Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Dec 1.
Published in final edited form as: Behav Res Ther. 2023 Nov 8;171:104428. doi: 10.1016/j.brat.2023.104428

Predictors of Dropout in Cognitive Processing Therapy for PTSD: An Examination of In-Session Treatment Processes

Danielle R Shayani a, Caroline A Canale a, Denise M Sloan b,c, Adele M Hayes a
PMCID: PMC10826797  NIHMSID: NIHMS1946574  PMID: 37952284

Abstract

Dropout rates for treatments for adult posttraumatic stress disorder (PTSD) are high. Process research can reveal client factors during treatment that predict dropout. An observational coding system was used to code client processes in audio-recorded early sessions of cognitive processing therapy (CPT), a gold-standard treatment for PTSD. Data are from a randomized controlled noninferiority trial of CPT and written exposure therapy (WET), with higher rates of dropout in CPT than WET (39.7% vs. 6.4%). Participants in this study were 53 treatment-seeking adults with PTSD who were in the CPT arm of the trial and completed the CAPS-5 at pretreatment and at least one session. Of these, 15 (28.3%) dropped out of CPT early (completing ≤ 9 sessions) and 38 (71.7%) completed treatment. Sessions were coded with an observational coding system on a four-point scale (0=absent to 3=high) for maladaptive trauma-related responses (overgeneralized beliefs, ruminative processing, avoidance), affective engagement (negative emotions, physiological distress), and adaptive processing (cognitive emotional processing). Binary logistic regressions showed that more physiological distress and cognitive emotional processing predicted lower dropout, whereas more avoidance predicted higher dropout. Negative emotion, ruminative processing, and overgeneralization were not significant predictors. These findings highlight potential early indicators of treatment engagement that could be targeted to reduce dropout and perhaps facilitate further therapeutic change.

Keywords: trauma, PTSD, dropout, therapy process, cognitive processing therapy

Posttraumatic Stress Disorder (PTSD) is a prevalent and debilitating condition that affects about 8% of adults in the United States (Calabrese et al., 2011; Kilpatrick et al., 2013) and approximately 1% to 9% worldwide (Atwoli et al., 2015). Fortunately, several trauma-focused treatments for PTSD have strong empirical support with demonstrated efficacy and effectiveness (American Psychological Association [APA], 2017; Cusack et al., 2016; Department of Veterans Affairs and Department of Defense [VA/DoD], 2017). According to the APA guidelines, treatments that are strongly recommended for adult PTSD include cognitive processing therapy (CPT: Resick et al., 2017), cognitive therapy (Ehlers & Clark, 2000), prolonged exposure (PE: Foa et al., 2019), and cognitive behavioral therapy (CBT) more broadly.

Although clinical trials of these evidence-based treatments demonstrate moderate to large effect sizes in both civilian (Cusack et al., 2016) and veteran populations (Monson et al., 2006; Schnurr et al., 2022), a significant minority of individuals do not respond (Holmes et al., 2019; Kline et al., 2018), and premature treatment dropout is a concern. Recent meta-analyses have reported dropout rates of approximately 34% in randomized trials of CPT and 29% in PE (Varker et al., 2021). Dropout rates are even higher among active duty and veteran populations, ranging from 31% to 50% (Eftekhari et al., 2013; Imel et al., 2013; Steenkamp et al., 2020). Although there is evidence that some who drop out of PTSD treatment do improve, the majority do not (Holmes et al., 2019; Larsen et al., 2023).

It would be helpful to identify client factors that predict dropout risk so that clinicians can tailor PTSD interventions to maximize treatment completion and outcomes. Current research on dropout in adult PTSD treatment has focused primarily on fixed and relatively immutable patient characteristics such as demographics (e.g., age, sex) and pretreatment symptoms and functioning, but this approach has not yielded consistent predictors of dropout (Cooper et al., 2018). Similarly, in the randomized noninferiority trial examined in the present study (Sloan et al., 2018), none of the pretreatment variables predicted dropout. Analyses included a range of demographic variables and pretreatment variables, such as IQ, age, educational level, household income, sex, military veteran status, baseline PTSD symptom severity, and treatment expectancy.

Specific clusters of PTSD symptoms (rather than total scores) and functional impairments at pretreatment have yielded useful information about predictors of dropout. For example, more pretreatment avoidance (Bryant et al., 2003; Garcia et al., 2011) and re-experiencing (Garcia et al., 2011) have been associated with higher dropout. Further research has investigated the relationship between physiological arousal and dropout; however, the findings are inconsistent. Higher pretreatment hyperarousal symptoms (Garcia et al., 2011; Zayfert et al., 2005), as well as expression of physiological distress in between-session written narratives (Alpert et al., 2020), have been identified as predictors of dropout in prior studies. In contrast, findings from a recent study examining pretreatment heart rate reactivity suggests that lower heart rate reactivity was associated with increased likelihood of dropout (Carpenter et al., 2021). Other variables that have been identified as pretreatment predictors of dropout are more catastrophic cognitions (Bryant et al., 2003), anger (Keefe et al., 2018; Rizvi et al., 2009), current partner conflict (Keefe et al., 2018), and impairment in social functioning (Zayfert et al., 2005).

Machine learning approaches have been used to identify groups of individuals with different trajectories of treatment dropout (Taubitz et al., 2022). One study used an iterative Random Forest algorithm to examine associations between individual characteristics and session attendance in a treatment for comorbid PTSD and substance use disorder (López-Castro et al., 2021). In older patients, a steeper decrease in PTSD symptoms was associated with completing more sessions, whereas in younger patients, this pattern of symptom reduction was associated with completing fewer sessions. Across all patients, full-time employment was associated with attending fewer sessions.

Cooper and colleagues (2018) urge researchers to rethink their approach to investigating dropout, highlighting the limits of focusing only on demographics and other pretreatment variables. They suggest that research efforts should shift towards investigating client factors that unfold during treatment, which might reveal more about the process of dropout. Thus, the interval between pre- and post-treatment has the potential to provide rich data and valuable insights on early predictors of dropout.

Two examples of changes during the course of treatment can be found in studies using observational coding. In trauma-focused cognitive behavioral therapy (TF-CBT) for youth, observational coding revealed that more in-session avoidance in the first seven sessions predicted dropout (Yasinski et al., 2018). In another study using the same observational coding system to code the CPT narratives from the present clinical trial (Sloan et al., 2018), Alpert et al. (2020) reported that more negative emotions and ruminative processing predicted less dropout, whereas more overgeneralized beliefs and physiological reactions predicted more dropout. Avoidance in the narratives did not predict dropout, but the authors note that in-session recordings might reveal more comprehensive and contextual information on avoidance than written narratives.

Other research has examined PTSD symptom trajectories as predictors of dropout. Using self-report measures during the course of Prolonged Exposure (PE; Foa et al., 2019) in a national sample of veterans, Eftekhari and colleagues (2020) found that symptom worsening did not predict dropout. Similarly, in a study among veterans receiving CPT, Larsen et al. (2022) found that self-reported PTSD symptom increases did not predict dropout. In another study with veterans enrolled in time-limited evidence-based psychotherapy for PTSD, Larsen et al. (2023) reported a dropout rate of 54%. The majority of those who dropped out either worsened or did not report symptom improvements. However, the authors note that 27.7% of the participants reported significant symptom improvement prior to dropping out.

Another line of qualitative research has identified clients’ reasons for discontinuing PTSD treatments. This research has examined clients’ self-reported reasons for dropping out of PTSD treatment prematurely, as well as perspectives from their therapists (Browne et al., 2021; Hundt et al., 2020; Kehle-Forbes et al., 2022; Meis et al., 2023). A study that involved interviewing 23 participants who did not complete PE or CPT at a VA clinic reported several reasons for drop out, including issues related to the therapist or therapy content (e.g., misunderstanding of therapy rationale, preferring a present-focused therapy) and emotional barriers (e.g., treatment was too stressful; Hundt et al., 2020). Another study used a mixed deductive/inductive approach to understand factors and processes that contributed to veterans’ decisions to dropout of PE and CPT and identified several differences between those who completed treatment and those who did not (Kehle-Forbes et al., 2022). In both treatments, those who completed treatment reported a stronger therapeutic alliance than those who discontinued early. A particularly interesting finding was that completers interpreted symptom worsening as part of treatment, whereas those who dropped out viewed symptom worsening as an indicator that treatment was not working. This suggests that the interpretation of treatment-related distress and emotional engagement might be important. Those who discontinued treatment also reported that everyday stressors interfered with their ability to engage in treatments, whereas completers did not report this experience. In the noninferiority trial from which the CPT sessions were drawn for the present study, a number of clients (82%) reported that the reason for dropout in CPT was that they found the treatment too distressing (Sloan et al., 2018). Taken together, these findings highlight additional factors that might contribute to dropout during trauma-focused treatments, rather than only at pretreatment.

The Current Study

The goal of the current study was to follow the recommendations of Cooper and colleagues (2018) for studying dropout. The data source for this study is a randomized controlled noninferiority trial (Sloan et al., 2018) of CPT (Resick et al., 2017; Resick & Schnicke, 1992), a 12-session, gold-standard treatment for PTSD, and written exposure therapy (WET; Sloan & Marx, 2019), a 5-session treatment. Our study focuses on dropout from the CPT condition, as the WET condition had a very low 6.4% dropout rate. The CPT condition had a 39.7% dropout rate, and 76.0% of the sample dropped out by session 5. All available audio-recorded sessions from the first seven sessions were coded using an observational coding system (CHANGE; Hayes, Feldman, & Goldfried, 2007) to identify predictors of dropout in this specific trauma-informed treatment.

This study examined whether the significant predictors from the Alpert et al. study (2020) on client narratives (negative emotions, overgeneralized beliefs, ruminative processing, and physiological distress) and two additional variables (cognitive-emotional processing and avoidance) predicted dropout in CPT. We coded full CPT therapy sessions rather than the written narratives in the Alpert et al. (2020) study. The therapy sessions occur more frequently than narratives (maximum of two impact statements and two trauma accounts) and capture more context, such as client verbalization and audible cues like crying. As in the narrative study (Alpert et al., 2020), we expected overgeneralization and rumination in the CPT therapy sessions to predict higher dropout. More negative emotion and physiological distress expressed in the session could reflect emotional engagement with the traumatic experiences, and with the scaffolding of therapy, we expected both to predict less dropout. The CPT narrative studies from this trial found that more negative emotion not only predicted less dropout (Alpert et al., 2020), but also sudden gains (substantial PTSD symptom improvement between consecutive sessions; Sloan et al., 2022). The few studies on the role of physiological arousal and dropout (Carpenter et al., 2021; Garcia et al., 2011; Zayfert et al., 2005) have reported mixed findings, but we conceptualize physiological distress as high levels of emotional arousal. Although avoidance was not a significant predictor of dropout in the Alpert et al. (2020) CPT narrative study, several previous studies suggest that more avoidance at pretreatment (Bryant et al., 2003; Garcia et al., 2011) and in session (Yasinski et al., 2018) did predict dropout. It is also possible that avoidance will be captured more accurately in session rather than in client written narratives. Thus, we hypothesized that avoidance expressed in session would predict dropout. Cognitive-emotional processing has not yet been explored as a predictor of dropout. However, cognitive theory informs CPT (Beck, 1979) and suggests that approaching and challenging trauma-related cognitions encourages expression of natural emotions and is crucial to treatment success (Resick et al., 2017). In CPT, the client’s ability to approach trauma-related beliefs and explore their meanings could also reflect treatment engagement. Therefore, we expected more cognitive-emotional processing to predict lower dropout.

Method

Participants

Participants were 126 treatment-seeking adults with a primary Diagnostic and Statistical Manual of Mental Disorders (DSM-5) diagnosis with PTSD (APA, 2013). Of those, 63 (50%) participants were randomized to receive the CPT. The present sample includes 53 of those 63 individuals (84.13%), who had at least one audio-recorded session of CPT and completed the Clinician-Administered PTSD Scale for DSM-5 (CAPS-5; Weathers et al., 2013) at pretreatment. One participant was administratively withdrawn from the study and therefore not included.

Participants were recruited as part of a randomized controlled noninferiority trial of written exposure therapy (WET; Sloan & Marx, 2019) and CPT, which has been described in detail elsewhere (see Sloan et al., 2018). The trial was conducted at the VA Boston Healthcare System and had IRB oversight at that institution (ClinicalTrials.gov identifier NCT01800773; see Sloan et al., 2018 and Thompson-Hollands et al., 2018 for trial outcomes and detailed study procedures).

Participants who had experienced any type of traumatic event were recruited from the greater Boston community via flyers, Craigslist posts, and clinician referrals. Inclusion criteria were a primary PTSD diagnosis based on the Clinician-Administered PTSD Scale for DSM-5 (CAPS-5; Weathers et al., 2013), a duration of three months since the index trauma, and a stable pharmacotherapy regimen for at least 4 weeks, if on medication. Exclusion criteria were current substance dependence, psychotic symptoms, unstable bipolar disorder, severe cognitive impairment, current involvement in an abusive relationship if the index trauma was intimate partner violence, and high risk for suicide.

Participants were 27 men (50.9%) and 26 women (49.1%) with a mean age of 42.9 (SD = 14.1). Twenty-nine (54.7%) reported that they were White, 16 (30.2%) African American/Black, two (3.8%) American Indian or Alaska Native, one (1.9%) Asian, and five (9.4%) Other. Eight (15.1%) reported their ethnicity as Hispanic/Latino, 45 (84.9%) as Not Hispanic/Latino. For reported household income, 22 clients (41.5%) reported $15,000 or less, seven (13.2%) reported $15,000-$25,000, six (11.3%) reported $25,001-$35,000, eight (15.1%) reported $35,001-$50,000, three (5.7%) reported a $50–001-$75,000, three (5.7%) reported $75,001-$100,000, and four (7.5%) reported over $100,000.

Of the 53 participants analyzed in the present sample, 38 (71.7%) completed all 12 treatment sessions, and 15 (28.3%) dropped out before the end of treatment. The number of sessions completed by the 15 individuals who dropped out were: three sessions (n = 3), four sessions (n = 6), five sessions (n = 1), six sessions (n = 2), seven sessions (n = 1), and nine sessions (n = 2). Sloan and colleagues (2018) collected information about why participants dropped out of CPT prematurely. Dropout reasons included: treatment was too distressing (n=9), too busy for treatment (n=1), not motivated for treatment (n=1), and could not be reached to provide a reason (n=4). No one endorsed the option that they dropped out because their symptoms had improved.

Procedure

After consenting to the study and completing eligibility assessments, participants were assessed for PTSD symptom severity using the CAPS-5 at pre-treatment. All CPT sessions were audio recorded and archived. As we are interested in predicting dropout, only the first seven sessions of CPT were examined, which is the time frame during which nearly all dropouts occurred.

Treatment

CPT (Resick et al., 2017) is a trauma-focused therapy that includes 12, 60-minute sessions. The first session includes psychoeducation about common reactions and responses to trauma and PTSD, followed by the treatment rationale. The subsequent sessions focus on reviewing cognitive-behavioral conceptualization of PTSD using Socratic questioning to challenge and resolve clients’ negative beliefs and help clients process emotions related to those beliefs. The version of CPT used in this trial was the protocol that includes written accounts (now an optional component in the most recent version of CPT; Resick et al., 2017).

Therapists

Therapists were masters and doctoral level clinicians who delivered both treatments in the study. The therapists on the trial were trained and supervised by the developer of CPT (Patricia Resick, Ph.D.). Neither treatment expectations nor therapist effects predicted client dropout (Marx et al., 2021). Twenty percent of sessions were randomly selected to be rated for fidelity. Raters assessed adherence and competence in implementation according to a 7-point scale (1 = poor; 7 = excellent). Ratings of a four and above were considered satisfactory. Across all rated sessions in both conditions, 92.1% of adherence ratings were “good” to “excellent,” and none were below satisfactory. Additionally, ratings were very good for both adherence (M = 5.71) and competence (M = 5.91).

Definition of Dropout

In this dataset, those who did not complete the full 12 sessions of CPT dropped by session 9, and the majority dropped before that. Therefore, the threshold for defining dropout was discontinuing by session 9. This definition is consistent with that used in the Alpert et al. (2020) study, which examined the content of client narratives in this same clinical trial. Further, this cutoff for dropout (≤ 9 sessions) is similar to that used in other investigations of dropout in CPT (Jeffreys et al., 2014; Rizvi et al., 2009) and other trauma-focused treatments (Eftekhari et al., 2013; Goodson et al., 2017).

Measures

Clinician-Administered PTSD Scale for DSM-5

The Clinician Administered PTSD Scale for DSM-5 (CAPS-5; Weathers et al., 2013) was administered by four evaluators, who had at least a master’s degree in psychology and were masked to treatment conditions. The CAPS-5 was administered at pretreatment to assess PTSD severity. The CAPS-5 is a gold-standard diagnostic interview with good reliability and validity (Weathers et al., 2018). Each PTSD symptom is rated on a five-point scale from 0 (absent) to 4 (extreme/incapacitating), and total scores range from 0–80. Interrater reliability in this trial was very good (κ = .85). The CAPS-5 total score served as the primary outcome measure of PTSD severity in the parent trial (Sloan et al., 2018).

CHANGE

The CHANGE is an observational coding system designed to assess inhibitors and facilitators of therapeutic change (Hayes, Feldman, & Goldfried, 2007). The CHANGE was used to code all the session recordings from the first seven sessions in the CPT arm of the Sloan et al. (2018) trial. The CHANGE coding system has been used to identify predictors of outcome in narratives written by clients in exposure-based cognitive therapy for depression (Hayes et al., 2005; Hayes, Feldman, Beevers, et al., 2007) and from WET and CPT for adult PTSD (Alpert et al., 2020; Alpert, Hayes, Barnes, et al., 2023; Sloan et al., 2022), as well as audio-recorded sessions of treatments for youth PTSD (Alpert et al., 2021; Ready et al., 2015; Yasinski et al., 2016), PE for adult PTSD (Alpert, Hayes, & Foa, 2023), adult treatment-resistant depression (Abel et al., 2016), and personality disorders (Hayes & Yasinski, 2015). Each CHANGE coding variable is coded on a four-point scale from 0 (absent to very low) to 3 (high). Codes are not mutually exclusive and can co-occur.

A team of eight undergraduate and graduate students coded audio-recorded sessions from the CPT sessions. Coders were trained in CHANGE and then coded with experienced coders until they reached good inter-rater agreement (intraclass correlations (ICCs) ≥ .75). Two coders rated each audio-recorded session, and weekly consensus meetings were held to prevent rater drift and reach consensus on discrepancies of two or more points on the four-point (0–3) scale. Consensus ratings replaced items with discrepant scores (discrepancies of ≥ two points), and the original raw scores for each coder pair were retained for nondiscrepant scores. The ratings of both coders (with consensus and original scores) were then averaged. Coders were masked to study hypotheses and session number.

For each session, variables were coded to reflect participant’s responses to their traumatic experiences. The variables from the CHANGE that we included were significant predictors of dropout in CPT narratives (Alpert et al., 2020) in this trial. We also added two other content areas: avoidance and cognitive emotional processing.

To capture participants’ maladaptive trauma-related responses, each session was coded for overgeneralized beliefs (exaggerated thinking or beliefs that spread across situations, people, or time), ruminative processing (approaching trauma-related content but becoming stuck in a repetitive thought loop), and avoidance (difficulty engaging with or remaining in contact with aversive experiences). To identify clients’ affective engagement, each session was coded for negative emotions (e.g., fear, guilt, shame, anger, sadness) and physiological distress responses (e.g., heart racing, problems sleeping, feeling tense, restless, or numb). Finally, to assess a more adaptive mode of processing, each session was coded for cognitive emotional processing (approaching trauma-related content and beginning to make meaning of it). Table 1 presents more detailed descriptions of the variables, examples of in-session content that was coded highly, and ICCs for each code. Inter-rater agreement was good to excellent for all CHANGE variables (ICCs range: .74-.86; Cicchetti, 1994).

Table 1.

Descriptions of CHANGE Coding Categories with Examples of High Levels of Each Variable and Intraclass Correlations (ICCs) of Inter-Rater Agreement

Coding Category Description Example ICC
Negative Emotions Number and intensity of negative emotion words (e.g., anxious, sad, angry, ashamed, guilty) described in the session. Emotional tone is also considered. “I feel so sad and depressed right now. I want to take a break.” .86
Physiological Distress Negative responses in the body related to a person’s level of emotional distress (e.g., heart racing, agitation, dizziness, feeling numb). “I feel more jittery, like my heart is always racing. I am more jumpy and nervous all the time.” .80
Avoidance Difficulty engaging or remaining with aversive emotions, thoughts, memories, somatic sensations, or situations. Includes pulling away, withdrawing, shutting down, or emotional blunting. “I am trying not to think about the event or think at all.”; “I can’t get myself to leave the house because I feel the lack of control.” .83
Cognitive Emotional Processing Approaching, exploring, and making meaning of a problem. This
can include thinking about, questioning a problem, and exposing oneself to new information and is followed by insights or shifts in perspective or meaning.		 “As I grow older and become more self-assured, my dad’s words can’t hurt me as much as they used to. I will not let
that affect me”		 .79
Ruminative Processing Approaching, exploring, and attempting to make meaning of a problem area but becoming stuck repeatedly thinking about or analyzing the issue without significant insight. Includes analysis without progress, emotional venting, worry, or intrusive re-experiencing. “I should’ve done better in school. I should have sought help earlier when I was first so depressed. What is wrong with me? Why do I mess up every opportunity I get?” .74
Overgeneralization Exaggerating and applying beliefs about the self, others, or the world across time, people, and situations. “After this experience, I will never be the same.” .75
Open in a new tab

Note. ICC = intraclass correlation.

Session Selection for CHANGE Coding

Most CPT dropout in our sample occurred by session 7 (87%), so we coded all available sessions for each participant from across the first seven sessions. Again, dropout was defined as discontinuing by session 9. We coded sessions before session 9 to create a clear temporal separation between sampling and the point by which all drop out had occurred (session 9).

Data Analytic Plan

First, descriptive statistics were examined, including means and standard deviations of the predictor and outcome variable for CPT participants, as well as intercorrelations among predictor and outcome variables.

Next, binary logistic regression analyses were conducted in SPSS to examine predictors of dropout (0 = completion; 1 = dropout), consistent with most prior investigations of dropout in PTSD treatments (e.g., Garcia et al., 2011; Ormhaug & Jensen, 2018; Rizvi et al., 2009; Yasinski et al., 2018; Zayfert et al., 2005) and with the Alpert et al. (2020) study of narratives from the same trial of CPT used in the current study.

Each CHANGE coding score was averaged across the available early in-session recordings. This approach makes use of all available data and was also used in the Alpert et al (2020) study of CPT narratives to identify predictors of dropout in this trial. Selecting the first seven sessions allowed us to investigate in-session variables that unfold early in the course of CPT. Age was unrelated to dropout or any predictor variable in the present sample, therefore it was not included in the analyses. Although pretreatment PTSD symptom severity was not a predictor of dropout in this sample, this variable was associated with some of the predictor variables. Therefore, we controlled for PTSD symptom severity at pretreatment in the model, as we were interested in the effect of the predictor variables beyond pretreatment PTSD symptom severity.

Results

Power Analysis

Post-hoc power analysis was conducted using G*Power. For the binary logistic regression analyses, using a medium effect size (d = .40), two-tailed test with α = .05, and six predictors (overgeneralized beliefs, ruminative processing, avoidance, negative emotions, physiological distress, and cognitive emotional processing), a sample size of 45 participants is required to achieve power of .80. Thus, the sample size in this study was acceptable.

Descriptive Analyses and Intercorrelations

Means and standard deviations of the study variables are presented in Table 2. These descriptive statistics are presented for the total sample, as well as separated by treatment completion and dropout status. Intercorrelations among the predictor variables are presented in Table 3.

Table 2.

Descriptive Statistics for Predictors of Dropout Averaged Across Early Sessions

Group Total Sample Treatment Completion Dropout
Mean SD Range Mean SD Range Mean SD Range
Control Variable
 Pretreatment CAPS-5 37.45 8.91 16.00 to 59.00 36.95 9.37 16.00 to 55.00 38.73 9.20 25.00 to 59.00
CHANGE Predictor Variables
 Negative Emotion 1.29 .63 0.13 to 2.88 1.22 .64 0.13 to 2.83 1.45 .59 0.67 to 2.50
 Physiological Distress 1.20 .62 0.00 to 2.83 1.25 .64 0.00 to 2.83 1.08 .55 0.10 to 2.00
 Avoidance 1.90 .63 0.50 to 3.00 1.78 .60 0.50 to 2.75 2.20 .61 0.80 to 3.00
 Ruminative Processing 1.97 .58 0.50 to 3.00 1.93 .64 0.50 to 3.00 2.08 .42 1.25 to 3.00
 Cognitive Emotional Processing 1.42 .49 0.17 to 2.50 1.54 .45 0.50 to 2.50 1.15 .49 0.17 to 2.13
 Overgeneralization 2.05 .51 1.00 to 3.00 2.02 .52 1.00 to 3.00 2.14 .50 1.17 to 3.00
Open in a new tab

Note. Early sessions = first seven sessions, CAPS-5 = Clinician-Administered PTSD Scale for DSM-5. The CHANGE predictor variables were rated on a scale of 0 (absent to very low) to 3 (high).

Table 3.

Correlations Among CHANGE Predictors of Dropout: Averages of Session Coding Across Early Sessions

1 2 3 4 5 6 7
Averaged across first seven sessions
1. Pretreatment PTSD (CAPS-5) – – – – – – –
2. Negative Emotion .26 – – – – – –
3. Physiological Distress .37** .29* – – – – –
4. Avoidance .25 .24 .41** – – – –
5. Ruminative Processing −.08 .38** .12 .10 – – –
6. Cognitive Emotional Processing −.32* −.27* −.18 −.24 −.08 – –
7. Overgeneralization .12 .33* .11 .33* .52*** −.01 –
Open in a new tab

Note. Early sessions = first seven sessions, PTSD = Posttraumatic Stress Disorder, CAPS-5 = Clinician-Administered PTSD Scale for DSM-5. The CHANGE predictor variables were rated on a scale of 0 (absent to very low) to 3 (high).

*

p < .05,

**

p < .01,

***

p< .001

Predictors of Dropout

In Table 4, results are presented from logistic regression analyses examining predictors of treatment dropout (1) vs. completion (0) using coding variables averaged across the first seven in-session recordings. This model controlled for pretreatment PTSD symptom severity.

Table 4.

Binary Logistic Regression Model Predicting Dropout, Controlling for Pretreatment PTSD Severity: Average of CHANGE Codes Across Early Sessions

Group B SE Wald p OR 95% CI
Predictor (averaged across five sessions) Lower Upper
CAPS-5 Pretreatment .01 .05 .05 .825 1.01 .93 1.10
Negative Emotions .39 .71 .31 .580 1.48 .37 6.00
Physiological Distress −1.79 .82 4.74 .030 .17* .03 .84
Avoidance 1.42 .70 4.09 .043 4.12* 1.05 16.26
Ruminative Processing .45 .88 .26 .613 1.56 .28 8.77
Cognitive Emotional Processing −2.07 1.01 4.25 .039 .13* .02 .90
Overgeneralization −.18 .98 .03 .857 .84 .12 5.69
Open in a new tab

Note. Each group of predictors entered separately in a multiple logistic regression model predicting therapy dropout (1) vs. completion (0), controlling for pretreatment PTSD severity. Early sessions = first seven sessions, CAPS-5 = Clinician Administered PTSD Scale for DSM-5, CPT = cognitive processing therapy, SE = Standard Error, OR = Odds Ratio or Exponentiated B, CI = confidence interval.

*

p < .05

The model included the CHANGE variables ruminative processing, overgeneralization, avoidance, negative emotions, physiological distress, and cognitive emotional processing. The full model containing all predictors was statistically significant, χ2 (6, n = 53) = 16.96, p = .018. As predicted, more avoidance was associated with higher risk of dropout (OR = 4.12, 95% CI [1.05–16.26], p=.043), with one-point increase associated with a 4.12 times higher likelihood of dropping out, but ruminative processing and overgeneralization were not significant predictors of dropout.

Although higher levels of negative emotion did not predict lower risk for dropout as hypothesized, higher levels of physiological distress responses did (OR = .17, 95% CI [.03-.84], p=.030). A one-point increase in physiological distress on the 0–3 scale of CHANGE corresponded to being 83% times less likely to dropout. Finally, more cognitive emotional processing predicted lower risk for dropout (OR = .13, 95% CI [.02-.90], p=.039), such that a one-point increase in cognitive emotional processing corresponded to being 87% less likely to dropout.

Discussion

Although CPT is a recommended treatment for PTSD (APA, 2017), like other frontline treatments for PTSD, it has notably high dropout rates that range from 31% to 50% (Imel et al., 2013; Kline et al., 2018). The current study investigated predictors of dropout in CPT by coding the content of early session recordings of CPT (Resick et al., 2017). Session recordings were coded with the CHANGE coding system (Hayes, Feldman, & Goldfried, 2007) for emotional engagement (negative emotion and physiological distress), maladaptive trauma-related responses (overgeneralized beliefs, ruminative processing, avoidance), and also a more adaptive mode of processing traumatic experiences (cognitive emotional processing). Mean values of the CHANGE variables over the early sessions were examined as predictors of dropout status. On average, more physiological distress and cognitive emotional processing predicted lower risk of dropout in CPT, whereas more avoidance predicted higher risk. More overgeneralized beliefs, ruminative processing, and negative emotion were not significant predictors of treatment completion.

As hypothesized, more avoidance in session predicted more dropout. This differs from Alpert et al.’s (2020) finding that avoidance expressed in client written narratives in CPT did not predict dropout. However, other studies have found that avoidance at both pretreatment (Bryant et al., 2003; Garcia et al., 2011) and in session (Yasinski et al., 2018) predicted early discontinuation in trauma-focused treatments. Additionally, avoidance measured across narratives in CPT predicted worse symptom outcomes at posttreatment (Alpert, Hayes, Barnes, et al., 2023). It is possible that avoidance can be captured in a more nuanced and contextualized way in early therapy sessions than in written narratives.

Avoidance in trauma-focused treatment can be expressed in various ways, including not completing the written assignments as homework, minimizing natural emotions (e.g., sadness), and dropping out. A particularly interesting behavioral finding in the current trial was that clients who avoided completing the written narratives for homework were more likely to drop out of treatment than those who completed the narratives (Alpert et al., 2020). Avoidance of trauma-focused homework in CPT might indicate low engagement or difficulty approaching the material, which may predict premature dropout (Stirman et al., 2018). Our analysis of in-session avoidance highlights that there might be various opportunities for clinicians to carefully monitor and address avoidance behaviors early in CPT, in order to both decrease the likelihood of dropout and improve treatment outcomes. Although CPT therapists are trained to address client avoidance in general, it could be useful to identify the specific ways in which avoidance can manifest early in CPT before dropout occurs, as well as strategies clinicians can use to mitigate early signs of disengagement (Kehle-Forbes et al., 2022).

Inconsistent with prior findings from client narratives (Alpert et al., 2020) and with our hypotheses, overgeneralized beliefs and rumination were not significant predictors of risk for dropout. In CPT sessions, therapists can help to shift the client from overgeneralized beliefs and rumination to more constructive processing through targeted feedback, Socratic questioning, and hypothesis-testing exercises. More overgeneralization in narratives written between sessions might have predicted more risk of dropout (Alpert et al., 2020) because these maladaptive beliefs could proliferate without the therapist to keep them in check. In contrast, rumination in the between-session narratives in that study predicted less dropout. The authors proposed that when clients initiate meaning-making attempts on their own and early in treatment, this might reflect interest in understanding their traumatic experiences that could encourage them to complete treatment. In CPT sessions, therapists are likely to redirect clients from rumination to more constructive processing. In addition, it is possible that rumination did not predict dropout in the CPT sessions in our study because cognitive emotional processing was also included in the analyses and was an important predictor of less dropout risk. The latter captures constructive processing (Watkins et al., 2008) and separates out unproductive brooding and churning (rumination).

Cognitive emotional processing is the opposite of the CHANGE variable, ruminative processing, and instead captures a range of constructive meaning-making attempts, including questioning and exploring the traumatic experiences, gaining insight and new perspectives, and shifting to healthy, balanced beliefs (accommodation). Accommodated beliefs (adapting pre-existing beliefs to incorporate new information in a healthy way) have been found to predict better posttreatment symptom outcomes in CPT (Alpert, Hayes, Barnes, et al., 2023; Dondanville et al., 2016; Iverson et al., 2015; Scher et al., 2017), and the current study is the first to our knowledge to examine cognitive emotional processing as a predictor of dropout. As predicted, we found that higher average levels of cognitive emotional processing predicted lower risk of dropout. Clients who are beginning to make sense of their traumatic experiences and respond early in treatment are likely be motivated by such progress and might be less likely to drop out early. Thus, therapists may be able to use this as a positive indication of treatment progress. Further, future clinician trainings in CPT might benefit from focusing on how to promote adaptive cognitive emotional processing, even in the early sessions of CPT.

An unexpected finding was that more expression of negative emotion in the CPT sessions did not predict dropout. This is inconsistent with our prediction and with other findings from this same trial that examined client narratives rather than therapy sessions. Specifically, more negative emotion in client narratives predicted sudden gains (early symptom improvement; Sloan et al., 2022) and lower risk of dropout (Alpert et al., 2020), whereas when emotional responses were so strong that they were associated with expressions of physiological distress (e.g., sweaty hands, shaking, crying), the risk of dropout was higher. In contrast, we found in the current study that more negative emotion did not predict dropout, and more physiological distress in CPT sessions predicted lower rather than higher risk of dropout. Consistent with our prediction, it is possible that physiological responses that occur in the context of a therapy session might be an indicator of emotional activation and engagement with the treatment material. With the support and scaffolding of the therapist, these physiological responses are likely to be therapeutic. This might be different from coding based on CPT narratives, which are written as homework between sessions.

Physiological distress at home could be overwhelming and contribute to dropout without the therapist to help the client explore and contextualize the distress. In session, the therapist might help to frame the increase in distress as an important part of therapy, as suggested by Kehle-Forbes and colleagues’ (2022). Clients in their study who completed PE or CPT interpreted symptom worsening (increased distress) as part of treatment, whereas those who dropped out viewed distress as an indicator that treatment was not working. Together, these findings highlight that it is important for clinicians to help clients engage emotionally in trauma-focused treatments, but to do so at optimal levels and with an understanding of the role of distress tolerance in the change process (Foa et al., 2006). The processing of the physiological responses that occur in CPT sessions might not only be associated with better treatment completion, but also with treatment outcomes.

Strengths, Limitations, and Future Directions

The current study included a diverse sample of both civilians and veterans, with a primary diagnosis of PTSD. Yet, one limitation is the relatively small sample size. Another limitation is that we did not collect information on participants’ sexual orientation or gender identity. Future studies might collect these demographics, as gender minority individuals are an important at-risk group for PTSD (Beckwith et al., 2019).

The CHANGE observational coding system (Hayes, Feldman, & Goldfried, 2007) provides ratings by independent coders on trauma-related emotions, cognitions, and processes and yields information beyond clients’ reports of their own functioning. Another strength of this study is that we used CPT session recordings over the course of an RCT to identify predictors of dropout. The use of session audiotapes allowed for the study of variables in treatment that might be more amenable to change than the baseline characteristics and demographics most often examined as predictors of dropout. Further, our analysis of in-session content also enabled us to capture more detailed information and context than what clients can express in written narratives. Another strength in using in-session recordings and average CHANGE scores over the sessions was that clients were included in analyses even if they attended only one session, thereby maintaining the sample size and minimizing sampling bias. It might also be useful in future studies with larger samples to examine change in process variables from first to last available session, and if all sessions are coded, to conduct more detailed analyses of session-by-session changes.

Although in-session audio recordings capture extensive client information and context, coders rely on client verbalizations and audible cues such as crying. Other nonverbals (e.g., frowning, clenching fists) are not captured. Additionally, avoidance related to dropping out before treatment started or choosing not to complete narratives and other homework were not captured in the current study. Additional measurement modalities such as lab tasks, physiological assessments, and behavioral assessments over the course of treatment could supplement the CHANGE codes and provide useful and more multi-modal information in future investigations of dropout. Given the mixed findings regarding negative emotions and physiological responses predicting dropout (Alpert et al., 2020), it is possible that additional nonverbal measures might better capture the role of affective engagement in predicting dropout in trauma-focused interventions.

The current investigation focused only on client variables. Therapist factors (e.g., years of experience) and therapist-client factors (e.g., therapeutic alliance) might also be important for understanding dropout. Future studies could utilize the CHANGE coding system to examine how the therapeutic alliance contributes client variables (e.g., avoidance, negative physiological responses, cognitive and emotional processing) and to dropout. For instance, previous studies have found that better therapist-child relationship predicted lower dropout in youth PTSD treatment (Ormhaug & Jensen, 2018; Yasinski et al., 2018). In the Sloan et al. (2018) original RCT, the therapeutic alliance was only assessed at the dropout session or at post-treatment and therefore could not be examined as a predictor of dropout. However, those who did drop out in this trial reported lower alliance scores rated immediately after dropout (Marx et al., 2021). It might also be helpful to capture factors outside of therapy that contribute to dropout. For instance, a previous qualitative study reported that those who did not complete treatment were more affected by day-to-day stressors (Kehle-Forbes et al., 2022). It is possible that one might discontinue treatment due to additional stressful life events or low social support outside of therapy. These could be captured through self-report measures or observational coding.

In the current study, we focused on the CPT arm of the trial to identify predictors of dropout from a gold-standard treatment with a large evidence base but notably high dropout rates. In line with Cooper and colleagues’ (2018) approach to investigating dropout, we used an observational coding system to quantify early in-session content across the first seven sessions and were able to identify important variables that predicted risk for dropout. Additionally, our study builds on Alpert et al.’s (2020) study of written narratives that also identified several variables associated with increased risk for dropout. In addition to pretreatment variables and self-report measures, coding the content of CPT sessions provides an additional tool for identifying risk factors that emerge over the course of treatment. This research also highlights a point of early intervention to enhance treatment engagment and potentially minimize dropout.

This study contributes to the growing literature examining dropout in treatments with strong evidence base but notably high dropout rates. Research on predictors of dropout is difficult, and the findings tend to be inconsistent across studies. For example, some predictors reported across multiple studies later show no relationship to dropout in other studies. Therefore, the present findings need to be replicated. Further investigation into the mechanistic process of dropout is still necessary. For example, an exciting future direction might be to conduct more fine-grained, session-by-session analyses to examine potential mediators and moderators of in-session predictors of dropout.

Although studies suggest that participants may not require the full 12 sessions of CPT to achieve favorable outcomes (Galovski et al., 2012), current research suggests that individuals who complete more sessions of CPT have better PTSD and well-being outcomes (Holmes et al., 2019; Larsen et al., 2023). Recent efforts to minimize dropout include modifications to CPT, such as intensive-CPT treatment programs and virtual delivery of CPT (Szoke et al., 2023). Other briefer trauma-informed treatments have been developed, such as WET, a five-session intervention with minimal therapist involvement (Sloan & Marx, 2019). Our study provides some evidence that clinicians conducting CPT can attend to their client’s engagement, as indicated by avoidance, physiological responses, and cognitive and emotional processing, to gauge dropout risk early in treatment. This is especially important and timely, as evidence-based treatments urge clinicians to emphasize patient-centered treatment. In doing this, clinicians might be able to better tailor therapy to the individual and encourage clients to complete treatment and receive optimal benefit.

Highlights.

  • Dropout rates in cognitive processing therapy (CPT) are high.

  • We examined predictors of dropout in audio-recorded sessions of CPT.

  • More physiological distress and adaptive processing predicted lower dropout.

  • More avoidance predicted higher dropout.

  • Targeting early indicators of treatment engagement might reduce risk of dropout.

Funding:

This research was supported by a grant from the National Institute of Mental Health (NIMH: R01-MH095737) awarded to Denise M. Sloan.

Footnotes

CRediT authorship contribution statement

Danielle Shayani: Conceptualization, Formal analysis, Writing – Original draft, Writing – Review and editing. Caroline Canale: Investigation, Data curation, Project administration, Writing – Review and editing. Denise Sloan: Funding acquisition, Writing – Review and editing. Adele Hayes: Conceptualization, Methodology, Writing – Original draft, Writing – Review and editing, Supervision, Developer of the CHANGE coding system.

Declaration of interest: none

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Abel A, Hayes AM, Henley W, & Kuyken W. (2016). Sudden gains in cognitive–behavior therapy for treatment-resistant depression: Processes of change. Journal of Consulting and Clinical Psychology, 84(8), 726–737. 10.1037/ccp0000101 [DOI] [PubMed] [Google Scholar]
  2. Alpert E, Hayes AM, Barnes JB, & Sloan DM (2020). Predictors of dropout in cognitive processing therapy for PTSD: An examination of trauma narrative content. Behavior Therapy, 51(5), 774–788. 10.1016/j.beth.2019.11.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Alpert E, Hayes AM, Barnes JB, & Sloan DM (2023). Using client narratives to identify predictors of outcome in written exposure therapy and cognitive processing therapy. Behavior Therapy, 54(2), 185–199. 10.1016/j.beth.2022.09.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Alpert E, Hayes AM, & Foa EB (2023). Examining emotional processing theory and predictions of outcome in prolonged exposure for PTSD. Behaviour Research and Therapy. 10.1016/j.brat.2023.104341 [DOI] [PubMed] [Google Scholar]
  5. Alpert E, Hayes AM, Yasinski C, Webb C, & Deblinger E. (2021). Processes of change in trauma-focused cognitive behavioral therapy for youth: An emotional processing theory informed approach. Clinical Psychological Science: A Journal of the Association for Psychological Science, 9(2), 270–283. 10.1177/2167702620957315 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). 10.1176/appi.books.9780890425596 [DOI] [Google Scholar]
  7. American Psychological Association. (2017). Clinical Practice Guideline for the Treatment of PTSD. Washington, DC. https://www.apa.org/ptsd-guideline/ptsd.pdf [Google Scholar]
  8. Atwoli L, Stein DJ, Koenen KC, & McLaughlin KA (2015). Epidemiology of posttraumatic stress disorder: Prevalence, correlates and consequences. Current Opinion in Psychiatry, 28(4), 307–311. 10.1097/YCO.0000000000000167 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Beck AT (1979). Cognitive therapy of depression. Guilford Press. [Google Scholar]
  10. Beckwith N, McDowell MJ, Reisner SL, Zaslow S, Weiss RD, Mayer KH, & Keuroghlian AS (2019). Psychiatric epidemiology of transgender and nonbinary adult patients at an urban health center. LGBT Health, 6(2), 51–61. 10.1089/lgbt.2018.0136 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Browne KC, Chen JA, Hundt NE, Hudson TJ, Grubbs KM, & Fortney JC (2021). Veterans self-reported reasons for non-attendance in psychotherapy for posttraumatic stress disorder. Psychological Services, 18, 173–185. 10.1037/ser0000375 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bryant RA, Moulds ML, Guthrie RM, Dang ST, & Nixon RDV (2003). Imaginal exposure alone and imaginal exposure with cognitive restructuring in treatment of posttraumatic stress disorder. Journal of Consulting and Clinical Psychology, 71, 706–712. 10.1037/0022-006X.71.4.706 [DOI] [PubMed] [Google Scholar]
  13. Calabrese JR, Prescott M, Tamburrino M, Liberzon I, Slembarski R, Goldmann E, Shirley E, Fine T, Goto T, Wilson K, Ganocy S, Chan P, Serrano MB, Sizemore J, & Galea S. (2011). PTSD comorbidity and suicidal ideation associated with PTSD within the Ohio Army National Guard. The Journal of Clinical Psychiatry, 72(8), 1072–1078. 10.4088/JCP.11m06956 [DOI] [PubMed] [Google Scholar]
  14. Carpenter JK, Suzanne Pineles L, Werner K, Kecala NM, Resick PA, & Galovski TE (2021). Physiological reactivity to trauma memories as a predictor of PTSD treatment response [Conference Presentation]. International Society for Traumatic Stress Studies (ISTSS), Virtual. [Google Scholar]
  15. Cicchetti DV (1994). Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychological Assessment, 6(4), 284–290. 10.1037/1040-3590.6.4.284 [DOI] [Google Scholar]
  16. Cooper AA, Kline AC, Baier AL, & Feeny NC (2018). Rethinking research on prediction and prevention of psychotherapy dropout: A mechanism-oriented approach. Behavior Modification. 10.1177/0145445518792251 [DOI] [PubMed] [Google Scholar]
  17. Cusack K, Jonas DE, Forneris CA, Wines C, Sonis J, Middleton JC, Feltner C, Brownley KA, Olmsted KR, Greenblatt A, Weil A, & Gaynes BN (2016). Psychological treatments for adults with posttraumatic stress disorder: A systematic review and meta-analysis. Clinical Psychology Review, 43, 128–141. 10.1016/J.CPR.2015.10.003 [DOI] [PubMed] [Google Scholar]
  18. Department of Veterans Affairs and Department of Defense. (2017). VA/DOD clinical practice guideline for the management of posttraumatic stress disorder and acute stress disorder. [Google Scholar]
  19. Dondanville KA, Blankenship AE, Molino A, Resick PA, Wachen JS, Mintz J, Yarvis JS, Litz BT, Borah EV, Roache JD, Young-McCaughan S, Hembree EA, & Peterson AL (2016). Qualitative examination of cognitive change during PTSD treatment for active duty service members. Behaviour Research and Therapy, 79, 1–6. 10.1016/j.brat.2016.01.003 [DOI] [PubMed] [Google Scholar]
  20. Eftekhari A, Crowley JJ, Mackintosh M-A, & Rosen CS (2020). Predicting treatment dropout among veterans receiving prolonged exposure therapy. Psychological Trauma: Theory, Research, Practice, and Policy, 12, 405–412. 10.1037/tra0000484 [DOI] [PubMed] [Google Scholar]
  21. Eftekhari A, Ruzek JI, Crowley JJ, Rosen CS, Greenbaum MA, & Karlin BE (2013). Effectiveness of national implementation of prolonged exposure therapy in Veterans Affairs care. JAMA Psychiatry, 70(9), 949–955. 10.1001/jamapsychiatry.2013.36 [DOI] [PubMed] [Google Scholar]
  22. Ehlers A, & Clark DM (2000). A cognitive model of posttraumatic stress disorder. Behaviour Research and Therapy, 38(4), 319–345. 10.1016/S0005-7967(99)00123-0 [DOI] [PubMed] [Google Scholar]
  23. Foa EB, Hembree EA, Rothbaum BO, & Rauch SAM (2019). Prolonged exposure therapy for PTSD: Emotional processing of traumatic experiences: Therapist guide (2nd ed.). Oxford University Press. 10.1093/med-psych/9780190926939.001.0001 [DOI] [Google Scholar]
  24. Foa EB, Huppert JD, & Cahill SP (2006). Emotional processing theory: An update. In Rothbaum BO (Ed.). Pathological anxiety: Emotional processing in etiology and treatment (pp. 3–24). The Guilford Press. [Google Scholar]
  25. Forbes D, Lloyd D, Nixon RDV, Elliott P, Varker T, Perry D, Bryant RA, & Creamer M. (2012). A multisite randomized controlled effectiveness trial of cognitive processing therapy for military-related posttraumatic stress disorder. Journal of Anxiety Disorders, 26(3), 442–452. 10.1016/j.janxdis.2012.01.006 [DOI] [PubMed] [Google Scholar]
  26. Galovski TE, Blain LM, Mott JM, Elwood L, & Houle T. (2012). Manualized therapy for PTSD: Flexing the structure of cognitive processing therapy. Journal of Consulting and Clinical Psychology, 80(6), 968–981. 10.1037/a0030600 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Garcia HA, Kelley LP, Rentz TO, & Lee S. (2011). Pretreatment predictors of dropout from cognitive behavioral therapy for PTSD in Iraq and Afghanistan war veterans. Psychological Services, 8, 1–11. 10.1037/a0022705 [DOI] [Google Scholar]
  28. Goodson JT, Helstrom AW, Marino EJ, & Smith RV (2017). The impact of service-connected disability and therapist experience on outcomes from prolonged exposure therapy with veterans. Psychological Trauma: Theory, Research, Practice, and Policy, 9, 647–654. 10.1037/tra0000260 [DOI] [PubMed] [Google Scholar]
  29. Hayes AM, Beevers CG, Feldman GC, Laurenceau J-P, & Perlman C. (2005). Avoidance and processing as predictors of symptom change and positive growth in an integrative therapy for depression. International Journal of Behavioral Medicine, 12(2), 111. 10.1207/s15327558ijbm1202_9 [DOI] [PubMed] [Google Scholar]
  30. Hayes AM, Feldman GC, Beevers CG, Laurenceau J-P, Cardaciotto L, & Lewis-Smith J. (2007). Discontinuities and cognitive changes in an exposure-based cognitive therapy for depression. Journal of Consulting and Clinical Psychology, 75(3), 409–421. 10.1037/0022-006X.75.3.409 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Hayes AM, Feldman GC, & Goldfried MR (2007). The change and growth experiences scale: A measure of insight and emotional processing. In Castonguay LG, & Hill C. (Eds.). Insight in Psychotherapy (pp. 231–253). Washington, DC: American Psychological Association. 10.1037/11532-011 [DOI] [Google Scholar]
  32. Hayes AM, & Yasinski C. (2015). Pattern destabilization and emotional processing in cognitive therapy for personality disorders. Frontiers in Psychology, 6. 10.3389/fpsyg.2015.00107 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Holmes SC, Johnson CM, Suvak MK, Sijercic I, Monson CM, & Wiltsey Stirman S. (2019). Examining patterns of dose response for clients who do and do not complete cognitive processing therapy. Journal of Anxiety Disorders, 68, 102120. 10.1016/j.janxdis.2019.102120 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Hundt NE, Ecker AH, Thompson K, Helm A, Smith TL, Stanley MA, & Cully JA (2020). “It didn’t fit for me:” A qualitative examination of dropout from prolonged exposure and cognitive processing therapy in veterans. Psychological Services, 17, 414–421. 10.1037/ser0000316 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Imel ZE, Laska K, Jakupcak M, & Simpson TL (2013). Meta-analysis of dropout in treatments for posttraumatic stress disorder. Journal of Consulting and Clinical Psychology, 81(3), 394–404. 10.1037/a0031474 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Iverson KM, King MW, Cunningham KC, & Resick PA (2015). Rape survivors’ trauma-related beliefs before and after cognitive processing therapy: Associations with PTSD and depression symptoms. Behaviour Research and Therapy, 66, 49–55. 10.1016/J.BRAT.2015.01.002 [DOI] [PubMed] [Google Scholar]
  37. Jeffreys MD, Reinfeld C, Nair PV, Garcia HA, Mata-Galan E, & Rentz TO (2014). Evaluating treatment of posttraumatic stress disorder with cognitive processing therapy and prolonged exposure therapy in a VHA specialty clinic. Journal of Anxiety Disorders, 28(1), 108–114. 10.1016/j.janxdis.2013.04.010 [DOI] [PubMed] [Google Scholar]
  38. Keefe JR, Wiltsey Stirman S, Cohen ZD, DeRubeis RJ, Smith BN, & Resick PA (2018). In rape trauma PTSD, patient characteristics indicate which trauma-focused treatment they are most likely to complete. Depression and Anxiety, 35(4), 330–338. 10.1002/da.22731 [DOI] [PubMed] [Google Scholar]
  39. Kehle-Forbes SM, Ackland PE, Spoont MR, Meis LA, Orazem RJ, Lyon A, Valenstein-Mah HR, Schnurr PP, Zickmund SL, Foa EB, Chard KM, Alpert E, & Polusny MA (2022). Divergent experiences of U.S. veterans who did and did not complete trauma-focused therapies for PTSD: A national qualitative study of treatment dropout. Behaviour Research and Therapy, 154, 104123. 10.1016/j.brat.2022.104123 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Kehle-Forbes SM, Meis LA, Spoont MR, & Polusny MA (2016). Treatment initiation and dropout from prolonged exposure and cognitive processing therapy in a VA outpatient clinic. Psychological Trauma: Theory, Research, Practice, and Policy, 8, 107–114. 10.1037/tra0000065 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Kilpatrick DG, Resnick HS, Milanak ME, Miller MW, Keyes KM, & Friedman MJ (2013). National estimates of exposure to traumatic events and PTSD prevalence using DSM-IV and DSM-5 criteria. Journal of Traumatic Stress, 26(5), 537–547. 10.1002/JTS.21848 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Kline AC, Cooper AA, Rytwinksi NK, & Feeny NC (2018). Long-term efficacy of psychotherapy for posttraumatic stress disorder: A meta-analysis of randomized controlled trials. Clinical Psychology Review, 59, 30–40. 10.1016/j.cpr.2017.10.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Larsen SE, Hamrick LA, Thomas KB, Hessinger JD, Melka SE, Khaled M, Hove MC, & Maieritsch K. (2023). Symptom change prior to treatment discontinuation (dropout) from a naturalistic Veterans Affairs evidence-based psychotherapy clinic for PTSD and depression. Psychological Services. 10.1037/ser0000757 [DOI] [PubMed] [Google Scholar]
  44. Larsen SE, Mackintosh M-A, La Bash H, Evans WR, Suvak MK, Shields N, Lane JEM, Sijercic I, Monson CM, & Wiltsey Stirman S. (2022). Temporary PTSD symptom increases among individuals receiving CPT in a hybrid effectiveness-implementation trial: Potential predictors and association with overall symptom change trajectory. Psychological Trauma: Theory, Research, Practice, and Policy, 14, 853–861. 10.1037/tra0000545 [DOI] [PubMed] [Google Scholar]
  45. López-Castro T, Zhao Y, Fitzpatrick S, Ruglass LM, & Hien DA (2021). Seeing the forest for the trees: Predicting attendance in trials for co-occurring PTSD and substance use disorders with a machine learning approach. Journal of Consulting and Clinical Psychology, 89, 869–884. 10.1037/ccp0000688 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Marks I, Lovell K, Noshirvani H, Livanou M, & Thrasher S. (1998). Treatment of posttraumatic stress disorder by exposure and/or cognitive restructuring: A controlled study. Archives of General Psychiatry, 55(4), 317–325. 10.1001/archpsyc.55.4.317 [DOI] [PubMed] [Google Scholar]
  47. Marx BP, Thompson-Holland J, Lee DJ, Resick PA, & Sloan DM (2021). Estimated intelligence moderates cognitive processing therapy outcome for posttraumatic stress symptoms. Behavior Therapy, 52(1), 162–169. 10.1016/j.beth.2020.03.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Meis LA, Polusny MA, Kehle-Forbes SM, Erbes CR, O’Dougherty M, Erickson EPG, Orazem RJ, Burmeister LB, & Spoont MR (2023). Making sense of poor adherence in PTSD treatment from the perspectives of veterans and their therapists. Psychological Trauma: Theory, Research, Practice, and Policy, 15, 715–725. 10.1037/tra0001199 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Monson CM, Schnurr PP, Resick PA, Friedman MJ, Young-Xu Y, & Stevens SP (2006). Cognitive processing therapy for veterans with military-related posttraumatic stress disorder. Journal of Consulting and Clinical Psychology, 74, 898–907. 10.1037/0022-006X.74.5.898 [DOI] [PubMed] [Google Scholar]
  50. Ormhaug SM, & Jensen TK (2018). Investigating treatment characteristics and first-session relationship variables as predictors of dropout in the treatment of traumatized youth. Psychotherapy Research, 28(2), 235–249. 10.1080/10503307.2016.1189617 [DOI] [PubMed] [Google Scholar]
  51. Ready CB, Hayes AM, Yasinski CW, Webb C, Gallop R, Deblinger E, & Laurenceau J-P (2015). Overgeneralized beliefs, accommodation, and treatment outcome in youth receiving trauma-focused cognitive behavioral therapy for childhood trauma. Behavior Therapy, 46(5), 671–688. 10.1016/j.beth.2015.03.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Reger GM, Koenen-Woods P, Zetocha K, Smolenski DJ, Holloway KM, Rothbaum BO, Difede J, Rizzo AA, Edwards-Stewart A, Skopp NA, Mishkind M, Reger MA, & Gahm GA (2016). Randomized controlled trial of prolonged exposure using imaginal exposure vs. virtual reality exposure in active duty soldiers with deployment-related posttraumatic stress disorder (PTSD). Journal of Consulting and Clinical Psychology, 84, 946–959. 10.1037/ccp0000134 [DOI] [PubMed] [Google Scholar]
  53. Resick PA, Monson CM, & Chard KM (2017). Cognitive processing therapy for PTSD: A comprehensive manual. The Guilford Press. [Google Scholar]
  54. Resick PA, & Schnicke MK (1992). Cognitive processing therapy for sexual assault victims. Journal of Consulting and Clinical Psychology, 60(5), 748–756. 10.1037/0022-006X.60.5.748 [DOI] [PubMed] [Google Scholar]
  55. Rizvi SL, Vogt DS, & Resick PA (2009). Cognitive and affective predictors of treatment outcome in cognitive processing therapy and prolonged exposure for posttraumatic stress disorder. Behaviour Research and Therapy, 47(9), 737–743. 10.1016/j.brat.2009.06.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Scher CD, Suvak MK, & Resick PA (2017). Trauma cognitions are related to symptoms up to 10 years after cognitive behavioral treatment for posttraumatic stress disorder. Psychological Trauma: Theory, Research, Practice, and Policy, 9(6), 750–757. 10.1037/tra0000258 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Schnurr PP, Chard KM, Ruzek JI, Chow BK, Resick PA, Foa EB, Marx BP, Friedman MJ, Bovin MJ, Caudle KL, Castillo D, Curry KT, Hollifield M, Huang GD, Chee CL, Astin MC, Dickstein B, Renner K, Clancy CP, … Shih M-C (2022). Comparison of prolonged exposure vs cognitive processing therapy for treatment of posttraumatic stress disorder among US veterans: A randomized clinical trial. JAMA Network Open, 5(1), e2136921. 10.1001/jamanetworkopen.2021.36921 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Sloan DM, Lee DJ, Litwack SD, Sawyer AT, & Marx BP (2013). Written exposure therapy for veterans diagnosed with PTSD: A pilot study. Journal of Traumatic Stress, 26(6), 776–779. 10.1002/jts.21858 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Sloan DM, & Marx BP (2019). Written exposure therapy for PTSD: A brief treatment approach for mental health professionals. American Psychological Association. 10.1037/0000139-000 [DOI] [Google Scholar]
  60. Sloan DM, Marx BP, Bovin MJ, Feinstein BA, & Gallagher MW (2012). Written exposure as an intervention for PTSD: A randomized clinical trial with motor vehicle accident survivors. Behaviour Research and Therapy, 50(10), 627–635. 10.1016/j.brat.2012.07.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Sloan DM, Marx BP, Lee DJ, & Resick PA (2018). A brief exposure-based treatment vs cognitive processing therapy for posttraumatic stress disorder: A randomized noninferiority clinical trial. JAMA Psychiatry, 75(3), 233–239. 10.1001/jamapsychiatry.2017.4249 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Sloan DM, Thompson-Hollands J, Hayes AM, Lee DJ, Alpert E, & Marx BP (2022). Sudden gains in two trauma-focused treatments for posttraumatic stress disorder. Behavior Therapy, 53(2), 255–266. 10.1016/j.beth.2021.08.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Steenkamp MM, Litz BT, & Marmar CR (2020). First-line psychotherapies for military-related PTSD. JAMA, 323(7), 656–657. 10.1001/jama.2019.20825 [DOI] [PubMed] [Google Scholar]
  64. Stirman SW, Gutner CA, Suvak MK, Adler A, Calloway A, & Resick P. (2018). Homework completion, patient characteristics, and symptom change in cognitive processing therapy for PTSD. Behavior Therapy, 49(5), 741–755. 10.1016/j.beth.2017.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Szafranski DD, Gros DF, Acierno R, Brady KT, Killeen TK, & Back SE (2019). Heterogeneity of treatment dropout: PTSD, depression, and alcohol use disorder reductions in PTSD and AUD/SUD treatment noncompleters. Clinical Psychology & Psychotherapy, 26(2), 218–226. 10.1002/cpp.2344 [DOI] [PubMed] [Google Scholar]
  66. Szoke D, Klassen B, & Held P. (2023) Achieving low levels of dropout during intensive PTSD treatment: Data from multiple length and multiple format intensive CPT interventions. In D. R. Shayani (Chair), The dropout dilemma: Identifying and overcoming barriers to treatment completion in trauma-focused PTSD interventions. Symposium to be presented at the 57th Annual Convention of the Association for Behavioral and Cognitive Therapies, Seattle, WA. [Google Scholar]
  67. Taubitz F-S, Büdenbender B, & Alpers GW (2022). What the future holds: Machine learning to predict success in psychotherapy. Behaviour Research and Therapy, 156, 104116. 10.1016/j.brat.2022.104116 [DOI] [PubMed] [Google Scholar]
  68. Thompson-Hollands J, Marx BP, Lee DJ, Resick PA, & Sloan DM (2018). Long-term treatment gains of a brief exposure-based treatment for PTSD. Depression and Anxiety, 35(10), 985–991. 10.1002/da.22825 [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Varker T, Jones KA, Arjmand H-A, Hinton M, Hiles SA, Freijah I, Forbes D, Kartal D, Phelps A, Bryant RA, McFarlane A, Hopwood M, & O’Donnell M. (2021). Dropout from guideline-recommended psychological treatments for posttraumatic stress disorder: A systematic review and meta-analysis. Journal of Affective Disorders Reports, 4, 100093. 10.1016/j.jadr.2021.100093 [DOI] [Google Scholar]
  70. Watkins ER (2008). Constructive and unconstructive repetitive thought. Psychological Bulletin, 134(2), 163–208. 10.1037/0033-2909.134.2.163 [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Weathers FW, Blake DD, Schnurr PP, Kaloupek DG, Marx BP, & Keane T. (2013). Clinician-administered PTSD scale for DSM–5 (CAPS-5). Boston, MA: National Center for PTSD. [Google Scholar]
  72. Yasinski C, Hayes AM, Alpert E, McCauley T, Ready CB, Webb C, & Deblinger E. (2018). Treatment processes and demographic variables as predictors of dropout from trauma-focused cognitive behavioral therapy (TF-CBT) for youth. Behaviour Research and Therapy, 107, 10–18. 10.1016/j.brat.2018.05.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Yasinski C, Hayes AM, Ready CB, Cummings JA, Berman IS, McCauley T, Webb C, & Deblinger E. (2016). In-session caregiver behavior predicts symptom change in youth receiving trauma-focused cognitive behavioral therapy (TF-CBT). Journal of Consulting and Clinical Psychology, 84(12), 1066–1077. 10.1037/ccp0000147 [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Zayfert C, DeViva JC, Becker CB, Pike JL, Gillock KL, & Hayes SA (2005). Exposure utilization and completion of cognitive behavioral therapy for PTSD in a “real world” clinical practice. Journal of Traumatic Stress, 18(6), 637–645. 10.1002/jts.20072 [DOI] [PubMed] [Google Scholar]

RESOURCES