Impact of Intensive Treatment Programs for Posttraumatic Stress Disorder on Suicidal Ideation in Veterans and Service Members

Abstract

Intensive treatment programs (ITPs) are treating veterans with posttraumatic stress disorder (PTSD) and suicidal ideation (SI). The reduction of SI is a target to the abatement of suicide risk. This study examined whether ITPs utilizing PTSD treatments reduce SI and whether SI reduction is associated with PTSD symptom improvement. Veterans (N = 684) enrolled in a 2-week Prolonged Exposure (PE)-ITP or a 3-week Cognitive Processing Therapy (CPT)-ITP. Study data were drawn from self-report measures [PTSD Checklist for DSM-5 (PCL-5); item 9 of the Patient Health Questionnaire-9 (PHQ-9)] administered at intake and throughout treatment. The ITPs produced large treatment effects for PTSD. SI scores also decreased over time. Lower PTSD symptom severity was associated with less severe SI in both the PE-ITP and CPT-ITP. In conclusion, both PE- and CPT-ITPs effectively treat PTSD and reduce SI among veterans in as little as 2 weeks of intensive PTSD treatment.

Increasing suicide rates among U.S. service members and veterans are a major concern. For instance, veteran suicide deaths per year increased from 5,787 in 2005 to 6,139 in 2017 (Department of Veterans Affairs, 2019). An examination of suicide rates among patients receiving health care services in Veterans Health Administration (VHA) facilities between FY2000 and FY2007 demonstrated that suicide is more common among VHA patients than the general U.S. population (standardized mortality ratios for all VHA users ranged from 1.42 to 1.66; Blow et al., 2012). In 2017, the suicide rate for veterans was 1.5 times the rate for nonveteran adults (Department of Veterans Affairs, 2019).

Different variables influence the association between suicide attempts and suicidal ideation (SI) within service member and veteran populations. Within a military sample, specific aspects of SI (i.e., recency of ideation onset with the presence of a suicide plan and worst-point duration and difficulty controlling suicidal thoughts) are associated with an increased risk of suicide attempts (Nock et al., 2018). Importantly, the relationship between mental health disorders and suicidal self-directed violence (SDV) is largely mediated by SI (Borges et al., 2008; Millner et al., 2019), supporting the need to decrease SI over the course of treatment in order to reduce risk of suicidal SDV.

One particularly relevant mental health disorder affecting service members and veterans is posttraumatic stress disorder (PTSD). Between 2004 and 2012, the rate of PTSD diagnosis increased from 1.2% to 7.0% in active-duty service members, 1.8% to 6.7% in reservists, and 4.0% to 20.7% in veterans (Institue of Medicine, 2014). Among Operation Enduring Freedom (OEF), Operation Iraqi Freedom (OIF), and Operation New Dawn (OND) veterans, rates of PTSD are estimated to be 23% (Fulton et al., 2015). PTSD has been shown to be a risk factor for SI and suicidal SDV among service members and veterans (Haney et al., 2012; Ilgen et al., 2010; Naifeh et al., 2018; Ursano et al., 2016). Soldiers with SI who were identified as at risk for suicide attempts had higher odds of current PTSD (Naifeh et al., 2018). The odds of endorsing SI among OEF/OIF veterans were four times higher among those screening positive for PTSD compared to those with negative PTSD screens (Jakupcak et al., 2009). Thus, treatments that effectively target SI to decrease suicide risk among those with PTSD are imperative.

The use of specific cognitive-behavioral suicide prevention interventions to reduce SI and suicidal SDV is strongly supported (Tarrier et al., 2008). Researchers continue to develop and examine effective suicide prevention interventions for active duty and veteran populations such as supplemental brief cognitive-behavioral therapy (CBT) for recent SI with intent and/or recent suicide attempt in active-duty military personnel (Rudd et al., 2015). Importantly, interventions that address SI by highlighting the contribution of specific direct and indirect drivers of suicidality (e.g., Collaborative Assessment and Management of Suicidality [CAMS]), such as trauma-related distress, have been shown to be effective in reducing SI (Ryberg et al., 2019). The provision of evidence-based PTSD treatments alone among individuals with SI (without severe intent or planning and no recent suicide attempt) may still result in decreases in SI and reductions in future suicide risk (Bryan, 2016). Evidence-based treatments for specific mental health disorders associated with increased rates of suicide attempts and deaths by suicide, such as borderline personality disorder with and without co-occurring PTSD diagnosis (Harned et al., 2014), have been shown to reduce SI and lower suicide risk. However, studies examining the effects of evidence-based treatments for PTSD, such as Prolonged Exposure (PE) and Cognitive Processing Therapy (CPT), on reducing SI and suicide risk in veterans are limited and often lack clear definitions of suicide risk and strategies for suicide risk assessment (Holliday et al., 2019).

PE and CPT are considered a gold standard, first-line psychotherapies for PTSD (American Psychological Association, 2017; Department of Veterans Affairs, 2017; Institute of Medicine, 2008, 2014). Both PE and CPT have demonstrated moderate-to-large effect sizes for PTSD symptom improvement and loss of diagnosis (Jonas et al., 2013). In examining evidence-based treatments for PTSD, several studies have analyzed changes in SI and the relationship between reductions in PTSD symptoms and SI over the course of weekly therapy. A randomized clinical trial examining the comparative effectiveness of CPT and PE for the treatment of PTSD among a sample of rape survivors demonstrated reductions in SI (measured with the Beck Depression Inventory (BDI; Beck et al., 1961) suicide item) and that PTSD symptom severity over the course of treatment was associated with reductions in SI for both treatments, even when controlling for major depressive disorder and hopelessness (Gradus et al., 2013). When directly compared, CPT demonstrated a larger effect in the decline of SI over the course of treatment than PE. Further, although PE resulted in reduced SI, the changes in SI were not significantly related to PTSD symptom change, but rather associated only with a general improvement in functioning. Bryan et al. (2016) found that group CPT-cognitive only version and group Present Centered Therapy resulted in significant and comparable SI (measured with the Beck Scale for Suicide Ideation (BSSI; Beck et al., 1997) reductions in active-duty military personnel with PTSD. Another study utilized data from primarily male veterans who received PE as part of routine care in three VA PTSD specialty clinics to investigate the change in SI over the course of PTSD treatment (Cox et al., 2016). Results demonstrated that PE was associated with reductions in SI (measured with the BDI-II (Beck et al., 1996) suicide item) and that reductions in PTSD symptoms occurred prior to and partially explained SI reductions. The aforementioned studies delivered treatment sessions weekly. With ongoing research and implementation of intensive PTSD treatments, it is critical to understand whether these condensed treatment formats have the same effect on SI.

Intensive treatment programs (ITPs), in which evidence-based PTSD treatments are delivered daily over the course of 2–3 weeks, have begun to receive empirical support (Ehlers et al., 2014; Foa et al., 2018; Held et al., 2020; Yasinski et al., 2018; Zalta et al., 2018). For instance, in a randomized controlled trial comparing massed-PE (i.e., daily sessions for two weeks) to a minimal-contact control condition among a sample of active-duty military personnel, massed-PE led to significantly steeper reductions in SI (measured with the BSSI (Beck et al., 1979) and the BDI-II suicide item (Beck et al., 1996)) during treatment and reduction in PTSD symptoms was significantly associated with a reduction of SI (Brown et al., 2019). ITPs have been shown to be acceptable to patients (Sherrill et al., 2020), highly feasible possibly due to lower logistical barriers and limited opportunities for avoidance compared to weekly treatment sessions (Held et al., 2019a), and to produce large reductions in PTSD symptom severity comparable to well-controlled clinical trials while also showing high retention and treatment completion rates (Held et al., 2019b, 2020). ITPs that integrate PTSD focused psychotherapy and complementary interventions may be particularly useful for patients with severe, complex presentations, including elevated risk for suicide, given the programs demonstrate excellent retention and large, clinically significant reduction in symptoms in a short amount of time (e.g., daily sessions for 2 weeks; Held et al., 2019b; Rauch et al., 2020). Yet, little is currently known about the effect of intensive PTSD treatment on SI and whether PTSD symptom reductions precede reductions in SI. Using data from the PE-based ITP (PE-ITP) at the Emory Healthcare Veterans Program and CPT-based ITP (CPT-ITP) at the Road Home Program at Rush University Medical Center, the present study was the first to examine PTSD symptoms and SI over the course of 2 and 3 weeks of intensive PTSD treatment. Moreover, this study examined whether changes in PTSD symptom severity preceded changes in SI. Based on the existing literature, we hypothesized participation in both the PE-ITP and CPT-ITP would result in (a) a significant decrease in PTSD symptoms, (b) a significant decrease in SI, and (c) PTSD symptom reduction predicting a reduction in next session SI.

Materials and Method

Sample and Treatment Setting

The present study is part of a retrospective effectiveness study. The data for the present study were drawn from the 2-week PE-ITP at the Emory Healthcare Veterans Program (n = 376) and from the 3-week CPT-ITP at the Road Home Program (n = 308) as part of the program evaluation. The analyses were approved by both Emory University and Rush University Institutional Review Boards. The 684 veterans and service members (hereafter collectively referred to as “veterans”) sought treatment at each site between February 2016—April 2019. Veterans were referred from Wounded Warrior Project, VHA and community providers, and self-referral. Veterans with imminent risk of suicide (i.e., the level of suicide risk that requires intervention to maintain safety and/or suicide attempt within the past month) were excluded from the ITPs at the intake evaluation. Veterans did not receive compensation for treatment.

The average age of the sample was 40.71 years (SD = 9.43). The majority of the veterans were male (67.69%), served following the September 11th terrorist attacks (93.86%), and were active duty (9.94%) or discharged or retired from the military (85.82%). A breakdown of the demographic characteristics by ITP can be found in Table 1; both samples were comparable in age, sex, race, and ethnicity.

Table 1.

Sample Characteristics for the PE-ITP and CPT-ITP

Demographic Variable	PE-ITP, No. (%)	CPT-ITP, No. (%)	ITP differences χ2
Sex			p = .683
Female	119 (31.65)	102 (33.12)
Male	257 (68.35)	206 (66.88)
Race			p = .313
White	217 (57.71)	214 (69.48)
Black/African	122 (32.45)	58 (18.83)
American
Other	37 (9.84)	36 (11.69)
Ethnicity			p = .125
Hispanic	44 (11.70)	65 (21.10)
Non-Hispanic	317 (84.31)	242 (78.57)
Not reported	15 (3.99)	1 (0.32)

Note. PE-ITP = Prolonged Exposure-based Intensive Treatment Program; CPT-ITP = Cognitive Processing Therapy-based Intensive Treatment Program.

Procedure

PE-ITP—Emory Healthcare Veterans Program

The Emory Healthcare Veterans Program is a 2-week program utilizing a modified intensive outpatient approach to PE (Blount et al., 2014; Foa et al., 2019) that was adapted from a group-based version (Smith et al., 2015). PE-ITP involves daily 90-min individual imaginal exposure (nine sessions total) and daily 120-min group in vivo exposure (nine sessions total). Patients also complete daily out-of-session in vivo exposures and listen to audio recordings of imaginal exposures. In addition to the active treatment, veterans receive a number of other services based on their interest and treatment plan, including daily case management, medication management, relapse prevention, skills training in family/relationship management, recreational activities, career/financial support, and promotion of physical health and wellness through sleep education, nutrition education, yoga, massage, and acupuncture. See Rauch et al. (2020) for a full description of the program. The program uses a cohort model with up to 12 veterans per week.

CPT-ITP—Road Home Program

The Road Home Program is a 3-week program combining daily individual and group CPT (Resick et al., 2017). During the 3-week ITP, veterans receive daily 50-min individual CPT (14 sessions total) and daily 120-min group CPT (13 sessions total). Veterans also participate in daily mindfulness training (13 sessions total) and yoga (12 sessions total), and psychoeducation groups on a range of topics, including medication management, sleep hygiene, and communication skills. Veterans also receive case management and have the option to receive additional skills coaching to improve their interpersonal effectiveness, emotion regulation, and distress tolerance, or relapse prevention. The program utilizes a cohort model with each cohort consisting of up to 12 veterans. See Zalta et al. (2018) and Held et al. (2019b) for more detailed descriptions of the program and treatment outcomes.

Measures

All measures were administered at the intake assessment and on day 1 of treatment and again on days 3, 5, 8, and 10 (PE-ITP and CPT-ITP), 13, and 14 (CPT-ITP). Veterans completed self-report measures via iPads that utilized the secure web applications REDCap or Qualtrics.

PTSD Symptom Severity

The PTSD Checklist for DSM-5 (PCL-5; Blevins et al., 2015) is a 20-item self-report measure of the DSM-5 PTSD symptoms. All items are rated on a scale from 0 = Not at all to 4 = Extremely; higher scores indicate greater PTSD symptom severity. The PCL-5 has well-established psychometric properties in samples of service members and veterans (Bovin et al., 2016; Weathers et al., 2013; Wortmann et al., 2016). For the present study, veterans completed the PCL-5 in relation to an identified index trauma for symptoms present over the past month at the intake assessment and for symptoms present over the past week at each between session assessment time point. The internal consistency of the PCL-5 total scores across all time points for the two ITPs were .924–.925.

Depression Symptom Severity

The Patient Health Questionnaire-9 (PHQ-9; Kroenke et al., 2001) is a 9-item self-report measure of depression symptoms over the past 2 weeks. All items are rated on a scale from 0 = Not at all to 3 = Nearly every day; higher total scores indicate greater depression symptom severity. Several studies support excellent psychometrics (Löwe et al., 2004; Manea et al., 2012). For the present study, veterans completed the PHQ-9 at the intake assessment and at each between session assessment time point. The internal consistency of the PHQ-9 total scores across all time points for the two ITPs were .909–.928. The PHQ-8, which did not include the suicidal ideation item, was used in the analyses.

Suicidal Ideation. SI was assessed using item 9 of the PHQ-9 (“Thoughts that you would be better off dead or of hurting yourself in some way”; 0 = Not at all, 1 = Several days, 2 = More than half the days, 3 = Nearly every day). The PHQ-9 item 9 is commonly utilized as a standalone item to assess SI (Bauer et al., 2013; Simon et al., 2013; Uebelacker et al., 2011).

Data Analytic Plan

The sample was initially characterized using descriptive statistics, and change in depression and PTSD symptoms over time. The ordinal logistic mixed-model analysis was used to examine the trajectory of PTSD symptoms and SI severity over the course of the ITPs and to determine whether changes in PTSD symptom severity predicted SI severity in the following session. This approach can be conceptualized in a manner similar to a traditional ordinal logistic regression equation as:

$$y_{ij}=x_{ij}^{\text{'}}b+u_{0i}+e_{ij}$$

with $x_{ij}^{\text{'}}b$ representing the covariate vector and slope parameters, and u_0i representing random effects or participant-level variation. Slopes may be conceptualized as the odds of an increase in suicidal ideation response category per unit increase in X. Models initially included time, age, sex, race, and ethnicity and were further adjusted for PHQ-8 depression score, which did not include the suicidal ideation item. Next, models were adjusted for demographic covariates as well as autocorrelation in SI by including lagged-SI as a covariate. Analyses were initially conducted for the entire sample. A subsample of individuals with elevated SI was examined as a sensitivity analysis. Elevated SI was defined as a score greater than zero (i.e., 1 = Several days, 2 = More than half the days, or 3 = Nearly every day) on the PHQ-9 item nine administered at the intake assessment. All coefficients displayed in the results represent odds ratios. Effect sizes for pre–post comparisons utilize within-subjects variants of Cohen’s d (Gibbons et al., 1993) for continuous variables, and Cliff’s Delta for ordinal variables. Approximately 12.17% of all SI observations were missing in the data set overall. Mixed-effects models assume data are missing at random, and this was verified in early cleaning/screening stages as general practice (missingness was not related to any other variable of interest, including the outcome, SI, or the predictors, PCL-5 or PHQ-8). Analyses were conducted in Stata 14 (Statacorp) and Supermix 1.1 (Scientific Software International). Figures were created in Sigmaplot 13 (Systat Software).

Results

Statistically significant decreases in PTSD and depression symptom severity from the intake assessment to the completion of treatment were observed for both the PE- and CPT-ITPs, consistent with previously published subsamples (see Rauch et al., 2020 and Zalta et al., 2018). Each ITP produced large treatment effects for PTSD (PE-ITP: d = 1.25; CPT-ITP: d = 1.52) and depression (PE-ITP: d = 1.09; CPT-ITP: d = 1.05) symptom severity, and small-to-moderate decreases in SI (PE-ITP: Cliff’s Δ = .20; CPT-ITP: Cliff’s Δ = 0.27).¹ Importantly, in both ITPs only a very small percentage of individuals reported an increase in SI from intake to treatment completion (PE-ITP–3.03%; CPT-ITP–5.36%). The mean amount of change in PTSD symptom severity and SI over the course of the program was similar across ITPs (SI: PE-ITP = 0.23, CPT-ITP = 0.34; PCL-5: PE-ITP = 20.78, CPT-ITP = 21.92). Additional treatment outcome statistics are displayed in Table 2.

Table 2.

Change in PTSD and Depression Symptoms During PE-ITP and CPT-ITP

	PE-ITP				CPT-ITP
Measure	Baseline M(SD)	Completion M(SD)	t	d	Baseline M(SD)	Completion M(SD)	t	d
PCL-5	48.97 (15.69)	27.11 (19.07)	21.44***	1.26	56.56 (11.51)	32.72 (19.04)	22.08***	1.52
PHQ-9	15.96 (5.56)	9.58 (6.13)	20.68***	1.10	18.45 (4.70)	12.63 (6.24)	16.40***	1.05
PHQ-9 Item 9	.59 (.88)	.27 (.62)			.91 (1.03)	.41 (.74)

Note. PE-ITP = Prolonged Exposure-based Intensive Treatment Program; CPT-ITP = Cognitive Processing Therapy-based Intensive Treatment Program; PCL-5 = PTSD Checklist for DSM-5; PHQ-9 = Patient Health Questionnaire-9.

^***p < .001.

Ordinal Logistic Mixed-Model Analysis

Full Sample

Likelihood ratio tests and Akaike information criterion (AIC) values indicated that random intercepts models were a considerably better fit than single-level ordinal logistic models (p < .001), but random slopes components did not improve model fit. Thus, all models included only random intercept components, and unstructured covariance structure was utilized. Intraclass Correlation values of .62 for CPT and .67 for PE further supported the uility of the mixed-model approach. All reported relationships between PCL-5 and SI severity controlled for age, sex, race, and ethnicity. SI scores decreased over time in both the PE-ITP (b = 0.755, p < .001, 95% CI: 0.717–0.795) and CPT-ITP (b = 0.807, p < .001, 95% CI: 0.778–0.837). Lower PCL-5 scores were associated with less severe SI overall across time in the PE-ITP (b = 1.06, p < .001, 95% CI: 1.038–1.085) and the CPT-ITP (b = 1.02, p = .006, 95% CI: 1.007–1.042). Depression was also a significant predictor of SI over time in both the PE-ITP (b = 1.26, p < .001, 95% CI: 1.176–1.348) and the CPT-ITP (b = 1.414, p < .001, 95% CI: 1.331–1.501). In lagged models, both the PE-ITP and CPT-ITP, prior session SI severity was associated with SI severity over time (p < .001). Even after adjusting for prior session SI in lagged models, lower PCL-5 scores remained associated with less severe SI in the PE-ITP (b = 1.02, p < .001, 95% CI: 1.013–1.033) and CPT-ITP (b = 1.036, p < .001, 95% CI: 1.025–1.047).

Within-subjects and between-subjects effects for the PCL-5 were disaggregated to examine their contributions to SI. Separation of these effects resulted in a greater model fit according to likelihood-ratio tests (p < .001), suggesting invariance in the contribution of these effects. Further analysis revealed that both within- and between-subjects effects of PCL-5 were significant predictors of SI over time in the PE-ITP and CPT-ITP (p < .001). In the PE-ITP, the association between within-subjects changes in PCL-5 and SI over time (b = 1.13, 95% CI: 1.096–1.169) was similar to corresponding cross-sectional between-subjects associations (b = 1.10, 95% CI: 1.078–1.119). Similarly, in the CPT-ITP, the association between within-subjects changes in PCL-5 and SI over time (b = 1.15, 95% CI: 1.117–1.184) was also comparable to corresponding cross-sectional between-subjects associations (b = 1.04, 95% CI: 1.027–1.060).

Veterans With Elevated SI

Repeating the above analyses using only patients who reported elevated SI during the intake assessment (43.79%) as a sensitivity analysis yielded similar results to those reported above. Reductions in PCL-5 were significantly associated with reductions in SI for both PE-ITP (b = 1.09, p < .001, 95% CI: 1.063–1.126; see Figure 1) and CPT-ITP (b = 1.11, p < .001, 95% CI: 1.081–1.145; see Figure 2) when adjusting for age, sex, race, and ethnicity. When also including previous-session SI in lagged models, this relationship remained significant for both PE-ITP (b = 1.05, p < .001, 95% CI: 1.024–1.078) and CPT-ITP (b = 1.03, p = .008, 95% CI: 1.007–1.049). A total of 67.82% (196/289) of the veterans who endorsed elevated SI reported a reduction in SI.

Figure 1. Associations Between PTSD Symptom and SI Change Across PE-ITP Among Individuals with Elevated SI.

Note. Error bars represent standard error; PTSD = posttraumatic stress disorder; SI = suicidal ideation; PE-ITP = Prolonged Exposure-based Intensive Treatment Program; PCL = PTSD Checklist; PHQ-9 = Patient Health Questionnaire-9. A median split for PCL-5 change was used to identify groups for those high or low in PCL change during the program for illustrative purposes.

Figure 2. Associations Between PTSD Symptom and SI Change Across CPT-ITP Among Individuals with Elevated SI.

Note. Error bars represent standard error; PTSD = posttraumatic stress disorder; SI = suicidal ideation; CPT-ITP = Cognitive Processing Therapy-based Intensive Treatment Program; PCL = PTSD Checklist; PHQ-9 = Patient Health Questionnaire-9. A median split for PCL-5 change was used to identify groups for those high or low in PCL change during the program for illustrative purposes.

Discussion

The findings from the present study demonstrate that both PE- and CPT-ITPs effectively treat PTSD while also reducing SI among veterans. Specifically, completion of two weeks of treatment in either a PE-ITP or a CPT-ITP was associated with significant reductions in PTSD symptom severity (PE-ITP: d = 1.26; CPT-ITP: d = 1.52), as well as reductions in depression symptom severity (PE-ITP: d = 1.09; CPT-ITP: d = 1.05). Importantly, over half of the veterans who endorsed elevated SI at the ITP intake evaluation reported a reduction in SI following only two weeks of treatment. These findings are similar to studies that have evaluated the effect of traditionally delivered (weekly) PE and CPT on SI (Cox et al., 2016; Gradus et al., 2013). Given elevated SI is a key risk factor for suicidal SDV, it is imperative to reduce SI in the shortest amount of time possible. The demonstrated ability to reduce SI among veterans in as little as two weeks of intensive PTSD treatment is of importance and further supports the utilization of condensed PTSD treatment formats for veterans with PTSD and co-occurring SI.

Further strengthening these findings, reductions in PTSD symptom severity predicted subsequent decreases in SI. Although neither ITP directly targeted SI, veterans reported reduced SI following improvements in PTSD severity. This lagged effect has previously been documented for traditionally delivered PE and CPT (Cox et al., 2016; Gradus et al., 2013) and suggests that unaddressed PTSD symptoms are directly associated with SI (Naifeh et al., 2018). The findings from the present study are particularly important as they provide further support for the notion that effective treatment of PTSD can lead to significant reductions in suicide risk factors even when the treatment does not directly target SI. The mechanisms by which PTSD and suicidal SDV are related have been addressed in empirically supported conceptual models of suicide, such as the fluid vulnerability theory in which PTSD facilitates SDV because it acts on the three processes (i.e., set points such as avoidance and anxiety; thresholds of activation such as self-blame and emotion regulation deficits; and the capacity to recover from acute episodes such as impaired parasympathetic response when aroused) that determine which individuals will become suicidal (Bryan et al., 2017). Thus, it is possible that as these PTSD symptom presentations improve through PE- and CPT-ITPs, reductions in suicidal SDV naturally follow. Further, ITP models may be especially well-suited for co-occurring SI in this patient population given the high rates of treatment retention in ITP models (91%–96% treatment completion; Rauch et al., 2020; Zalta et al., 2018) compared to standard outpatient PTSD care (60% treatment completion; Kehle-Forbes et al., 2016).

Limitations

The present study has several limitations. As is typical of research in naturalistic settings, this study did not have randomization or comparison groups. Relatedly, given the structure of the ITPs, we were unable to control for the effects of other treatment components such as medication management, psychoeducation, and wellness services on PTSD symptom severity and SI. Additionally, the use of the PHQ-9 item 9 provides a limited measurement of SI. We examined the change in SI across the study sample and among those reporting clinically significant ideation at the ITP intake assessment. The sample as a whole did include many veterans who did not report SI at baseline, but results were consistent across both the full group and subsample. Finally, increasing access to evidence-based treatment through ITPs such as those featured in this study is only part of the solution to suicide prevention. While both programs demonstrated effects, a subset of veterans continued to report SI even after two weeks of treatment in the ITP, suggesting additional interventions, such as suicide-specific therapies (e.g., CAMS; Ryberg et al., 2019) and supplemental brief CBT for recent SI with intent and/or recent suicide attempt (Rudd et al., 2015), are warranted for these individuals.

Conclusion

The findings of this study support the use of both PE- and CPT-ITPs to target PTSD symptom severity and associated SI in an effective and efficient manner. Future research can expand on these results using more comprehensive measures to assess SI (i.e., assessment of suicidal intent, plan for suicide, and recent suicide attempt) and examining the maintenance of treatment gains over time. Moreover, by using more nuanced measures of behavior, future studies can determine the impact of ITPs on suicidal SDV. Closer examination of veterans with elevated acute suicide risk and patterns and predictors of treatment response in PTSD, comorbid disorders, and SI and behaviors (i.e., assess behaviors and attempts) are also warranted. Finally, additional research is needed to explore the specific mechanisms by which reduction in PTSD symptoms may reduce SI and risk for suicide.

Impact Statement.

This study supports that intensive treatment programs providing evidence-based PTSD therapies that do not directly target suicidal ideation can effectively treat PTSD while also reducing co-occurring suicidal ideation among veterans in as little as two weeks. These effective intensive treatment programs for PTSD may lead to significant reductions in important suicide risk factors among veterans in an efficient manner.