Military personnel, police officers, firefighters, and other first responders must prepare for and respond to life-threatening crises on a daily basis. This lifestyle places stress on personnel, particularly so on military troops who may be isolated from support systems and other resources. Combat stressors include traumas, such as injury, attempted attack on one's unit or camp, killing, witnessing death, and death of a unit member. Operational stressors include being away from family, close quarters, difficulty acclimating to weather, and other environmental changes. Such stress can result in physical, behavioral, and psychological sequelae, such as posttraumatic stress disorder (PTSD).
Combat and operational stress control (COSC) is the U.S. military's multifaceted approach that consists of all programs developed and actions taken by military leadership to prevent, identify, and manage stress reactions in active duty troops in all branches of service. COSC programs target entire units, specific areas of operation, and individuals identified as exhibiting stress-related behaviors. COSC works across the deployment cycle to prepare service members for combat and deployment stressors, provide support in-theater, and assist reintegration upon the return home.
This study describes preliminary results of a systematic review investigating the efficacy and comparative effectiveness of interventions designed to prevent, identify, and manage stress reactions in military, law enforcement, and first responders, including those evaluated as part of COSC and similar programs for nonmilitary populations.
The following questions guided the systematic review:
- What are the effects of interventions on physiological, psychological, behavioral, occupational, and acceptability outcomes?
- Do these effects vary by intervention components, intensity, and modality?
- Do these effects vary by setting?
- Do these effects vary by population?
We searched the electronic databases PsycINFO, PubMed, PTSDpubs, Cochrane Central Register of Controlled Trials, and the U.S. Defense Technical Information Center, as well as bibliographies of existing systematic reviews, to identify English-language studies evaluating the efficacy or comparative effectiveness of interventions, including stress inoculation, resilience training, traumatic event management, and psychological first aid, among others. For applicability to the modern context, studies published in 1990 or later were included. Controlled trials and cohort comparisons of interventions with military, law enforcement, and first responders were included. In addition, the Psychological Health Center of Excellence requested that we include case studies (with no comparison group) for military populations. Two independent reviewers screened literature using predetermined eligibility criteria. Researchers individually abstracted study-level information and outcome data and assessed the quality and risk of bias of each included study; data were reviewed for accuracy by the project leader.
Continuous outcomes, such as PTSD symptom scores, were converted to standardized mean differences for comparison across studies. Risk ratios were calculated for dichotomous outcomes, such as percentage of personnel diagnosed with PTSD postdeployment. When several studies that compared an intervention group with a control group or active comparator reported the same outcome category (i.e., depression) and measure type (e.g., categorical or continuous outcome), we conducted meta-analysis. We conducted two types of meta-analyses, comparing (1) the effects of COSC interventions versus no intervention and (2) the effects of COSC interventions versus another active stress control intervention (usually standard training or education). To assess whether certain settings or populations were associated with an outcome, we reported direct and indirect comparisons. In the absence of head-to-head trials, we conducted meta-regression by adding a categorical variable, representing setting (i.e., in theater) or population (military versus law enforcement or first responders) to the meta-analysis model to assess whether this variable was associated with the outcome across studies. (Because of insufficient data, we were unable to conduct meta-regressions for intervention type or components.) The quality of the body of evidence was assessed using the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach, which considers study limitations (risk of bias), directness, consistency, precision, and publication bias.
Key Question 1: What Are the Effects of Interventions on Physiological, Psychological, Behavioral, Occupational, and Acceptability Outcomes?
An extensive search for studies of stress control interventions for active duty military, law enforcement, and first responders identified 4,742 potentially relevant publications. After a review of abstracts, we obtained 566 full texts published from 1990 to 2020. One hundred and fifteen studies reported in 136 publications met the inclusion criteria. The body of evidence consisted of 38 controlled trials, 35 cohort comparisons, and 42 case studies or case series with no comparison group. In addition, we identified 14 relevant systematic reviews. Interventions studied included resilience training, stress inoculation with biofeedback, mindfulness, psychological first aid, frontline mental health centers, embedded mental health staff, two- to seven-day restoration programs, debriefing (including critical incident stress debriefing), third-location decompression, postdeployment mental health screening, reintegration programs, and family-centered programs.
Meta-analyses found that COSC programs, on the whole, showed no significant difference in effect on sleep, alcohol misuse, anxiety, depression, PTSD symptoms (usually measured by PTSD Checklist score), or help-seeking stigma, compared with active interventions, such as a standard stress-management class. Meta-analyses comparing COSC programs with no intervention showed no significant effect on alcohol misuse, depression, PTSD symptoms, PTSD case rate, stress level, or help-seeking stigma. Against active comparators, the quality of evidence of no difference in effect was moderate for sleep, depression, and PTSD symptoms and low for alcohol misuse, anxiety, and help-seeking stigma. Quality of evidence was moderate for no effect on PTSD symptoms and PTSD case rate and low for no effect on alcohol misuse, depression, stress level, and help-seeking stigma versus no intervention.
Quality of evidence was insufficient to form conclusions regarding the effect on stress and distress levels versus active comparators and the effect on heart rate, marriage or family outcomes, and occupational performance versus any comparison group.
Positive outcomes were found in several areas. The COSC programs we studied, on the whole, had positive effects on return to duty (moderate quality of evidence), absenteeism (low quality of evidence), and distress (moderate quality of evidence versus no intervention). In general, these outcomes are important targets of COSC programs. However, although COSC programs may reduce distress enough for active duty troops to function, in studies that include control or comparison groups, COSC appears to have no significant impact on psychological outcomes, such as symptoms of PTSD and depression.
Most COSC programs reported high levels of acceptability and satisfaction.
Results for specific approaches and intervention types are summarized in the section for key question 1a. Results for specific settings and populations are summarized under key question 1b and 1c, respectively. These sections summarize evaluations of efficacy or effectiveness; focus group studies and case studies that reported only acceptability or usability are discussed in the main report, with findings displayed in tables.
Key Question 1a: Do the Effects Vary by Intervention Components, Intensity, or Modality?
Because of insufficient data, it was not possible to conduct meta-regression analysis for intervention types or components. No intervention types or components have high or moderate strength evidence of efficacy for any outcome. There is low-strength evidence that trauma risk management (TRiM), a psychological first aid program developed by the UK Royal Navy, decreases help-seeking stigma; two- to seven-day restoration programs have a significant positive effect on distress, PTSD symptoms, homefront issues and return to duty; eye movement desensitization and reprocessing for subclinical stress reduces stress and PTSD symptoms and improves marital adjustment; group psychological debriefing (not incident-specific) at the end of deployment has a positive effect on alcohol misuse and depression; decompression at a location other than the area of operations or home (third-location decompression) has a positive effect on depression but a possible negative effect on alcohol use; end-of-deployment screening has a positive effect on alcohol and substance misuse; and family interventions have positive effects on marriage and parenting outcomes. There is insufficient evidence to conclude that any type of predeployment stress-control training is effective in preventing or reducing psychological symptoms.
Key Question 1b: Do These Effects Differ by Setting?
Meta-regression analyses to assess possible effects of setting (in theater versus not in theater) were not possible for sleep, alcohol misuse, anxiety, PTSD case rate, stress, marriage and family outcomes, or help-seeking stigma. The quality of evidence was insufficient to determine the effect of in-theater setting on those outcomes. Meta-regression analyses found no significant effect of in-theater setting on results for depression, distress, or PTSD symptoms; quality of evidence was rated low. Two studies compared locations head-to-head; a small-cohort study comparing frontline reintegration in Iraq at the end of deployment with standard reintegration postdeployment found no differences in outcomes, whereas a 20-year follow-up of Israeli Defense Forces found that frontline mental health treatment had a moderate-size effect on PTSD intensity compared with a rear-echelon treatment that bordered on statistical significance.
Key Question 1c: Do These Effects Differ by Population?
Meta-regression analyses to assess possible effects of population (military versus law enforcement or other first responders) were not possible for sleep, alcohol misuse, anxiety, PTSD case rate, stress, marriage and family outcomes, or help-seeking stigma. The quality of evidence is insufficient to draw conclusions about any possible population differences in effects on those outcomes. Meta-regression analyses found no effect of study population on results for depression, distress, or PTSD symptoms; the quality of evidence is low. Although there is some evidence from two cohort studies that COSC programs are more useful for junior-ranked personnel, the quality of evidence is insufficient to formulate conclusions.
Discussion and Conclusions
Most COSC programs reported high levels of acceptability and satisfaction. On the whole, in-theater and postdeployment programs had significant positive effects on return to duty, absenteeism, and distress. However, COSC programs, in general, appear to have little significant impact on symptoms of psychological disorders, such as PTSD.
Because of study limitations, inconsistency of results, indirectness, and possible publication bias, there was insufficient evidence to form conclusions about the efficacy of many specific intervention types and components. We found low-strength evidence for positive effects on some outcome areas for the UK's TRiM program, eye movement desensitization and reprocessing for subclinical stress, group psychological debriefing (not incident-specific) at the end of deployment, third-location decompression, end-of-deployment screening, and marriage and family interventions. Additional studies of these interventions are suggested to increase the quality of evidence.
Regarding setting, the quality of evidence was either insufficient to formulate conclusions or low for no effect of setting (in theater versus not in theater), depending on the outcome area. Although there is some evidence from two cohort studies that COSC programs are more useful for junior-ranked personnel, the quality of evidence is insufficient to formulate conclusions. Future studies of COSC should report stratified results by rank and military occupation.
Only case studies with no comparison group reported on return to duty. These were large government reports on in-theater mental health intervention during Operation Iraqi Freedom and Operation Enduring Freedom. Such evaluations rarely reported psychological outcomes or the effect of specific components. If possible, retrospective cohort studies should use official individual service member records to assess the effect of the amount and type of mental health services used during deployment on psychological measures. Ideally, these evaluations should report longitudinal postdeployment data on service use and mental health outcomes.