Share this article on:

Disaster-Related Exposures and Health Effects Among US Coast Guard Responders to Hurricanes Katrina and Rita: A Cross-Sectional Study

Rusiecki, Jennifer A. PhD; Thomas, Dana L. MD, MPH; Chen, Ligong MD, MPH; Funk, Renée DVM, MPH&TM; McKibben, Jodi PhD; Dayton, Melburn R. MS

Journal of Occupational & Environmental Medicine: August 2014 - Volume 56 - Issue 8 - p 820–833
doi: 10.1097/JOM.0000000000000188
Original Articles

Objective: Disaster responders work among poorly characterized physical and psychological hazards with little understood regarding health consequences of their work.

Methods: A survey administered to 2834 US Coast Guard responders to Hurricanes Katrina and Rita provided data on exposures and health effects. Prevalence odds ratios (PORs) evaluated associations between baseline characteristics, missions, exposures, and health effects.

Results: Most frequent exposures were animal/insect vector (n = 1309; 46%) and floodwater (n = 817; 29%). Most frequent health effects were sunburn (n = 1119; 39%) and heat stress (n = 810; 30%). Significant positive associations were for mold exposure and sinus infection (POR = 10.39); carbon monoxide and confusion (POR = 6.27); lack of sleep and slips, trips, falls (POR = 3.34) and depression (POR = 3.01); being a Gulf-state responder and depression (POR = 3.22).

Conclusions: Increasing protection for disaster responders requires provisions for adequate sleep, personal protective equipment, and access to medical and psychological support.

From the Department of Preventive Medicine and Biometrics (Drs Rusiecki, Chen, and McKibben), Uniformed Services University of the Health Sciences, Bethesda, Md; Centers for Disease Control and Prevention (Dr Thomas), National Center for Emerging Zoonotic and Infectious Diseases, Division of Vector Borne Diseases, Dengue Branch, San Juan, Puerto Rico; United States Coast Guard (Drs Rusiecki and Thomas and Mr Dayton), Directorate of Health, Safety, and Work Life, Washington, DC; Centers for Disease Control and Prevention (Dr Funk), National Institute for Occupational Safety and Health, Emergency Preparedness and Response Office, Atlanta, Ga; and Department of Psychology (Dr McKibben), West Chester University, Pa.

Address correspondence to: Jennifer A. Rusiecki, PhD, Department of Preventive Medicine and Biometrics, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Rd, Bethesda, MD 20814 (Jennifer.rusiecki@usuhs.edu).

The findings and conclusions expressed in this scientific article are those of the authors and do not reflect the official policy or positions of the Uniformed Services University, the Department of Defense, the United States Coast Guard, the Department of Homeland Security, or the Centers for Disease Control and Prevention.

The authors declare no conflicts of interest.

Hurricane Katrina, which struck the Gulf Coast on August 29, 2005, was the most destructive natural disaster in US history.1 More than 1300 people lost their lives, more than a million people were driven from their homes, widespread damage stretched more than a 90,000 mile2 area, and millions of gallons of oil were spilled into the environment.1 The US Coast Guard played a key role in the planning, response, and recovery efforts, carrying out missions such as search and rescue and marine pollution response. Coast Guard rescue operations commenced as the hurricane moved along the Gulf Coast to major population centers in Louisiana, Alabama, and Mississippi and as it moved inland. Of the estimated 60,000 people left stranded by Hurricane Katrina, more than 33,500 were rescued by the Coast Guard.1,2 Only a few weeks after Hurricane Katrina struck, Hurricane Rita made landfall in Texas on September 24. In Louisiana, the storm surge from Rita inundated low-lying communities near the coast, worsening effects caused by Hurricane Katrina. More than 5600 Coast Guard personnel participated in the initial response and aftermath restoration to the areas affected by Katrina and Rita.1

The environmental challenges presented by Katrina and Rita were vast, and the potential for environmental exposures for the Katrina/Rita responders significant. During September and October of 2005, experts from the Centers for Disease Control and Prevention's National Institute for Occupational Safety and Health were deployed to New Orleans and Baton Rouge to assist federal, state, and local agencies in addressing occupational safety and health issues. A health hazard evaluation report was published, which addressed occupational exposure issues faced by emergency responders and the public in the wake of Katrina and Rita.3 Various potential chemical exposure and safety hazards were identified, including exposures to floodwaters, dust and dried flood sediment, carbon monoxide, mold, and physical safety hazards. In addition, given the high pace of the response and resulting long hours for responders, as well as the intensity of heat during the early weeks of the response, mental and physical stressors were potentially significant influences on the health and well-being of responders.

In an effort to better document these potential hazards and to understand adverse health effects experienced by the responders, the Coast Guard Office of Safety & Environmental Health in the Directorate of Health, Safety and Work-Life designed and launched the Significant Event/Acute Exposure Tracking Tool soon after the response began. This Coast Guard intranet-accessible tool, hereafter referred to as “the responder exit survey” or “the survey,” served as a registry for personnel to document their exposures and adverse health effects (ie, illnesses, injuries, and symptoms), while providing a mechanism for occupational health surveillance and for service members to access immediate required/desired health care through the Coast Guard. The data gathered provided a unique opportunity to describe the self-reported exposure/health effect experiences of the Katrina/Rita Coast Guard responders and to investigate possible associations between exposures and adverse health effects experienced during the response. More than 2800 Coast Guard members completed the responder exit survey. Although this count represents approximately 51% of the total estimated responder population, the data obtained from the survey provide some much needed insight into exposures and health of first responders during disaster response and may ultimately assist in understanding how to support frontline responders during disasters and aid in prevention planning.

Back to Top | Article Outline

METHODS

Study Population

Subjects for this cross-sectional study were all Coast Guard members who responded to the Hurricane Katrina and Rita disasters (hereafter referred to as “Katrina/Rita responders”) between 27 August, 2005, and 12 November, 2006, and who completed the responder exit survey. This survey captured self-reported deployment information, such as time and location of deployment, mission(s) performed, exposures, health effects, medical treatment sought, and lifestyle information. The Coast Guard issued a Coast Guard–wide communication (ALCOAST 546-05), which provided a link to the survey and stated that it was required for all Katrina and Rita responders to fill out. For the small number of responders who had multiple deployments and completed a survey for each deployment, we combined their survey responses by using an inclusive system such that positive responses (ie, to an exposure, health effect, etc) from any of the surveys resulted in a positive response for the combined survey data. This study was approved by the institutional review boards at the Uniformed Services University of the Health Sciences and the United States Coast Guard.

Back to Top | Article Outline

Exposure Assessment

Exposure was assessed via self-report, on the basis of responses from the survey. Participants indicated ever having been involved in various missions, for example, general job categories, during their response or having been exposed to various environmental stressors. A responder could have been involved in more than one of the following missions during their response: search and rescue (SAR) air, SAR boat, SAR ground, hazardous oils response, firefighting, building clean up, body recovery, administrative work, incident command system (ICS) support, work life support, law enforcement, medical/dental support, safety and environmental health support, rescue swimmer, and “other” missions (specified via free text). Regarding specific, self-reported environmental exposures, those were assessed via indicating ever or never exposed to the following five exposures: (1) animal or insect bite/scratch (hereafter referred to as “animal/insect vector”; subcategories were dog bite/scratch, mosquito bite, fire ant bite, cat bite/scratch, spider bite, rodent bite/contact, snake bite, and other), (2) carbon monoxide (CO), (3) floodwater (subcategories were exposure via submersion, via ingestion, and via inhalation), (4) infectious agents (subcategories were exposure to deceased body, exposure to human waste, exposure to animal carcass, exposure to bloodborne pathogen, and other), and (5) mold exposure (subcategories were as manifested via mild respiratory symptoms, as manifested via severe respiratory symptoms). Participants were asked to report all that applied, so that there could be multiple exposures reported for each responder and within each broad exposure category, multiple subcategories reported.

In this study, we used two major exposure metrics: (1) individual mission (which we also categorized into higher-intensity missions [ie, operational-based missions], lower-intensity missions [ie, administrative-based missions], and other missions [as indicated by “other” in the exit survey]) and (2) self-reported exposure (ever/never) of the five general exposure categories given earlier. We categorized each self-reported exposure into ever/never, on the basis of responses to all subcategories of each of the exposures. For example, if someone reported ever having a fire ant bite but no other animal/insect vector exposures, he or she would be assigned a level of “ever” for animal/insect vector exposure. Although we do report some descriptive statistics on the various subcategories of each self-reported exposure, our analyses of exposure–health effect relationships focused on the five broadly defined exposures given earlier.

Back to Top | Article Outline

Assessment of Health Effects

Health effects, defined as self-reported illnesses, injuries, or symptoms, were reported by participants in the survey. For our analyses, we focused on the following self-reported health effects: confusion, dehydration, depression, diarrhea, digestive symptoms (nausea, vomiting, or loss of appetite), difficulty concentrating, heat stress, infected skin, joint pain, muscle strain, sinus infection, skin rash, penetrating injury, slips, trips, and falls, and sunburn. Participants were asked to report all that applied. Each health effect was recorded as ever/never. Responders were also queried on reasons for which they sought medical treatment. There were nine categories on the exit survey of reasons for which medical treatment was sought: animal/insect vector exposure, CO exposure, floodwater exposure, infectious agent exposure, respiratory symptoms deemed to be associated with mold exposure, sunburn, heat stress symptoms, penetrating injuries, and injuries from slips, trips, and falls.

Back to Top | Article Outline

Other Risk Factors Assessed

We also assessed other potential risk factors on the likelihood of reporting health effects and reasons for seeking medical treatment. These included demographic and lifestyle factors, use of personal protective equipment (PPE), duration of deployment/response, and timing of deployment/response. All of these factors were derived from the responder exit survey, as well as linkage to the Coast Guard's personnel administrative system, Direct Access.

Back to Top | Article Outline

Statistical Analyses

Chi-square tests were used to assess the association between exposures and specific type of PPE used. To investigate specific risk factor–health effects relationships cross-sectionally, we used logistic regression analyses in which the dependent variable was a specific health effect (ever/never) and the independent variable was a specific baseline factor (eg, demographic, lifestyle, PPE use, etc), mission, or exposure (ever/never) to calculate prevalence odds ratios (PORs) and 95% confidence intervals (CIs). In the analyses using a baseline factor as the main independent variable, we adjusted models for the five self-reported exposures. In the analyses using mission as the main independent variable, we adjusted models by age (categorized at the median, 31 years), sex, average hours of sleep per night (less than or equal to 5 hours; more than 5 hours), responder residence (“local” responders, ie, residing in any of the five Gulf States at the time of the hurricanes, and “nonlocal” responders), and tobacco use (yes/no). In the analyses using a particular self-reported exposure as the main independent variable, we adjusted each model for the four other self-reported exposures, as well as the covariates listed earlier. For a few of the health effects, that is, heat stress, dehydration, muscle strain, and sunburn, we did not carry out logistic regressions investigating the associations with the various self-reported exposures. Rather, we felt it made more sense to evaluate those with respect to baseline factors and mission(s) performed. Using medical treatment as a proxy for severity of exposure or health effect, we calculated PORs and 95% CIs for models in which the dependent variable was medical treatment sought and the independent variable was a baseline factor, a mission, a self-reported exposure, or a self-reported health effect. Those models were not adjusted for any other variables, because the number of people reporting medical treatment for any of the nine reasons queried in the exit survey was small.

We also carried out stratified analyses of the self-reported exposure models, on the basis of the following strata: age (31 years or younger; older than 31 years), sex, average amount of sleep (more than 5 hours; 5 hours or less), days off during deployment (1 in 7 or 1 in 10; none or 1 in 14), tobacco use (ever/never), nonlocal/local responder status, active duty/reserve status of responder, enlisted/officer status of responder, intensity of mission (low risk, eg, administrative mission; high risk, eg, operational), PPE use (ever/never), duration of deployment (30 days or less; more than 30 days), and timing of deployment start (August/September 2005; any time after September 2005). We tested the interaction of each of these stratifying variables with the main effect via a multiplicative model and considered a P value significant if it was less 0.05.

Back to Top | Article Outline

RESULTS

The ALCOAST 546-05 message required all Coast Guard Katrina/Rita responders to complete the survey; however, there was no mechanism to ensure 100% compliance, and the response rate among Katrina/Rita responders was approximately 51% (n = 2834). This is approximate, because the exact number of Coast Guard Katrina/Rita responders is unknown but was estimated by a Government Accountability Office Report to be approximately 5600.1 Table 1 presents baseline characteristics of the study population. The mean age was 32.4 (SD = 8.6) years. Men comprised 90% (n = 2541) of the responders in this study, females 10% (n = 281). The vast majority (95%; n = 2695) of responders in this study performed one deployment, and 5% (n = 139) performed two or more deployments during the period between 27 August, 2005, and 12 November, 2006. Although Hurricane Katrina made landfall on 29 August, some responders were on scene in advance of the disaster, thus the period start date was earlier than August 29. Nearly two thirds of the population performed response work for less than 1 month. Most responders (n = 2407; 85%) started their deployment in August or September, 2005. A large majority of the study population comprised Coast Guard active duty (83%; n = 2358) and reserve (11%; n = 326) personnel. Another approximately 5% (n = 150) were civilians (eg, federal employees, contractors, Coast Guard Auxiliary, and other). Enlisted personnel comprised approximately 75% of the military study population; officers (including warrant officers), 25%. Approximately 50% of the study population resided in states along the coast of the Gulf of Mexico (eg, Louisiana, Alabama, Mississippi, Texas, and Florida, thus “local”); the other approximately 50% were deployed from states beyond the Gulf Coast, thus “nonlocal.” About one third (32%; n = 897) reported sleeping an average of 5 or fewer hours per night during their deployment (n = 897). Time off was categorized into 1 in 7 or 1 in 10 days off (41%; n = 1174) and 1 in 14 days off or no days off during deployment (42%; n = 1193). One third of the population reported using tobacco products, most of whom reported smoking cigarettes only.

Data on the various missions carried out by Coast Guard responders are presented in Table 2. Because each responder could have been involved in more than one mission during their deployment/response, the total count for missions does not add up to the total count of responders in the study population. The most common missions performed by the study population were ICS support (13%; n = 668), SAR boat (12%; n = 622), law enforcement (11%; n = 568), and building cleanup (10%; n = 517). Fewer than 100 people each reported having carried out the following missions: work life (n = 88), rescue swimmer (n = 73), body recovery (n = 70), safety and environmental health (n = 42), and firefighting (n = 14). We categorized all the missions reported into “lower intensity” (eg, support and administrative work–related), “higher intensity” (eg, operational), and “other missions.” Table 2 provides additional detail on these definitions. Briefly, 37% (n = 1057) of the study population were involved in lower-intensity missions, 46% (n = 1299) were involved in higher-intensity missions, the rest reported “other” missions, which were specified via free text in the survey and were not further analyzed in this study. More than half of the responders (52%) reported carrying out only one mission, 26% reported carrying out two missions, and 22% reported carrying out three or more missions.

Figure 1 presents the number of exposures (of five general categories assessed via the responder exit survey) responders reported. Among all responders in the study, 38% reported no exposures of the five exposures queried. Approximately 25% reported one exposure only. Figure 2 presents the distribution of specific exposures. Because a responder could indicate multiple exposures during a given deployment, the counts for this figure sum to more than the total number of responders. The most commonly reported exposure by responders was animal/insect vector (n = 1309; 46%), 73% of which were from mosquito bites and 16% of which were from fire ant bites. The next most commonly reported exposure was floodwater exposure (n = 817; 29%); approximately 75% reported submersion in floodwaters, and 25% reported floodwater exposure from ingestion/inhalation. Infectious agents were reported by 749 people; 40% of those were exposed to human waste, 34% to an animal carcass, 18% to a deceased human body, and 7% to bloodborne pathogens. Mold exposure was reported by 595 responders, about 90% of all reports of this exposure based on mild respiratory problems, the other 10% based on severe respiratory problems. Carbon monoxide exposure was reported by 276 responders. Figure 3 presents the proportion of exposed responders who sought medical treatment for each exposure. Mold exposure was the top reason for seeking medical treatment for an exposure; 12.6% (n = 75) of responders who ever reported mold exposure sought medical treatment for it. We broke this down further and found that 42% of responders who reported severe respiratory symptoms they perceived were associated with mold exposure sought medical treatment for it; of those reporting mild respiratory symptoms, 10% sought medical treatment. Other exposures had low frequencies of reports of medical treatment affiliated with them, the next highest being floodwater (just more than 5% [n = 46] of those reporting exposure to floodwater sought medical treatment for it) and carbon monoxide (5% [n = 14] of those reporting carbon monoxide sought medical treatment for it).

Figure 4 presents the frequency of self-reported health effects in the population. Of the study responder population, 43% (n = 1222) reported having none of the health effects queried. Approximately 20% (n = 578) reported having one health effect. Figure 5 presents the distribution of specific health effects queried. Because a responder could indicate multiple health effects during a given deployment, the counts for this figure sum to more than the total number of responders. For sunburn, heat stress, penetrating injury, and injury from slips, trips, or falls, there is a corresponding bar indicting the distribution of reports of medical treatment sought for those health effects. The most commonly reported health effect by responders was sunburn (n = 1119; 39% of responders reported sunburn; 11% of all responders [n = 314] reported severe sunburn), followed by heat stress symptoms (n = 810; 30% of responders reported heat stress symptoms), penetrating injury (n = 470; 17% reported penetrating injury), and sinus infection (n = 333; 12% reported sinus infection). Depression, the only mental health effect directly queried, was reported by 5% of responders (n = 146). Figure 6 presents the proportion of responders who sought medical treatment for specific health effects. As described earlier, medical treatment sought was queried for four of the health effects—penetrating injury; heat stress; slips, trips, and falls; and sunburn. The highest frequencies of medical treatment sought for health effects were for injuries: 28% (n = 42) of people who reported an injury from a slip, trip, or fall reported seeking medical treatment for it; 11% (n = 51) of people who reported a penetrating injury sought medical treatment for it. Medical treatment rates for the two other health effects queried were substantially lower: heat stress (6%; n = 48) and sunburn (2%; n = 24).

Most responders (82%) reported use of at least one type of PPE, though this total is based on the inclusion of commonly worn uniform items, such as safety boots (65%), and commonly used items such as hand sanitizer (64%), sunscreen (43%), and insect repellent (41%). Safety glasses were worn by nearly 40% of the responder population surveyed. Close to 70% wore gloves, whether they were nitrile-latex gloves or leather work gloves. Approximately 20% wore a “boonie hat” to protect their necks from sunburn. Smaller numbers reported use of respirators (8%) and Tyvek suits (3%). Positive associations via chi-square tests between exposures reported and specific type of PPE were (1) animal/insect vector with hand sanitizer, insect repellant, nitrile-latex gloves, and leather work gloves; (2) carbon monoxide with respirator use; (3) floodwater with personal floatation device and waders; (4) infectious agents with hand sanitizer, nitrile-latex gloves, and Tyvek suits; and (5) mold exposure with nitrile-latex gloves, leather work gloves, respirator, and Tyvek suits (data not shown).

Logistic regression results for select baseline factors associated with health effects, adjusted for self-reported exposures, are presented in Table 3. Older people were more likely than younger people to report sinus infection (POR = 1.75; 95% CI, 1.34 to 2.29) and skin rash (POR = 1.44; 95% CI, 1.06 to 1.94), and they were less likely to report heat-/sun-related health effects. Those who responded for more than 1 month had elevated PORs for most health effects, most notably depression, which they were nearly three times (POR = 2.70; 95% CI, 1.86 to 3.92) more likely to report than those who responded for a month or less. As expected, responders whose response work started on October 1, 2005, or later were less likely to report the heat-/sun-related health effects queried in the survey, compared with those whose response began in either August or September 2005. Local responders were more than three times more likely to report depression than nonlocal responders (POR = 3.32; 95% CI, 2.22 to 4.98). They were also about two times more likely to report confusion (POR = 2.13; 95% CI, 1.11 to 4.07), difficulty concentrating (POR = 1.87; 95% CI, 1.28 to 2.73), and muscle strain (POR = 1.92; 95% CI, 1.33 to 2.76) than nonlocal responders. Sleeping 5 or fewer hours per night was associated with elevated PORs for most health effects, compared with sleeping at least 5 hours per night, most notably for depression (POR = 1.97; 95% CI, 1.34 to 2.90), slips, trips, and falls (POR = 2.12; 95% CI, 1.46 to 3.09), muscle strain (POR = 2.18; 95% CI, 1.50 to 3.19), and dehydration (POR = 2.00; 95% CI, 1.41 to 2.86). Responders who reported tobacco use had an increased likelihood of reporting heat stress (POR = 1.28; 95% CI, 1.05 to 1.55) and sunburn (POR = 1.33; 95% CI, 1.11 to 1.59) compared with those who reported not using tobacco. Responders in higher-intensity missions were about half as likely as those in lower-intensity missions to report difficulty concentrating (POR = 0.33; 95% CI, 0.15 to 0.71), depression (POR = 0.46; 95% CI, 0.24 to 0.89), and sinus infection (POR = 0.54; 95% CI, 0.33 to 0.88), but they had elevated PORs for penetrating injury (POR = 1.62; 95% CI, 1.02 to 2.58) and sunburn (POR = 1.86; 95% CI, 1.38 to 2.51).

Logistic regression results for select missions associated with each reported health effect, adjusted for age, sex, average sleep, responder residence, and tobacco use, are presented in Table 4. Responders involved in SAR boat, SAR ground, building cleanup, and body recovery had elevated PORs for almost all the health effects queried. Of particular note, they were approximately two to three times more likely to report heat stress and sunburn than people not carrying out those missions. Responders performing hazardous oils response work had elevated PORs for skin rash (POR = 1.70; 95% CI, 1.09 to 2.64) and were more than twice as likely to report sunburn (POR = 2.47; 95% CI, 1.89 to 3.23) compared with responders in other missions. Firefighters, a group composed of only 14 responders, had consistently high PORs for neurological health effects, mental health, gastrointestinal health effects, and injuries, in particular penetrating injuries (POR = 11.05; 95% CI, 2.80 to 43.57). They also had the highest POR for sinus infection (POR = 3.91; 95% CI, 0.99 to 15.49). Responders involved in administrative missions and ICS missions had a reduced likelihood compared with those in other missions for dermal health effects, injuries, musculoskeletal health effects, and heat-/sun-related health effects. Nevertheless, responders reporting administrative missions were almost three times more likely to report difficulty concentrating (POR = 2.97; 95% CI, 1.96 to 4.51) than those reporting other missions. Rescue swimmers had the highest POR for dehydration (6.24; 95% CI, 3.47 to 11.23), as well as the highest POR for infected skin (4.81; 95% CI, 1.79 to 12.91). They also had from two- to nearly fourfold increased likelihoods of having gastrointestinal health effects, skin rash, injuries, musculoskeletal health effects, and heat stress. Responders involved in SAR from the air had a greater than twofold positive association with dehydration (POR = 2.39; 95% CI, 1.62 to 3.59). Although not included in the table, medical and dental workers had a twofold increased risk of skin rash, but they had no other significant elevations or deficits.

Table 5 presents the results of logistic regressions to investigate the associations between exposures reported and health effects. Animal/insect vector exposure was positively associated with most of the health effects queried; the strongest associations were found for infected skin (POR = 6.76; 95% CI, 2.55 to 17.95), skin rash (POR = 3.22; 95% CI, 2.14 to 4.85), and penetrating injuries (POR = 3.75; 95% CI, 2.83 to 4.98). Carbon monoxide exposure was most strongly associated with confusion (POR = 6.27; 95% CI, 295 to 13.32). It was also the exposure most strongly associated with slips, trips, and falls (POR = 2.01; 95% CI, 1.29 to 3.12). Floodwater exposure was associated with a more than twofold likelihood of reporting confusion (POR = 2.56; 95% CI, 1.10 to 5.95) and skin rash (POR = 2.04; 95% CI, 1.39 to 3.00). Exposure to infectious agents was associated more than twofold with depression (POR = 2.85; 95% CI, 1.78 to 4.56) and penetrating injury (POR = 2.33; 95% CI, 1.77 to 3.07). The association between mold exposure and sinus infection was of a very high magnitude (POR = 10.39; 95% CI, 7.75 to 13.95). When we categorized mold exposure into three levels, (1) no mold exposure, (2) mold exposure reported based on mild respiratory symptoms, and (3) mold exposure reported based on severe respiratory symptoms, there was a monotonic increase in PORs for sinus infection, such that PORmild = 14.49; 95% CI, 11.03–18.79 and PORsevere = 21.21; 95% CI, 10.96–41.05; p-test for trend <0.001 (data not shown). Mold exposure was also associated, nearly threefold, with difficulty concentrating (POR = 2.89; 95% CI, 1.91–4.39), depression (POR = 2.90; 95% CI, 1.93–4.35), diarrhea (POR = 3.01; 95% CI, 2.17–4.17), digestive health effects (POR = 2.61; 95% CI, 1.90–3.60), and joint pain (POR = 2.52; 95% CI, 1.70–3.73).

We evaluated the potential effect modification of various baseline factors on the significant associations reported earlier between self-reported exposures and health effects. The few statistically significant interaction terms for these stratified analyses are presented hereafter (data not shown). Regarding the association between CO exposure and neurological health effects, responders who were younger (31 years or less) had statistically significant higher PORs for CO and confusion and difficulty concentrating than older (more than 31 years) responders (confusion: PORyounger = 9.82; PORolder = 3.19; P-interaction = 0.04; difficulty concentrating: PORyounger = 3.89; PORolder = 0.75; P-interaction = 0.01). Age also modified the association between CO and depression in that younger responders had a higher POR (3.54) than older responders (POR = 1.35); P-interaction = 0.04. Regarding the association between infectious agent exposure and penetrating injuries, responders who reported never using any PPE had a higher POR (6.02) than those who ever used a PPE (POR = 2.17); P-interaction = 0.03. Responders who reported never using PPE had a stronger association between mold exposure and skin rash (POR = 31.62) than those who ever used PPE (OR = 9.24); P-interaction = 0.03. We also evaluated the association between mold exposure and sinus infection further, stratifying by respirator use; however, the PORs for both the respirator use and the respirator nonuse groups were similar, and respirator use was not an effect modifier for this association.

Because numbers of those seeking medical treatment for particular exposures or health effects were small, unadjusted logistic regression models were used to evaluate the associations between the nine reported reasons for seeking medical treatment and baseline factors, missions, and health effects (data not shown). Seeking medical treatment for animal/insect vector exposure was most strongly associated with reporting skin rash (POR = 4.92; 95% CI, 1.91 to 12.72). Medical treatment for CO was most strongly associated with carrying out law enforcement (POR = 14.90; 95% CI, 4.14 to 53.58) or SAR boat (POR = 13.26; 95% CI, 369 to 47.67) missions, and with having reported confusion (POR = 25.83; 95% CI, 7.80 to 85.60). Medical treatment for floodwater exposure was most strongly associated with being a rescue swimmer (POR = 8.82; 95% CI, 3.96 to 19.65) or doing SAR boat mission (POR = 4.78; 95% CI, 2.65 to 8.62), as well reporting infected skin (POR = 12.55; 95% CI, 5.55 to 28.39) and skin rash (POR = 9.84; 95% CI, 5.37 to 18.04). Medical treatment for infectious agent exposure was most strongly associated with reporting infected skin (POR = 10.16; 95% CI, 2.87 to 35.96) and skin rash (POR = 12.03; 95% CI, 4.83 to 29.94). Medical treatment for mold exposure was most strongly associated with being a local responder (POR = 1.76; 95% CI, 1.10 to 2.83), carrying out administrative mission (POR = 2.11; 95% CI, 1.26 to 3.53), and reporting sinus infection (POR = 12.7; 95% CI, 7.98 to 20.76). Medical treatment for heat stress was most strongly associated with reporting dehydration (POR = 15.88; 95% CI, 8.83 to 28.58), skin rash (POR = 13.16; 95% CI, 7.33 to 23.65), and digestive health effects (POR = 8.77; 95% CI, 4.88 to 15.76). Medical treatment for sunburn was most strongly associated with reporting heat stress (POR = 28.23; 95% CI, 6.62 to 120.31). Medical treatment for penetrating injury was most strongly associated with reporting animal/insect bite/scratch (POR = 4.90; 95% CI, 2.44 to 9.82) and infected skin (POR = 7.60; 95% CI, 3.09 to 18.66). Medical treatment for slips, trips, and falls was most strongly associated with reporting joint pain (POR = 17.48; 95% CI, 9.32 to 32.80) and muscle strain (POR = 13.38; 95% CI, 7.06 to 25.38); in addition, people who slept less were nearly four times more likely (POR = 3.82; 95% CI, 1.89 to 7.71) to seek medical treatment for slips, trips, and falls than those who slept more than 5 hours per night.

Back to Top | Article Outline

DISCUSSION

The National Response Team recently published, Emergency Responder Health Monitoring and Surveillance: Technical Assistance Document,4 which stresses the importance of assessing exposures and health effects of responders during disasters. Few studies, however, have focused on disaster responder populations, thus they face poorly characterized risks and unknown short- and long-term health consequences.5 Most of the research that has been carried out on these populations has been described in terms of psychological risk factors6,7 and mental health consequences suffered by disaster responders.5,8,9 This cross-sectional study evaluated the self-reported exposures and health effects (ie, illnesses, injuries, and symptoms) of approximately 51% of Coast Guard members who responded to hurricanes Katrina and Rita. Although it is unclear how representative this sample is of the entire Coast Guard Katrina/Rita responder population, baseline characteristics of this study population are similar to those of the entire Coast Guard in 2005 (mean age among those in active duty was 30.7 years; active duty 85%, reserve 17%, civilian 15%; officer 20%, enlisted 80%; men 90%, women 10%).10 The exit survey was mandated by the Coast Guard for all responders; therefore, it was completed not only by responders who thought they were exposed to something hazardous or had health issues, as is reflected in the statistic that about 40% of the study population reported no exposures, 40% of the study population reported no health effects, and about 30% reported neither exposures nor health effects.

The most commonly reported exposures were animal/insect vector (46%, with 73% of those reports being mainly attributed to mosquito bites) and floodwater (29%). Mold exposure accounted for the greatest proportion (12.6%) of medical treatment sought of all the exposures, 42% of those indicating severe respiratory problems from mold reporting that they sought medical treatment. Sunburn was the most commonly reported health effect, reported by about 39% of the population, 11% reporting severe sunburn. Next, heat stress symptoms were reported by 30% of the population. Slips, trips, and falls accounted for the greatest proportion (28%) of medical treatment sought of all the health effects. About one third of the responders reported getting 5 or fewer hours of sleep on average per night, and this statistic did not differ by mission. About one third reported some use of tobacco, which is slightly higher than the rate estimated for active duty Coast Guard personnel (28%) in a 2008 report.11 Most responders (82%) reported some use of PPE, although this metric included some commonly worn uniform items, such as safety boots.

A number of interesting risk factor/mission/exposure–health effect associations were found in this study. Responders with longer deployments were more likely to report depression, as were local responders, firefighters, and responders involved in building cleanup and body recovery missions. Sleeping 5 or fewer hours per night was associated with depression and injury from slips, trips, and falls. As would be expected, responders in higher-intensity missions, such as SAR boat, SAR ground, building cleanup, and body recovery had elevated PORs for many of the health effects, in particular heat stress and sunburn. Firefighters and rescue swimmers, two groups characterized by involvement in some of the most-intense emergency response tasks, also had highly elevated PORs for most health effects. The number of firefighters in this study was small (n = 14); however, so their POR estimates, though of high magnitude, were potentially unstable. Responders involved in hazardous oils response had elevated PORs for adverse dermal health effects. Adverse health effects were not confined to only those in operational, higher-intensity missions. For example, responders reporting administrative mission had threefold elevated PORs for difficulty concentrating and nearly a 70% increased likelihood of reporting depression compared with responders not reporting administrative mission. In addition, seeking medical treatment for mold exposure was most strongly associated with those carrying out administrative mission. Some of the strongest associations we found between self-reported exposures and health effects were for mold exposure and sinus infections and for CO exposure and confusion and difficulty concentrating, the CO associations being particularly elevated among younger responders. Floodwater exposure was most strongly associated with gastrointestinal health effects and skin rash. Both animal/insect vector and infectious agent exposures were strongly associated with dermal health effects.

The excess standing water and warmer temperatures after hurricanes Katrina/Rita provided optimal conditions for mosquito breeding, a scenario typical of posthurricane environments. This study found that exposures to animal/insect vectors were most strongly associated with infected skin, skin rash, penetrating injury, and seeking medical treatment for skin rash. A previous study among Katrina relief workers found that 44% reported mosquitos bites,5 similar to the 46% reported by responders in our study. This high prevalence of mosquito bites may suggest that repellant was perhaps not utilized effectively to afford adequate protection. Fire ants were the second-most common animal/insect vector reported by Coast Guard responders. Stinging insects are a common postdisaster exposure,14 and emerging zoonotic infections in the aftermath of a hurricane can incapacitate first responders. Protection against insect bites must be a priority for disaster planners, in at-risk locations.

After hurricanes strike, portable generators and other gasoline-powered appliances for electrical power and cleanup are often used. These devices produce CO, and improper use can lead to CO poisoning, which can be difficult to diagnose, because the common symptoms are nonspecific and are often mistaken for the flu, for example, headache, dizziness, weakness, nausea, vomiting, chest pain, and confusion.12 The state poison control center in Louisiana found that 2 weeks after Katrina, there were seven observed CO exposures compared with one expected.13 In our study, exposure to CO was assessed via responder self-report of ever having been exposed during deployment. Although the exact sources of CO exposure are unknown, likely sources would include exhaust from gasoline boat engines, other vehicles, and emergency generators used in the aftermath of the hurricanes. We found that responders involved in SAR boat and law enforcement missions were most likely to report seeking medical treatment for CO exposure. We found a more than sixfold association with reporting confusion, which was even stronger in younger responders, and a more than fivefold association with seeking medical treatment for slips, trips, and falls. It is worrisome that 14 responders who reported CO exposure also reported that they sought medical treatment for it, though we do not have details on their medical treatment.

Many first responders were exposed to floodwaters over the course of days and weeks, with potential pollutants in the floodwater, including toxic chemical constituents and potentially pathogenic microorganisms.14 Our study found a positive association between floodwater exposure and digestive symptoms (nausea, vomiting, and loss of appetite). In addition, floodwater exposure and being a rescue swimmer were both strongly associated with seeking medical treatment for infectious agents. These findings may be attributed to elevated fecal coliform bacteria, an infectious agent, found in the surface floodwater shortly after the hurricanes.14 We found that responders seeking medical treatment for floodwater exposure and for infectious agents were most likely to report infected skin and skin rash. A prior study among firefighters responding to Katrina found that those with floodwater contact with skin and either nose/mouth or eyes had significantly increased rates of new-onset skin rash.15 Recommendations from that study were that response workers should use appropriate PPE to minimize direct skin and mucosal contact with floodwater. Other nonfloodwater infectious agents of concern were human waste, bloodborne pathogens, deceased bodies, and animal carcasses. In this study, rescue swimmers were the most likely group to seek medical treatment for infectious agent exposures.

Mold exposure was a significant problem in the aftermath of Hurricanes Katrina/Rita. In this study, seeking medical treatment for mold exposure was strongly associated with being a local responder and reporting administrative mission. We found an association more than 10-fold between mold exposure and sinus infection and an association more than 30-fold between reporting mold based on severe respiratory symptoms and sinus infection. The experience of new-onset, allergic response to mold is often characterized as “sinusitis” because of the increased fullness and pressure in the frontal and maxillary sinuses. In keeping with use of lay terminology on the survey, the description “sinus infection” was used to characterize what many seeking medical care called their perceived condition. Mold exposure has been shown to be associated with several adverse health effects, such as upper and lower respiratory tract irritation, asthma, allergic rhinitis, aspergillosis, sinusitis, conjunctivitis, hypersensitivity pneumonitis, posttraumatic skin infection, and ocular pain/redness.16–18 Mold has also been found to exacerbate asthma in sensitized asthmatics, and emerging evidence suggests that asthma can develop in the occupational setting as a result of exposure to mold.19,20 Both the indoor and outdoor mold spores like those found in flooded New Orleans were approximately double those found in nonflooded areas.16 Our assessment of mold exposure was based on self-report of respiratory symptoms (mild or severe) perceived to be associated with mold, so if responders considered sinus infection a respiratory symptom, then the association we found could overestimate the true associate between mold and sinus infection. However, our position is that lay persons would not be very likely to consider sinuses as part of the respiratory system, since they do not relate directly to oxygen/carbon dioxide exchange. The Centers for Disease Control and Prevention recommends clinicians to have a “high index of suspicion” to recognize the symptoms of mold exposure among their patients after major hurricanes and floods16 and also recommends workers use National Institute for Occupational Safety and Health–certified respirators that have been fit-tested according to the Occupational Safety and Health Administration respiratory protection standard21 when working in areas of mold contamination.17 Respirator use in our study did not modify the association between mold exposure and sinus infection. Nevertheless, it is unclear whether responders included surgical masks in their interpretation of respirators, which are insufficient to protect from mold exposure. Among the 517 responders doing building cleanup, putting aside the issue of appropriate use, the rate of respirator use was only 21%.

Sleeping 5 or fewer hours per night was associated, at least threefold each, with slips, trips, and falls, muscle strain, dehydration, and depression and was positively associated with all of the health effects measured in this study. In addition, seeking medical treatment for slips, trips, and falls was strongly associated with lack of sleep. First responders may need to endure short sleep duration or erratic sleep patterns that could lead to sleep disturbance; however, it has been demonstrated that both short sleep duration and sleep disturbance have significant and deleterious impacts on both function and quality of life. Specifically, short sleep duration has been shown to be associated with an increased risk of a number of physical health issues, including pain, acute and chronic respiratory conditions, diabetes onset, cardiovascular disease, and mortality.22–24 Furthermore, reduced sleep has been shown to be associated with poor mental health, including depression and anxiety,24 and more specifically with major depressive disorder and posttraumatic stress disorder (PTSD).25 These relationships have been demonstrated in a number of populations, including public health workers,23,26,27 first responders,26 US military service members,22 and US military veterans.25 Similarly, sleep disturbance has been shown to be related to occupational impairment, including poor concentration,26,28 and occupational injuries in public sector employees,27 as well as to poor physical, mental health, and day-to-day function,26,29 including the onset of diabetes,22 being at increased risk for major depressive disorder, PTSD, and suicidal ideation.25 These findings and those of our study underscore the importance of adequate sleep for emergency responders, and leadership personnel in emergency responses should be sensitive to this issue. Although this may seem an unattainable goal given the nature of emergency responders' jobs/duties, it is also a critical issue to address, given the serious short- and long-term consequences.

Another important finding from this study is that responders who lived local to the disaster site had an at least threefold greater likelihood of reporting depression, difficulty concentrating, and confusion. Local responders are typically responding to a disaster in their own backyard, and despite the stability of employment with the Coast Guard, local responders had to deal with other major stressors, such as the destruction of their own homes and neighborhoods and injury or relocation of family members. According to the Coast Guard historian's office, approximately 30% of Coast Guard personnel stationed in the New Orleans area lost their homes and numerous personnel had to work in temporary facilities because of extensive damage to work stations.2

The rate of reporting depression in this study was 5%, slightly higher than the 2005 rate of major depression (4.5%) in the active duty Coast Guard.30 In addition to being associated with local residence, as described earlier, depression was also more likely among those with deployments longer than 1 month and those performing firefighting, building cleanup, and body recovery (twofold elevation, though not statistically significant) missions. A number of previous studies have shown that emergency responders are at increased risk of developing PTSD and depression after natural disasters.31–37 Although emergency responders may be expected to be more psychologically resilient to traumatic stressors, compared with community members, the level of destruction from Katrina/Rita required many responders to conduct missions outside the scope of previous training, including deceased body transportation, and may have overwhelmed their typical coping strategies. It should be noted, however, that during disasters, grief and depression may be a normal short-term response when bearing witness to previously unimaginable loss and suffering. Longer-term follow-up of responder populations would help address whether the depression was a chronic condition and evaluate whether there were elevated rates of PTSD.

Heat stress was one of the most commonly reported health effects in this study population. We found that heat stress, dehydration, and seeking medical treatment for heat stress, not surprisingly, were reported more frequently by responders in high-intensity missions, and particularly in rescue swimmers. Despite their high prevalence in disaster relief situations, little research has investigated heat stress and heat-related injury among disaster responders.38–40 One of these studies found that heat-related injuries resulted in most of the medical claims among the National Guard in the 1993 midwest floods.40 Implementation of a few simple preventive measures may decrease the impact of this problem,42 such as attention to sufficient hydration and implementation of a comprehensive program against heat stress.38,39

Although we did not carry out an in-depth analysis of PPE use in this study, in some cases, PPE use was identified as a protective effect modifier in the association between exposure and health effect. In stratified analyses, we found that responders who used any PPE had lower likelihood of reporting penetrating injuries if exposed to infectious agents and of reporting skin rash if exposed to mold. First responders typically rush in with more regard for others and property than themselves, thus firm establishment of PPE required for the operation will accentuate the response and aid in protection of responders.

This study represents the first large-scale effort to evaluate exposures and health effects in deployed Coast Guard disaster responders. This research offers an important examination of the common exposures and health effects encountered and, given this, the preventive tactics that should prove most effective. Nonetheless, the findings of this study must be interpreted in terms of several methodologic considerations. First, because this is a cross-sectional study, further research using longitudinal designs is needed to better determine the course of exposures and health effects during deployments, as well as the association between these and specific missions. A longitudinal study design could also illuminate the long-term health effects of deployment exposures, which we were not able to do in this study. While a strong response rate of 51% was achieved, it would be ideal to reach the remaining responders for future research. The survey employed self-report techniques, which may raise validity concerns. Nevertheless, procedures to promote honesty on self-report studies, such as ensuring respondents' anonymity (confidentiality) and allowing persons to complete the survey in private via computer, were applied. Although problems with recall could result in overreporting or underreporting of exposures, health effects, and medical treatment usage, we do not have a particular reason to believe that these were biased in any specific direction, especially given our large distribution of responders who reported either no exposures or no health effects. There is also the potential limitation of overestimating the association between mold exposure and sinus infection if the responders considered sinus infection a respiratory symptom, though, as we discussed earlier, we think that the potential for this was quite small.

When disasters strike, the Federal Emergency Management Administration activates the worker safety and health provisions of the National Response Plan, which describes measures needed to ensure that safety and health factors are recognized, evaluated, and controlled so that responders are adequately protected. In reality, though, given the operational tempo of a response to natural disasters, such as Hurricanes Katrina/Rita, responder safety and health may often be compromised to fulfill missions. Without this type of operational tempo, the Coast Guard may not have been able to save the number of lives or evacuate the number of victims it did.1,2 Nevertheless, agencies deploying personnel to disasters have a responsibility to assess their responders' exposures and health effects4 to adequately follow up these populations for longer-term health consequences of disaster response work and to enable implementation of appropriate safety/health interventions on the basis of both real-time and postdisaster data. As such, identification of at-risk emergency responder cohorts and appropriate “rostering” of each cohort member are imperative. Once a response site has been identified and characterized, maintaining site integrity and managing the response are essential for assuring responder safety. Responders arriving, departing, and integrating into the response should be cleared through a centralized checkpoint to assure that they have been accounted for and briefed on the hazards and correct PPE use. Incident management tools such as ICS can greatly assist with performing this function as long as the check-in process is cited as a performance objective. Establishing this goal is an incident command decision that should be paramount. It derives from the overarching objectives to keep responders safe and to save lives.

During disaster response and recovery, the ability to triage injury or illness and adequate sleep hygiene must become a focus of emergency response leadership. In addition, length of deployment, the number of deployments, time off, and special consideration for the effects on local responders must also be considered. Utilization of a disaster–response buddy system for improved self-care, from use of sunscreen and insect repellant to reminders about required rest/time off, to assisting each other with the development of a narrative to better understand one's emotions during disaster response may prove similarly supportive to what is noted for those in combat during military deployments.41 In recent years, the Coast Guard has focused on raising awareness of its available work life programs and has implemented the CG SUPRT Program, an expanded Employee Assistance Program. CG SUPRT, which includes expanded counseling services and an e-counseling option for members in remote locations (or at home), is available to active duty, Selected Reserve, and civilian employees, and their family members. Although no two disasters are identical, use of technology to improve accountability of when and where personnel are deployed, and to capture which missions are performed and types of PPE used, as well as preparatory disaster-specific training based on lessons learned could provide potential avenues for protecting a workforce knowingly dispatched into harm's way.

Back to Top | Article Outline

CONCLUSIONS

Caring for emergency responders relies partly on the evaluation and understanding of shortfalls of previous responses and implementing the lessons learned. Continued surveillance and evaluation of the risks, hazards, and health outcomes in disaster responders such as Coast Guard responders is critical to promote and ensure their continued health and by extension, that of the population at large.

Back to Top | Article Outline

REFERENCES

1. Government Accounting Office to Congressional Committees. Observations on the preparation, response, and recovery missions related to Hurricane Katrina: In: Report Observations on the Preparation, Response, and Recovery Missions Related to Hurricane Katrina. Washington, DC: United States Government Accountability Office; 2006:1–8.
2. The Coast Guard's Katrina Documentation Project. Available at http://www.uscg.mil/history/katrina/docs/karthistory.asp. Accessed May 7, 2014.
3. Centers for Disease Control and Prevention, Ed. Hurricane Katrina response. In: [book] Hurricane Katrina Response. Atlanta, GA: CDC; 2007.
4. National Response Team, Ed. Emergency responder health monitoring and surveillance: technical assistance document. In: [report] Emergency Responder Health Monitoring and Surveillance: Technical Assistance Document. Washington, DC: National Response Team; 2012.
5. Swygard H, Stafford RE. Effects on health of volunteers deployed during a disaster. Am Surg. 2009;75:747–752; discussion 752–743.
6. Dolce A, Ricciardi M. Impact of psychological risk factors on disaster rescue operations: the case of Italian volunteers. Disasters. 2007;31:91–103.
7. Perrin MA, DiGrande L, Wheeler K, Thorpe L, Farfel M, Brackbill R. Differences in PTSD prevalence and associated risk factors among World Trade Center disaster rescue and recovery workers. Am J Psychiatry. 2007;164:1385–1394.
8. Centers for Disease Control and Prevention. Mental health status of World Trade Center rescue and recovery workers and volunteers—New York City, July 2002–August 2004. MMWR Morb Mortal Wkly Rep. 2004;53:812–815.
9. Bills CB, Levy NA, Sharma V, et al. Mental health of workers and volunteers responding to events of 9/11: review of the literature. Mt Sinai J Med N Y. 2008;75:115–127.
10. United States Coast Guard Workforce Forcasting and Analysis Group (CG-12a). Coast guard specific demographics, 2005. In: [report] Coast Guard Specific Demographics, 2005.
11. RTI International. 2008 Department of Defense survey of health-related behaviors among active duty military personnel. In: [report] 2008 Department of Defense Survey of Health Related Behaviors Among Active Duty Military Personnel. 2009:75.
12. Centers for Disease Control and Prevention. Carbon monoxide poisoning after hurricane Katrina—Alabama, Louisiana, and Mississippi, August–September 2005. MMWR Morb Mortal Wkly Rep. 2005;54:996–998.
13. Ely EW, Moorehead B, Haponik EF. Warehouse workers' headache: emergency evaluation and management of 30 patients with carbon monoxide poisoning. Am J Med. 1995;98:145–155.
14. Pardue JH, Moe WM, McInnis D, et al. Chemical and microbiological parameters in New Orleans floodwater following Hurricane Katrina. Environ Sci Technol. 2005;39:8591–8599.
15. Tak S, Bernard BP, Driscoll RJ, Dowell CH. Floodwater exposure and the related health symptoms among firefighters in New Orleans, Louisiana 2005. Am J Indust Med. 2007;50:377–382.
16. Bush RK, Portnoy JM, Saxon A, Terr AI, Wood RA. The medical effects of mold exposure. J Allergy Clin Immunol. 2006;117:326–333.
17. Brandt M, Brown C, Burkhart J, et al. Mold prevention strategies and possible health effects in the aftermath of hurricanes and major floods. MMWR Recommend Rep. 2006;55:1–27.
18. Institute of Medicine. Damp indoor spaces and health. In: [report] Damp Indoor Spaces and Health. National Academies of Science; 2004.
19. Cox-Ganser JM, White SK, Jones R, et al. Respiratory morbidity in office workers in a water-damaged building. Environ Health Perspect. 2005;113:485–490.
20. Cummings KJ, Cox-Ganser J, Riggs MA, Edwards N, Hobbs GR, Kreiss K. Health effects of exposure to water-damaged New Orleans homes six months after Hurricanes Katrina and Rita. Am J Public Health. 2008;98:869–875.
21. Occupational Respiratory Protection Standard. 29 CFR 1910.134.
22. Boyko EJ, Seelig AD, Jacobson IG, et al. Sleep characteristics, mental health, and diabetes risk: a prospective study of U.S. military service members in the Millennium Cohort Study. Diabetes Care. 2013;36:3154–3161.
23. Ferrie JE, Shipley MJ, Cappuccio FP, et al. A prospective study of change in sleep duration: associations with mortality in the Whitehall II cohort. Sleep. 2007;30:1659–1666.
24. Krueger PM, Friedman EM. Sleep duration in the United States: a cross-sectional population-based study. Am J Epidemiol. 2009;169:1052–1063.
25. Swinkels CM, Ulmer CS, Beckham JC, Buse N, Calhoun PS. The Association of Sleep Duration, Mental Health, and Health Risk Behaviors among U.S. Afghanistan/Iraq Era Veterans. Sleep. 2013;36:1019–1025.
26. McKibben JB, Fullerton CS, Ursano RJ, et al. Sleep and arousal as risk factors for adverse health and work performance in public health workers involved in the 2004 Florida hurricane season. Dis Med Public Health Preparedness. 2010;4(suppl 1):S55–S62.
27. Salminen S, Oksanen T, Vahtera J, et al. Sleep disturbances as a predictor of occupational injuries among public sector workers. J Sleep Res. 2010;19:207–213.
28. Swanson LM, Arnedt JT, Rosekind MR, Belenky G, Balkin TJ, Drake C. Sleep disorders and work performance: findings from the 2008 National Sleep Foundation Sleep in America poll. J Sleep Res. 2011;20:487–494.
29. Lee M, Choh AC, Demerath EW, et al. Sleep disturbance in relation to health-related quality of life in adults: the Fels Longitudinal Study. J Nutr Health Aging. 2009;13:576–583.
30. Armed Forces Health Surveillance Center. Mental health conditions among active component U.S. Coast Guard; Defense Medical Surveillance System accessed 5 June, 2012 In: [report] Mental Health Conditions Among Active Component U.S. Coast Guard; Defense Medical Surveillance System accessed 5 June, 2012. 2012.
31. Bryant RA, Harvey AG. Posttraumatic stress in volunteer firefighters. Predictors of distress. J Nervous Ment Dis. 1995;183:267–271.
32. Fullerton CS, Ursano RJ, Wang L. Acute stress disorder, posttraumatic stress disorder, and depression in disaster or rescue workers. Am J Psychiatry. 2004;161:1370–1376.
33. Ginexi EM, Weihs K, Simmens SJ, Hoyt DR. Natural disaster and depression: a prospective investigation of reactions to the 1993 midwest floods. Am J Community Psychol. 2000;28:495–518.
34. Guo YJ, Chen CH, Lu ML, Tan HK, Lee HW, Wang TN. Posttraumatic stress disorder among professional and non-professional rescuers involved in an earthquake in Taiwan. Psychiatry Res. 2004;127:35–41.
35. Marmar CR, Weiss DS, Metzler TJ, Ronfeldt HM, Foreman C. Stress responses of emergency services personnel to the Loma Prieta earthquake Interstate 880 freeway collapse and control traumatic incidents. J Trauma Stress. 1996;9:63–85.
36. Ozen S, Sir A. Frequency of PTSD in a group of search and rescue workers two months after 2003 Bingol (Turkey) earthquake. J Nerv Ment Dis. 2004;192:573–575.
37. West C, Bernard B, Mueller C, Kitt M, Driscoll R, Tak S. Mental health outcomes in police personnel after Hurricane Katrina. J Occup Environ Med. 2008;50:689–695.
38. Brearley MB, Heaney MF, Norton IN. Physiological responses of medical team members to a simulated emergency in tropical field conditions. Prehosp Disaster Med. 2013;28:139–144.
39. Mori K, Tateishi S, Hiraoka K, et al. How occupational health can contribute in a disaster and what we should prepare for the future—lessons learned through support activities of a medical school at the Fukushima Daiichi Nuclear Power Plant in Summer 2011. J Occup Health. 2013;55:6–10.
40. Dellinger AM, Kachur SP, Sternberg E, Russell J. Risk of heat-related injury to disaster relief workers in a slow-onset flood disaster. J Occup Environ Med. 1996;38:689–692.
41. Adler AB, Castro CA, McGurk D. Time-driven battlemind psychological debriefing: a group-level early intervention in combat. Mil Med. 2009;174:21–28.
42. Diaz JH. The impact of hurricanes and flooding disasters on hymenopterid-inflicted injuries. Am J Dis Med. 2007;2:257–269.
43. Tak S, Driscoll R, Bernard B, West C. Depressive symptoms among firefighters and related factors after the response to Hurricane Katrina. J Urban Health. 2007;84:153–161.
44. Depression and physical symptoms: the mind–body connection. J Clin Psychiatry. 2004;65:867–876.
45. Centers for Disease Control and Prevention. Health hazard evaluation of police officers and firefighters after Hurricane Katrina—New Orleans, Louisiana, October 17–28 and November 30–December 5, 2005. MMWR Morb Mortal Wkly Rep. 2006;55:456–458.
46. Alexander DA, Klein S. First responders after disasters: a review of stress reactions, at-risk, vulnerability, and resilience factors. Prehosp Disaster Med. 2009;24:87–94.
47. Henningsen P, Zimmermann T, Sattel H. Medically unexplained physical symptoms, anxiety, and depression: a meta-analytic review. Psychosomat Med. 2003;65:528–533.
Copyright © 2014 by the American College of Occupational and Environmental Medicine