Testing adaptive interventions to improve PTSD treatment outcomes in Federally Qualified Health Centers: Protocol for a randomized clinical trial

Abstract

Background:

Posttraumatic stress disorder (PTSD) disproportionately affects low-income individuals and is untreated in 70% of those affected. One third of low-income Americans are treated in Federally Qualified Health Centers (FQHCs), which do not have the capacity to provide all patients with first-line treatments such as Prolonged Exposure (PE). To address this problem, FQHCs could use low-intensity interventions (e.g., Clinician-Supported PTSD Coach: CS PTSD Coach) and medium-intensity interventions (e.g., PE for Primary Care: PE-PC) to treat PTSD with fewer resources. However, some patients will still require high-intensity treatments (e.g., full-length PE) for sustained clinical benefit. Thus, there is a critical need to develop stepped-care models for PTSD in FQHCs.

Method:

We are conducting a Sequential, Multiple Assignment, Randomized Trial (SMART) with 430 adults with PTSD in FQHCs. Participants are initially randomized to CS PTSD Coach or PE-PC. After four sessions, early responders step down to lower frequency interaction within their assigned initial treatment strategy. Slow responders are re-randomized to either continue their initial treatment strategy or step up to Full PE for an additional eight weeks. The specific aims are to test the effectiveness of initiating treatment with PE-PC versus CS PTSD Coach in reducing PTSD symptoms and to test the effectiveness of second-stage strategies (continue versus step-up to Full PE) for slow responders.

Conclusions:

This project will provide critical evidence to inform the development of an effective stepped-care model for PTSD. Testing scalable, sustainable sequences of PTSD treatments delivered in low-resource community health centers will improve clinical practice for PTSD.

1. Introduction

Posttraumatic stress disorder (PTSD) affects 6% of the adult U.S. population^1,2 and is associated with decreased quality of life,³ increased risk of later-life physical disability,⁴ lost productivity,⁵ increased suicidal ideation,⁶ suicide attempts,⁷ and suicide mortality.⁸ Unfortunately, only 30% of individuals with PTSD receive treatment in a given year.² Moreover, although traumatic events and their sequelae disproportionately affect low-income individuals,⁹ treatment receipt is even less common among this subset of the population.^2,10,11

One third of the nation’s low-income individuals are treated in Federally Qualified Health Centers (FQHCs), which receive federal funds to provide primary care services in underserved areas.¹² Highlighting health disparities in access to services, 91% of FQHC patients are low-income, 63% are racial/ethnic minorities, and 20% screen positive for PTSD.^13,14 Despite this elevated prevalence compared to the general population, treatment for PTSD is severely limited at FQHCs.^15,16 FQHCs do not have the capacity to provide all patients with first-line, resource-intensive treatments for PTSD such as Prolonged Exposure (PE).¹⁷ Emerging data from mobile app-based interventions (e.g., Clinician-Supported PTSD Coach; CS PTSD Coach) and modified versions of evidence-based trauma-focused treatments (e.g., Prolonged Exposure for Primary Care; PE-PC) suggest that low- and medium-intensity interventions can effectively treat PTSD within about four sessions in the primary care setting.^18,19 Nonetheless, approximately two thirds of patients will require more than four sessions,^18,19 and some patients will still require high-intensity first-line treatments (e.g., PE) for sustained clinical benefit.²⁰ What remains unknown is how to effectively treat PTSD in low-resource settings with limited capacity to provide high-intensity treatment. Furthermore, no established guidelines suggest how to sequence evidence-based PTSD treatments after a suboptimal response to an initial treatment strategy.²¹ Thus, there is a critical need to develop stepped-care models to treat PTSD in FQHCs using efficacious treatment strategies. A stepped-care model has the potential to substantially impact clinical practice for PTSD treatment by improving clinical benefit, scalability, efficiency, and cost savings. This approach can be readily applied in real-world practice and community settings, which will expand access to efficacious PTSD treatment among underserved individuals.

The Sequential, Multiple Assignment, Randomized Trial (SMART) design can help develop and test stepped-care interventions.^22-24 These trials can be used to answer scientific questions about how to best initiate a stepped-care intervention, and how to best step-up the intensity of treatment for those showing early signs of suboptimal response.^25,26 Here, we employ the SMART to inform the construction of an empirically-based stepped-care model for PTSD in FQHCs. The specific aims of the project are to (1) test the effectiveness of initiating treatment with PE-PC versus CS PTSD Coach and (2) test the effectiveness of second-stage strategies (continue or step up to Full PE) for slow responders. The exploratory aims will identify the optimal stepped-care treatment sequence, conduct cost-effectiveness analysis of the stepped-care sequences, identify key mediators and moderators of treatment response, and prepare for future implementation through process evaluation.

2. Material and Methods

2.1. Study Design.

Figure 1 displays the SMART design. All eligible participants are initially randomized (with 0.5 probability) to one of the two first-stage options, either (a) PE-PC, or (b) CS PTSD Coach. Therapists provide four weekly sessions of treatment to allow sufficient time for participants to experience clinically significant improvement.^27,28 After four sessions or nine weeks (whichever occurs first), PTSD symptom change is assessed via the PCL-5.²⁹ Slow responders (see criteria below) are re-randomized (with 0.5 probability) to one of two second-stage strategies, either (a) continue in their first-stage treatment strategy, or (b) step up to full PE. Early responders step down to a lower intensity version of their current treatment strategy (i.e., early responders are not re-randomized). For those in PE-PC, this entails moving from weekly sessions to twice monthly sessions. For those in CS PTSD Coach, this entails moving into self-managed PTSD Coach. The primary outcome is PTSD severity measured by the Clinician-Administered PTSD Scale for DSM-5 (CAPS-5),³⁰ administered by independent blinded evaluators at baseline, 3-, 6-, and 9-months post-randomization.

Figure 1.

SMART PTSD study design. CS=Clinician Support, PE=Prolonged Exposure, PE-PC=Prolonged Exposure for Primary Care, SM=Self-managed

2.2. Setting and Participants.

We are recruiting participants from 8-10 FQHCs to achieve a balance of rural, urban, and suburban settings. All participating FQHCs provide both primary care-based behavioral health and longer-term behavioral health services. We plan to enroll 430 adults. The study is conducted in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans. The University of Michigan Institutional Review Board approved this study.

2.3. Inclusion/Exclusion Criteria.

Inclusion criteria for PTSD symptoms is a score of 33 or above on the 20-item PCL-5, when anchored to the index trauma.³¹ Individuals must also own a mobile device that they can use for the PTSD Coach app. Rates of smartphone ownership in the FQHC population are expected to exceed 75%.³² Exclusion criteria include: acute suicide risk, moderate-severe alcohol use disorder on Alcohol Use Disorders Identification Test (AUDIT)³³ (scores ≥ 15), high risk on NIDA-modified Alcohol, Smoking, and Substance Involvement Screening Test (NM-ASSIST)³⁴ (scores ≥ 27), unstable housing, severe cognitive impairment that in the judgment of the investigators makes it unlikely that the participant can adhere to the study regimen, psychosis or unmanaged bipolar disorder, and current engagement in another trauma-focused therapy (e.g., PE, Cognitive Processing Therapy, Eye Movement Desensitization and Reprocessing). Concurrent supportive therapy is permitted. We also exclude patients with a new prescription or dose change of a psychotropic medication in the past four weeks. Patients who do not speak English are excluded for logistical reasons.

2.4. Recruitment and Enrollment.

FQHC staff screen participants for PTSD using a 6-item PCL screener³⁵ or PC-PTSD-5 screener³⁶, which have good sensitivity and specificity in primary care settings^35,36 and correspond well to CAPS-5 scores in the FQHC setting.¹³ Patients may also self-refer to the study after learning of it through flyers or social media advertisements. Participants who screen positive receive a warm handoff to or are scheduled for a subsequent visit with research staff, who complete the remaining screening items with the participant and conduct written informed consent via SignNow, a HIPAA-compliant E-Signature service. The study team then notifies the FQHC therapist that a new participant is ready to initiate treatment. The therapist schedules the participant once they have a new opening for weekly therapy sessions. Enrolled participants are scheduled for a CAPS-5 the week before their initial appointment with the FQHC therapist.

2.5. Randomization Procedures and Decision Rule.

Following CAPS-5 administration, we randomize participants to PE-PC or CS PTSD Coach, using a secure centralized interactive web-based randomization application. Randomization is stratified by site, gender, and initial PTSD severity (PCL-5 < 48 or ≥ 48; based on the mean initial PCL-5 score in our pilot), and is computer-generated using a minimization allocation method.^37,38 After four sessions or nine weeks (in order to give participants ample time to complete four sessions), PTSD symptom change is assessed via the PCL-5, which is highly correlated with CAPS-5 (r = .878).³⁹ Participants who have experienced less than a clinically significant (15-point⁴⁰) reduction, have a PCL-5 score greater than 60 (“very severe”), or cannot be reached for assessment are classified as slow responders and are re-randomized (with 0.5 probability) to one of two second-stage strategies, either (a) continue in their first-stage treatment strategy, or (b) step up to full PE. Participants who have experienced at least a 15-point symptom reduction or have a PCL-5 score less than 29 are classified as early responders and step down to a lower intensity version of their current treatment strategy (i.e., early responders are not re-randomized). We opted to include these severity benchmarks in addition to the 15-point drop criterion in order to avoid inappropriately stepping down care for individuals with ongoing severe symptoms or intensifying care for individuals with mild symptoms. The therapist administers the PCL-5 at the beginning of session four and communicates the score to the study team – during the session if possible. The study team determines response status, conducts the second randomization if necessary, and communicates the treatment plan (i.e. second stage strategy) to the therapist during the session.

2.6. Treatment Strategies.

The three treatment strategies used in this study are: Full PE, PE-PC, and CS PTSD Coach, corresponding to high, medium, and low intensity. FQHC therapists provide therapy sessions by telehealth or in person.

2.6.1. CS PTSD Coach (Low Intensity)

is a brief psychotherapy that uses the PTSD Coach mobile app.^19,41 The PTSD Coach app incorporates evidence-based assessment, psychoeducation, and cognitive behavioral therapy and self-management strategies that are customizable to the user. Clinician support (CS) is a series of four, 30-minute sessions that teach participants to use and personalize the app to their specific concerns.¹⁹ Augmenting PTSD Coach with brief clinical support (CS PTSD Coach) results in greater daily app use, a higher likelihood of patients who still need care stepping up to full-length PTSD treatments, and greater reduction in PTSD symptoms as compared to when the app is self-managed.¹⁹ Given that these support sessions require minimal time and training, CS PTSD Coach overcomes many of the most common barriers to receiving PTSD treatment in FQHCs. CS PTSD Coach sessions focus on instructions for app use and encouraging specific PTSD Coach activities (i.e., assessments, symptom management strategies, psychoeducational readings) for completion between sessions. After four weekly sessions, therapists encourage early responders to continue app use and clinician support sessions are discontinued. Slow responders are randomized to continue with clinician support sessions (but with frequency reduced to twice monthly) or step up to Full PE.

2.6.2. PE-PC (Medium Intensity).

Treatment content for PE-PC is drawn from the Full PE model and condensed so as to deliver the most efficacious components of PE in a brief format.⁴² PE-PC is delivered in four weekly 30-45 minute sessions consisting of in-vivo and imaginal exposure. PE-PC significantly reduces PTSD and other mental health symptoms, as compared to treatment as usual, and retention is high.⁴² In our FQHC pilot, the average number of completed sessions among individuals who initiated treatment was 4.1.¹⁸ Treatment follows the standardized PE-PC manual and workbook. In vivo and imaginal exposure are introduced at the first session and reviewed at Sessions 2-4. To conduct in vivo exposure, participants repeatedly and systematically approach objectively safe people, places, objects, and situations they avoid because these stimuli remind them of their trauma. To conduct imaginal exposure, participants repeatedly and systematically approach their trauma memory and related thoughts and feelings. At the 4-session assessment point, early responders step down to every-other-week sessions for the duration of the second stage. Slow responders are randomized to continue with 8 weekly sessions of PE-PC or step up to 8 weekly sessions of Full PE.

2.6.3. Full PE (High Intensity).

Full PE is a trauma-focused PTSD treatment with substantial evidence supporting its efficacy.^17,43-45 PE includes the following elements: (a) in vivo exposure; (b) imaginal exposure to the trauma memory; and (c) facilitated emotional processing of the trauma. Treatment follows the PE manual.⁴⁶ Participants receive eight weekly 60-minute sessions. Sixty-minute sessions are non-inferior to 90-minute sessions.⁴⁷

2.7. FQHC Therapist and Evaluator Training and Consultation.

Each FQHC nominates at least two therapists to participate in the project. All participating therapists receive manualized training in PE, PE-PC, and CS PTSD Coach. Full PE training consists of a self-paced web-based PE course (PE Web; http://pe.musc.edu/) followed by a two-day virtually-delivered didactic training course. PE-PC training consists of a half-day virtually-delivered didactic course, followed by a standardized roleplay of sessions 1-4. CS PTSD Coach training consists of a half-day virtually-delivered didactic course, followed by a standardized roleplay of all sessions.

Therapists are trained to a minimal 85% competency on fidelity manual elements prior to certification in a given modality. Following the didactic training, therapists receive weekly video consultation while they complete their initial two training cases in each modality. All therapy sessions for the first two training cases are audio recorded. Certified trainers review the audio recordings and provide feedback to therapists on each session.

2.7.4. Evaluator Training:

Independent evaluators complete CAPS-5 training through the VA National Center for PTSD web-based training and achieve at least 90% agreement prior to conducting assessments.⁶⁵ Consultation calls occur biweekly and include discussion of cases, session audio review, and discussion of fidelity if indicated.⁴⁸ Interrater reliability for CAPS-5 is assessed throughout the study period on 20% of randomly selected CAPS-5 recordings.⁴⁸

2.8. Fidelity Measures.

Therapists conduct each session with the aid of Session Content Checklists to ensure fidelity. All sessions are audio recorded and treatment fidelity is assessed by Session Audio Fidelity Monitoring, in which we review 100% of session recordings for the first two training cases in each modality and thereafter a randomly selected 20%. Trained fidelity evaluators review sessions for fidelity (adherence and competence) using a fidelity manual containing checklists of essential treatment components, and also monitor sessions for proscribed content.⁴⁸ Incidents of inadequate fidelity are addressed in weekly consultation calls.⁴⁸

2.9. Risk Assessment:

We assess suicide risk at each therapy session and assessment point. We manage suicidal ideation via a structured protocol. If a participant endorses suicidal ideation, we ask about plan and intent. Depending on the participant’s responses, and in consultation with the PI, the study team member or FQHC therapist determines the appropriate disposition (e.g., create safety plan; request wellness check; activate emergency medical services).

2.10. Data Collection Procedures.

Assessments are performed at five time points: baseline, session 4, endpoint (3-months), and 6-, and 9-month follow-up. The endpoint assessment is performed approximately 8 weeks following the initiation of the second treatment stage (see Figure 1). Self-report assessments are administered via Qualtrics. Participants are reimbursed for their time by mailed gift card, at the rate of $20 for completing the screen, $40 for the baseline, $25 for the 4-session assessment, $40 for the 3-month, $50 for the 6-month, and $60 for the 9-month assessment.

2.11. Variables and Measures.

Table 1 shows the study measures. The primary outcome is 3-month score on the Clinician-Administered PTSD Scale for DSM-5 (CAPS-5),³⁰ an interview measure of PTSD severity with excellent psychometrics.⁴⁹ The CAPS-5 is anchored to the index trauma. The secondary outcome is score on the PCL-5,²⁹ a 20-item self-report measure of PTSD symptoms as defined by the DSM-5, with strong internal consistency, test-retest reliability, and convergent and discriminant validity.³¹ Participants are prompted to respond to PCL-5 items in reference to their index trauma, as identified via the Trauma History Screen,⁵⁰ a self-report measure of exposure to traumatic events.

Table 1.

Assessment Schedule and Measures

	Screening	Baseline	Intervention - Session 4	Follow-up – 3-month	Follow-up – 6-month	Follow-up – 9-month
Screen
The Trauma History Screen (THS)50	X
Outcome Measures
Clinician rated PTSD symptoms (Clinician-Administered PTSD Scale for DSM-5; CAPS-5)30		X		X	X	X
Self-report PTSD symptoms (PTSD Checklist for DSM-5; PCL-5)29	X	X	X	X	X	X
Depression (Patient Health Questionnaire; PHQ-9)70	X			X	X	X
Suicidality (Columbia Suicide Severity Rating Scale; C-SSRS)71	X	X		X	X	X
Quality of Life/Functioning (EQ-5D-5L)51		X		X	X	X
Patient Satisfaction (Client Satisfaction Questionnaire; CSQ-8)72				X	X	X
Modified Cornell Service Index52		X		X	X	X
Mediators/Mechanisms
Posttraumatic cognitions (Posttraumatic Cognitions Inventory; PTCI-9)73		X	X	X	X	X
Coping Self-efficacy (PhenX Toolkit Self-Efficacy Scale)74		X	X	X	X	X
Moderators
Demographics	X	X
Alcohol and substance use (Alcohol Use Disorders Identification Test; AUDIT33), National Institute on Drug Abuse (NIDA) Quick Screen and the NIDA-modified Alcohol, Smoking, and Substance Involvement Screening Test (NM-ASSIST34)	X			X	X	X
Diagnostic Interview for Anxiety, Mood, and Obsessive-Compulsive and Related Neuropsychiatric Disorders (DIAMOND)75		X

2.12. Cost Effectiveness.

We will estimate the direct costs from the health system perspective for the embedded treatment sequences. Labor input will include the time needed to train the therapists in the treatment strategies, assess patients for PTSD, and conduct the sessions. We will apply appropriate wage rates (including fringe benefits) for staff to the labor time to derive labor cost. We will divide the total labor cost by the number of participants in the treatment strategies to derive per-participant treatment cost. We will calculate quality adjusted life years (QALYs) lost or gained due to the interventions based on the EQ-5D-5L.⁵¹ Receiving PTSD treatment may lead to a change in participants’ healthcare utilization. Therefore, we will collect cost data for all health services received by the participants one year before study enrollment, as well as during the period of study participation, using a modified version of the Cornell Services Index.⁵² This survey measures use of outpatient (general medical health, mental health), inpatient, emergency department services, and other supportive services. We will analyze both within-patient and between-patient changes in resource utilization patterns for each treatment strategy.

2.13. Process Evaluation.

We will use process evaluation to identify barriers and facilitators to implementation efforts.⁵³ We will gather information from 10-20 FQHC therapists, 10-20 patients, and 5-10 administrators on perceptions of the embedded treatment strategies and implementation barriers.⁵⁴ Interviews will encourage reflection on the implementation of the embedded stepped-care sequences, help provide insight into their effectiveness, improve our understanding of mechanisms of change, and generate ideas for sustainability.⁵⁵ We will assess selected constructs from the Consolidated Framework for Implementation Research (CFIR) including Outer Setting factors (e.g. external policy and incentives, patient needs and resources), Inner Setting factors (e.g. culture, compatibility, implementation climate, leadership engagement, and available resources), Intervention Characteristics (e.g. relative advantage, adaptability), and Characteristics of Individuals (e.g. knowledge and beliefs about treatment).⁵⁶

2.14. Advisory Panel.

We constructed an advisory panel to help prepare for future implementation. The advisory panel meets semi-annually to refine study procedures, address organizational factors that affect delivery of the embedded treatment sequences, and share best practices between FQHCs for providing PTSD treatment. Each FQHC has contributed at least two members to the panel, including frontline therapists, leaders, and administrators.

2.15. Analysis Strategy.

Baseline variables will be examined for comparability between groups. Variables found to be confounded with either treatment or site will be considered as covariates in outcome analyses. The analytic cohort is intent-to-treat.

Specific aim 1: Test the effectiveness of initiating treatment with PE-PC versus CS PTSD Coach in reducing PTSD symptoms at 3 months (endpoint), 6 months, and 9 months.

We hypothesize that initiating treatment with PE-PC (vs. CS PTSD Coach) will result in greater PTSD symptom reduction.

The analysis of the primary endpoint will compare the first-stage strategies, namely the participants randomized to PE-PC (i.e., cells A+B+C in Figure 1) vs. CS PTSD Coach (cells D+E+F in Figure 1) using a linear mixed model (LMM).⁵⁷ The model will include PTSD symptoms (CAPS-5 score) at baseline and Month 3 as the outcome variable and will include a PE-PC (vs. CS PTSD Coach) group indicator, a 3-month (vs. baseline) time indicator and their interaction. We will also visualize the PTSD symptom data over all assessment times and model the longitudinal data by extending the LMM model with appropriate parametrization of time to reflect the long-term trends seen over the follow-up time between the two groups.

Specific aim 2: Test the effectiveness of second-stage strategies (continue or step up to Full PE) for slow responders.

We hypothesize that stepping up to Full PE (vs. continuing with initial strategy) will result in greater symptom reduction.

The analysis of the Secondary Aim will compare the two second-stage strategies among slow responders, namely Continue (i.e., cells B+E in Figure 1) vs. Step-Up (cells C+F in Figure 1) on change in CAPS-5 score from baseline to Month 3. This analysis will be similar to the Aim 1 analysis using LMM, but (a) will include only slow responders to the initial intervention; and (b) the group indicator will be defined as Step-up (cells C+F in Figure 1) vs. Continue (i.e., cells B+E in Figure 1).

Exploratory Aim A: Identify the optimal stepped-care treatment sequence.

We will compare the four stepped-care treatment sequences embedded in the SMART in terms of PTSD symptom reduction at 3-, 6-, and 9-month follow-up. Table 2 describes the 4 embedded stepped-care interventions and their corresponding experimental conditions (cells). We will use the Weight and Replicate (W&R) method,^58,59 which employs standard regression analysis, with a small adjustment involving weighting and replication. Robust standard errors are used to account for the “double use” of the early responders.⁶⁰ We will use the Multiple Comparison with the Best (MCB) approach to compare embedded adaptive stepped-care treatment sequences.⁶¹

Table 2.

Embedded Adaptive Interventions

#	Initial Treatment	Early Responder?	Second Randomization	See Design Figure 1
1	PE-PC	Yes	PE-PC 2x month	A+B
		No	Continue
2	PE-PC	Yes	PE-PC 2x month	A+C
		No	Full PE
3	CS PTSD Coach	Yes	SM PTSD Coach	D+E
		No	Continue
4	CS PTSD Coach	Yes	SM PTSD Coach	D+F
		No	Full PE

Exploratory Aim B: Conduct cost-effectiveness analysis.

We will determine the cost-effectiveness of each embedded stepped-care treatment sequence relative to the other embedded stepped-care treatment sequences in terms of quality-adjusted life-years (QALYs). As recommended by the Panel on Cost-Effectiveness in Health Care,⁶² we will use QALYs saved as the effectiveness measure. We will measure utility values using the EQ-5D-5L at baseline and follow-up interviews to calculate QALYs during the study period. We will calculate the incremental cost-effectiveness ratio (ICER) as the difference in the cost per participant between the embedded treatment sequence groups (see Table 2) divided by the difference in the QALYs per participant in each group.

Exploratory Aim C: Confirm the mechanism and identify key mediators/moderators.

We hypothesize that the interventions will engage the target mechanism of coping self-efficacy.^63,64 We will also test whether effects are mediated by treatment adherence or posttraumatic cognitions,^63,65,66 and whether they are moderated by initial severity, gender, and age. For moderator analysis, we will examine whether candidate moderators (see Table 1) predict (a) who will benefit more from initiating with PE-PC versus CS PTSD Coach, and (b) among slow responders, who will benefit more from stepping up to Full PE versus continuing in their initial intervention. To do this, we will apply Q-learning,⁶⁷ which is a generalization of moderated regression analysis to sequences of treatments.

Exploratory Aim D: Prepare for future implementation through process evaluation.

We will conduct interviews with patients, therapists, and administrators and analyze qualitative data using Rapid Analysis.⁶⁸ For therapist and administrator interviews, CFIR domains will act as the preliminary set of codes but will be modified and expanded based on the data. We will analyze patient interviews using a coding template developed in our pilot. The advisory panel will review process evaluation findings and provide input on relevant and appropriate implementation strategies to overcome barriers and capitalize on facilitators for future research and implementation activities.⁶⁹

2.16. Power Analysis.

Aim 1.

We will enroll 430 participants. This sample size was planned to provide 80% power to detect a small to moderate effect size with a 0.05 two-sided test for the outcome of PTSD symptoms. Assuming 30% lost to follow-up at the primary endpoint of 3 months post-randomization, we expect complete data for the primary outcome at 3 months from a total of ~300 study enrollees. Assuming within-person correlation of 0.56, Aim 1 analyses based on an LMM model will provide 80% power to detect standardized mean difference in CAPS-5 score from baseline to Month 3 of at least Cohen’s d=0.30 (considered small to moderate) between participants initially assigned to PE-PC vs. CS PTSD Coach. This anticipated effect size is consistent with prior work by our team.

Aim 2.

We anticipate an overall 67% to be slow responders. With the anticipated 10% lost to follow-up at the second stage randomization time and another 20% lost at 3 Months, we expect 259 slow responders randomized at 4 sessions/9 weeks (= 430*0.67*0.9) with n=207 providing 3-month data (about 103 in each group of B+E and C+F in Figure 1). This will provide 80% power for detecting a standardized difference of at least Cohen’s d=0.39 (considered moderate) between Step up vs. Continue among slow responders averaged across the initial treatment. Thus, we are adequately powered to meet this aim given that this anticipated effect size is consistent with prior work by our team.

3. Discussion.

Upon the successful completion of this project, we expect that we will have determined which stepped-care approach is most effective for reducing PTSD symptoms among patients in the FQHC setting. We expect to determine: (1) whether it is better to initiate treatment with PE-PC or CS PTSD Coach; (2) whether it is better for slow responders to step up to Full PE or continue in a given treatment strategy; (3) whether or not the treatment strategies exert their effects through coping self-efficacy; (4) the relative cost-effectiveness of the embedded treatment sequences; and (5) the barriers and facilitators to future implementation. Answering these questions will yield an empirically-based stepped-care model for PTSD that can be readily deployed in the FQHC setting. This contribution will improve clinical practice for PTSD by establishing an effective treatment sequence that can be delivered in a scalable, sustainable manner. Initiating treatment for PTSD with low- or medium-intensity options will impose less burden on the patient and on the healthcare system. Knowing at an early stage of treatment whether it is beneficial to continue, decrease intensity, or increase intensity will empower patients and practitioners to select an empirically-based course of action. Understanding the moderators of treatment can lead to more deeply tailored interventions that can be tested in future work. For these reasons, developing an empirically-based stepped-care intervention will have a major impact for low-income patients suffering from PTSD. Future research will compare the optimized stepped-care intervention developed in this project to a suitable control (to confirm its effectiveness), and test implementation strategies based on our process evaluation.