TY - JOUR
AU - USN, Mark S. Riddle, MC
AB - ABSTRACT Background and Methods: Vector-borne diseases are known threats to deployed troops. We performed a cross-sectional study of troops deployed to Southwest Asia between January 2005 and February 2007 to evaluate practices of personal protective measures and their relationship to self-report of Old World cutaneous leishmaniasis (CL), a marker of vector-borne disease threat. Results: Regular or always N, N-diethyl-m-toluamide (DEET) use was low (2–5%). Associations for DEET use were command emphasis, branch of service, uniform treatment with permethrin, and duty station. Uniform treatment with permethrin was associated with branch of service, command emphasis, and use of DEET. We identified 22 cases of CL (incidence density of 1.8–3.7 per 100 person-years) with increased risk among Reserve/National Guard components, Air Force and Marine personnel. Conclusions: Commanders can influence the use of the military insect repellent system. Unit-based treatment of uniforms improves prevalence. CL incidence may be higher than previously reported. INTRODUCTION Deployed military personnel are often exposed to various vector-borne diseases (VBD). The range of insect vectors and associated diseases, the potential severe morbidity, and lack of available vaccines for many of these diseases make the reliance on environmental modification and personal protective measures (PPM) of utmost importance. A previous serosurvey has shown evidence of exposure to arboviruses in U.S. troops deployed in Iraq.1 Old World cutaneous leishmaniasis (CL), a parasitic infection caused by Leishmania spp. (most commonly Leishmania major and Leishmania tropica) obtained from the bite of a female Phlebotomus sand fly, has been a significant problem for U.S. personnel recently deployed to Iraq and Afghanistan.2,5 The numerous reported cases occurring among troops have refocused attention on this health threat and current prevention strategies. Leishmaniasis is endemic to both Afghanistan and Iraq. According to World Health Organization authorities cited by the Associated Press, Afghanistan and Iraq now have ∼200,000 cases and 3,000 cases of leishmaniasis per year, respectively.6 Although developed infrastructure (e.g., screened housing) and vector control are critical mainstays of reducing vector-related deployment health threats, PPM in the form of uniform pretreatment with permethrin and the use of topical insect repellants (e.g., N, N-diethyl-m-toluamide [DEET]) are considered to be effective (particularly in combination) and an important tool for integrated management.7,10bib Although deployment conditions and scenarios may not always allow for use of environmental modification, use of PPM should be a portable technology in most every situation, but it seems that these measures are not regularly used. A study of disease and nonbattle injury in deployed troops to Iraq and Afghanistan conducted from January through March 2004 found that despite a 2.1% prevalence of self-reported leishmaniasis diagnosis, few personnel reported using DEET more than occasionally and over half reported never using it.11 Furthermore, Coleman et al.12 showed only 5% of Army units and 15% of Air Force units deployed to Tallil Air Base had insect repellent, permethrin-treated uniforms or bed nets. When available, these measures were poorly used. An exception was noted in the flying and security forces personnel, who showed high PPM use.12 Previous studies of attitudes toward and use of DEET by military personnel have shown reluctance to use11 and preference for commercial products.12 This study was designed to measure the prevalence of PPM use, determine what factors affect/promote the use of PPM, and assess the impact of these factors on the self-reported incidence of CL among U.S. military personnel deployed to the Central Command Area of Responsibility (CENTCOM AOR) from January 2005 through February 2007. METHODS This was a repeated cross-sectional survey among U.S. troops deployed to Iraq, Afghanistan, and the surrounding region using two survey instruments. Survey 1 was administered between January 2005 and May 2006. Survey 2 was administered between March 2006 and February 2007. Respondents participated in only one survey. Study Sample and Survey Construct The survey instruments were administered to a convenience sample of U.S. military personnel participating in a rest and recuperation (R&R) program at Camp As Sayliyah in Doha, Qatar, or in transit through Incirlik Air Base, Turkey or Camp Arifjan, Kuwait. They contained a total of 23 and 20 multiple-choice questions, respectively. Completion of the survey was voluntary and anonymous. Response rate was not tracked for this survey. Historical response rate to similar surveys in the CENTCOM AOR has been 65 to 80% (M.S. Riddle, personal communication). Answers to questions were completed by a multiple-choice answer. Primary outcomes of interest were: use of DEET/insect repellent and uniform treatment with permethrin both before and while deployed; prevalence of vector exposure; and incidence density of CL, a marker for VBDs. CL was selected as a surrogate VBD marker, albeit insensitive, because it has specific regional occurrence and is less prone to self-reporting bias due to the clinical manifestations of skin lesions (as opposed to other less frequent or “nonspecific” in-theater VBD threats including but not limited to sandfly fever, Sindbis virus, West Nile virus, Crimean-Congo hemorrhagic fever, and malaria), as well as recent focused attention increasing troop awareness of CL (e.g., “Bagdad boil”) in current deployments. Predictor variables included age and gender, branch and component of service, rank and military occupation, attitudes toward DEET, use and availability of DEET/ insect repellent, uniform treatment with permethrin, frequency of insect bites, impression of command emphasis on use of PPM, knowledge of insect-borne diseases and PPM, and remoteness of duty location (measured by type of toilet facilities). A knowledge score variable was created as a composite of six questions regarding PPM and VBD. Data Analysis Student's t test (two tailed) for independent samples was used to compare characteristics between the two survey samples. The PPM use measures included DEET/insect repellent use and treatment of uniforms with permethrin. These endpoints were assessed with both χ2 tests of independence and multi-variable logistic regression. The χ2 test of independence and Poisson regression analysis were used to study relationships of variables of interest with outcomes of CL and insect/sand fly bite. Incidence density was estimated using the following methods. CL events (numerator) were obtained from a question asking, “Did you get leishmaniasis during this deployment?” Respondents were also asked the following Likert-type questions, “How often were you bitten by any insect?” and “How often were you bitten by sand flies?” Denominator data were based on person time provided by date for arrival in-theater and date of questionnaire completion. All assumptions for Poisson regression analysis were met. Incidence density was approximated using full-time and mid-time in deployment to estimate actual incidence density. The full-time estimate results in the most conservative estimate by assuming that disease took place on the day the form was filled out. The mid-time estimate will likely overestimate some and underestimate others, but is a reasonable compromise. Five hundred thirty-eight people (24%) and 361 (31%) did not answer the “date arrived in theater” question posed in surveys 1 and 2, respectively This made calculating “time in theater” impossible for these individuals. Missing time data were addressed in the following manner. Because this was a cross-sectional survey, it is more likely that these data were missing at random and not due to any effect of outcome, treatment, or follow-up. A stochastic multiple regression model was developed to deal with the missing data in accord with the work of Little and Rubin.13 Univariable and multi-variable logistic regression for statistical significance of predictors was carried out on all demographic variables from both surveys, as well as the form completion date. The statistically significant variables from the final regression equation were then included as predictor variables in the impute function of Stata version 10.0 (StataCorp LC, College Station, Texas) to generate the missing time data.14 This method was thought valid since military members tend to deploy/redeploy in unit cohorts. Therefore, those of the same service, similar rank, coming from same duty location, of the same unit type, and completing the survey on the same day are highly likely to have arrived in theater together. Model building for multivariable Poisson regression of CL events and multivariable logistic regression for vector exposure, “DEET/repellent use” and “uniform treatment with permethrin” was performed by a backward-stepwise technique, for inclusion p ≤ 0.2. Predictor variables were retained in the final model if they were statistically significant, confounders, or thought to be in the protective pathway from previous literature (DEET use, permethrin uniform treatment). Age was not linearly associated with the outcomes and was modeled categorically. The Hosmer-Lemeshow “goodness-of- fit” test was run on all regression models.15 All statistical tests were run with Stata version 10.0 (StataCorp LC) to a significance level of 0.05. RESULTS Study Population and Demographics Between January 2005 to February 2007, 3,446 surveys were completed by personnel on rest and recuperation in Doha, Qatar (n = 3,003, 87%) and during redeployment from Kuwait (n = 333, 10%) and Turkey (n = 110, 3%). Table I shows demographic characteristics of the two samples. Survey 2 respondents differed from survey 1 respondents on the following characteristics: slightly more female respondents (18% vs. 15%, p = 0.05), more regular component respondents (59% vs. 51%. p < 0.001), fewer ground troops (22% vs. 29%, p < 0.001), more likely to be supporting OIF (72% vs. 66%, p = 0.001), and less likely to report no previous deployment (26.9% vs. 70.4%, p < 0.001). Compared to demographic data, requested by the authors and supplied by the Defense Manpower Data Center for the time frame of this survey, the following groups were oversampled: women, National Guard and Reserve components, and Army members. Air Force and Navy members were undersampled. No significant differences were found for age or rank structure. TABLE I. Comparison of Survey Respondent Characteristics Characteristic  Survey 1 n (%)  Survey 2 n (%)  P  No. of respondents  2,277  1,169    Survey dates  January 2005–May 2006  March 2006–February 2007    Mean age (years)  28.75  28.91  0.6  Female gender  341 (15%)  210 (18%)  0.05  Location completed survey          Qatar  1,855 (81%)  1,148 (98%)      Kuwait  333 (15%)  0      Turkey  89 (4%)  21 (2%)    Military component      <0.001    Regular  1,161 (51%)  690 (59%)      Reserve  523 (23%)  222 (19%)      National Guard  568 (25%)  245 (21%)    Branch of service      0.9    Army  1,752 (77%)  887 (76%)      Air Force  68 (3%)  35 (3%)      Marine Corps  317 (14%)  174 (15%)      Navy  137 (6%)  82 (7%)    Military grade      0.06    E1–4  1,138 (50%)  573 (49%)      E5–6  751 (33%)  362(31%)      E7–9  158 (7%)  94 (8%)      Warrant  Warrant 18 (1%)  23 (2%)      01–3  115 (5%)  83 (7%)      04–6  89 (4%)  47 (4%)    Unit type      <0.001    Ground  659 (29%)  257 (22%)      Air  205 (9%)  129 (11%)      Support  1,046 (46%)  538 (46%)      Command  157 (7%)  118(10%)      Special Operations  47 (2%)  35 (3%)      Other  134 (6%)  71 (6%)    Operation supported      0.001    OIF  1,499 (66%)  842 (72%)      OEF  525 (23%)  197 (17%)      Both  203 (9%)  105 (9%)    Mean No. of days in theater (SD)  199 (87)  199 (79)  0.4  Reporting no prior deployment  1,603 (70.4%)  314(26.9%)  <0.001  Characteristic  Survey 1 n (%)  Survey 2 n (%)  P  No. of respondents  2,277  1,169    Survey dates  January 2005–May 2006  March 2006–February 2007    Mean age (years)  28.75  28.91  0.6  Female gender  341 (15%)  210 (18%)  0.05  Location completed survey          Qatar  1,855 (81%)  1,148 (98%)      Kuwait  333 (15%)  0      Turkey  89 (4%)  21 (2%)    Military component      <0.001    Regular  1,161 (51%)  690 (59%)      Reserve  523 (23%)  222 (19%)      National Guard  568 (25%)  245 (21%)    Branch of service      0.9    Army  1,752 (77%)  887 (76%)      Air Force  68 (3%)  35 (3%)      Marine Corps  317 (14%)  174 (15%)      Navy  137 (6%)  82 (7%)    Military grade      0.06    E1–4  1,138 (50%)  573 (49%)      E5–6  751 (33%)  362(31%)      E7–9  158 (7%)  94 (8%)      Warrant  Warrant 18 (1%)  23 (2%)      01–3  115 (5%)  83 (7%)      04–6  89 (4%)  47 (4%)    Unit type      <0.001    Ground  659 (29%)  257 (22%)      Air  205 (9%)  129 (11%)      Support  1,046 (46%)  538 (46%)      Command  157 (7%)  118(10%)      Special Operations  47 (2%)  35 (3%)      Other  134 (6%)  71 (6%)    Operation supported      0.001    OIF  1,499 (66%)  842 (72%)      OEF  525 (23%)  197 (17%)      Both  203 (9%)  105 (9%)    Mean No. of days in theater (SD)  199 (87)  199 (79)  0.4  Reporting no prior deployment  1,603 (70.4%)  314(26.9%)  <0.001  Totals may not reach 100% due to non-respondents to various questions and rounding. Repeated cross-sectional survey among U.S. troops deployed to Iraq, Afghanistan, and the surrounding region utilizing two survey instruments. Survey 1 was administered between January 2005 and May 2006. Survey 2 was administered between March 2006 and February 2007. Respondents participated in only one survey. View Large TABLE I. Comparison of Survey Respondent Characteristics Characteristic  Survey 1 n (%)  Survey 2 n (%)  P  No. of respondents  2,277  1,169    Survey dates  January 2005–May 2006  March 2006–February 2007    Mean age (years)  28.75  28.91  0.6  Female gender  341 (15%)  210 (18%)  0.05  Location completed survey          Qatar  1,855 (81%)  1,148 (98%)      Kuwait  333 (15%)  0      Turkey  89 (4%)  21 (2%)    Military component      <0.001    Regular  1,161 (51%)  690 (59%)      Reserve  523 (23%)  222 (19%)      National Guard  568 (25%)  245 (21%)    Branch of service      0.9    Army  1,752 (77%)  887 (76%)      Air Force  68 (3%)  35 (3%)      Marine Corps  317 (14%)  174 (15%)      Navy  137 (6%)  82 (7%)    Military grade      0.06    E1–4  1,138 (50%)  573 (49%)      E5–6  751 (33%)  362(31%)      E7–9  158 (7%)  94 (8%)      Warrant  Warrant 18 (1%)  23 (2%)      01–3  115 (5%)  83 (7%)      04–6  89 (4%)  47 (4%)    Unit type      <0.001    Ground  659 (29%)  257 (22%)      Air  205 (9%)  129 (11%)      Support  1,046 (46%)  538 (46%)      Command  157 (7%)  118(10%)      Special Operations  47 (2%)  35 (3%)      Other  134 (6%)  71 (6%)    Operation supported      0.001    OIF  1,499 (66%)  842 (72%)      OEF  525 (23%)  197 (17%)      Both  203 (9%)  105 (9%)    Mean No. of days in theater (SD)  199 (87)  199 (79)  0.4  Reporting no prior deployment  1,603 (70.4%)  314(26.9%)  <0.001  Characteristic  Survey 1 n (%)  Survey 2 n (%)  P  No. of respondents  2,277  1,169    Survey dates  January 2005–May 2006  March 2006–February 2007    Mean age (years)  28.75  28.91  0.6  Female gender  341 (15%)  210 (18%)  0.05  Location completed survey          Qatar  1,855 (81%)  1,148 (98%)      Kuwait  333 (15%)  0      Turkey  89 (4%)  21 (2%)    Military component      <0.001    Regular  1,161 (51%)  690 (59%)      Reserve  523 (23%)  222 (19%)      National Guard  568 (25%)  245 (21%)    Branch of service      0.9    Army  1,752 (77%)  887 (76%)      Air Force  68 (3%)  35 (3%)      Marine Corps  317 (14%)  174 (15%)      Navy  137 (6%)  82 (7%)    Military grade      0.06    E1–4  1,138 (50%)  573 (49%)      E5–6  751 (33%)  362(31%)      E7–9  158 (7%)  94 (8%)      Warrant  Warrant 18 (1%)  23 (2%)      01–3  115 (5%)  83 (7%)      04–6  89 (4%)  47 (4%)    Unit type      <0.001    Ground  659 (29%)  257 (22%)      Air  205 (9%)  129 (11%)      Support  1,046 (46%)  538 (46%)      Command  157 (7%)  118(10%)      Special Operations  47 (2%)  35 (3%)      Other  134 (6%)  71 (6%)    Operation supported      0.001    OIF  1,499 (66%)  842 (72%)      OEF  525 (23%)  197 (17%)      Both  203 (9%)  105 (9%)    Mean No. of days in theater (SD)  199 (87)  199 (79)  0.4  Reporting no prior deployment  1,603 (70.4%)  314(26.9%)  <0.001  Totals may not reach 100% due to non-respondents to various questions and rounding. Repeated cross-sectional survey among U.S. troops deployed to Iraq, Afghanistan, and the surrounding region utilizing two survey instruments. Survey 1 was administered between January 2005 and May 2006. Survey 2 was administered between March 2006 and February 2007. Respondents participated in only one survey. View Large PPM Use When analyzed by military operation supported, regular, or always DEET use, although still low, was higher for OIF than for OEF (6% vs. 3%, p = 0.05). Reporting that command adequately emphasized repellent use differed by country: Iraq, 43%; Afghanistan, 39%; and Kuwait, 30% (p = 0.001). Reported use of any insect repellent paralleled this pattern: Iraq, 62%; Afghanistan, 52%; and Kuwait, 49% (p = 0.002). Marine Corps and Navy members reported more permethrin-treated uniforms before deployment than the Army and Air Force. Air Force and Marine Corps members reported adequate command emphasis more than Army and Navy members. Most service members reported DEET was available when needed and believe in its safety and efficacy. The Army had the fewest respondents believing in DEET safety (56%). A minority of service members (16% overall) reported using only military-supplied insect repellent. Air Force members were at the high end with 26% of respondents reporting military-only insect repellent use. Table II shows the univariable and multivariable logistic regression analysis for outcome variables: uniform treatment before deployment, uniform treatment while deployed, and any repellent use (military issue, commercial, or both). The multivariable logistic regression model for “uniform treatment with permethrin before deployment” (survey 2) included the following statistically significant variables: branch of service, command emphasis on use of insect repellents, repellent use, rank, and knowledge score (goodness-of- fit: χ2, p = 0.13). The Air Force, Navy, and Marine Corps had a five times greater prevalence odds ratio (POR) for reporting a treated a uniform before deployment compared to Army respondents (p < 0.001). Increasing levels of command emphasis on insect repellent use led to increasing POR of treating a uniform. Reporting any repellent use or military-only repellent use doubled the POR of treating a uniform before deployment (p < 0.01). Whereas reporting commercial repellent-only use halved the POR of treating a uniform (p = 0.008). The POR of reporting a permethrin-treated uniform increased by 20% for each 1 point increase in knowledge score (p = 0.004). TABLE II. Univariable and Multivariable Logistic Regression Analysis for Uniform Treatment with Permethrin and Repellent Use   Uniform Treated Before Deployment (POR)  Uniform Treated While Deployed (POR)  Any Repellent Use (POR)          Characteristic  Univariable (95% CI)  Multivariable (95% CI)  Univariable (95% CI)  Multivariable (95% CI)  Univariable (95% CI)  Multivariable (95% CI)  Branch of service                  Army  Reference  Reference  Reference  —  Reference  Reference      Air Force  3.2(1.5–7.1)  5.4 (2.2–13)  1.4(0.6–3.1)  —  0.5 (0.2–1.1)  0.4 (0.2–0.9)      Marine Corps  4.7 (3.2–6.9)  5.8 (3.7–9.3)  1.5 (1.1–2.2)  —   0.7 (0.5–0.98)  0.4 (0.3–0.7)      Navy  3.4 (2.1–5.7)  5.3 (2.9–9.5)  1.2(0.7–2.1)  —  0.5 (0.3–0.8)  0.3 (0.2–0.6)  Command emphasis                  Too much  7.3 (3.9–14)  5.0 (2.5–10)  9.6 (5.0–18)  4.0(1.9–8.2)  3.5 (1.9–6.3)  2.7 (1.3–5.1)      Adequate  2.8 (2.1–3.9)  2.5 (1.7–3.5)  3.7 (2.6–5.3)  2.1 (1.4–3.2)  2.8 (2.1–3.7)  2.2 (1.5–2.9)      Some, not enough  1.6(1.1–2.4)  1.2(0.8–1.9)  2.4(1.6–3.7)  2.0(1.2–3.3)  2.7 (1.8–3.8)  2.4 (1.6–3.6)      Not at all  Referencec a  Reference  Referencea  Reference  Referencea  Reference  Repellent use                  None  Reference  Reference  Reference  Reference  —  —      Any  1.9 (1.4–2.7)  2.2 (1.2–4.2)  2.7 (1.8–3.8)  —  —  —      Military  4.1 (2.8–6.2)  1.9 (1.2–3.0)  5.4 (3.6–8.2)  2.4(1.7–3.3)  —  —      Commercial  1.3 (0.9–1.8)  0.5 (0.3–0.8)  1.2(0.8–1.8)  —  —  —  Grade                  E1–4  Reference  Reference  Reference  —  Reference  —      E5–6  0.7 (0.5–0.9)  0.7 (0.5–1.1)  0.8 (0.6–1.1)  —  1.04 (0.8–1.4)  —      E7–9  0.6 (0.4–1.0)  0.6 (0.3–1.3)  1.0(0.6–1.6)  —  1.2(0.7–1.9)  —      Warrant      0.3 (0.08–0.95)  0.2 (0.4–1.0)  0.3 (0.6–1.1)  —  0.5 (0.2–1.3)  —      01–3  0.6(0.4–1.1)  0.7 (0.4–1.4)  0.5 (0.3–0.9)  —  0.7 (0.4–1.2)  —      04–6  0.4 (0.2–0.9)  0.3 (0.1–0.8)  0.7 (0.3–1.3)  —  0.7 (0.4–1.3)  —  Knowledge score  1.2(1.1–1.3)  1.2(1.1–1.3)  1.3 (1.2–1.4)  1.2 (1.0–1.3)  1.2 (1.1–1.3)  1.1 (1.0–1.3)  Treated uniform prior to deployment  —  —   10.5 (7.4–15)  8.4 (5.9–12)  2.0(1.6–2.7)  2.1 (1.5–2.9)  No. of previous deployments  1.04 (0.9–1.2)  —  1.1 (1.0–1.2)  —  1.1 (1.0–1.2)  1.2(1.1–1.4)    Uniform Treated Before Deployment (POR)  Uniform Treated While Deployed (POR)  Any Repellent Use (POR)          Characteristic  Univariable (95% CI)  Multivariable (95% CI)  Univariable (95% CI)  Multivariable (95% CI)  Univariable (95% CI)  Multivariable (95% CI)  Branch of service                  Army  Reference  Reference  Reference  —  Reference  Reference      Air Force  3.2(1.5–7.1)  5.4 (2.2–13)  1.4(0.6–3.1)  —  0.5 (0.2–1.1)  0.4 (0.2–0.9)      Marine Corps  4.7 (3.2–6.9)  5.8 (3.7–9.3)  1.5 (1.1–2.2)  —   0.7 (0.5–0.98)  0.4 (0.3–0.7)      Navy  3.4 (2.1–5.7)  5.3 (2.9–9.5)  1.2(0.7–2.1)  —  0.5 (0.3–0.8)  0.3 (0.2–0.6)  Command emphasis                  Too much  7.3 (3.9–14)  5.0 (2.5–10)  9.6 (5.0–18)  4.0(1.9–8.2)  3.5 (1.9–6.3)  2.7 (1.3–5.1)      Adequate  2.8 (2.1–3.9)  2.5 (1.7–3.5)  3.7 (2.6–5.3)  2.1 (1.4–3.2)  2.8 (2.1–3.7)  2.2 (1.5–2.9)      Some, not enough  1.6(1.1–2.4)  1.2(0.8–1.9)  2.4(1.6–3.7)  2.0(1.2–3.3)  2.7 (1.8–3.8)  2.4 (1.6–3.6)      Not at all  Referencec a  Reference  Referencea  Reference  Referencea  Reference  Repellent use                  None  Reference  Reference  Reference  Reference  —  —      Any  1.9 (1.4–2.7)  2.2 (1.2–4.2)  2.7 (1.8–3.8)  —  —  —      Military  4.1 (2.8–6.2)  1.9 (1.2–3.0)  5.4 (3.6–8.2)  2.4(1.7–3.3)  —  —      Commercial  1.3 (0.9–1.8)  0.5 (0.3–0.8)  1.2(0.8–1.8)  —  —  —  Grade                  E1–4  Reference  Reference  Reference  —  Reference  —      E5–6  0.7 (0.5–0.9)  0.7 (0.5–1.1)  0.8 (0.6–1.1)  —  1.04 (0.8–1.4)  —      E7–9  0.6 (0.4–1.0)  0.6 (0.3–1.3)  1.0(0.6–1.6)  —  1.2(0.7–1.9)  —      Warrant      0.3 (0.08–0.95)  0.2 (0.4–1.0)  0.3 (0.6–1.1)  —  0.5 (0.2–1.3)  —      01–3  0.6(0.4–1.1)  0.7 (0.4–1.4)  0.5 (0.3–0.9)  —  0.7 (0.4–1.2)  —      04–6  0.4 (0.2–0.9)  0.3 (0.1–0.8)  0.7 (0.3–1.3)  —  0.7 (0.4–1.3)  —  Knowledge score  1.2(1.1–1.3)  1.2(1.1–1.3)  1.3 (1.2–1.4)  1.2 (1.0–1.3)  1.2 (1.1–1.3)  1.1 (1.0–1.3)  Treated uniform prior to deployment  —  —   10.5 (7.4–15)  8.4 (5.9–12)  2.0(1.6–2.7)  2.1 (1.5–2.9)  No. of previous deployments  1.04 (0.9–1.2)  —  1.1 (1.0–1.2)  —  1.1 (1.0–1.2)  1.2(1.1–1.4)  Adjusted for age and gender. Survey 2 data. Knowledge score was generated as a composite measure of six “agree/disagree” survey questions. a Positive test for trend. View Large TABLE II. Univariable and Multivariable Logistic Regression Analysis for Uniform Treatment with Permethrin and Repellent Use   Uniform Treated Before Deployment (POR)  Uniform Treated While Deployed (POR)  Any Repellent Use (POR)          Characteristic  Univariable (95% CI)  Multivariable (95% CI)  Univariable (95% CI)  Multivariable (95% CI)  Univariable (95% CI)  Multivariable (95% CI)  Branch of service                  Army  Reference  Reference  Reference  —  Reference  Reference      Air Force  3.2(1.5–7.1)  5.4 (2.2–13)  1.4(0.6–3.1)  —  0.5 (0.2–1.1)  0.4 (0.2–0.9)      Marine Corps  4.7 (3.2–6.9)  5.8 (3.7–9.3)  1.5 (1.1–2.2)  —   0.7 (0.5–0.98)  0.4 (0.3–0.7)      Navy  3.4 (2.1–5.7)  5.3 (2.9–9.5)  1.2(0.7–2.1)  —  0.5 (0.3–0.8)  0.3 (0.2–0.6)  Command emphasis                  Too much  7.3 (3.9–14)  5.0 (2.5–10)  9.6 (5.0–18)  4.0(1.9–8.2)  3.5 (1.9–6.3)  2.7 (1.3–5.1)      Adequate  2.8 (2.1–3.9)  2.5 (1.7–3.5)  3.7 (2.6–5.3)  2.1 (1.4–3.2)  2.8 (2.1–3.7)  2.2 (1.5–2.9)      Some, not enough  1.6(1.1–2.4)  1.2(0.8–1.9)  2.4(1.6–3.7)  2.0(1.2–3.3)  2.7 (1.8–3.8)  2.4 (1.6–3.6)      Not at all  Referencec a  Reference  Referencea  Reference  Referencea  Reference  Repellent use                  None  Reference  Reference  Reference  Reference  —  —      Any  1.9 (1.4–2.7)  2.2 (1.2–4.2)  2.7 (1.8–3.8)  —  —  —      Military  4.1 (2.8–6.2)  1.9 (1.2–3.0)  5.4 (3.6–8.2)  2.4(1.7–3.3)  —  —      Commercial  1.3 (0.9–1.8)  0.5 (0.3–0.8)  1.2(0.8–1.8)  —  —  —  Grade                  E1–4  Reference  Reference  Reference  —  Reference  —      E5–6  0.7 (0.5–0.9)  0.7 (0.5–1.1)  0.8 (0.6–1.1)  —  1.04 (0.8–1.4)  —      E7–9  0.6 (0.4–1.0)  0.6 (0.3–1.3)  1.0(0.6–1.6)  —  1.2(0.7–1.9)  —      Warrant      0.3 (0.08–0.95)  0.2 (0.4–1.0)  0.3 (0.6–1.1)  —  0.5 (0.2–1.3)  —      01–3  0.6(0.4–1.1)  0.7 (0.4–1.4)  0.5 (0.3–0.9)  —  0.7 (0.4–1.2)  —      04–6  0.4 (0.2–0.9)  0.3 (0.1–0.8)  0.7 (0.3–1.3)  —  0.7 (0.4–1.3)  —  Knowledge score  1.2(1.1–1.3)  1.2(1.1–1.3)  1.3 (1.2–1.4)  1.2 (1.0–1.3)  1.2 (1.1–1.3)  1.1 (1.0–1.3)  Treated uniform prior to deployment  —  —   10.5 (7.4–15)  8.4 (5.9–12)  2.0(1.6–2.7)  2.1 (1.5–2.9)  No. of previous deployments  1.04 (0.9–1.2)  —  1.1 (1.0–1.2)  —  1.1 (1.0–1.2)  1.2(1.1–1.4)    Uniform Treated Before Deployment (POR)  Uniform Treated While Deployed (POR)  Any Repellent Use (POR)          Characteristic  Univariable (95% CI)  Multivariable (95% CI)  Univariable (95% CI)  Multivariable (95% CI)  Univariable (95% CI)  Multivariable (95% CI)  Branch of service                  Army  Reference  Reference  Reference  —  Reference  Reference      Air Force  3.2(1.5–7.1)  5.4 (2.2–13)  1.4(0.6–3.1)  —  0.5 (0.2–1.1)  0.4 (0.2–0.9)      Marine Corps  4.7 (3.2–6.9)  5.8 (3.7–9.3)  1.5 (1.1–2.2)  —   0.7 (0.5–0.98)  0.4 (0.3–0.7)      Navy  3.4 (2.1–5.7)  5.3 (2.9–9.5)  1.2(0.7–2.1)  —  0.5 (0.3–0.8)  0.3 (0.2–0.6)  Command emphasis                  Too much  7.3 (3.9–14)  5.0 (2.5–10)  9.6 (5.0–18)  4.0(1.9–8.2)  3.5 (1.9–6.3)  2.7 (1.3–5.1)      Adequate  2.8 (2.1–3.9)  2.5 (1.7–3.5)  3.7 (2.6–5.3)  2.1 (1.4–3.2)  2.8 (2.1–3.7)  2.2 (1.5–2.9)      Some, not enough  1.6(1.1–2.4)  1.2(0.8–1.9)  2.4(1.6–3.7)  2.0(1.2–3.3)  2.7 (1.8–3.8)  2.4 (1.6–3.6)      Not at all  Referencec a  Reference  Referencea  Reference  Referencea  Reference  Repellent use                  None  Reference  Reference  Reference  Reference  —  —      Any  1.9 (1.4–2.7)  2.2 (1.2–4.2)  2.7 (1.8–3.8)  —  —  —      Military  4.1 (2.8–6.2)  1.9 (1.2–3.0)  5.4 (3.6–8.2)  2.4(1.7–3.3)  —  —      Commercial  1.3 (0.9–1.8)  0.5 (0.3–0.8)  1.2(0.8–1.8)  —  —  —  Grade                  E1–4  Reference  Reference  Reference  —  Reference  —      E5–6  0.7 (0.5–0.9)  0.7 (0.5–1.1)  0.8 (0.6–1.1)  —  1.04 (0.8–1.4)  —      E7–9  0.6 (0.4–1.0)  0.6 (0.3–1.3)  1.0(0.6–1.6)  —  1.2(0.7–1.9)  —      Warrant      0.3 (0.08–0.95)  0.2 (0.4–1.0)  0.3 (0.6–1.1)  —  0.5 (0.2–1.3)  —      01–3  0.6(0.4–1.1)  0.7 (0.4–1.4)  0.5 (0.3–0.9)  —  0.7 (0.4–1.2)  —      04–6  0.4 (0.2–0.9)  0.3 (0.1–0.8)  0.7 (0.3–1.3)  —  0.7 (0.4–1.3)  —  Knowledge score  1.2(1.1–1.3)  1.2(1.1–1.3)  1.3 (1.2–1.4)  1.2 (1.0–1.3)  1.2 (1.1–1.3)  1.1 (1.0–1.3)  Treated uniform prior to deployment  —  —   10.5 (7.4–15)  8.4 (5.9–12)  2.0(1.6–2.7)  2.1 (1.5–2.9)  No. of previous deployments  1.04 (0.9–1.2)  —  1.1 (1.0–1.2)  —  1.1 (1.0–1.2)  1.2(1.1–1.4)  Adjusted for age and gender. Survey 2 data. Knowledge score was generated as a composite measure of six “agree/disagree” survey questions. a Positive test for trend. View Large The logistic regression model for “uniform treatment while deployed” included treating uniform before deployment, command emphasis, military repellent use, and knowledge score (goodness-of-fit: χ2, p = 0.65). The most important predictor variable was treating the uniform before deployment (POR, 8.4; 95% confidence interval [CI], 5.9–12.0). Command emphasis was again an important predictor of uniform treatment (p < 0.006). Only the “military-only repellent use” variable significantly predicted uniform treatment while deployed (odds ratio [OR], 2.4; 95% CI, 1.7–3.3). The knowledge score showed the same effect as seen in uniform treatment before deployment (p = 0.007). The multivariable logistic regression model for use of any insect repellent (military, commercial, or both) included the following variables: branch of service, command emphasis, knowledge score, number of previous deployments, and having treated a uniform before deployment (goodness-of-fit: χ2, p = 0.37). In this study, all services reported less use of insect repellent than the Army (p < 0.03). Command emphasis was again important for this outcome (p < 0.003). Treating a uniform before deployment doubled the POR for reporting repellent use (p < 0.001). For each previous deployment, there was a 20% increase in POR of repellent use (p = 0.004). Knowledge score again increased POR of repellent use but only 10% for each step increase (p = 0.02). A total of 107 (5%) reported “regular or always use of DEET” among the 2,222 respondents to this question. This low DEET use was reported even though 72% reported that DEET was available when needed. No service had frequent DEET use (Table III). On univariable analysis, the statistically significant associated variables were unit type, previous deployment to AOR, primary duty location, uniform treatment, attitudes toward DEET, and smoking status (this effect was not consistent across levels of smoking and only significant at the one pack per day level). DEET availability was not associated with regular or always use of DEET. TABLE III. Univariable Analysis of Nondemographic Questions by Branch of Service   Army (%)  Air Force (%)  Marine Corps (%)  Navy (%)  p  Personal protective measure use                Regular/always DEET use (1)  5  2  5  3  0.5      No DEET use (1)  69  80  66  73  0.35      No use of insect repellent (2)  39  56  48  55  0.003      Use of military repellent only (2)  15  26  17  14  0.003      DEET available when needed (1)  73  83  72  66  0.1        Treated uniform with permethrin (1)  37  38  70  60  <0.001      Treated uniform prior to deployment (2)  31  59  68  61  <0.001      Treated uniform while deployed (2)  29  37  39  31  0.1  Personal protective measure perception                DEET is safe (1)  56  63  72  71  <0.001      DEET is effective (1)  70  80  70  77  0.01      Military repellent is 33% DEET (2)  82  91  82  87  0.5      33% DEET is better than 75% (2)  46  33  50  63  0.04      Command adequately emphasized repellent use (2)  40  48  47  30  0.01      Permethrin is for field uniform (2)  84  96  80  92  0.07      Permethrin is NOT for skin (2)  79  72  82  89  0.2  VBD Knowledge                Leishmaniasis is from insect bites (2)  85  83  79  87  0.3      Malaria is from insect bites (2)  92  96  93  96  0.6  Vector exposure                Never or rarely bitten by sand flies (1)  81  94  77  85  0.02      Never or rarely bitten by insects (1)  68  87  72  71  0.02      Never or one insect bite (2)  36  74  3%  69  <0.001    Army (%)  Air Force (%)  Marine Corps (%)  Navy (%)  p  Personal protective measure use                Regular/always DEET use (1)  5  2  5  3  0.5      No DEET use (1)  69  80  66  73  0.35      No use of insect repellent (2)  39  56  48  55  0.003      Use of military repellent only (2)  15  26  17  14  0.003      DEET available when needed (1)  73  83  72  66  0.1        Treated uniform with permethrin (1)  37  38  70  60  <0.001      Treated uniform prior to deployment (2)  31  59  68  61  <0.001      Treated uniform while deployed (2)  29  37  39  31  0.1  Personal protective measure perception                DEET is safe (1)  56  63  72  71  <0.001      DEET is effective (1)  70  80  70  77  0.01      Military repellent is 33% DEET (2)  82  91  82  87  0.5      33% DEET is better than 75% (2)  46  33  50  63  0.04      Command adequately emphasized repellent use (2)  40  48  47  30  0.01      Permethrin is for field uniform (2)  84  96  80  92  0.07      Permethrin is NOT for skin (2)  79  72  82  89  0.2  VBD Knowledge                Leishmaniasis is from insect bites (2)  85  83  79  87  0.3      Malaria is from insect bites (2)  92  96  93  96  0.6  Vector exposure                Never or rarely bitten by sand flies (1)  81  94  77  85  0.02      Never or rarely bitten by insects (1)  68  87  72  71  0.02      Never or one insect bite (2)  36  74  3%  69  <0.001  (1) Indicates question from survey 1 (N = 2,277); (2) indicates question from survey 2 (N = 1,169). See Table I for explanation of surveys. View Large TABLE III. Univariable Analysis of Nondemographic Questions by Branch of Service   Army (%)  Air Force (%)  Marine Corps (%)  Navy (%)  p  Personal protective measure use                Regular/always DEET use (1)  5  2  5  3  0.5      No DEET use (1)  69  80  66  73  0.35      No use of insect repellent (2)  39  56  48  55  0.003      Use of military repellent only (2)  15  26  17  14  0.003      DEET available when needed (1)  73  83  72  66  0.1        Treated uniform with permethrin (1)  37  38  70  60  <0.001      Treated uniform prior to deployment (2)  31  59  68  61  <0.001      Treated uniform while deployed (2)  29  37  39  31  0.1  Personal protective measure perception                DEET is safe (1)  56  63  72  71  <0.001      DEET is effective (1)  70  80  70  77  0.01      Military repellent is 33% DEET (2)  82  91  82  87  0.5      33% DEET is better than 75% (2)  46  33  50  63  0.04      Command adequately emphasized repellent use (2)  40  48  47  30  0.01      Permethrin is for field uniform (2)  84  96  80  92  0.07      Permethrin is NOT for skin (2)  79  72  82  89  0.2  VBD Knowledge                Leishmaniasis is from insect bites (2)  85  83  79  87  0.3      Malaria is from insect bites (2)  92  96  93  96  0.6  Vector exposure                Never or rarely bitten by sand flies (1)  81  94  77  85  0.02      Never or rarely bitten by insects (1)  68  87  72  71  0.02      Never or one insect bite (2)  36  74  3%  69  <0.001    Army (%)  Air Force (%)  Marine Corps (%)  Navy (%)  p  Personal protective measure use                Regular/always DEET use (1)  5  2  5  3  0.5      No DEET use (1)  69  80  66  73  0.35      No use of insect repellent (2)  39  56  48  55  0.003      Use of military repellent only (2)  15  26  17  14  0.003      DEET available when needed (1)  73  83  72  66  0.1        Treated uniform with permethrin (1)  37  38  70  60  <0.001      Treated uniform prior to deployment (2)  31  59  68  61  <0.001      Treated uniform while deployed (2)  29  37  39  31  0.1  Personal protective measure perception                DEET is safe (1)  56  63  72  71  <0.001      DEET is effective (1)  70  80  70  77  0.01      Military repellent is 33% DEET (2)  82  91  82  87  0.5      33% DEET is better than 75% (2)  46  33  50  63  0.04      Command adequately emphasized repellent use (2)  40  48  47  30  0.01      Permethrin is for field uniform (2)  84  96  80  92  0.07      Permethrin is NOT for skin (2)  79  72  82  89  0.2  VBD Knowledge                Leishmaniasis is from insect bites (2)  85  83  79  87  0.3      Malaria is from insect bites (2)  92  96  93  96  0.6  Vector exposure                Never or rarely bitten by sand flies (1)  81  94  77  85  0.02      Never or rarely bitten by insects (1)  68  87  72  71  0.02      Never or one insect bite (2)  36  74  3%  69  <0.001  (1) Indicates question from survey 1 (N = 2,277); (2) indicates question from survey 2 (N = 1,169). See Table I for explanation of surveys. View Large The multivariable logistic regression model for “DEET use” included the following predictor variables: age, gender, attitudes toward and availability of DEET, military operation supported (OIF vs. all others), and treatment of uniform with permethrin (goodness-of-fit: χ2, p = 0.99). The only predictors reaching statistical significance were treatment of uniform (strong positive association) and military operation supported (negatively associated for those supporting OEF/both/other) (Table IV). TABLE IV. Univariable and Multivariable Regression IRRs for Leishmaniasis and PORs for DEET Use and Uniform Treatment   Personal Protective Measures  VBD Exposure        Regression Covariates  DEET Use POR (1) (95% CI)  DEET Use POR (2) (95% CI)  Uniform Treatment POR (1) (95% CI)  Uniform Treatment POR (2) (95% CI)  Leishmaniasis POR (1) (95% CI)  Leishmaniasis IRR (3) (95% CI)a  Military component                  Regular  Reference  —  Reference  —  Reference  Reference      Reserve or guard  1.05 (0.7–1.6)  —  0.98 (0.8–1.2)  —  3.7 (1.3–10)  5.6(1.7–18)  Branch of service                  Army  Reference  —  Reference  Reference  Reference  Reference      Marines  0.95 (0.5–1.7)  —  3.8 (2.9–5.0)  4.2 (3.4–6.1)  3.0 (1.1–8.2)  4.5 (1.3–15)      Navy  0.57 (0.2–1.6)  —  2.6 (1.8–3.6)  2.5 (1.7–3.9)  1.1 (0.1–8.9)  NE      Air Force    0.35 (0.05–2.6)  —  1.04 (0.6–1.8)  1.3 (0.8–2.7)  8.5 (2.3–32)  32 (3.7–300)  Military operation                  OIF  Reference  Reference  Reference  —  Reference  —      OEF/both/other  0.54 (0.3–0.9)  0.49 (0.3–0.9)  0.83 (0.7–1.0)  —  1.5 (0.5–3.6)  —  Toilet facility                  Flush/chemical  Reference  —  Reference  —  Reference  Reference      All other  1.2(0.5–2.6)  —  1.7 (1.1–2.7)  —  3.6(1.2–11)  3.05 (1.0–12)  Treated uniform                  No  Reference  Reference  —  —  Reference  Reference      Yes  4.5 (2.8–7.2)  3.5 (2.3–5.7)  —  —  1.8 (0.7–4.2)  2.4 (0.8–7.2)  DEET use                  Never  —  —  Reference  Reference  Reference  Reference      Occasionally  —  —  3.3 (2.7–4.1)  3.4 (2.8–4.5)  0.32 (0.1–1.4)  0.51 (0.2–1.7)      Regularly  —  —  6.3 (3.8–10)  4.9 (3.3–9.8)  2.2 (0.5–9.5)        Always  —  —  7.0 (2.3–21)  14 (3.3–68)  0 (p = 0.7)    DEET available when needed  1.3 (0.8–2.1)  0.89 (0.6–1.6)  1.8 (1.5–2.2)  1.3 (1.1–1.8)  0.42 (0.2–1.0)  0.42 (0.2–1.2)  Believe DEET effective  1.9 (1.2–3.2)  1.5 (0.8–2.5)  1.6(1.3–1.9)  1.2(0.9–1.5)  0.97 (0.4–2.5)  —  Believe DEET safe  1.8 (1.1–2.8)  1.3 (0.8–2.2)  1.8 (1.5–2.2)  1.1 (0.9–1.4)  0.84 (0.3–2.0)  —    Personal Protective Measures  VBD Exposure        Regression Covariates  DEET Use POR (1) (95% CI)  DEET Use POR (2) (95% CI)  Uniform Treatment POR (1) (95% CI)  Uniform Treatment POR (2) (95% CI)  Leishmaniasis POR (1) (95% CI)  Leishmaniasis IRR (3) (95% CI)a  Military component                  Regular  Reference  —  Reference  —  Reference  Reference      Reserve or guard  1.05 (0.7–1.6)  —  0.98 (0.8–1.2)  —  3.7 (1.3–10)  5.6(1.7–18)  Branch of service                  Army  Reference  —  Reference  Reference  Reference  Reference      Marines  0.95 (0.5–1.7)  —  3.8 (2.9–5.0)  4.2 (3.4–6.1)  3.0 (1.1–8.2)  4.5 (1.3–15)      Navy  0.57 (0.2–1.6)  —  2.6 (1.8–3.6)  2.5 (1.7–3.9)  1.1 (0.1–8.9)  NE      Air Force    0.35 (0.05–2.6)  —  1.04 (0.6–1.8)  1.3 (0.8–2.7)  8.5 (2.3–32)  32 (3.7–300)  Military operation                  OIF  Reference  Reference  Reference  —  Reference  —      OEF/both/other  0.54 (0.3–0.9)  0.49 (0.3–0.9)  0.83 (0.7–1.0)  —  1.5 (0.5–3.6)  —  Toilet facility                  Flush/chemical  Reference  —  Reference  —  Reference  Reference      All other  1.2(0.5–2.6)  —  1.7 (1.1–2.7)  —  3.6(1.2–11)  3.05 (1.0–12)  Treated uniform                  No  Reference  Reference  —  —  Reference  Reference      Yes  4.5 (2.8–7.2)  3.5 (2.3–5.7)  —  —  1.8 (0.7–4.2)  2.4 (0.8–7.2)  DEET use                  Never  —  —  Reference  Reference  Reference  Reference      Occasionally  —  —  3.3 (2.7–4.1)  3.4 (2.8–4.5)  0.32 (0.1–1.4)  0.51 (0.2–1.7)      Regularly  —  —  6.3 (3.8–10)  4.9 (3.3–9.8)  2.2 (0.5–9.5)        Always  —  —  7.0 (2.3–21)  14 (3.3–68)  0 (p = 0.7)    DEET available when needed  1.3 (0.8–2.1)  0.89 (0.6–1.6)  1.8 (1.5–2.2)  1.3 (1.1–1.8)  0.42 (0.2–1.0)  0.42 (0.2–1.2)  Believe DEET effective  1.9 (1.2–3.2)  1.5 (0.8–2.5)  1.6(1.3–1.9)  1.2(0.9–1.5)  0.97 (0.4–2.5)  —  Believe DEET safe  1.8 (1.1–2.8)  1.3 (0.8–2.2)  1.8 (1.5–2.2)  1.1 (0.9–1.4)  0.84 (0.3–2.0)  —  1 = univariable analysis; 2 = multivariable logistic regression analysis; 3 = multivariable Poisson regression analysis. Toilet facility “other” comprised of: burn latrine, bag latrine, slit trench, or none. NE, Not evaluable. Survey 1 data. a Adjusted for age, gender, and country of service, others age and gender adjusted. View Large TABLE IV. Univariable and Multivariable Regression IRRs for Leishmaniasis and PORs for DEET Use and Uniform Treatment   Personal Protective Measures  VBD Exposure        Regression Covariates  DEET Use POR (1) (95% CI)  DEET Use POR (2) (95% CI)  Uniform Treatment POR (1) (95% CI)  Uniform Treatment POR (2) (95% CI)  Leishmaniasis POR (1) (95% CI)  Leishmaniasis IRR (3) (95% CI)a  Military component                  Regular  Reference  —  Reference  —  Reference  Reference      Reserve or guard  1.05 (0.7–1.6)  —  0.98 (0.8–1.2)  —  3.7 (1.3–10)  5.6(1.7–18)  Branch of service                  Army  Reference  —  Reference  Reference  Reference  Reference      Marines  0.95 (0.5–1.7)  —  3.8 (2.9–5.0)  4.2 (3.4–6.1)  3.0 (1.1–8.2)  4.5 (1.3–15)      Navy  0.57 (0.2–1.6)  —  2.6 (1.8–3.6)  2.5 (1.7–3.9)  1.1 (0.1–8.9)  NE      Air Force    0.35 (0.05–2.6)  —  1.04 (0.6–1.8)  1.3 (0.8–2.7)  8.5 (2.3–32)  32 (3.7–300)  Military operation                  OIF  Reference  Reference  Reference  —  Reference  —      OEF/both/other  0.54 (0.3–0.9)  0.49 (0.3–0.9)  0.83 (0.7–1.0)  —  1.5 (0.5–3.6)  —  Toilet facility                  Flush/chemical  Reference  —  Reference  —  Reference  Reference      All other  1.2(0.5–2.6)  —  1.7 (1.1–2.7)  —  3.6(1.2–11)  3.05 (1.0–12)  Treated uniform                  No  Reference  Reference  —  —  Reference  Reference      Yes  4.5 (2.8–7.2)  3.5 (2.3–5.7)  —  —  1.8 (0.7–4.2)  2.4 (0.8–7.2)  DEET use                  Never  —  —  Reference  Reference  Reference  Reference      Occasionally  —  —  3.3 (2.7–4.1)  3.4 (2.8–4.5)  0.32 (0.1–1.4)  0.51 (0.2–1.7)      Regularly  —  —  6.3 (3.8–10)  4.9 (3.3–9.8)  2.2 (0.5–9.5)        Always  —  —  7.0 (2.3–21)  14 (3.3–68)  0 (p = 0.7)    DEET available when needed  1.3 (0.8–2.1)  0.89 (0.6–1.6)  1.8 (1.5–2.2)  1.3 (1.1–1.8)  0.42 (0.2–1.0)  0.42 (0.2–1.2)  Believe DEET effective  1.9 (1.2–3.2)  1.5 (0.8–2.5)  1.6(1.3–1.9)  1.2(0.9–1.5)  0.97 (0.4–2.5)  —  Believe DEET safe  1.8 (1.1–2.8)  1.3 (0.8–2.2)  1.8 (1.5–2.2)  1.1 (0.9–1.4)  0.84 (0.3–2.0)  —    Personal Protective Measures  VBD Exposure        Regression Covariates  DEET Use POR (1) (95% CI)  DEET Use POR (2) (95% CI)  Uniform Treatment POR (1) (95% CI)  Uniform Treatment POR (2) (95% CI)  Leishmaniasis POR (1) (95% CI)  Leishmaniasis IRR (3) (95% CI)a  Military component                  Regular  Reference  —  Reference  —  Reference  Reference      Reserve or guard  1.05 (0.7–1.6)  —  0.98 (0.8–1.2)  —  3.7 (1.3–10)  5.6(1.7–18)  Branch of service                  Army  Reference  —  Reference  Reference  Reference  Reference      Marines  0.95 (0.5–1.7)  —  3.8 (2.9–5.0)  4.2 (3.4–6.1)  3.0 (1.1–8.2)  4.5 (1.3–15)      Navy  0.57 (0.2–1.6)  —  2.6 (1.8–3.6)  2.5 (1.7–3.9)  1.1 (0.1–8.9)  NE      Air Force    0.35 (0.05–2.6)  —  1.04 (0.6–1.8)  1.3 (0.8–2.7)  8.5 (2.3–32)  32 (3.7–300)  Military operation                  OIF  Reference  Reference  Reference  —  Reference  —      OEF/both/other  0.54 (0.3–0.9)  0.49 (0.3–0.9)  0.83 (0.7–1.0)  —  1.5 (0.5–3.6)  —  Toilet facility                  Flush/chemical  Reference  —  Reference  —  Reference  Reference      All other  1.2(0.5–2.6)  —  1.7 (1.1–2.7)  —  3.6(1.2–11)  3.05 (1.0–12)  Treated uniform                  No  Reference  Reference  —  —  Reference  Reference      Yes  4.5 (2.8–7.2)  3.5 (2.3–5.7)  —  —  1.8 (0.7–4.2)  2.4 (0.8–7.2)  DEET use                  Never  —  —  Reference  Reference  Reference  Reference      Occasionally  —  —  3.3 (2.7–4.1)  3.4 (2.8–4.5)  0.32 (0.1–1.4)  0.51 (0.2–1.7)      Regularly  —  —  6.3 (3.8–10)  4.9 (3.3–9.8)  2.2 (0.5–9.5)        Always  —  —  7.0 (2.3–21)  14 (3.3–68)  0 (p = 0.7)    DEET available when needed  1.3 (0.8–2.1)  0.89 (0.6–1.6)  1.8 (1.5–2.2)  1.3 (1.1–1.8)  0.42 (0.2–1.0)  0.42 (0.2–1.2)  Believe DEET effective  1.9 (1.2–3.2)  1.5 (0.8–2.5)  1.6(1.3–1.9)  1.2(0.9–1.5)  0.97 (0.4–2.5)  —  Believe DEET safe  1.8 (1.1–2.8)  1.3 (0.8–2.2)  1.8 (1.5–2.2)  1.1 (0.9–1.4)  0.84 (0.3–2.0)  —  1 = univariable analysis; 2 = multivariable logistic regression analysis; 3 = multivariable Poisson regression analysis. Toilet facility “other” comprised of: burn latrine, bag latrine, slit trench, or none. NE, Not evaluable. Survey 1 data. a Adjusted for age, gender, and country of service, others age and gender adjusted. View Large The prevalence of uniform treatment with permethrin for survey 1 respondents was 43% (969 of 2,237 respondents). The Marines and Navy were more likely than the Air Force and Army to report having permethrin-treated uniforms. On univariable analysis, the statistically significant associated variables were branch of service, warrant officer status (OR = 0.09), attitudes toward, availability of and use of DEET, and smoking status (more than one pack per day). Although lower smoking levels did not reach statistical significance, there was a significant test for trend. The multivariable logistic regression model for “treatment of uniform with permethrin” included the following predictor variables: age, gender, the DEET available, safe and effective variables, use of DEET, branch of service, and cigarette smoking before deployment (goodness-of-fit: χ2, p = 0.42). Including the smoking variable limited the model to 585 observations; therefore, this was rerun without the smoking variable for 1,780 observations. Use of DEET predicted treatment of uniform significantly across all levels of DEET use, with a positive trend (Table IV). Vector Exposure More OEF respondents reported never or rare/one insect bite than OIF respondents across both surveys (survey 1: 77% vs. 67% and survey 2: 60% vs. 33%, p < 0.001 for both). OEF respondents also reported fewer sand fly bites. Analyzing insect bites by country of deployment (survey 2), 30% of Iraq respondents reported being bitten daily or almost daily compared to 12% for Afghanistan and 7% in Kuwait (p < 0.001). Army members were least likely to report never or rare/one insect bite. Both insect and fly bites were more frequent with increasing use of DEET/repellent on univariable analysis. This relationship held in the multivariable models and across both surveys (data not shown). The age- and gender-adjusted Poisson model for sand fly bites resulted in only one significant association: “regular/always DEET use” (incidence rate ratio [IRR]: 1.8, p = 0.003). The corresponding insect bite model resulted in the following associations: “regular/always DEET use” (IRR: 1.5, p = 0.01), branch of service (Navy vs. Army, IRR: 1.7, p = 0.005), rank/grade (E7–9 vs. E1–4, IRR: 0.6, p = 0.01); and unit (support and command vs. ground, IRR: 0.8 and 0.6 respectively, p = 0.03 each). Vector-Borne Disease Outcome Twenty-two (1%) of 2,179 respondents reported developing CL during this deployment. The full-time incidence density of CL was 1.8 per 100 person-years (95% CI, 1.2–2.8). Half-time incidence density was 3.7 per 100-person years (95% CI, 2.4–5.6). Upon univariable analysis, the statistically significant associated variables were military component, branch of service, type of toilet facilities, and being frequently bitten by sand flies. DEET use and availability, treatment of uniform with permethrin, and attitudes toward DEET were not significant on univariable analysis. DEET availability did approach significance, those without access had a disease OR of 2.4 (p = 0.06). The multivariable Poisson regression model evaluating self-reported leishmaniasis incidence included age, gender, DEET use and availability, uniform treatment with permethrin, military component, type of toilet facility, country of service and branch of service and is shown in Table IV (goodness-of-fit test: χ2, p = 1.0). Reserve and National Guard members had more than a 5-fold increase in incidence rate (IRR: 5.6, p = 0.004) compared to regular members. Compared to the Army, the Air Force personnel (IRR: 32.8, p = 0.002) and Marines (IRR: 4.5, p = 0.02) had a higher incidence rate of self-reported leishmaniasis. Both DEET availability and use (any level compared to none) tended toward a protective effect but neither achieved statistical significance (IRR: 0.4, p = 0.1 and 0.5, p = 0.3, respectively). Neither the insect bite nor the sand fly bite variables were significantly associated with the disease outcome. DISCUSSION Prevention of insect vector bites remains the best method to reduce the burden of VBD on the military and military health care system. At the base-wide level, this effort includes vector control measures as described by Coleman et al.12 and Croft et al.9 At the individual level, education on the safety and efficacy of DEET and permethrin-treated uniforms, emphasis on proper and regular use of PPM, and availability of supplies is crucial. Our results reveal the importance of “command emphasis” on subsequent troop use of PPM. This study failed to detect a statistically significant protective association between PPM and CL. The most likely reason for this is the relatively few numbers of individuals actually reporting regular or always use of PPM and the rarity of CL reported. Previous prospective controlled studies have shown a protective benefit of PPM and this study does not cast doubt on their validity.7,10 The lack of significant association between DEET availability and use is likely attributable to the low overall reported use and high overall reported availability, and thus the study was probably underpowered to detect an epidemiologically important difference. We report here service-specific differences in utilization of personal protective measures with the Navy and Marine Corps demonstrating higher rates of use compared to the other services. These differences may be due to differences in the service-specific populations, health promotion messages, and/or service-specific attitudes and behaviors or by chance. An additional factor to consider: uniform treatment with permethrin is a unit level function in the Marine Corps and Navy, while it is generally a member level responsibility in the Air Force and the Army. Nearly 10 times the people reported treatment of uniforms with permethrin than reported regular or always DEET use. Uniform treatment with permethrin still remained far short of 100%. This large difference between uniform treatment and Deet use is understandable in that a uniform needs to be treated once and the protective effect is considered good for the life of the uniform. Use of DEET is a behavior that requires greater effort with multiple daily applications. An examination of Navy and Marine Corps procedures is indicated since they are the most successful in getting uniforms treated with permethrin. This may explain the finding that fewer Army and Air Force members reported uniform treatment than Navy and Marine Corps members across both surveys. Further examination of uniform treatment procedures is warranted to improve force health protection from VBD. To mitigate the burden of VBD, a redoubling of efforts is needed to increase the use of DEET and treatment of uniforms with permethrin. Further research to affect modifiable factors associated with PPM use, i.e., attitudes, knowledge, and beliefs, is indicated. Our results show that individuals that use DEET more often are also more likely to have treated uniforms, suggesting their belief in the effectiveness of PPM against VBD. They also tend to have higher “knowledge scores” and more past deployments. The association of higher insect/sand fly bites with increasing level of DEET/repellent use was not expected. Because this study in cross-sectional in nature, we cannot assess any causal association. It is unlikely that repellent use causes vector exposure, rather it is more likely that those with higher vector exposure use more repellent. CL is a risk for deployed personnel in the current operational regions of Iraq and Afghanistan and a marker for the larger VBD category. Our estimates of CL incidence density go from as low as 1.8 to as high as 3.7 cases per 100 person-years. Although our person-years calculation has its limitations, including the self-reported diagnosis, it is the first time that CL incidence density has been reported for all U.S. military components. We do believe that our person-time is most likely an overestimation since cases are unlikely to all present on the day of survey; therefore, our incidence density is likely a conservative estimate of the true value. This incidence density is 10 times greater than that recently reported in another article that examined regular component U.S. military members between 2001 and 2006 in Iraq and Afghanistan.16 However, our data included National Guard and Reserve components which revealed a higher incidence rate than the regular component addressed in this earlier publication. Seventy-seven percent of the cases we found occurred in Guard or Reserve troops. Also, we used self-report which is likely to be more sensitive than the medical surveillance system data in the previous report.16 Anecdotally, this reporting bias is supported in discussions with a number of soldiers who stated that they did not seek care for CL because it was not very painful, treatment was considered worse than the disease, and they did not want to leave their units (J. P. Vickery, personal communication). The findings of this study must be interpreted while understanding its limitations. Our primary outcome was self-reported leishmaniasis, which is not as accurate as physician-diagnosed disease. Self-reported cases avoid the case finding problem of data access seen in the Army Medical Surveillance Activity report.16 Self-reporting lends itself to bias and misclassification. The missing time data are another limitation; however, our solution to this problem follows a reasonable and accepted process. On the other hand, this is the largest study of leishmaniasis incidence reported to date and its results are in-line with those of previous incidence studies, with the exception of the report by Aliaga and Aronson16 and Aronson.17 VBD has and continues to be a threat to deployed forces around the world. The control of these diseases should be seen as an integrated approach which combines best practices in environmental modification, personal protective behavior, and effective treatment. This study found that there is definite room for improvement in the area of personal protective measures. Strategies that successfully supply military members with permethrin-treated uniforms should be examined and disseminated. Interventions based on the Health Belief model may be useful to increase future PPM use. One initial step indicated by our data is to increase command emphasis on the use of insect repellent. ACKNOWLEDGMENTS We thank the following individuals for their assistance with statistical modeling and data management: Cara H. Olsen, MS DrPH, Ann I. Scher, PhD, Hanan Raafat, Yasmine Farid, Mohamed Fakhry, Noha Effat, Manal Moustafa, and Jamie Bland. This study was conducted under the support of the Military Infectious Disease Research Program and DoD-GEIS funding. REFERENCES 1. Riddle MS, Althoff JM, Earhart K et al. Serological evidence of arbo-viral infection and self-reported febrile illness among U.S. troops deployed to Al Asad, Iraq. Epidemiol Infect  2007; 25: 1– 5. 2. Weina PJ, Neafie RC, Wortmann G, Polhemus M, Aronson NE Old world leishmaniasis: an emerging infection among deployed U.S. military and civilian workers. Clin Infect Dis  2004; 39: 1674– 80. Google Scholar CrossRef Search ADS PubMed  3. Aronson N, Coleman R, Coyne P, et al.   Cutaneous leishmaniasis in U.S. military personnel-Southwest/Central Asia, 2002–2003. MMWR Morb Mortal Wkly Rep  2003; 52: 1009– 12. Google Scholar PubMed  4. Jappe U Unusual skin infections in military personnel. Clin Dermatol  2002; 20: 425– 34. Google Scholar CrossRef Search ADS PubMed  5. Aronson NE, Sanders JW, Moran KA In harm's way: infections in deployed American military forces. Clin Infect Dis  2006; 43: 1045– 51. Google Scholar CrossRef Search ADS PubMed  6. Cheng M UN health officials: skin diseases rising in southern Iraq with children at risk. North County Times , March 2, 2008, p 12. 7. Asilian A, Sadeghinia A, Shariati F, Imam Jorne M, Ghoddusi A Efficacy of permethrin-impregnated uniforms in the prevention of cutaneous leishmaniasis in Iranian soldiers. J Clin Pharm Ther  2003; 28: 175– 8. Google Scholar CrossRef Search ADS PubMed  8. Soto J, Medina F, Dember N, Berman J Efficacy of permethrin-impregnated uniforms in the prevention of malaria and leishmaniasis in Colombian soldiers. Clin Infect Dis  1995; 21: 599– 602. Google Scholar CrossRef Search ADS PubMed  9. Croft AM, Baker D, von Bertele MJ An evidence-based vector control strategy for military deployments: the British Army experience. Med Trop (Mars)  2001; 61: 91– 8. Google Scholar PubMed  10. Schreck CE, Kline DL, Chaniotis BN, Wilkinson N, McGovern TP, Weidhaas DE Evaluation of personal protection methods against phlebotomine sand flies including vectors of leishmaniasis in Panama. Am J Trop Med Hyg  1982; 31: 1046– 53. Google Scholar CrossRef Search ADS PubMed  11. Sanders JW, Putnam SD, Frankart C, et al.   Impact of illness and non-combat injury during Operations Iraqi Freedom and Enduring Freedom (Afghanistan). Am J Trop Med Hyg  2005; 73: 713– 9. Google Scholar PubMed  12. Coleman RE, Burkett DA, Putnam JL, et al.   Impact of phlebotomine sand flies on U.S. Mlitary operations at Tallil Air Base, Iraq: 1. Background, military situation, and development of a “Leishmaniasis Control Program.” J Med Entomol  2006; 43: 647– 62. Google Scholar PubMed  13. Little RJ, Rubin DB Causal effects in clinical and epidemiological studies via potential outcomes: concepts and analytical approaches. Annu Rev Public Health  2000; 21: 121– 45. Google Scholar CrossRef Search ADS PubMed  14. Mander A, Clayton D. Update to Hotdeck Imputation . College Station, TX, Stata Press, 2000. 15. Woodward M Epidemiology: Study Design and Data Analysis , Ed 2. Boca Raton, FL: Chapman & Hall/CRC Press; 2005. 16. Aliaga P, Aronson N Leishmaniasis in relation to service in Iraq/ Afghanistan, U.S Armed Forces, 2001–2006. Med Surveill Monthly Rep  2007; 14: 2– 5. 17. Aronson NE Leishmaniasis in American soldiers: parasites from the front. In: Emerging Infections , pp 325– 42. Edited by Scheid W, Hooper D, Hughes J Washington, DC, ASM Press, 2007. Footnotes 3 This is a primary data analysis of survey responses collected from U.S. military personnel deployed to Iraq, Afghanistan, and the surrounding region under NAMRU-3 DoD approved protocol. There are no personal identifiers included in these data. Surveys were limited to one page and were completely voluntary and anonymous. This study received institutional review board exemption from the Uniformed Services University of the Health Sciences. Reprint & Copyright © Association of Military Surgeons of the U.S.
TI - Factors Associated with the Use of Protective Measures against Vector-Borne Diseases among Troops Deployed to Iraq and Afghanistan
JF - Military Medicine
DO - 10.7205/MILMED.173.11.1060
DA - 2008-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/factors-associated-with-the-use-of-protective-measures-against-vector-DLZtprYfou
SP - 1060
EP - 1067
VL - 173
IS - 11
DP - DeepDyve
ER -