The role of measurement in moral injury care: latent profiles on the moral injury and distress scale

Introduction: Although studies of moral injury proliferated over the past decade, few studies have examined common moral injury symptom presentations.

Methods: Data were analyzed from a population-based sample (N = 645) of combat veterans, healthcare workers, and first responders. All participants endorsed exposure to a potentially morally injurious event (PMIE) and completed the Moral Injury and Distress Scale (MIDS).

Results: Latent profile analysis revealed three distinct symptom profiles based on MIDS items assessing psychological, emotional, social, and spiritual symptoms of moral injury. The majority of participants (74.7%, n = 482) reported minimal moral injury symptoms. Approximately one in five participants (20.2%, n = 130) endorsed moderate levels of moral injury symptoms, and one in twenty (5.1%, n = 33) reported severe moral injury symptoms. All participants with the severe profile screened positive for probable moral injury on the MIDS (score ≥ 27); however, 9.1% had scores within normal limits on measures of posttraumatic stress and depression. After adjusting for demographics, those with moderate or severe moral injury symptoms were more likely than those with minimal moral injury symptoms to be younger, have fewer years of education, and identify as Hispanic.

Discussion: These findings support the use of the MIDS in measurement-based care to identify distinct clinical presentations of moral injury, including those with moderate to severe presentations that warrant further evaluation and potentially treatment.

Introduction

Individuals internalize beliefs about morally acceptable conduct from various sources, including culture, occupational norms, family values, social groups, and personal experiences (1). Unfortunately, those employed in high stress, service-oriented occupations may be exposed to events in their workplaces that violate these beliefs. For example, military veterans might participate in combat operations involving civilian casualties, and healthcare workers may limit access to potentially life-saving treatments when they triage patients during mass-casualty events. Even when consistent with the relevant ethical standards (e.g., military rules of engagement, medical codes of ethics), they may perceive their own or others’ conduct as violations of their internalized moral beliefs (e.g., care for vulnerable others). In these cases, if they believe that they participated in an event that violated their beliefs because of what they did or failed to do, or if they witness or learn about wrongs committed by others in a group or institution to which they belong, they may sustain a moral injury. Symptoms of moral injury include but are not limited to intractable moral emotions (e.g., guilt and shame), withdrawal from close relationships or valued groups, and loss of meaning or purpose in life or work (2–4).

In a recent population-based survey of military veterans, healthcare workers, and first responders (N = 1232), nearly half (47.3%) reported exposure to at least one potentially morally injurious event (5). Additionally, about one in twenty individuals (6.0%) screened positive for probable moral injury, indicating the presence of moral injury symptoms severe enough to negatively impact their ability to function in daily life. In another nationally representative study of U.S. combat veterans, 5.9% screened positive for moral injury, with an additional 4.1% reporting subclinical levels of moral distress (6). Exposure to potentially morally injurious events and moral injury are both linked to elevated rates of psychiatric symptoms, with posttraumatic stress and depression being the most commonly examined (7–10). Similarly, moral injury is associated with increased risk of suicidal thoughts and behaviors, even after accounting for psychiatric comorbidities (11, 12).

Conceptualization and measurement of moral injury

Like other stress-related syndromes elicited by exposure to a traumatic event (e.g., posttraumatic stress disorder [PTSD]), moral injury stems from exposure to a highly stressful and morally salient event, referred to as a potentially morally injurious event (PMIE). In fact, most studies on moral injury conducted to date employ questionnaires that assess the extent to which respondents endorse being exposed to PMIEs, such as the Moral Injury Events Scale (MIES), rather than symptoms related to the psychological, emotional, social, and spiritual sequelae of exposure (i.e., moral injury (13, 14). Although information obtained from measures of exposure may facilitate case conceptualization and treatment planning, these questionnaires are not efficient screeners for identifying those in greatest need of care because most individuals exposed to a PMIE will not experience clinically significant problems. Rather, moral distress occurs on a spectrum. It can be acute and adaptive when it motivates reparation or advocacy, but at more severe levels, it can be chronic and disabling (15–17). For this reason, recently published questionnaires that screen for the psychological, emotional, social, and spiritual symptoms of moral injury rather than mere PMIE exposure are the optimal measurement tools to identify individuals likely to benefit from targeted treatment (for reviews, see (7, 18)). In the current study, we utilized one of these measures, the Moral Injury and Distress Scale (MIDS) (19).

The MIDS is a valuable tool to screen for moral injury for several reasons (19, 20). First, the MIDS anchors reported distress to a specific PMIE exposure by including an open-response item for respondents to describe the event, alongside forced-choice items that assess whether the exposure occurred because of something the respondent did, failed to do, or witnessed. This helps to eliminate ambiguity about the source of the respondent’s distress, ensuring that moral distress is distinguished from general psychological distress. Second, the MIDS is easy to use for patients and clinicians. Similar to other measures of mental health symptom severity, patients rate the extent to which they were bothered by an array of potential problems in the past month from “not at all” to “extremely.” Because the MIDS is a unidimensional tool, clinicians may sum these responses into a continuous total score, with higher values indicating greater symptom severity. Scores equal to or greater than 27 also indicate probable moral injury, signifying a need for further evaluation and potentially treatment (20). Finally, the MIDS was developed and validated using a diverse population-based sample of U.S. military veterans, healthcare workers, and first responders. In contrast, most other measures were developed solely with military veterans (7), potentially excluding civilian populations who may also be exposed to PMIEs and experience moral injury.

The current study

The past decade has seen a surge in studies of moral injury. Specifically, the recent development of questionnaires to assess the psychological, emotional, social, and spiritual features of moral injury is a notable improvement upon prior work that assessed PMIE exposure but not the impacts of exposure (19, 21). A next step in this program of research is to empirically test for distinct symptom presentations based on these questionnaires. Doing so could enable clinicians and researchers to better identify those at risk of adverse outcomes due to moral injury and tailor interventions to develop personalized treatment plans. To that end, the objectives of this study were to: 1) identify groups of participants characterized by distinct profiles of responding to the MIDS items; 2) compare mental health symptoms and psychosocial functioning across the identified groups; and 3) examine the extent to which membership in the observed profiles differs by sex, age, race, ethnicity, education level, and religious affiliation.

Materials and methods

Participants and procedures

Participants in the parent study (N = 1,232) were combat veterans, healthcare workers, and first responders, drawn from KnowledgePanel, an online survey panel with a nationally representative sample of over 50,000 households (19). For military veterans, combat status was determined by providing affirmative responses to two prescreening questions about prior military service and deployment to a combat zone. Healthcare workers and first responders were identified as having worked in their field for at least six months and either still work in or have retired no more than five years prior. Table 1 displays sociodemographic characteristics.

Table 1

Table 1. Sample characteristics for full sample and by profile.

Because the purpose of the parent study was to recruit a nationally representative sample of individuals including those with and without PMIE exposure and PMIE exposure is a prerequisite of moral injury, we added an additional inclusion criterion for this secondary analysis focused on moral injury symptoms. Specifically, participants were asked if they have been exposed to a PMIE either by participating in an event because of what they did (i.e., commission), what they failed to do (i.e., omission), or by witnessing others’ wrongful acts. Two authors (S.M., S.B.N.) reviewed the descriptions prior to statistical analysis; 279 cases were considered invalid for providing insufficient information, which made it unclear whether they had experienced a PMIE or not (e.g., “I wish I knew the things I now know”) or for giving an example that did not describe a PMIE (e.g., “Personal goals that I did not achieve”). The final sample included only respondents who endorsed exposure to a valid PMIE (n = 645). Those who denied PMIE exposure or provided insufficient information to validate the exposure were excluded from the current study.

Measures

Participants completed the MIDS alongside mental health symptom measures for posttraumatic stress, depression, and psychosocial functioning.

Moral injury and distress scale

The MIDS is a comprehensive measure of PMIE exposure and outcomes (19). The measure consists of two parts. Part 1 includes 6 items which assess the following three types of PMIE exposure: 1) participating in a PMIE by commission, 2) participating in a PMIE by omission, and 3) witnessing a PMIE. Participants were asked about the extent to which they were exposed to and bothered by each type of PMIE exposure using a 5-point Likert scale from 0 (not at all) to 4 (extremely). If respondents endorsed any MIDS Part 1 item, they were prompted to complete the MIDS Part 2 items. The MIDS Part 2 includes 18 items that assess how bothered the respondent was in the past month by possible cognitive, emotional, social, and spiritual reactions indexed to a specific PMIE. Example items include “I think about how I should have been able to do more” and “I don’t seek support because I worry others would not understand.” They responded to each item using a 5-point response format (0 = not at all, 4 = extremely). We aggregated item responses into a sum score, with higher scores indicating greater severity of moral distress, and we used the recommended cut score to identify probable moral injury (score = 27; 20). In the current sample, internal consistency of scores on the MIDS Part 2 items was excellent (α = .93).

Posttraumatic stress symptoms

The PTSD Checklist for DSM-5 (PCL-5; (22) is a 20-item self-report measure of severity of PTSD symptoms, based on criteria for PTSD outlined in the 5^th edition of the Diagnostic and Statistical Manual of Mental Disorders (23). The questionnaire asks how bothered a person was by each potential symptom over the past month. Scores are on a Likert scale of 0 (not at all) to 4 (extremely). Responses were summed, with possible total scores ranging from 0-80. Higher scores indicated more severe PTSD symptoms. A clinical cut-off score of 33 indicated probable PTSD (24). In the current sample, internal consistency of scores on the PCL-5 items was excellent (α = .97).

Depressive symptoms

The Patient Health Questionnaire-9 (PHQ-9; (25) is a nine-item measure of major depressive symptoms, which were based on criteria for major depressive disorder as outlined in the DSM-IV (26). Participants were asked about how frequently they experienced each depressive symptom over the past 2 weeks, with responses on a Likert scale from 0 (not at all) to 3 (nearly every day). Responses were summed, with possible total scores ranging from 0-27. Higher scores indicated more severe depressive symptoms. Consistent with prior literature, we used a cut-off score of 10 to indicate positive screen for a probable depressive disorder (25). In the current sample, internal consistency of scores on the PHQ-9 items was excellent (α = .92).

Psychosocial functioning

The Brief Inventory of Psychosocial Functioning (B-IPF; (27) is an abridged version of the 80-item Inventory of Psychosocial Functioning (28). The B-IPF is a 7-item measure of PTSD-related functional impact endorsed across seven domains (romantic relationships, relationships with children, family relationships, friendships, work, training/education, and daily activities). Each item represents one of the functional domains. Responses are on a 6-point Likert scale from 0 (never) to 6 (always). A total score is calculated by summing items, dividing the total score by the maximum possible score based on number of items scored, and multiplying by 100. Higher scores indicate greater functional impairment, and scores equal to or greater than 31 indicate at least a moderate level of impairment (27). In the current sample, internal consistency of scores on the B-IPF items was excellent (α = .88).

Data analysis

We conducted a latent profile analysis (LPA) to identify latent groups distinguished by unique response patterns to the 18 MIDS Part Two items assessing symptoms of moral injury. Using Mplus version 8.4 (29), we specified a series of models wherein items assessing psychological, emotional, social, and spiritual symptoms of moral injury were regressed onto a discrete latent variable representing the number of unique response patterns (i.e., profiles). The number of profiles specified for the discrete latent variable was increased iteratively in successive models. To determine the model with the optimal number of profiles, we evaluated multiple fit indices, including the Lo-Mendell-Rubin Likelihood Ratio Test (LMR) and Bootstrapped Likelihood Ratio Test (BLRT). Statistically significant p values for the LMR and BLRT suggested that the specified model fit the data better than the model with one less profile (30). Lower values for the Akaike Information Criterion (AIC), Bayesian Information Criteria (BIC), and sample size adjusted Bayesian Information Criteria (aBIC) also indicated better fit (31). Additionally, we considered entropy values and average posterior probabilities (APPs) to evaluate the quality of fit across models, with values closer to 1 and no less than .70 suggesting greater accuracy of classification (32). Finally, models containing profiles to which more than 5% of participants were assigned and models with profiles that consistently emerged across solutions were prioritized over models containing profiles with fewer participants and exhibiting less stability across solutions (33).

To validate the identified profiles by examining between group differences in mental health symptom severity, we tested for equality of mean scores on measures of posttraumatic stress, depression, and psychosocial functioning between the latent profiles using Wald chi-square tests, as recommended by Asparouhov and Muthén in their three-step approach for comparing distal outcomes in mixture models (34). We also examined factors associated with likelihood of assignment to the observed profiles by estimating a multivariable binary logistic regression. In these models, we specified profile assignment as the dependent variable, with assignment to the modal profile (minimal symptoms) as the reference category compared to all other profiles. For the regression model, we planned to aggregate symptomatic profiles for this analysis to reduce the likelihood of biased model parameters for profiles containing a small percentage of the overall sample. Independent variables, including demographic characteristics, were simultaneously entered into the model. These characteristics included sex, age, race and ethnicity, highest educational attainment, and religious affiliation. Levels of independent variables that included less than 10% of the sample were collapsed to facilitate interpretation.

Results

Model selection

Table 2 displays fit indices for models specifying one to five profiles. We evaluated the fit of each solution, from the most to least complex model, to identify the solution with the fewest number of profiles and best overall fit. The five- and four-profile models were not ideal because less than 5.0% of the sample was assigned to one or more profiles. The LMR also indicated that the five-profile solution did not fit the data better than the four-profile solution, and the four-profile solution did not fit the data better than the three-profile solution. Taken together, we concluded that the five- and four-profile models were over-extracted, given that too small a percentage of the total sample was assigned to at least one profile that exhibited only minor differences in response patterns when compared to another profile (Supplementary Figure 1). Next, the information criterion indices and BLRT suggested that the three-profile solution fit better than the two-profile solution; however, the LMR did not indicate that the three-profile solution was superior. To adjudicate these findings, we followed Nylund, Asparouhov, and Muthen’s recommendation to stop increasing the number of classes when the LMR becomes non-significant, given that the true solution can still be identified in a model with an extra profile but not in a model containing too few profiles (30). Entropy and average posterior probability values also suggested high classification accuracy for the three-profile solution; thus, participants were a good fit for the profile to which they were assigned and not any of the remaining profiles. In summary, we determined the three-profile solution to be optimal.

Table 2

Table 2. Fit indices for latent profile analysis using moral injury and distress scale part 2 items.

Profile assignment and labelling

As illustrated in Figure 1, visual inspection of the estimated item-level mean scores for the three-profile solution identified three clinically relevant profiles of moral injury symptoms (Supplementary Figure 2). Approximately three out of every four participants (74.7%, n = 482) were categorized into a minimal moral injury profile, with mean scores on each MIDS item between “not at all” and “a little bit.” About one in five participants (20.2%, n = 130) were assigned to a moderate moral injury profile, with mean scores on most MIDS items falling between “a little bit” and “moderate” levels of distress. To further characterize this group and provide clinically relevant information, we visually inspected item-level scores and found that average scores were highest on MIDS items 1 (“I think about how I should have been able to do more”) and 3 (“I feel guilty”) for respondents assigned to this profile. Next, one in twenty participants (5.1%, n = 33) were assigned to a severe moral injury profile, with mean scores indicating greater than moderate levels of distress on each MIDS item. Item-level scores were again highest on items 1 and 3, with participants also reporting elevated symptom severity on item 2 (“withdrawn from others”), item 7 (“feel helpless”), and item 13 (“not seek support because others would not understand”).

Figure 1

Figure 1. Estimated item-level mean scores for the 3-profile solution indicating severity of moral injury symptoms. Symptom severity scores on the X axis include 0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit, and 4 = extremely. The Moral Injury and Distress Scale was used to assess psychological, emotional, social, and spiritual features of moral injury in a sample of military Veterans, healthcare workers, and first responders exposed to a potentially morally injurious event (N = 645).

We then compared MIDS total scores between the profiles and found that participants assigned to the severe moral injury profile exhibited the highest levels of morally injurious psychological, emotional, social, and spiritual sequelae (M = 45.99, SE = 1.65, Range = 33.00-68.00; see Figure 2A), followed by those assigned to the moderate moral injury profile (M = 22.99, SE = 0.58, Range = 11.00-41.00; χ² = 174.05, df = 1, p <.001), and finally those assigned to the minimal moral injury profile (M = 4.90, SE = 0.19, Range = 0.00-16.00; χ² = 609.53, df = 1, p <.001). Participants assigned to the moderate moral injury profile also reported greater levels of morally injurious sequelae than those assigned to the minimal moral injury profile (χ² = 894.20, df = 1, p <.001). Using the established cut score on the MIDS for identifying probable moral injury, 100.0% of those assigned to the severe moral injury profile screened positive on the MIDS, followed by 25.4% of those assigned to the moderate moral injury profile, and none of those assigned to the minimal moral injury profile.

Figure 2

Figure 2. Differences in mental health symptom severity on measures of moral injury (A), depression (B), posttraumatic stress (C), and functional impairment (D) for participants exhibiting minimal, moderate, and severe moral injury. Scales used included the Moral Injury and Distress Scale to assess moral injury symptoms, Patient Health Questionnaire to assess depressive symptoms, Posttraumatic Stress Disorder Checklist – 5 to assess posttraumatic stress symptoms, and Brief Inventory of Psychosocial Functioning to assess functional impairment.

Mental health outcomes by profile of moral injury symptoms

We next compared mean levels of mental health symptom severity between the profiles. Participants assigned to the severe moral injury profile reported the highest level of depression (M = 14.51, SE = 1.34, see Figure 2B) and posttraumatic stress symptoms (M = 50.12, SE = 3.15, see Figure 2C). By comparison, participants assigned to the moderate moral injury profile reported significantly less severe depression (M = 9.05, SE = 0.49; χ² = 14.50, df = 1, p <.001) and posttraumatic stress symptoms (M = 27.43, SE = 1.29; χ² = 43.72, df = 1, p <.001). Those assigned to the minimal moral injury profile reported the least severe depression (M = 2.03, SE = 0.17) and posttraumatic stress (M = 6.31, SE = 0.50), relative to those assigned to the severe moral injury profile (χ^2PHQ = 85.00, df_PHQ = 1, p_PHQ <.001; χ^2PCL = 188.36, df_PCL = 1, p_PCL <.001) and those assigned to the moderate moral injury profile(χ^2PHQ = 175.71, df_PHQ = 1, p_PHQ <.001; χ^2PCL = 225.43, df_PCL = 1, p_PCL <.001).

Findings on a measure of functional impairment were similar. Participants assigned to the severe moral injury profile endorsed greater impairment (M = 52.73, SE = 4.25, see Figure 2D) compared to those assigned to the moderate moral injury profile (M = 35.86, SE = 2.06; χ² = 12.66, df = 1, p <.001) and those assigned to the minimal moral injury profile (M = 10.62, SE = 0.91; χ² = 94.04, df = 1, p <.001). Those assigned to the moderate moral injury profile also reported worse impairment relative to those assigned to the minimal moral injury profile (χ² = 105.37, df = 1, p <.001).

Over two-thirds of those assigned to the severe moral injury profile screened positive for probable depression (75.8%) or posttraumatic stress (84.8%), and about one-third of those assigned to the moderate moral injury profile screened positive for probable depression (35.4%) or posttraumatic stress (27.7%). In contrast, none of those assigned to the minimal moral injury profile scored higher than the threshold for moral injury, and very few screened positive for probable depression (4.3%) or posttraumatic stress (3.5%). Furthermore, 78.8% of those assigned to the severe moral injury profile screened positive for at least a moderate level of functional impairment, compared to 50.0% of those assigned to the moderate moral injury profile, and 12.2% of those assigned to the minimal moral injury profile.

Risk correlates of assignment to moderate and severe moral distress profiles

Finally, we examined correlates of assignment to the moderate and severe moral injury profiles compared to the minimal moral injury profile (Table 3). After accounting for all other variables in the model, Hispanic respondents of any race (n = 62) had a predicted probability (PP) of 0.468 (SE = .011) for being assigned to the moderate or severe moral injury profile versus the minimal moral injury profile. White, non-Hispanic respondents (n = 499) had a predicted probability of .220 (SE = 003), corresponding to a significantly elevated odds of assignment to the moderate/severe moral injury profiles for Hispanic respondents (adjusted odds ratio [AOR] = 2.96, 95% confidence interval [CI] = 1.70, 5.16). Odds of assignment to the moderate or severe profiles versus the minimal profile did not differ between White, Non-Hispanic respondents and non-Hispanic respondents of other races. Each additional year of age was associated with slightly lower odds of assignment to the moderate or severe profile versus the minimal profile (AOR = .987, 95% CI = .974,.999). Relative to respondents with an associate’s degree or fewer years of education (n = 286, PP = .283, SE = .005), those with a master’s degree or higher (n = 158, PP = .190, SE = .006) also had lower odds of assignment to the moderate or severe profiles (AOR = .56, 95% CI = .34,.91). No difference was observed between those with an associate’s degree or fewer years of education and those with a bachelor’s degree. Similarly, odds of assignment to the moderate or severe profiles did not differ by sex or religious affiliation.

Table 3

Table 3. Models predicting moral injury symptom profile membership.

Discussion

The purpose of this study was to identify groups of participants with distinct moral injury symptom presentations by examining response patterns to the Moral Injury and Distress Scale (MIDS) items. Overall, in our population-based sample of military veterans, healthcare workers, and first responders, three out of four individuals did not report a detectable level of moral injury symptoms, suggesting that these individuals never experienced any moral injury symptoms despite being exposed to a PMIE or their symptoms were acute and resolved by the time of the survey. This finding aligns with research on other stress-related syndromes, where most individuals exposed to highly stressful or potentially traumatic events do not experience significant symptoms and resilience is the norm (35).

However, one in 20 participants (5.1%) reported severe moral injury symptoms. These individuals reported a range of clinical features, including intractable moral emotions like guilt (36); social withdrawal (37); and a loss of meaning or purpose in life (38) – all of which are known risk factors for suicidal thoughts and behaviors (39–41). An additional one in five participants (20.2%) reported moderate levels of moral injury, which was characterized by similar symptoms of guilt and hindsight bias but with potentially less emphasis on perceived rejection by close others or valued groups and feelings of powerless related to the perceived inevitability of transgression or impossibility of repair. These individuals may also benefit from clinical interventions that target moral injury or ongoing monitoring, depending on symptom progression and associated impairment. Importantly, before the development of tools like the MIDS, there was no way to efficiently screen individuals for moral injury symptoms and determine who may benefit from targeted treatment. The MIDS fills this gap by measuring PMIE exposure, and the psychological, emotional, social, and spiritual impacts of PMIE exposure.

The MIDS also detects distress not accounted for by other commonly used screening measures like the PCL-5 (22) and PHQ-9 (25), which respectively measure posttraumatic stress and depression symptom severity. While all participants in the severe moral injury profile screened positive for probable moral injury on the MIDS, scores on measures of posttraumatic stress and depression were within normal limits for about one in ten of these participants (9.1%). Without including the MIDS in an assessment battery, these individuals may not be flagged for evaluation and treatment. This would be a missed opportunity given that screening positive for probable moral injury on the MIDS has been associated increased risk of adverse outcomes, including 3-to-6-fold elevated odds of suicidal ideation and attempt (42). Taken together, these findings suggest that individuals in high-stress, service-oriented occupations seeking care for PTSD and depression might also be evaluated and treated for moral injury, especially in cases where residual symptoms persist following delivery of frontline mental health treatments. Also, future research is needed to further explore areas of overlap and distinction between moral injury and common mental health comorbidities, especially across different samples given that rates of exposure to specific types of occupational hazards (e.g., situations involving risk of serious injury or death) vary across groups included in our sample.

Finally, those at particularly elevated risk of assignment to the moderate and severe moral injury symptom presentations included Hispanic individuals of any race versus White, Non-Hispanic individuals, younger versus older individuals, and individuals with fewer versus more years of education. These results highlight the need for more research on moral injury in racial and ethnic minority groups, who may have distinct moral frameworks or trauma histories that heighten vulnerability to moral injury (43). We also replicated prior findings showing that younger individuals report more severe moral injury symptoms than older individuals (44), and fewer years of education emerged as a risk factor in our sample. Further research is needed to explicate these differences, potentially indicating that age-related changes or years of education could prepare individuals to negotiate overlapping and potentially contradictory moral principles (45). Related, future research could broaden its scope beyond individual-level characteristics to examine other factors associated with risk of severe moral injury symptoms including features of PMIE exposure itself—for example, whether the person committed, failed to prevent, or witnessed a morally injurious act. We did not pursue this question in the current paper because the sample was limited to only those who endorsed PMIE exposure, which excluded individuals who denied PMIE exposure as a referent for comparison.

Limitations

These findings should be considered in light of three key limitations. First, while our population-based sample is more generalizable to community-based samples of military veterans, healthcare workers, and first responders, findings may differ in treatment-seeking samples. In treatment-seeking samples a higher percentage of the sample is likely to screen positive for probable moral injury, with more individuals falling into the moderate and severe profiles. This may also yield greater variability in clinical presentations of moral injury symptoms [e.g., internalizing and externalizing presentations (46, 47)]. Similarly, we aggregated groups that comprised a small percentage of our overall sample to facilitate analysis, which might conceal meaningful differences between groups. Future studies may oversample these individuals to examine differences between underrepresented groups, such as those belonging to minority racial and religious groups and the various occupations included among first responders (e.g., fire service, law enforcement service, etc.).

Second, all data were collected using self-report questionnaires, which are vulnerable to recall and social desirability biases (48). Future studies would benefit from multi-method approaches, including interviews and clinician assessments, as well as data collection immediately following when the event occurred. Notably, however, there is no clinician-administered interview specifically designed to assess moral injury severity—similar to the Clinician-Administered PTSD Scale for DSM-5 (CAPS-5; (49). Finally, the cross-sectional nature of the data limits our ability to draw conclusions about changes in moral injury symptom profiles over time. As longitudinal studies using new moral injury measures become available, person-centered methodologies (e.g., latent growth modeling) may help identify trajectories of moral injury symptoms.

Future directions for clinical research

Several psychotherapeutic treatments that address moral injury are supported by at least one randomized clinical trial (RCT) or pilot trial: Adaptive Disclosure [AD; (50)], Building Spiritual Strength [BSS; (51)], Impact of Killing [IOK; (52)], and Trauma-Informed Guilt Reduction therapy [TrIGR; (53)]. However, all of these trials were conducted before the development of scales designed to assess the symptoms of moral injury including the MIDS. Thus, eligibility decisions in these trials have depended on patients screening positive for a proxy for moral injury, most commonly PTSD. Unfortunately, to the extent that PTSD and moral injury are distinct syndromes, use of proxy measures as an eligibility criterion likely prevents some individuals from accessing needed treatments (e.g., if a person would have screened positive for moral injury when assessed with the MIDS but screened negative on a proxy like PTSD and was found ineligible). Effectiveness evaluations may also be diminished if change over the course of targeted moral injury treatment is assessed using proxy measures, especially considering that PTSD symptom severity focuses on biobehavioral fear responses that can differ substantially from moral injury symptoms. Taken together, identification of distinct moral injury clinical presentations may help to identify who needs treatment, what kind of treatment they need, and how to evaluate clinical response. It is essential that future clinical trials testing treatments theorized to support recovery from moral injury use scales specifically designed for assessing the psychological, emotional, social, and spiritual features of moral injury. The MIDS appears uniquely well-suited for use in RCTs, given that total scores equal to or greater than 27 indicate probable moral injury, which may be used as an eligibility criterion, and continuous total scores may be examined to assess change over the course of treatment.

Conclusion

In this population-based sample of veterans, healthcare professionals, and first responders who completed the Moral Injury and Distress Scale (MIDS), we identified three profiles – minimal, moderate, and severe moral injury symptoms. Although most participants reported minimal moral injury, one-quarter endorsed moderate (20.2%) to severe (5.1%) levels of moral injury symptom severity, thus highlighting the burden of moral injury in this population. Results of this study highlight the clinical utility of the MIDS as an effective tool to identify those experiencing severe levels of moral injury and those reporting subclinical levels of moral distress. Its use in measurement-based care may improve detection and treatment planning—ultimately expanding access to emerging interventions for moral injury across diverse care settings.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by University of California San Diego Institutional Review Board. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements.

Author contributions

BG: Supervision, Software, Writing – original draft, Conceptualization, Methodology, Visualization, Formal analysis. EW: Writing – original draft, Formal analysis, Visualization, Methodology, Conceptualization. MM: Formal analysis, Writing – review & editing, Methodology, Conceptualization. RP: Methodology, Conceptualization, Funding acquisition, Writing – review & editing, Formal analysis. CM: Methodology, Conceptualization, Writing – review & editing, Formal analysis. SN: Funding acquisition, Writing – original draft, Methodology, Conceptualization, Formal analysis, Project administration. SM: Methodology, Funding acquisition, Formal analysis, Writing – original draft, Project administration, Conceptualization.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Conflict of interest

The authors declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyt.2025.1691018/full#supplementary-material

References

1. Limone P and Toto GA. Origin and development of moral sense: A systematic review. Front Psychol. (2022) 13:887537. doi: 10.3389/fpsyg.2022.887537

PubMed Abstract | Crossref Full Text | Google Scholar

2. Litz BT, Stein N, Delaney E, Lebowitz L, Nash WP, Silva C, et al. Moral injury and moral repair in war veterans: A preliminary model and intervention strategy. Clin Psychol Rev. (2009) 29:695–706. doi: 10.1016/j.cpr.2009.07.003

PubMed Abstract | Crossref Full Text | Google Scholar

3. Griffin BJ, Purcell N, Burkman K, Litz BT, Bryan CJ, Schmitz M, et al. Moral injury: An integrative review. J Traumatic Stress. (2019) 32:350–62. doi: 10.1002/jts.22362

PubMed Abstract | Crossref Full Text | Google Scholar

4. Griffin BJ, Weber MC, Hinkson KD, Jendro AM, Pyne JM, Smith AJ, et al. Toward a dimensional contextual model of moral injury: a scoping review on healthcare workers. Curr Treat Options Psychiatry. (2023) 10:199–216. doi: 10.1007/s40501-023-00296-4

Crossref Full Text | Google Scholar

5. Maguen S, Griffin BJ, Pietrzak RH, McLean CP, Hamblen JL, and Norman SB. Prevalence of moral injury in nationally representative samples of combat veterans, healthcare workers, and first responders. J Gen Intern Med. (2025). doi: 10.1007/s11606-024-09337-x

PubMed Abstract | Crossref Full Text | Google Scholar

6. Litz BT, Walker HE, Pietrzak RH, and Rusowicz-Orazem L. The prevalence of moral distress and moral injury among U.S. veterans. J Psychiatr Res. (2025) 189:435–44. doi: 10.1016/j.jpsychires.2025.06.031

PubMed Abstract | Crossref Full Text | Google Scholar

7. Griffin BJ, Price LR, Jenkins Z, Childs A, Tong L, Raciborski RA, et al. A systematic review and meta–analysis of moral injury outcome measures. Curr Treat Options Psych. (2024) 12:7. doi: 10.1007/s40501-024-00342-9

Crossref Full Text | Google Scholar

8. Norman SB, Nichter B, Maguen S, Na PJ, Schnurr PP, and Pietrzak RH. Moral injury among U.S. combat veterans with and without PTSD and depression. J Psychiatr Res. (2022) 154:190–7. doi: 10.1016/j.jpsychires.2022.07.033

PubMed Abstract | Crossref Full Text | Google Scholar

9. Hall NA, Everson AT, Billingsley MR, and Miller MB. Moral injury, mental health and behavioural health outcomes: a systematic review of the literature. Clin Psychol Psychoth. (2022) 29:92–110. doi: 10.1002/cpp.2607

PubMed Abstract | Crossref Full Text | Google Scholar

10. Williamson V, Stevelink SAM, and Greenberg N. Occupational moral injury and mental health: systematic review and meta-analysis. Br J Psychiatry. (2018) 212:339–46. doi: 10.1192/bjp.2018.55

PubMed Abstract | Crossref Full Text | Google Scholar

11. Maguen S, Griffin BJ, Vogt D, Hoffmire CA, Blosnich JR, Bernhard PA, et al. Moral injury and peri- and post-military suicide attempts among post-9/11 veterans. Psychol Med. (2023) 53:3200–9. doi: 10.1017/S0033291721005274

PubMed Abstract | Crossref Full Text | Google Scholar

12. Jamieson N, Carey LB, Jamieson A, and Maple M. Examining the association between moral injury and suicidal behavior in military populations: A systematic review. J Relig Health. (2023) 62:3904–25. doi: 10.1007/s10943-023-01885-6

PubMed Abstract | Crossref Full Text | Google Scholar

13. Steen S, Law GU, and Jones C. The internal consistency of the Moral Injury Event Scale: A reliability generalization meta-analysis and systematic review. Eur J psychol Assess. (2024). doi: 10.1027/1015-5759/a000824

Crossref Full Text | Google Scholar

14. Nash WP, Marino Carper TL, Mills MA, Au T, Goldsmith A, and Litz BT. Psychometric evaluation of the moral injury events scale. Mil Med. (2013) 178:646–52. doi: 10.7205/MILMED-D-13-00017

PubMed Abstract | Crossref Full Text | Google Scholar

15. Litz BT and Kerig PK. Introduction to the special issue on moral injury: conceptual challenges, methodological issues, and clinical applications. J Traumatic Stress. (2019) 32:341–9. doi: 10.1002/jts.22405

PubMed Abstract | Crossref Full Text | Google Scholar

16. Norman S. Trauma-Informed Guilt Reduction Therapy: overview of the treatment and research. Curr Treat Options Psychiatry. (2022) 9:115–25. doi: 10.1007/s40501-022-00261-7

PubMed Abstract | Crossref Full Text | Google Scholar

17. Norman SB, Wilkins KC, Myers US, and Allard CB. Trauma informed guilt reduction therapy with combat veterans. Cogn Behav Pract. (2014) 21:78–88. doi: 10.1016/j.cbpra.2013.08.001

PubMed Abstract | Crossref Full Text | Google Scholar

18. Houle SA, Ein N, Gervasio J, Plouffe RA, Litz BT, Carleton RN, et al. Measuring moral distress and moral injury: A systematic review and content analysis of existing scales. Clin Psychol Rev. (2024) 108:102377. doi: 10.1016/j.cpr.2023.102377

PubMed Abstract | Crossref Full Text | Google Scholar

19. Norman SB, Griffin BJ, Pietrzak RH, McLean C, Hamblen JL, and Maguen S. The Moral Injury and Distress Scale: Psychometric evaluation and initial validation in three high-risk populations. psychol Trauma: Theory Res Pract Policy. (2024) 16:280. doi: 10.1037/tra0001533

PubMed Abstract | Crossref Full Text | Google Scholar

20. Maguen S, Griffin BJ, Pietrzak RH, McLean CP, Hamblen JL, and Norman SB. Using the Moral Injury and Distress Scale to identify clinically meaningful moral injury. J Traumatic Stress. (2024) 37:685–96. doi: 10.1002/jts.23050

PubMed Abstract | Crossref Full Text | Google Scholar

21. Litz BT, Plouffe RA, Nazarov A, Murphy D, Phelps A, Coady A, et al. Defining and assessing the syndrome of moral injury: initial findings of the Moral Injury Outcome Scale Consortium. Front Psychiatry. (2022) 13:923928. doi: 10.3389/fpsyt.2022.923928

PubMed Abstract | Crossref Full Text | Google Scholar

22. Weathers FW, Litz BT, Keane TM, Palmieri PA, Marx BP, and Schnurr PP. The PTSD checklist for DSM-5 (PCL-5) (2013). Available online at: www.ptsd.va.gov (Accessed August 1, 2025).

Google Scholar

23. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 5th edition: DSM-5. Washington, DC: American Psychiatric Publishing (2013). 991.

Google Scholar

24. Bovin MJ, Marx BP, Weathers FW, Gallagher MW, Rodriguez P, Schnurr PP, et al. Psychometric properties of the PTSD Checklist for Diagnostic and Statistical Manual of Mental Disorders–Fifth Edition (PCL-5) in veterans. Psychol Assess. (2016) 28:1379–91. doi: 10.1037/pas0000254

PubMed Abstract | Crossref Full Text | Google Scholar

25. Kroenke K and Spitzer RL. The PHQ-9: A new depression diagnostic and severity measure. Psychiatrics Ann. (2002) 32:509–15. doi: 10.3928/0048-5713-20020901-06

Crossref Full Text | Google Scholar

26. American Psychiatric Association. Diagnostic and statistical manual of mental disorders DSM-IV. 4th ed. American Psychiatric Association (1994).

Google Scholar

27. Kleiman SE, Bovin MJ, Black SK, Rodriguez P, Brown LG, Brown ME, et al. Psychometric properties of a brief measure of posttraumatic stress disorder-related impairment: The Brief Inventory of Psychosocial Functioning. Psychol Serv. (2020) 17:187–94. doi: 10.1037/ser0000306

PubMed Abstract | Crossref Full Text | Google Scholar

28. Bovin MJ, Black SK, Rodriguez P, Lunney CA, Kleiman SE, Weathers FW, et al. Development and validation of a measure of PTSD-related psychosocial functional impairment: the Inventory of Psychosocial Functioning. Psychol Serv. (2018) 15:216–29. doi: 10.1037/ser0000220

PubMed Abstract | Crossref Full Text | Google Scholar

29. Muthén LK and Muthén BO. Mplus User’s Guide Eighth Edition. Los Angeles, CA: Muthén & Muthén (2017).

Google Scholar

30. Nylund KL, Asparouhov T, and Muthén BO. Deciding on the number of classes in latent class analysis and growth mixture modeling: a monte carlo simulation study. Struct Equation Model: A Multidiscip J. (2007) 14:535–69. doi: 10.1080/10705510701575396

Crossref Full Text | Google Scholar

31. Jung T and Wickrama KAS. An introduction to latent class growth analysis and growth mixture modeling. Soc Pers Psychol Compass. (2008) 2:302–17. doi: 10.1111/j.1751-9004.2007.00054.x

Crossref Full Text | Google Scholar

32. Lubke G and Muthen BO. Performance of factor mixture models as a function of model size, covariate effects, and class-specific parameters. Struct Equation Model: A Multidiscip J. (2007) 14:26–47. doi: 10.1080/10705510709336735

Crossref Full Text | Google Scholar

33. Hipp JR and Bauer DJ. Local solutions in the estimation of growth mixture models. Psychol Methods. (2006) 11:36–53. doi: 10.1037/1082-989X.11.1.36

PubMed Abstract | Crossref Full Text | Google Scholar

34. Asparouhov T and Muthen B. Auxiliary variables in mixture modeling: 3-step approaches using mplus. Mplus (2013). Available online at: https://www.statmodel.com/download/3stepOct28.pdf (Accessed May 23, 2025).

Google Scholar

35. Kilpatrick DG, Resnick HS, Milanak ME, Miller MW, Keyes KM, and Friedman MJ. National estimates of exposure to traumatic events and PTSD prevalence using DSM-IV and DSM-5 criteria. J Trauma Stress. (2013) 26:537–47. doi: 10.1002/jts.21848

PubMed Abstract | Crossref Full Text | Google Scholar

36. Farnsworth JK, Drescher KD, Nieuwsma JA, Walser RB, and Currier JM. The role of moral emotions in military trauma: implications for the study and treatment of moral injury. Rev Gen Psychol. (2014) 18:249–62. doi: 10.1037/gpr0000018

Crossref Full Text | Google Scholar

37. Chesnut RP, Richardson CB, Morgan NR, Bleser JA, Perkins DF, Vogt D, et al. Moral injury and social well-being: a growth curve analysis. J Trauma Stress. (2020) 33:587–97. doi: 10.1002/jts.22567

PubMed Abstract | Crossref Full Text | Google Scholar

38. Currier JM, Holland JM, and Malott J. Moral injury, meaning making, and mental health in returning veterans: moral injury and meaning. J Clin Psychol. (2015) 71:229–40. doi: 10.1002/jclp.22134

PubMed Abstract | Crossref Full Text | Google Scholar

39. Corona CD, Van Orden K, Wisco BE, and Pietrzak RH. Meaning in life moderates the association between morally injurious experiences and suicide ideation among U.S. Combat veterans: results from the national health and resilience in veterans study. Psychol Trauma. (2019) 11:614–20. doi: 10.1037/tra0000475

PubMed Abstract | Crossref Full Text | Google Scholar

40. Levi-Belz Y, Levinstein Y, and Zerach G. The impact of moral injury on suicide risk among recently discharged Israeli veterans: A longitudinal moderated mediation model of trauma-related guilt and self-forgiveness. Psychol Trauma. (2024) 17:1023–32. doi: 10.1037/tra0001651

PubMed Abstract | Crossref Full Text | Google Scholar

41. Levi-Belz Y, Dichter N, and Zerach G. Moral injury and suicide ideation among Israeli combat veterans: The contribution of self-forgiveness and perceived social support. J Interpers Violence. (2022) 37:NP1031–57. doi: 10.1177/0886260520920865

PubMed Abstract | Crossref Full Text | Google Scholar

42. Griffin BJ, Maguen S, McCue M, Pietrzak R, McLean C, Hamblen JL, et al. Moral injury is independently associated with suicidal ideation and suicide attempt in high-stress, service-oriented occupations. npj Mental Health Research. (2025) 4:1–7. doi: 10.1038/s44184-025-00151-9

PubMed Abstract | Crossref Full Text | Google Scholar

43. Elbasheir A, Fulton TM, Choucair KC, Lathan EC, Spivey BN, Guelfo A, et al. Moral injury, race-related stress and post-traumatic stress disorder in a trauma-exposed Black population. J Psychiatr Res. (2024) 173:326–32. doi: 10.1016/j.jpsychires.2024.03.016

PubMed Abstract | Crossref Full Text | Google Scholar

44. Kim M, Jun J, Lambert J, Duah H, Tucker SJ, O’Mathúna DP, et al. Generational differences in moral injury, resilience, and well-being among nurses: predictors of intention to leave position and profession. West J Nurs Res. (2024) 46:909–18. doi: 10.1177/01939459241287458

PubMed Abstract | Crossref Full Text | Google Scholar

45. Borges LM, Desai A, Barnes SM, and Johnson JPS. The role of social determinants of health in moral injury: implications and future directions. Curr Treat Options Psych. (2022) 9:202–14. doi: 10.1007/s40501-022-00272-4

PubMed Abstract | Crossref Full Text | Google Scholar

46. Mensink B, van Schagen A, van der Aa N, and ter Heide FJJ. Moral injury in trauma-exposed, treatment-seeking police officers and military veterans: latent class analysis. Front Psychiatry. (2022) 13:904659. doi: 10.3389/fpsyt.2022.904659

PubMed Abstract | Crossref Full Text | Google Scholar

47. Mooren N, Boelen PA, Van Berlo A, and de la Rie SM. Moral injury appraisals and PTSD symptoms in treatment-seeking refugees: a latent profile analysis. Eur J Psychotraumatol. (2024) 15. doi: 10.1080/20008066.2024.2437957

PubMed Abstract | Crossref Full Text | Google Scholar

48. Chang EM, Gillespie EF, and Shaverdian. Truthfulness in patient-reported outcomes: factors affecting patients’ responses and impact on data quality. Patient related outcome measures. Patient Relat Outcome Meas. (2019) 10:171–86. doi: 10.2147/PROM.S178344

PubMed Abstract | Crossref Full Text | Google Scholar

49. Weathers FW, Bovin MJ, Lee DJ, Sloan DM, Schnurr PP, Kaloupek DG, et al. The Clinician-Administered PTSD Scale for DSM–5 (CAPS-5): Development and initial psychometric evaluation in military veterans. Psychol Assess. (2018) 30:383–95. doi: 10.1037/pas0000486

PubMed Abstract | Crossref Full Text | Google Scholar

50. Litz BT, Rusowicz-Orazem L, Doros G, Grunthal B, Gray M, Nash W, et al. Adaptive disclosure, a combat-specific PTSD treatment, versus cognitive-processing therapy, in deployed marines and sailors: a randomized controlled non-inferiority trial. Psychiatry Res. (2021) 297:113761. doi: 10.1016/j.psychres.2021.113761

PubMed Abstract | Crossref Full Text | Google Scholar

51. Harris JI, Usset T, Voecks C, Thuras P, Currier J, and Erbes C. Spiritually integrated care for PTSD: A randomized controlled trial of “Building Spiritual Strength. Psychiatry Res. (2018) 267:420–8. doi: 10.1016/j.psychres.2018.06.045

PubMed Abstract | Crossref Full Text | Google Scholar

52. Maguen S, Burkman K, Madden E, Dinh J, Bosch J, Keyser J, et al. Impact of killing in war: A randomized, controlled pilot trial. J Clin Psychol. (2017) 73:997–1012. doi: 10.1002/jclp.22471

PubMed Abstract | Crossref Full Text | Google Scholar

53. Norman SB, Capone C, Panza KE, Haller M, Davis BC, Schnurr PP, et al. A clinical trial comparing trauma-informed guilt reduction therapy (TrIGR), a brief intervention for trauma-related guilt, to supportive care therapy. Depress Anxiety. (2022) 39:262–73. doi: 10.1002/da.23244

PubMed Abstract | Crossref Full Text | Google Scholar