Past-Year Treatment Utilization Among Individuals Meeting DSM-5 PTSD Criteria: Results From a Nationally Representative Sample

Data were obtained from the National Epidemiologic Survey on Alcohol and Related Conditions–III (NESARC-III), a nationally representative population-based survey of the adult civilian population of the United States. The sample includes 36,309 persons living in households or noninstitutionalized group quarters and veterans not currently on active duty, who were selected via multistage probability sampling. Participants were interviewed by using the Alcohol Use Disorder and Associated Disabilities Interview Schedule–5 (AUDADIS-5; [11]), a semistructured diagnostic interview designed to measure lifetime and past-year DSM-5 criteria for psychiatric conditions. AUDADIS-5 concordance with DSM-5 PTSD has been shown to be fair to moderate (k range=.34–.46), and dimensional scale concordance for all symptoms combined was good (intraclass correlation [ICC]=.69; [12]). The present investigation utilized a subset of individuals from the NESARC-III who met DSM-5 criteria for lifetime (N=2,339) and past-year (N=1,779) PTSD.

PTSD symptom count was used as a proxy for symptom severity and was calculated by summing the number of symptoms of re-experiencing, avoidance, negative alterations in mood and cognition, and hyperarousal. The possible range for individuals meeting criteria was six to 20 symptoms.

Sociodemographic measures were obtained through the AUDADIS-5 and included age, sex, race/ethnicity, education level, marital status, ever active-duty military, and health insurance. The presence or absence of lifetime mood (major depressive disorder, dysthymia/persistent depressive disorder, bipolar I disorder, and bipolar II disorder), phobia (specific phobia and social phobia, agoraphobia), anxiety (generalized anxiety disorder and panic disorder), alcohol use, and drug use disorders (alcohol, cannabis, opioid, cocaine, stimulant, and heroin use disorders) were calculated and included in the modeling procedures.

The 12-Item Short-Form Health Survey (SF-12) is a modified and shortened version of the 36-Item Short Form Health Survey (SF-36), designed to measure eight constructs of functioning, including physical function, mental health, role functioning (emotional), role functioning (physical), social functioning, bodily pain, vitality, and general health (13). The present study included the mental component, mental health, social functioning, and role functioning (emotional) scales as factors related to treatment utilization. The SF-12 has demonstrated reliability and validity in both clinical and population-based applications (14).

Data analyses were conducted with R statistical software and the “survey” package for analysis of complex survey samples. Weighted percentages, with standard errors adjusting for the complex survey design using Taylor series linearization, were calculated and reported for all variables of interest. Logistic regression modeling also accounted for the complex survey design and was used to examine associations between the a priori specified independent variables of interest and the odds of initiating treatment for PTSD within the past year.

Of the 36,309 individuals in the NESARC-III sample, 2,339 individuals (6.1% weighted proportion of total sample) met the DSM-5 criteria for lifetime PTSD, and 1,779 (4.7% weighted) met the criteria for past-year PTSD. Of those meeting lifetime PTSD criteria, 554 (24%) utilized treatment for PTSD in the past year. Of those meeting PTSD criteria in the past year, 528 (30.3%) utilized treatment for PTSD in the past year.

Table 1 presents the results of the unadjusted and adjusted logistic regressions comparing past-year treatment utilization among the subsets of NESARC-III participants who met the criteria for PTSD in their lifetime and in the past year, respectively, by characteristic. In the unadjusted models, the following variables were related to greater odds of utilization of PTSD treatment for individuals who met the lifetime criteria: being between the ages of 30 and 44 versus ages 18 to 29; having some college education or having graduated college versus having less than a high school education; having health insurance versus no insurance; having a comorbid mood, phobia, anxiety, or alcohol use disorder versus not having those comorbidities; and endorsing a greater number of symptoms of PTSD. Factors related to lower odds of past-year PTSD treatment utilization for individuals meeting lifetime criteria included: being 65 or older versus ages 18 to 29; identifying as Asian versus white non-Hispanic; and having higher scores on the SF-12 component, mental health, social functioning, and role functioning (emotional) subscales. After adjustment, being ages 30 to 44, having lifetime mood and alcohol use disorders, and scores on the SF-12 mental component, mental health, and role functioning (emotional) subscales were no longer significant. However, identifying as non-Hispanic black was significantly associated with lower odds of past-year treatment utilization after adjustment. The adjusted model for individuals who met past-year criteria was similar to that for individuals who met lifetime criteria, except that identifying as Asian and having a comorbid phobic disorder were not significantly associated with treatment utilization.

In the present study, an estimated 6.1% of adults in the United States met DSM-5 criteria for PTSD in their lifetime (and were included in the study sample). Of the subset of individuals with lifetime PTSD, 23.9% reported having sought treatment for their trauma-related symptoms in the past year. This result is similar to a previous investigation that reported that rates of treatment for PTSD (and other anxiety disorders) tend to be less than that of mood disorders, which was estimated to be 37% (3). In general, our findings suggest that past-year PTSD treatment utilization may be more likely for those with a college education, health insurance, a comorbid phobia disorder, and greater PTSD symptom burden. Past-year treatment utilization may be less likely for individuals who are older than 65, those who identify as black or Asian, and those with greater social functioning. These findings highlight subgroups who may be at risk for not receiving or having access to treatment that may be beneficial for their PTSD.

Similar to previous research, health insurance was a significant factor in predicting past-year treatment utilization among individuals meeting the criteria for PTSD (7). Individuals with access to health insurance had nearly three times greater odds of utilizing treatment for PTSD than their counterparts without insurance. This finding highlights the importance of increasing access to health insurance for mental health treatment, given that lack of access may represent a significant barrier to treatment utilization.

In sum, these findings reinforce trends in the treatment utilization literature for PTSD that suggest that there are groups of individuals who tend to not utilize treatment that may benefit them. Although the types of barriers experienced by these various groups may not be homogenous, there are likely interventions that can increase the likelihood of treatment use. For example, improving access to health insurance may increase the odds of treatment utilization. Increasing resources devoted to mental health treatment at hospitals and community mental health centers may facilitate the hiring of additional qualified staff and provide training for evidence-based PTSD treatment. Additionally, training of frontline medical staff (e.g., nurses and physicians) in techniques such as motivational interviewing may be utilized to decrease a patient's ambivalence toward treatment and increase motivation for change. Finally, placement of mental health providers in integrated care settings to enable “warm hand-offs” between medical and mental health staff following positive screening may help facilitate transfer of care and increase the probability of follow-through of the referral (15).

There were several limitations in this study. First, because of the cross-sectional nature of the survey data, the associations reported should not be interpreted as causal. Additional research is needed to determine the directionality of these relationships. Second, although the AUDADIS-5 assesses symptoms based on DSM-5 criteria for PTSD, it is not a clinical interview and, thus, not equivalent to a formal diagnosis. Third, PTSD symptom count may be interpreted as a proxy variable for severity; however, the presence or absence of a symptom does not impart information regarding the frequency or intensity of that symptom, which are better indicators of severity. Gathering this type of data would allow a more fine-grained analysis of the symptom presentations that are associated with treatment seeking. Last, PTSD treatment utilization was a dichotomous variable and does not indicate what type of treatment was used, the setting in which it was obtained, or the outcome of the treatment. In future studies, researchers may want to gather additional information regarding the type, quality, and results of treatment obtained.

Our investigation examined the prevalence of PTSD that met DSM-5 criteria in a nationally representative survey and aimed to identify variables related to treatment utilization among individuals meeting criteria for the diagnosis. Results highlight important demographic and person-level variables that may be related to underutilization of mental health treatment services. Providers of these services should consider these factors when assessing for mental health needs or implementing treatment programs. Further, populations with a low likelihood of utilizing treatment may benefit from outreach programs, the development and provision of alternative treatments that better meet their needs, and investment of additional resources for mental health services. Future research should continue to explore and assess factors related to individuals engaging in treatment for PTSD to ensure that effective treatments are accessible and utilized by those who would benefit from them.

This report was prepared by using a limited-access data set obtained from the National Institute on Alcohol Abuse and Alcoholism (NIAAA) and does not reflect the opinions or views of NIAAA or the U.S. Government.