that occurred off duty; however, men were more likely to get injured due to sports and athletics (p⫽0.001) and due to fighting (p⫽0.017) while off duty compared to women.
Conclusions: Injury prevention messages for military personnel should focus on reducing risk factors for both on- and off-duty injuries. (Am J Prev Med 2007;33(6):464 – 470) © 2007 American Journal of Preventive Medicine
Introduction
I
njuries are the leading cause of death, hospitalizations, disabilities, outpatient visits, and manpower losses in the military compared to all other causes of morbidity and mortality.1,2 While the occupational duties performed by military personnel predispose them to risk factors such as intense physical activity, use of heavy machinery and vehicles, dangerous environments, and harmful substances, the leading causes of injury in the military are non-occupational and include falls, athletic injuries, and nonduty motor-vehicle crashes.2 Consequently, both occupational and non-occupational inju-
From the National Institute for Occupational Safety and Health, Division of Safety Research (Tiesman), Morgantown, West Virginia; Department of Occupational and Environmental Health, University of Iowa (Peek-Asa, Zwerling, Sprince), Iowa City, Iowa; Madigan Army Medical Center (Amoroso), Tacoma, Washington Address correspondence and reprint requests to: Hope Tiesman, MSPH, PhD, NIOSH, Division of Safety Research, 1095 Willowdale Road M/S 1811, Morgantown WV 26505. E-mail:
[email protected].
464
ries continue to negatively affect the health of United States soldiers. Studies have found significant correlations between having an injury at home and having a workplace injury.3,4 At least two studies have demonstrated that those who had an injury while on the job were more likely to experience a leisure-time injury compared to those who did not have an occupational injury.3,4 Reasons for this association are not well understood, but current research suggests that personal risk factors such as musculoskeletal fitness, lifestyle behaviors, and education not only increase the risk for non-occupational injuries, but for occupational injuries as well.5– 8 These trends in injury risk have not been well described with respect to gender or occupation.3– 8 Many current databases have limited ability to examine these trends because occupational and non-occupational injuries are not differentiated. The military’s unique databases offer an opportunity to explore both on- and off-duty injuries.9 The majority of injury research in the military has focused on the time period of initial entry training
Am J Prev Med 2007;33(6) © 2007 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/07/$–see front matter doi:10.1016/j.amepre.2007.07.034
(IET).10 –15 This work has demonstrated that military women are at a significantly increased risk for musculoskeletal training-related injuries compared to military men, prompting some to conclude that military service is detrimental to women.16 However, very little research to date has considered the broad range of injuries military women experience in periods other than IET or in non-occupational settings. This study compares occupational and non-occupational injuries between male and female U.S. Army soldiers by examining injury hospitalization rates and injury characteristics. More specifically, are female soldiers at an increased risk for both occupational and non-occupational injuries compared to male soldiers? Also, are female soldiers’ injury characteristics similar to that of their male counterparts? A better understanding of gender differences in military injuries is important as women now make up 14% of active-duty personnel in the U.S. armed forces and they are in a variety of occupations once closed to them.17
Methods Cohort Definition A cross-sectional analysis was performed in 2006 of an 11-year cohort (1992–2002) of active-duty U.S. Army personnel with a hospitalized injury in their first 11 months of service. Data originated from the Total Army Injury and Health Outcomes Database (TAIHOD), a large, relational Army administrative database.18 The TAIHOD was developed and is maintained by the U.S. Army Research Institute of Environmental Medicine (USARIEM) and has been used in prior studies on military injuries.9,18 The TAIHOD captures data on virtually all military hospitalizations of active-duty Army soldiers from 1971 to the present.18 Additionally, the TAIHOD collects a high percentage of soldiers’ hospitalizations within civilian facilities.18 The cohort was originally defined to study injuries occurring early in a soldier’s military career and the 11-month time window was a predetermined category within the TAIHOD database. To be included in this cohort, soldiers had to meet the following criteria: (1) have a record in the TAIHOD, indicating active-duty status, between the years 1992 and 2002, (2) been hospitalized for an acute, traumatic injury in their first 11 months of military service, as categorized by the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) (codes 800 –959.9),19 and (3) the injury hospitalization was unintentional in nature as defined by the military injury coding system. If the individual had multiple injury hospitalizations during this time period, only the first injury is included in this analysis. Injuries treated on an outpatient basis, without hospital admission, were not included in this study. An important advantage of the military’s injury coding system is its ability to determine the work-relatedness of an injury through the Standard NATO Agreement (STANAG).20 This coding system has two components: the Trauma Code that conveys both the intent and work-relatedness of an injury, and an Injury Code that indicates the activity leading
December 2007
to the injury.20 The Trauma Code has ten specific values further categorized into three trauma classes: (1) battle wound or injury, (2) intentionally inflicted nonbattle injuries, and (3) accidental injury.20 The analysis was limited to those events within the ‘accidental injury’ class and included the following injury types: (1) off-duty, (2) schemes and exercises, (3) scheduled training, (4) on-duty, and (5) unknown.20 Battle-related injuries and intentional injuries were not included in this analysis.
Variables and Measures Each hospitalized injury was given up to eight diagnoses and procedures coded according to the ICD-9-CM, in addition to STANAG codes for the external cause of injury. Injury variables examined included the external cause of injury and the duty status of the soldier when injured. For this analysis, the following duty-status categories were considered: (1) off-duty, (2) on-duty, (3) schemes and exercises, (4) scheduled training, and (5) unknown. Schemes and exercises include specialized training scenarios such as military simulations and other field exercises. The category of scheduled training includes the first 8 to 12 weeks of military service (IET) and the occupational training occurring immediately afterwards (advanced individual training [AIT]). Schemes and exercises and scheduled training were combined into a single category for this analysis due to small sample sizes. Injury Severity Scores (ISSs) were calculated using ICD– MAP (Johns Hopkins University and Tri-Analytics Inc., 1997) with the “low” assumption.21 If a medical record contained two or more specific injuries of different severity levels, the severity of the lowest ICD-9-CM code was assigned using the “low” assumption. ISS scores were calculated by summing the squares of scores of the most severe injuries in up to three body regions.21 The ISS takes on values from 1 to 75, with higher numbers representing more severely injured patients.21 The number of days spent in the hospital was calculated using hospital admission and discharge dates. Demographic variables examined in this analysis were gender, age, education (less than high school, high school graduate, more than high school, or alternative education certificate) and military pay grade (enlisted E1–E3, enlisted E4 –E6, and officers/warrant officers O1–O6/W1–W5). Occupation was categorized into broader groups using the Department of Defense servicewide classification system called career management fields (CMFs).21 These twelve CMF categories included infantry and gun crews, electrical equipment repair, communications and intelligence, health care, technical and allied specialties, support and administration, electronic and mechanical equipment repair, craft workers, service and supply, non-occupational enlisted, and officer.22 The non-occupational enlisted category included those soldiers who had not yet chosen a military career path. CMFs are recorded in the demographic file of the TAIHOD and updated twice a year. For this analysis, the TAIHOD demographic record closest to the date of the injury, but not after this date, was merged with the injury record.
Analysis Injury rates were calculated and compared among different strata of demographic and occupational variables with rate ratios (RRs) and 95% confidence intervals (95% CI). Denom-
Am J Prev Med 2007;33(6)
465
inator data were obtained from the Defense Manpower Data Center (DMDC) by year, with regards to gender, age, education level, CMFs, and pay grade. Midyear estimates from the DMDC were used for this analysis. Injury characteristics were compared among duty-status categories, stratified by gender, using Pearson chi-square tests and Fisher’s exact test for categorical variables. For continuous variables, means, standard deviations, and ttests were calculated. All p-values were adjusted for multiple comparisons using the Bonferroni correction and adjustment procedure. All analyses were performed with SAS, version 9.1 (SAS Institute, Cary, NC, 2004).
Results Injury Rates From January 1992 through December 2002, there were 5678 soldiers with an injury hospitalization in their first 11 months of service (792 women and 4879 men) (Table 1). This resulted in an overall injury rate of 14.6 per 1000 soldiers per year (95% CI⫽13.5–15.8). Men had an overall injury rate of 15.5 per 1000 soldiers per year (95% CI⫽14.0 –16.9). Women had an injury rate of 11.0 per 1000 soldiers (95% CI⫽8.5–13.5). Men
were at a significantly increased risk for hospitalized injury over women (RR⫽1.40, 95% CI⫽1.30 –1.51) and enlisted men and women in the lowest pay grade compared to officers (RR⫽1.84, 95% CI⫽1.60 –2.07). The injury rates did not consistently increase across age categories. Soldiers with less than a high school education had the highest injury rate compared to all other educational categories (RR⫽3.74, 95% CI⫽3.21– 4.28). The occupational group with the highest hospitalized injury rate was infantry and gun crews (20.7 per 1000 personnel), followed by non-occupational enlisted (17.0 per 1000 personnel). When compared to military officers, all occupational groups for enlisted soldiers had a significantly increased risk for a hospitalized injury in their first 11 months of service.
Hospitalized Injuries by Gender, Duty Status, and Cause of Injury Women had significantly more injuries occur during scheduled training or schemes and exercises than men (women 28% and men 18%, p⬍0.0001). Men were significantly more likely than women to experience an
Table 1. Number of soldiers injured, injury rates, and rate ratios for demographics and occupation, 1992–2002 (N⫽5678)
Total Gender Male Female Age (years) 17–18 19–20 21–22 23–24 25–older Education More than high school Less than high school High school graduate Alternative education cert. Unknown Pay grade E1–E3 (enlisted) E4–E6 (enlisted) W1–W5 & O1–O6 (officer) Occupational groups Infantry/gun crews Electrical equipment repair Communications & intelligence Health care Technical & allied special Support & administration Electronic & mechanical equipment repair Craftsworkers Service and supply Non-occupational enlisted Officers
Average hospitalized injury rate per 1000 (95% CI)
Number injured
Average mid-year service population
5678
35,839
14.6 (13.5–15.8)
—
4879 792
28,948 6,716
15.5 (14.0–16.9) 11.0 (8.5–13.5)
1.40 (1.30–1.51) 1.00
345 2563 1252 732 762
6,161 12,615 7,514 4,333 4,974
5.33 (3.51–7.15) 18.35 (16.01–20.69) 15.07 (12.32–17.82) 15.56 (11.87–19.25) 14.75 (11.40–18.10)
0.36 (0.32–0.41) 1.24 (1.14–1.34) 1.02 (0.93–1.11) 1.05 (0.95–1.16) 1.00
403 350 4077 405 443
3,719 1,987 23,233 2,929 3,969
9.5 (6.0–13.0) 35.7 (27.5–43.8) 16.5 (14.9–18.2) 16.4 (12.3–20.4) 13.8 (10.1–17.4)
1.00 3.74 (3.21–4.28) 1.73 (1.56–1.91) 1.71 (1.48–1.95) 1.44 (1.25–1.64)
5305 136 237
31,216 1,954 2,649
15.5 (14.1–16.9) 8.6 (4.5–12.6) 8.4 (5.0–11.9)
1.84 (1.60–2.07) 1.01 (0.80–1.23) 1.00
1693 246 487 374 139 537 716 129 675 318 237
7,450 1,992 3,610 2,265 862 4,140 4,795 718 4,402 2,373 2,650
20.7 (17.4–24.0) 12.0 (7.2–16.8) 12.6 (9.0–16.2) 14.7 (9.8–19.7) 16.0 (7.6–24.4) 11.9 (8.6–15.2) 13.3 (10.0–16.5) 15.4 (6.4–24.4) 14.2 (10.7–17.7) 17.0 (11.8–22.1) 8.4 (5.0–12.0)
2.45 (2.12–2.78) 1.43 (1.17–1.68) 1.49 (1.26–1.73) 1.74 (1.46–2.03) 1.90 (1.50–2.30) 1.41 (1.19–1.63) 1.57 (1.34–1.81) 1.82 (1.43–2.21) 1.69 (1.44–1.93) 2.01 (1.67–2.35) 1.00
Rate ratio (95% CI)
CI, confidence interval.
466
American Journal of Preventive Medicine, Volume 33, Number 6
www.ajpm-online.net
Table 2. External cause of injury, stratified by gender and duty statusa Scheduled training, schemes, and exercises
Off duty Cause of injury
Male n (%)
Female n (%)
Falls Nonmilitary vehicle crash Athletics & sports Other specified agents Fighting Ill-fitting shoes Twist, turn, slip Cutting & piercing Parachuting Military vehicle crash Other Total
119 (15) 237 (29) 145 (18) 39 (5) 70 (9) 24 (3) 15 (2) 53 (6) 5 (1) 5 (1) 92 (11) 804
22 (19) 43 (37) 7 (6) 10 (9) 3 (3) 8 (7) 5 (4) 5 (4) 0 (0) 0 (0) 14 (12) 117
p-value 0.26 0.11 0.001 0.10 0.017b 0.04 0.16b 0.33 1.0b 1.0b 0.46
Male n (%)
Female n (%)
168 (19) 15 (2) 95 (11) 59 (7) 18 (2) 74 (9) 74 (9) 16 (2) 162 (19) 9 (1) 176 (20) 866
110 (50) 2 (1) 15 (7) 15 (7) 3 (1) 23 (11) 22 (10) 0 (0) 4 (2) 2 (1) 24 (11) 220
p-value
