Open Label Pilot Study on Posttrauma Health Impacts of the Processing of Positive Memories Technique (PPMT)

Abstract

Processing of Positive Memories Technique (PPMT) was proposed as a novel intervention for posttraumatic stress disorder (PTSD). PPMT comprises of 5 sessions focused on identifying and processing positive memories. As an open label pilot study, we explored PPMT’s effects on PTSD severity, depression severity, affect and cognitive processes, and therapeutic alliance. A sample of 12 trauma-exposed participants seeking services at a University Psychology Clinic participated in 5 PPMT sessions (M_age=29.25 years; 58.30% women). We used the reliable change indices and clinically significant change score approach. The following number of participants showed statistically reliable changes: 9 participants for PTSD severity (8 recovered/improved); 6 participants for depression severity (5 improved); 5 participants for positive affect levels (2 recovered/improved); 9 participants for negative affect levels (8 recovered); 9 participants for posttrauma cognitions (7 recovered/improved); 5 participants for positive emotion dysregulation (4 recovered); 11 participants for number of retrieved positive memories (3 recovered); and 5 participants for therapeutic alliance (4 recovered). PPMT may impact certain posttrauma targets more effectively (PTSD, depression, negative affect, posttrauma cognitions). PPMT may be more helpful in improving regulation rather than levels of positive affect. PPMT, if supported in further investigations, may add to the clinician tool-box of PTSD interventions.

Introduction

Autobiographical memories can be specific memories of events occurring at a certain place and time within 24 hours, or can be overgeneral memories with abstract content (Williams & Broadbent, 1986). Both trauma and positive autobiographical memories are relevant to examine in the context of posttraumatic stress disorder (PTSD; Brewin, 2014; Brewin & Holmes, 2003), which is a clinical condition affecting ~8% of individuals (Kilpatrick et al., 2013). Most existing clinical practices address therapeutic engagement with trauma memories to improve PTSD symptoms (Contractor, Weiss, Forkus, et al., 2020). Although efficacious, these trauma-focused treatments only result in remission of PTSD among 44–66% of treatment recipients (Cusack et al., 2016; Schnurr, 2017) and have substantial dropout (Goetter et al., 2015; Lewis et al., 2020). Given that not all existing PTSD interventions work for everyone or appeal to every individual seeking treatment, we need to expand our clinician toolbox to target innovative mechanisms of change such as positive memory processes.

Notably, research suggests that deficits in positive memory processes follow trauma/PTSD or render PTSD more likely to develop after a traumatic experience (Williams et al., 2007); and fewer and/or less access to specific positive memories relate to more PTSD severity (Contractor et al., 2019; Contractor, Caldas, et al., 2021). Indeed, emerging experimental and clinical research suggests that targeting positive memories may improve PTSD symptoms (Contractor, Banducci, et al., 2020; Miguel-Alvaro et al., 2021). In this regard, Contractor and colleagues (2018) proposed a Positive Memory-PTSD Model. They suggested that retrieving and processing positive memories may help to improve affect, correct maladaptive beliefs, enhance positive memory retrieval, as well as facilitate benefits for subsequent trauma-focused treatments (e.g., increasing readiness to start trauma-focused treatments, less fear/hesitation towards discussing trauma memories, better therapeutic alliance). Cumulatively, processing both positive and trauma memories may decrease the possibility of dropout from trauma-focused treatments. In turn, all such processes may contribute to improved PTSD symptoms. Recently, this Positive Memory-PTSD Model was updated to clarify that the model addresses specific vs. overgeneral positive memories, to underscore the importance of considering the heterogeneity in, and transitionary nature of, affect processes following positive memory retrieval, and to highlight potential moderators of relations between positive memory processing and PTSD symptoms (e.g., trauma type and count; Contractor, Banducci, et al., 2021b). See Supplemental Figure 1 for the updated Positive Memory-PTSD Model.

Drawing from this Positive Memory-PTSD Model, a novel 5-session Processing of Positive Memories Technique (PPMT) was developed (Contractor, Weiss, & Shea, 2020), that aims to improve access to and quality of salient positive memories. Therapists help individuals rehearse, relive, and engage with several salient positive memories to access and strengthen central positive values, affect, strengths, and thoughts (VAST). Procedurally, individuals retrieve positive memories, are guided through an imaginal experiencing (narration) of that positive memory, and process the positive memory to elicit VAST elements. The focus on positive memories continues between sessions via homework assignments such as recording positive memories retrieved daily along with their VAST components as well as engaging in an activity that is aligned with VAST elements discussed in session. Hereby, PPMT is purported to have the following potential impacts.

Improved Affect and Cognitions

Individuals with PTSD experience deficits in affect experiences and regulation (e.g., Litz et al., 2000; Weiss et al., 2018), including less positive affect and an emotional numbing response (Frewen et al., 2010; Orsillo et al., 2007), and the experience of negative affect following the experience of positive affect (negative affect interference; Frewen et al., 2010). Further, these individuals may also report difficulties regulating positive emotions, wherein they are non-accepting of positive emotions and have difficulties controlling impulsive behaviors and engaging in goal-directed behaviors when experiencing positive emotions (Weiss, Contractor, Forkus, et al., 2020; Weiss, Contractor, Raudales, et al., 2020). Also, individuals with PTSD report more negative cognitions (Janoff-Bulman, 1992). Critically, such affect-related difficulties (Foa et al., 1995; Van Minnen et al., 2002) and negative cognitions (Alpert et al., 2020; Bryant et al., 2003) have been shown to predict non-/poor responses to or dropout from PTSD interventions. Positive memory retrieval may help with these difficulties in affect and cognitions, which may aid treatment effectiveness for PTSD and treatment retention.

To elaborate, both experimental (Rusting & DeHart, 2000) and intervention (Panagioti et al., 2012) research support using positive memory retrieval as an adaptive affect-regulation strategy. By retrieving positive memories, individuals focus less on negative memories, which helps to downregulate negative affect and upregulate positive affect (Quoidbach et al., 2015; Rusting & DeHart, 2000). Also, processing of positive memories may aid one’s ability to accept and effectively regulate positive emotions (beyond just increasing levels of positive affect) via exposure and repeated practice. This improved affect may help to reduce PTSD symptoms (Bryan et al., 2020) and may contribute to factors that help to reduce PTSD symptoms such as improved health (Steptoe et al., 2009). Further, retrieving positive memories may enhance adaptive cognitions: (1) improved affect may increase positive interpretations of events (Rusting & DeHart, 2000), and positive content in thoughts (Fredrickson, 2001); and (2) positive memory content may prime adaptive beliefs to counter trauma cognitions (Contractor et al., 2018).

Improved Positive Memory Retrieval Abilities

Autobiographical memory specificity (AMS) references the ability to retrieve specific memories (Williams & Broadbent, 1986). PPMT is proposed to enhance positive AMS by enhancing the practice and ease of processing specific positive memories over time. Such improved positive AMS may contribute to less PTSD severity (Contractor, Caldas, et al., 2021; de Decker et al., 2003) by improving affect and cognitions as elaborated above. Further, such improved positive AMS may aid contextualization and integration of trauma memories with existing beliefs (Resick & Schnicke, 1993) and other memories (Ehlers & Clark, 2000). Additionally, positive memories may replace trauma memories as reference points to influence self-concept (Bernsten & Rubin, 2007; Janoff-Bulman, 1992).

Improved Therapeutic Alliance

A weak therapeutic alliance has been shown to predict treatment dropout broadly (Samstag et al., 1998) and a poor treatment response for some trauma-focused interventions (McLaughlin et al., 2014). In PPMT, content from the positive memories and consequently improved affect and cognitions may increase self-efficacy, resilience, and effective coping (Philippe et al., 2009), reduce avoidance (Taylor & Stanton, 2007) in therapy, and facilitate homework completion (Contractor et al., 2018); all such factors may improve therapeutic alliance. Concurrently, it is possible that improved therapeutic alliance may contribute to better treatment outcomes and improved functioning (Buchholz & Abramowitz, 2020). In essence, the relation between improved therapeutic alliance and positive therapeutic outcomes may be reciprocal and bidirectional. Further, such improved therapeutic alliance may also facilitate engagement in later and as-needed trauma-focused interventions (Contractor et al., 2018).

Current Study Objectives

Existing literature indicates that patient and provider stakeholders wish to process positive memories in therapy, are receptive to discussing positive memories in therapy, and endorse that positive memory processing is feasible, practical, easy to learn, and can improve therapy satisfaction/tolerability (Caldas et al., 2020; Contractor, Caldas, et al., 2020). Further, evidence suggests that receiving positive memory interventions is associated with improved well-being (Miguel-Alvaro et al., 2021), and that positive psychology interventions as well as memory-based interventions, that are foundational to PPMT, have beneficial therapeutic outcomes (Barry et al., 2019; Carr et al., 2021). PPMT builds on this supporting evidence and provides a unique therapeutic technique for and tailored to PTSD’s unique symptomatology. In this regard, no existing PTSD intervention exclusively focuses on positive memories, which is different for PPMT. Also, PPMT is based on and integrates positive and symptom-focused treatment models, with a focus on increasing positive elements (e.g., affect, thoughts), while simultaneously decreasing negative elements such as PTSD severity (Contractor, Weiss, & Shea, 2020). Moreover, PPMT is a brief treatment and also potentially cost-effective due to its brevity as well as perceived ease in learning and implementing it (which may add to the possibility of PPMT’s dissemination by paraprofessionals). However, we are lacking information on PPMT’s feasibility and effects. To this end, this open label pilot explored PPMT’s effects on: (1) PTSD severity, (2) depression severity given its comorbidity with PTSD (Rytwinski et al., 2013) and established beneficial impacts of memory interventions for depression (Dalgleish & Werner-Seidler, 2014); (3) affect and cognitive processes (levels of positive and negative affect, positive emotion dysregulation, posttrauma cognitions, positive memory retrieval); and (4) therapeutic alliance.

Method

Procedure

The current open label pilot was approved by the University of North Texas Institutional Review Board. From September 2019, we recruited participants who were community members seeking assessment and therapy at the Psychology Clinic of the University. Inclusion criteria were 1) ≥18 years, 2) fluent in English, 3) experienced a traumatic event as assessed by the first question of the Primary Care PTSD Screen for DSM-5 (PC-PTSD-5; Prins et al., 2015), 4) no suicidal/homicidal plan or attempt in the past six months (including current), 5) no current experiences of auditory or visual hallucinations, and 6) no current utilization of psychotherapy services. The traumatic event could have occurred at any point in one’s lifetime. Further, we did not require endorsement of a certain level of PTSD symptom severity or a PTSD diagnosis as an inclusion criterion for this study considering evidence of functional impairment among individuals with sub-threshold PTSD (Zlotnick et al., 2002), a higher prevalence of sub-threshold vs. threshold PTSD (Franklin et al., 2018), and the debate surrounding a dichotomous PTSD diagnostic algorithm (Miller et al., 2014).

We made several adaptations and transitions to account for the Coronavirus Disease of 2019 (COVID-19) pandemic after March 2020. We paused study recruitment (including on-going sessions with active participants) until September 2020 while we were working on a telehealth-adaptation plan. Eventually, the in-person study transitioned to a Health Insurance Portability and Accountability Act (HIPAA)-compliant telehealth platform due to the on-going COVID-19 pandemic. For the telehealth platform, we scheduled sessions via Google Voice (confidentiality maintained; Google, 2020), and conducted assessments and therapy sessions via Zoom for Healthcare (HIPAA-compliant video conferencing application; Zoom Video Communications, 2020). Also, we included participants seeking assessments at the Psychology Clinic from December 2020 to accommodate reduced study recruitment attributed to the COVID-19 pandemic. Lastly, we adapted all study protocols and documents, including the informed consent and the study methodology (e.g., participants were emailed survey links; participants had to use their own electronic devices for PPMT sessions). Broadly, the study comprised of three phases (See Figure 1). Participants received a $10 gift card after each session evaluation.

Figure 1.

Diagrammatic representation of procedure and participant exclusion data

Recruitment and Screening Phase

Among individuals who contacted the Psychology Clinic for services, all adults reporting trauma experiences were briefly introduced to the study and asked about their interest to participate. Next, trained study personnel contacted interested participants to ensure that participants met all inclusion criteria, answered questions about the study, and scheduled eligible and interested participants for five weekly PPMT sessions.

Assessment and Orientation Phase

In Session 1, study procedures were reviewed and informed consent was obtained. Participants completed baseline measures via Qualtrics (Qualtrics, 2018). Participants were provided psychoeducation about PTSD and positive memories, and the rationale/overview of PPMT. Homework consisted of reviewing PPMT’s rationale and the Values, Affect, Strengths, and Thoughts (VAST) log.

Therapy Phase

In Session 2, study therapists defined positive memories by stating “I am first going to define positive memories - they reference important positive events that you personally have experienced. Positive memories make you feel good and bring up positive feelings (e.g., feeling safe, loved, joyful). Think about a specific, positive memory that is personally meaningful to you and that caused you to feel a lot of happiness and joy at the time that it occurred. This memory should be a specific experience that occurred within a 24-hour period.” These positive memories did not have to be memories prompted by the Autobiographical Memory Test cues (described later). After participants retrieved one positive memory, they were guided through an imaginal experiencing (narration) of that positive memory and they processed the positive memory to elicit relevant VAST elements. Participants were asked to retrieve and process a different positive memory for each session. Homework consisted of completing a VAST log (3–4 entries a week); participants wrote about an activating event that prompted them to retrieve a specific positive memory, details of the positive memory, and positive values, affect, strengths and thoughts associated with the described positive memory. Homework also consisted of engaging in a behavioral activity that the participant identified as important to them based on the discussed VAST elements in session (e.g., engaging in volunteering activities based on identified value of kindness). Sessions 3 and 4 involved the same procedure as Session 2, with different positive memories elicited.

Post-Therapy Phase

In Session 5, participants completed measures via Qualtrics (Qualtrics, 2018) to provide post-treatment data. Participants also provided quantitative and qualitative feedback on PPMT’s content, format, feasibility, and acceptability and about their subjective experiences throughout treatment via a survey. Study therapists provided feedback on participants’ progress, and shared perspectives on participant’s strengths and areas of growth. In reference to treatment continuity, all participants were informed that they were on the waitlist for therapeutic services at the Psychology Clinic and they would be contacted once a therapist from the Psychology Clinic is assigned to their case. Further, participants were provided with a treatment summary if they wished to have one; they could provide this treatment summary to their next therapy provider for treatment planning and case conceptualization purposes.

Participants and Exclusions

From September 2019-March 2020 (Pre-COVID), a total of 37 potential participants were screened (~6/month) by research personnel; of those, 13 participants met inclusion criteria and agreed to participate in the study. Of those, three participants were not available during the offered study timeslots and one participant could not take part in the study due to COVD-19-related closures. Thus, 9 participants were scheduled for the Assessment and Orientation Phase. Among the eight participants who attended the Assessment and Orientation Phase, three dropped out of the study - one participant was deemed ineligible due to active suicidal ideation during the Assessment and Orientation Phase, one participant completed two sessions before withdrawing due to scheduling conflicts, and one participant discontinued the study after completing two sessions due to COVD-19-related closures. Thus, five participants completed all PPMT sessions.

From March 2020-May 2021 (COVID Adaptations), a total of 46 potential participants were screened (~5/month) by study research personnel, and 13 participants met inclusion criteria and agreed to participate in the study. Of those, two participants were not available during the offered study timeslots. Thus, 11 participants were scheduled for the Assessment and Orientation Phase. Among the 10 participants who attended the Assessment and Orientation Phase (one canceled), three participants dropped out of the study - one participant was deemed ineligible due to active suicidal ideation during Session 3 (had indicated not reporting these concerns during the Assessment and Orientation Phase), one participant completed three sessions and then no-showed thereafter, and one participant completed three sessions before withdrawing due to subjective discomfort with memory tasks. For the participant who reported active suicidal ideation, we followed a standard protocol that was established for this study: study therapist conducted a thorough suicidal risk assessment that they were trained in, study therapist consulted with the principal investigator of this study (licensed clinical psychologist) immediately to conclude the lack of imminent risk of harming oneself, study therapist created a safety plan with the participant, and study therapist provided mental health resources (including the crisis text line) to the participant. Thus, seven participants completed all PPMT sessions.

The final sample for the current study included 12 participants who completed all PPMT sessions. Participants averaged 29.25 years of age (SD = 10.91); majority were women (n = 7; 58.30%) and 50% of participants identified as heterosexual (n = 6). All participants identified as White (n = 12; 100%). In terms of treatment history, the majority of participants reported a history of receiving psychotherapy in the past (n = 9; 75%), and were not currently taking medications for mental health concerns (n = 8, 66.70%). The most frequently endorsed traumatic events by participants were transportation accident (83.30%), other unwanted or uncomfortable sexual experience (83.30%), natural disaster (66.70%), and physical assault (66.70%). Table 1 has demographic data.

Table 1.

Descriptive information on demographics and psychopathology constructs

	Mean	SD
Age	29.25	10.91
		n (%)*
Gender	Man	4(33.30)
	Woman	7 (58.30)
	Other - Demifemale	1 (8.30)
Sexual Orientation	Heterosexual	6 (50)
	Lesbian/Gay	1 (8.30)
	Bisexual	2 (16.70)
	Pansexual	2 (16.70)
	Asexual	1 (8.30)
Education Level	Twelfth Grade	1 (8.30)
	First Year of College	2 (16.70)
	Second Year of College	2 (16.70)
	Third Year of College	1 (8.30)
	Fourth Year of College	1 (8.30)
	Seventh Year of College	1 (8.30)
	Tenth Year of College	3 (25)
	Other (Mid Masters)	1 (8.30)
Employment Status	Full-time	3 (25)
	Part-time	5 (41.70)
	Unemployed	2 (16.70)
	Not in Labor Force	2 (16.70)
Current Annual Household Income	Less than $15,000	4 (33.30)
	$15,000 to $24,999	4 (33.30)
	$25,000 to $34,999	2 (16.70)
	$35,000 to $44,999	0 (0)
	$45,000 to $54,999	1 (8.30)
	$55,000 to $64,999	0 (0)
	$65,000 or $79,999	1 (8.30)
Current Relationship Status	Single	6 (50)
	Dating	4 (33.30)
	Married	1 (8.30)
	Divorced, Separated, or Widowed	1 (8.30)
Ethnicity	Hispanic or Latino	2 (16.70)
	Not Hispanic or Latino	9 (75)
	Prefer Not to Respond	1 (8.30)
Race (Could Endorse Multiple Responses)	White	12 (100)
	Black/African American	1 (8.30)
	Asian	0 (0)
	American Indian/Alaskan Native	1 (8.30)
	Native Hawaiian/Other Pacific Islander	0 (0)
Received Psychotherapy in the Past**	Yes	9 (75)
	No	3 (25)
Currently Taking Medications for Problems with Emotions, Nerves, and/or Mental Health?	Yes	4 (33.30)
	No	8 (66.70)
Traumatic Event Frequencies (Could Endorse Multiple Responses)	Natural Disaster	8 (66.70)
	Fire or Explosion	6 (50)
	Transportation Accident	10 (83.30)
	Serious Accident at Work, Home, or During Recreational Activity	7 (58.30)
	Exposure to Toxic Substance	2 (16.70)
	Physical Assault	8 (66.70)
	Assault with a Weapon	6 (50)
	Sexual Assault	7 (58.30)
	Other Unwanted or Uncomfortable Sexual Experience	10 (83.30)
	Combat or Exposure to a War-Zone	3 (25)
	Captivity	2 (16.70)
	Life Threatening Illness or Injury	3 (25)
	Severe Human Suffering	4 (33.30)
	Sudden Violent Death	6 (50)
	Sudden Accident	6 (50)
	Serious Injury, Harm, or Death You Caused to Someone Else	3 (25)
	Other Stressful Event	5 (41.70)

Note.

^*Valid percentages to account for missing data;

^**Question was “Not counting the times you were an overnight patient in the hospital, did you ever in your lifetime go to see any professional for problems with your emotions, nerves, mental health, sleep or your use of alcohol or drugs?”

Study Therapists, Training, and Treatment Fidelity

Three clinical psychology doctoral students (3^rd to 4^th years of training) were study therapists. As recommended (Borrelli, 2011), a one day intensive training was provided to study therapists to increase the likelihood of delivering PPMT systematically. Study therapists were required to follow standardized session protocols. Further, sessions were video-recorded. The first author reviewed all PPMT sessions for one study participant for each of the study therapists; later, they reviewed 20–50% of video-recorded sessions (Borrelli, 2011). The first author (licensed clinical psychologist) provided weekly, one-hour group supervision to all study therapists.

Additionally, we created fidelity checklists to record whether proscribed PPMT components occurred or did not occur. Examples of domains assessed by these fidelity checklists were whether assessments were administered, whether study therapists provided the active ingredients of PPMT in each session (e.g., imaginal experiencing of the positive memory, processing the positive memory), whether study therapists provided each appropriate homework assignment, and whether study therapists obtained emotion and positive memory ratings. Using these fidelity checklists, two trained research assistants independently coded a random subset of 18 video-recorded sessions (~27% of total recorded sessions) for inter-rater reliability (IRR) consistent with recommended guidelines (Borrelli, 2011) as well as with procedures used in clinical trials (Holder et al., 2018; Stirman et al., 2021). Percent agreement across raters for treatment fidelity ratings averaged 90% across Sessions 1 through 5. Discrepancies were resolved between the raters. Then, one independent rater coded the remainder of the sessions for treatment fidelity. Treatment fidelity for study therapists across sessions ranged from 96.78–99.58%. Such estimates reflect high treatment fidelity (Borrelli, 2011).

Measures (Supplemental Table 1 indicates Cronbach’s α for each measure).

Experience of Traumatic Events

The Life Events Checklist for DSM-5 (LEC-5; Weathers, Blake, et al., 2013) is a 17-item self-report measure assessing experiences of lifetime traumatic events. Participants rated LEC-5 items on a 6-point nominal scale: happened to me, witnessed it, learned about it, part of my job, not sure, doesn’t apply.

PTSD Severity

Administered across all sessions, the PTSD Checklist for DSM-5 (PCL-5; Weathers, Litz, et al., 2013) is a 20-item self-report measure of PTSD symptoms using a 0 (not at all) to 4 (extremely) scale. The PCL-5 total score ranges from 0 to 80, and a cut-off score of ≥31 indicates probable PTSD (Bovin et al., 2016). Participants referenced the most traumatic event endorsed on the LEC-5 while responding to the PCL-5. Participants referenced the past month in Session 1, and “since the first session” in Sessions 2–5 to rate symptoms. The PCL-5 has excellent psychometrics including high internal consistency (Cronbach’s α = .91–.95), good test-retest reliability (r = .84), and excellent convergent and discriminant validity (Bovin et al., 2016; Wortmann et al., 2016). In this study, Cronbach’s α for the PCL-5 total score was .93 (Session 1) and .87 (Session 5).

Depression Severity

Administered across all sessions, the Patient Health Questionnaire-9 (PHQ-9; Kroenke & Spitzer, 2002) is a nine-item self-report measure of depression symptoms using a 0 (not at all) to 3 (nearly every day) scale. The PHQ-9 total score ranges from 0 to 27, and a cut-off score of ≥10 indicates probable depression (Kroenke & Spitzer, 2002). In Session 1, participants rated symptoms referencing the past two weeks. In Sessions 2–5, participants rated symptoms “since the first session” of this study. The PHQ-9 has good psychometrics such as good internal consistency (Cronbach’s α = .89), strong construct validity, and good criterion validity (Kroenke et al., 2001). In this study, Cronbach’s α for the PHQ-9 total score was .91 (Session 1) and .85 (Session 5).

Levels of Positive and Negative Affect

Administered during Sessions 1, 3, and 5, the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988) is a 20-item self-report measure of current positive and negative affect experiences cued to the present moment. Response options range from 1 (not at all/very slightly) to 5 (extremely); total score ranges from 10 to 50 for each of the positive affect and negative affect subscales. The PANAS demonstrates good psychometrics including high internal consistency (Cronbach’s α = .85–.89), good test-retest reliability (r_s = .79 - .81), and adequate convergent/discriminant validity (Watson et al., 1988). In this study, Cronbach’s α for the positive affect subscale score was .89 (Session 1) and .88 (Session 5); Cronbach’s α for the negative affect subscale score was .87 (Session 1) and .81 (Session 5).

Positive Emotion Dysregulation

Administered during Sessions 1, 3, and 5, the Difficulties in Emotion Regulation Scale–Positive (DERS-P; Weiss et al., 2015) is a 13-item self-report measure of positive emotion dysregulation (i.e., dysregulation stemming from experiencing positive emotions). The DERS-P does not reference any specific time frame. Sample items include “When I’m happy, I feel out of control” and “When I’m happy, I feel ashamed with myself for feeling that.” Response options range from 1 (almost never) to 5 (almost always); total score ranges from 13 to 65. The DERS-P has excellent psychometrics including excellent internal consistency (Cronbach’s α = .96), and strong convergent and discriminant validity (Weiss et al., 2019; Weiss et al., 2015). In this study, Cronbach’s α for the total score was .91 (Session 1) and .90 (Session 5).

Posttrauma Cognitions

Administered during Sessions 1, 3, and 5, the Posttraumatic Cognitions Inventory (PTCI; Foa et al., 1999) is a 36-item self-report measure of trauma-related thoughts. The PTCI does not reference any specific time frame. Response options range from 1 (totally disagree) to 7 (totally agree); total score ranges from 33 to 231 (we used 33 items of the PTCI as recommended by Foa et al., 1999). The PTCI consists of three subscales: Negative Cognitions about the Self, Negative Cognitions about the World, and Self-Blame. The PTCI demonstrates good psychometrics including excellent internal consistency (Cronbach’s α = .97), good test-retest reliability (r = .85), and good convergent and discriminant validity (Foa et al., 1999). In this study, Cronbach’s α for the total score was .97 in both Sessions 1 and 5.

Number of Retrieved Specific Positive Memories

Administered during Sessions 1 and 5, the Autobiographical Memory Test (AMT; Williams & Broadbent, 1986) was used to measure the quantity of specific positive memories retrieved by participants. Instructions for the AMT were adapted from prior studies (Henderson et al., 2002; Williams et al., 1996; Zinbarg et al., 2006). Procedurally, participants received cue words with a positive emotional valence followed by a prompt to retrieve a positive, personally meaningful, and specific memory of an event. We presented 10 positive memory cues in Session 1: friendly, happy, honest, kind, humorous, cheer, pleased, relieved, lively, and glorious (Kleim & Ehlers, 2008; McNally et al., 1995). We presented a different set of 10 positive cue words in Session 5: peaceful, loyal, intelligent, helpful, brave, safe, strong, important, love, and special (Harvey et al., 1998; Kleim & Ehlers, 2008; McNally et al., 1995). Participants were presented with one positive cue word at a time and had 60 seconds to retrieve one positive memory in response to each positive cue word. They had additional time to type details of each retrieved memory.

Using recommended AMT coding practices (Griffith et al., 2009; Sutherland & Bryant, 2005; Williams et al., 2007), each AMT response was coded as specific (memories of incidents that occurred within 24 hours), extended (memories of incidents which lasted longer than 24 hours), categoric (memories that were summations of repeated incidents), semantic associate (information that did not describe a personal memory), omission (failure to retrieve a memory or did not retrieve the memory within 60 seconds), or non-positive (memory with a negative valence or tone). To code memories as positive or non-positive, we modified coding criteria from the Coding and Assessment System for Narratives of Trauma (CASNOT; Fernández-Lansac & Crespo, 2017). Raters coded a memory as non-positive if the emotional tone and/or valence was coded as primarily negative (i.e., ratings of 3 or 4 using a scale from 0 [completely positive] to 4 [completely negative]). Raters also used the Affective Norms for English Words list (Bradley & Lang, 2017) to code emotional valence. In this manner, each response to each positive cue word received one final code from a rater. The variable used in the current study was the total count of AMT responses coded as specific memories; total score ranges from 0 to 10. Percent agreement for coding 220 AMT responses provided across all cue words as specific vs. overgeneral (i.e., categoric or extended) ranged from 50% (responses to the cue word “helpful”) to 100% (responses to the cue words “honest,” “kind,” “pleased,” and “peaceful”). The range in percent agreement was a function of the different responses provided to the different cue words. Discrepancies in coding were resolved with discussion across all raters and the first author. As a complement to this measure, we examined ease of processing specific positive memories using one item rated on a 0 (not at all) to 4 (extremely) scale by participants.

Therapeutic Alliance

Administered during Sessions 3 and 5, the Working Alliance Inventory (WAI; Horvath & Greenberg, 1989) is a 36-item measure assessing three aspects of the clinician-client relationship: 1) agreement on therapeutic tasks, 2) agreement on therapeutic goals, and 3) therapeutic bond. The WAI is scored on a 7-point scale ranging from 1 (never) to 7 (always); total score ranges from 36 to 252. The WAI demonstrates good psychometrics such as excellent internal consistency (Cronbach’s α = .93), good content validity, and adequate construct validity (Horvath & Greenberg, 1989; Tracey & Kokotovic, 1989). In this study, Cronbach’s α for the total score was .91 (Session 5).

Data Analyses

To examine study aims, we used the reliable change indices (RCI) and clinically significant change score (CS) approach – a two-fold criterion to determine clinically significant change following treatment (Jacobson et al., 1984; Jacobson & Truax, 1992). We did not conduct intention-to-treat analysis because there was no randomization component in our study design (Gupta, 2011), and each RCI/CS analysis was conducted ideographically on data obtained from a single participant. We examined whether a change in scores over time between two assessments for each participant was reliable and significant - whether each participant changed significantly more than would be expected given measurement error. Then, we examined whether a participant has transitioned from a functional to a dysfunctional population or vice versa (Jacobson et al., 1984; Jacobson & Truax, 1992). A functional population references individuals who have been discharged from treatment and/or are not clinically distressed (Jacobson & Truax, 1992). In comparison, a dysfunctional population references individuals from a clinical population (e.g., individuals receiving inpatient care; Jacobson & Truax, 1992) and individuals whose scores on psychological measures fall outside of a functional/normal population score range (> 1 or 2 SD; Jacobson et al., 1984; Nietzel et al., 1987; Wise, 2004).

First, we used the formula proposed by Jacobson et al. (1984) and Jacobson and Truax (1992) for the RCI computation. We used estimates of standard deviation (SD) for normal/pre-treatment/control population and each scale scores’ internal consistency estimates from prior studies (Supplemental Table 2; Lambert & Ogles, 2009) to compute standard error of the difference between pre- and post-PPMT scores, which is then multiplied by 1.28 (Wise, 2003, 2004). The generated range of change scores includes 90% of the change scores that would occur by chance. If RCI > ±1.28, it is likely that the post-PPMT score reflected real change (Jacobson et al., 1984). Next, we determined if participants met criteria for clinically significant change (Jacobson et al., 1984; Jacobson & Truax, 1992). We determined if the post-PPMT score fell within the 1 SD range (Nietzel et al., 1987; Wise, 2004) of the functional or normal population (in the direction of dysfunctionality; Jacobson et al., 1984; Jacobson & Truax, 1992). Differently, for the AMT, we determined if the post-PPMT score was 1 SD above the mean obtained from a functional group; we had difficulties identifying studies using similar AMT methodology with a dysfunctional population. Lastly, using both criteria, we classified each participant as Recovered (passed both CS and RCI criteria), Improved (passed RCI criterion alone), Unchanged/Indeterminate (passed neither), Deteriorated (passed RCI criterion in the negative direction), or No Change (did not pass RCI criterion; Matthey, 2004; Wise, 2004).

Results

At the Assessment and Orientation Phase, scores on PTSD severity, depression severity, positive affect levels, negative affect levels, posttrauma cognitions, positive emotion dysregulation, number of specific positive memories, and therapeutic alliance averaged 33 (SD = 17.35; range from 3 to 67), 13.67 (SD = 7.39), 22.33 (SD = 8.28), 19.50 (SD = 7.15), 115.83 (SD = 40.77), 20.17 (SD = 8.31), 5.17 (SD = 2.62), and 5.89 (SD = .66), respectively (Supplemental Table 3). Also, we compared PTSD severity and basic demographic variables across subgroups recruited pre-COVID vs. those recruited during the COVID-19 pandemic. Results showed no significant subgroup differences on PTSD severity (pre-PPMT PTSD severity, t = 1.05, p = .318; post-PPMT PTSD severity, t = 1.48, p = .170) and on basic demographic variables including age (t = .01, p = .990), gender (χ² = 1.86, p = .394), ethnicity (χ² = 1.40, p = .237), education (χ² = 7.20, p = .408), and household income (χ² = 3.77, p = .438).

Supplemental Table 3 indicates correlations; Table 2 indicates RCI and CS estimates for each participant. Referencing PTSD severity, a 6.53-point change in the PCL-5 score was needed to be 90% confident that a real change had occurred; nine participants (75%) demonstrated a reliable change in PTSD severity from pre- to post-PPMT. According to CS estimates, nine participants (75%) demonstrated a clinically significant change in PTSD severity from pre- to post-PPMT. Referencing depression severity, a 2.28-point change in the PHQ-9 scores was needed to be 90% confident that a real change had occurred; six participants (50%) demonstrated a reliable change in depression severity from pre- to post-PPMT. According to CS estimates, three participants (25%) demonstrated a clinically significant change in depression severity from pre- to post-PPMT.

Table 2.

Classifications based on Reliability Change Index (RCI) and Clinical Significance (CS) estimate computations

Participant	Pre-PPMT Score	Post-PPMT Score	RCI Score	RCI Determination	CS Estimate Determination	Categorization based on RCI and CS estimates
PTSD Severity
1t	16.00	22.00	1.18	No Change	Changed	No Change
2	41.00	12.00	−5.69	Reliable Change	Changed	Recovered
3*	67.00	30.00	−7.26	Reliable Change	Changed	Recovered
4	38.00	33.00	−.98	No Change	No Change	Indeterminate
5	30.00	22.00	−1.57	Reliable Change	Changed	Recovered
6t	24.00	34.00	1.96	Reliable Change	No Change	Deteriorated
7t	36.00	28.00	−1.57	Reliable Change	Changed	Recovered
8	18.00	8.00	−1.96	Reliable Change	Changed	Recovered
9t	33.00	8.00	−4.90	Reliable Change	Changed	Recovered
10t	26.00	19.00	−1.37	Reliable Change	Changed	Recovered
11t	60.00	34.00	−5.10	Reliable Change	No Change	Improved
12	7.00	6.00	−.20	No Change	Changed	No Change
				Reliable Change: n = 9, 75%	Changed: n = 9, 75%	Recovered/Improved: n = 8, 66.67%
Depression Severity
1t	16.00	16.00	.00	No Change	No Change	Indeterminate
2	8.00	6.00	−1.12	No Change	Changed	No Change
3t	23.00	15.00	−4.49	Reliable Change	No Change	Improved
4	7.00	15.00	4.49	Reliable Change	No Change	Deteriorated
5	18.00	12.00	−3.37	Reliable Change	No Change	Improved
6t	17.00	19.00	1.12	No Change	No Change	Indeterminate
7t	22.00	13.00	−5.05	Reliable Change	No Change	Improved
8	1.00	1.00	.00	No Change	Changed	No Change
9t	19.00	8.00	−6.17	Reliable Change	No Change	Improved
10t	16.00	16.00	.00	No Change	No Change	Indeterminate
11t	15.00	11.00	−2.24	Reliable Change	No Change	Improved
12	2.00	3.00	.56	No Change	Changed	No Change
				Reliable Change: n = 6, 50%	Changed: n = 3, 25%	Recovered/Improved: n = 5, 41.67%
Levels of Positive Affect
1t	13.00	15.00	.54	No Change	No Change	Indeterminate
2	36.00	40.00	1.08	No Change	Changed	No Change
3t	21.00	18.00	−.81	No Change	No Change	Indeterminate
4	33.00	26.00	−1.89	Reliable Change	Changed	Deteriorated
5	29.00	20.00	−2.43	Reliable Change	No Change	Deteriorated
6t	17.00	13.00	−1.08	No Change	No Change	Indeterminate
7t	22.00	21.00	−.27	No Change	No Change	Indeterminate
8	19.00	19.00	.00	No Change	No Change	Indeterminate
9t	14.00	21.00	1.89	Reliable Change	No Change	Improved
10t	14.00	13.00	−.27	No Change	No Change	Indeterminate
11t	17.00	24.00	1.89	Reliable Change	Changed	Recovered
12	33.00	25.00	−2.16	Reliable Change	Changed	Deteriorated
				Reliable Change: n = 5, 41.67%	Changed: n = 4, 33.33%	Recovered/Improved: n = 2,16.67%
Levels of Negative Affect
1t	18.00	26.00	2.70	Reliable Change	No Change	Deteriorated
2	19.00	10.00	−3.04	Reliable Change	Changed	Recovered
3t	19.00	12.00	−2.37	Reliable Change	Changed	Recovered
4	15.00	13.00	−.68	No Change	Changed	No Change
5	30.00	15.00	−5.07	Reliable Change	Changed	Recovered
6t	15.00	11.00	−1.35	Reliable Change	Changed	Recovered
7t	25.00	17.00	−2.70	Reliable Change	Changed	Recovered
8	14.00	10.00	−1.35	Reliable Change	Changed	Recovered
9t	24.00	13.00	−3.72	Reliable Change	Changed	Recovered
10t	12.00	13.00	.34	No Change	Changed	No Change
11t	33.00	12.00	−7.10	Reliable Change	Changed	Recovered
12	10.00	10.00	.00	No Change	Changed	No Change
				Reliable Change: n = 9, 75%	Changed: n = 11, 91.67%	Recovered/Improved: n = 8, 66.67%
Posttrauma Cognitions
1t	136.00	132.00	−.69	No Change	No Change	Indeterminate
2	119.00	78.00	−7.12	Reliable Change	Changed	Recovered
3t	164.00	142.00	−3.82	Reliable Change	No Change	Improved
4	104.00	86.00	−3.12	Reliable Change	No Change	Improved
5	105.00	170.00	11.28	Reliable Change	No Change	Deteriorated
6t	167.00	128.00	−6.77	Reliable Change	No Change	Improved
7t	80.00	103.00	3.99	Reliable Change	No Change	Deteriorated
8	50.00	48.00	−.35	No Change	Changed	No Change
9t	101.00	43.00	−10.07	Reliable Change	Changed	Recovered
10t	141.00	119.00	−3.82	Reliable Change	No Change	Improved
11t	167.00	128.00	−6.77	Reliable Change	No Change	Improved
12	56.00	63.00	1.22	No Change	Changed	No Change
				Reliable Change: n = 9, 75%	Changed: n = 4, 33.33%	Recovered/Improved: n = 7, 58.33%
Positive Emotion Dysregulation
1t	17.00	16.00	−.34	No Change	Changed	No Change
2	17.00	15.00	−.69	No Change	Changed	No Change
3t	40.00	28.00	−4.12	Reliable Change	Changed	Recovered
4	15.00	18.00	1.03	No Change	Changed	No Change
5	13.00	13.00	.00	No Change	Changed	No Change
6t	24.00	22.00	−.69	No Change	Changed	No Change
7t	13.00	17.00	1.37	Reliable Change	Changed	Deteriorated
8	14.00	13.00	−.34	No Change	Changed	No Change
9t	23.00	13.00	−3.43	Reliable Change	Changed	Recovered
10t	23.00	16.00	−2.40	Reliable Change	Changed	Recovered
11t	30.00	23.00	−2.40	Reliable Change	Changed	Recovered
12	13.00	13.00	.00	No Change	Changed	No Change
				Reliable Change: n = 5, 41.67%	Changed: n = 12, 100%	Recovered/Improved: n = 4, 33.33%
Number of Specific Positive Memories
1t	3.00	1.00	−3.59	Reliable Change	No Change	Indeterminate
2	10.00	6.00	−7.18	Reliable Change	Changed	Deteriorated
3t	6.00	7.00	1.79	Reliable Change	Changed	Recovered
4	5.00	1.00	−7.18	Reliable Change	No Change	Indeterminate
5	2.00	6.00	7.18	Reliable Change	Changed	Recovered
6t	9.00	7.00	−3.59	Reliable Change	Changed	Deteriorated
7t	1.00	.00	−1.79	Reliable Change	No Change	Indeterminate
8	5.00	3.00	−3.59	Reliable Change	No Change	Indeterminate
9t	7.00	4.00	−5.38	Reliable Change	Changed	Deteriorated
10t	5.00	1.00	−7.18	Reliable Change	No Change	Indeterminate
11t	5.00	10.00	8.97	Reliable Change	Changed	Recovered
12	4.00	4.00	.00	No Change	Changed	No Change
				Reliable Change: n = 11, 91.67%	Changed: n = 7, 58.33%	Recovered/Improved: n = 3, 25%
Therapeutic Alliance*
1t	4.72	4.61	−.45	No Change	No Change	Indeterminate
2	6.47	6.61	.56	No Change	Changed	No Change
3t	6.78	6.78	.00	No Change	Changed	No Change
4	5.69	5.50	−.79	No Change	Changed	No Change
5	5.86	6.47	2.48	Reliable Change	Changed	Recovered
6t	5.56	6.44	3.60	Reliable Change	Changed	Recovered
7t	6.75	5.89	−3.49	Reliable Change	Changed	Deteriorated
8	5.47	5.83	1.46	Reliable Change	Changed	Recovered
9t	6.19	6.39	.79	No Change	Changed	No Change
10t	5.83	5.97	.56	No Change	Changed	No Change
11t	4.94	6.11	4.73	Reliable Change	Changed	Recovered
12	6.44	6.25	−.79	No Change	Changed	No Change
				Reliable Change: n = 5, 41.67%	Changed: n = 11, 91.67%	Recovered/Improved: n = 4, 33.33%

Note. PTSD is posttraumatic stress disorder;

^tParticipants received telehealth PPMT;

^*Utilized mean score for comparison (Hanson et al., 2002)

For levels of positive affect, a 4.75-point change in the PANAS positive subscale score was needed to be 90% confident that a real change had occurred; five participants (41.67%) demonstrated a reliable change in levels of positive affect from pre- to post-PPMT. According to CS estimates, four participants (33.33%) demonstrated a clinically significant change in levels of positive affect from pre- to post-PPMT. For levels of negative affect, a 2.68-point change in the PANAS negative subscale score was needed to be 90% confident that a real change had occurred; nine participants (75%) demonstrated a reliable change in levels of negative affect from pre- to post-PPMT. According to CS estimates, 11 participants (91.67%) demonstrated a clinically significant change in levels of negative affect from pre- to post-PPMT. Referencing positive emotion dysregulation (i.e., dysregulation stemming from experiencing positive emotions), a 3.72-point change in the DERS-P score was needed to be 90% confident that a real change had occurred; five participants (41.67%) demonstrated a reliable change in positive emotion dysregulation from pre- to post-PPMT. According to CS estimates, all participants (100%) demonstrated a clinically significant change in positive emotion dysregulation.

For posttrauma cognitions, a 7.37-point change in the PTCI score was needed to be 90% confident that a real change had occurred; nine participants (75%) demonstrated a reliable change in posttrauma cognitions from pre- to post-PPMT. According to CS estimates, four participants (33.33%) demonstrated a clinically significant change in posttrauma cognitions from pre- to post-PPMT. Referencing number of specific positive memories, a 0.71-point change in the AMT score was needed to be 90% confident that a real change had occurred; 11 participants (91.67%) demonstrated a reliable change in number of retrieved specific positive memories from pre- to post-PPMT. According to CS estimates, seven participants (58.33%) demonstrated a clinically significant change in number of retrieved specific positive memories from pre- to post-PPMT. We saw the following trend in the variable referencing ease of processing specific positive memories across sessions: M = 2.50 (SD = 1.24; Session 1), M = 2.58 (SD = .90; Session 3), and M = 2.83 (SD = .94; Session 5).

For therapeutic alliance, a 0.32-point change in the WAI score was needed to be 90% confident of a real change; five participants (41.67%) demonstrated a reliable change in therapeutic alliance from pre- to post-PPMT. Referencing CS estimates, 11 participants (91.67%) demonstrated a clinically significant change in therapeutic alliance.

Discussion

Results of the current open label pilot suggest that PPMT is a potentially promising intervention for PTSD that warrants further empirical investigation. To reiterate, participants who shifted from a dysfunctional to a healthy/functional classification (clinically significant change) and the difference between their pre-PPMT and post-PPMT scores reflected real change were categorized as recovered; participants who showed reliable improvement post-PPMT but whose post-PPMT scores were still within a clinical range based on a dysfunctional population were categorized as improved (Matthey, 2004; Wise, 2004). Our findings indicated that nine participants showed reliable changes in PTSD severity, of which eight recovered or improved. In addition, six participants showed reliable changes in depression severity, of which five improved. Such results are consistent with research indicating beneficial impacts of retrieving positive memories on posttrauma distress and well-being (Contractor, Banducci, et al., 2020; Miguel-Alvaro et al., 2021). PPMT, by virtue of retrieving and detailing components of positive memories, could help to improve mood (Contractor, Banducci, et al., 2020; Quoidbach et al., 2015; Rusting & DeHart, 2000); in turn, this improved mood may enhance adaptive thoughts (Blaney, 1986; Fredrickson, 2001; Rusting & DeHart, 2000). Such improved mood and cognitions may reduce posttrauma distress (Brown et al., 2019; Galatzer-Levy et al., 2013). In this regard, we found that nine participants showed reliable changes in levels of negative affect (eight recovered); and nine participants showed reliable changes in posttrauma cognitions (seven improved/recovered).

Of the five participants who showed reliable changes in levels of positive affect, only two participants reported increases in levels of positive affect as expected. However, of the five participants who showed reliable changes in positive emotion dysregulation (i.e., dysregulation stemming from experiencing positive emotions), four participants reported decreases in positive emotion dysregulation. Perhaps, PPMT may help individuals regulate positive emotions effectively rather than just influence changes in levels of positive affect. These results are critical given growing evidence that trauma-exposed individuals with PTSD symptoms experience dysregulation of positive emotions (Weiss, Contractor, Forkus, et al., 2020). Such positive emotion dysregulation may be the result of negative affect interference (Frewen et al., 2010), perhaps due to negative cognitions about positive affect (Foa et al., 1999; Frewen et al., 2010). In this regard, these individuals may believe that strong emotions lead to difficulties in self-control or that positive emotions are harmful (Farnsworth & Sewell, 2011; Salters-Pedneault et al., 2007). Further, such positive emotion dysregulation may be the result of negative affect cued by physiological arousal in the context of positive experiences (Salters-Pedneault et al., 2007; Weiss et al., 2018). Indeed, research has shown that more PTSD severity related to fewer positive memories when individuals reported positive emotion dysregulation (Contractor, Weiss, et al., 2021).

Further, as a check on whether PPMT is engaging the important target of enhancing positive memory retrieval, we found that 11 participants showed reliable changes in the number of retrieved specific positive memories. However, of those, only three participants showed increases in the number of retrieved specific positive memories from pre-PPMT to post-PPMT. For these three participants, PPMT functioned in ways similar to other memory-based interventions that target enhancing one’s ability to retrieve and process positive memories (Miguel-Alvaro et al., 2021). For the other participants, we could interpret results in a couple of ways. Perhaps, PPMT may better target improving the phenomenological qualities of retrieved positive memories such as vividness and accessibility beyond just count of retrieved positive memories. In this regard, we found that participants reported an increasing ease of processing specific positive memories across sessions. Alternatively, perhaps more PPMT sessions and/or greater engagement in homework may be needed for PPMT to truly impact one’s ability to retrieve positive memories. As a methodological consideration, we note that our AMT methodology was different than what is used in existing studies (e.g., a specific combination of 10 positive cue words); thus, the usage of RCI and CS estimates that draw from prior studies may have impacted current study findings. Additionally, perhaps the act of retrieving positive memories along with VAST elements may render them more central to one’s identity and life stories (i.e., event centrality); these positive memories may replace trauma memories as the reference point to interpret life experiences and to influence self-concept (Berntsen et al., 2011). Thus, future studies may examine event centrality in reference to positive memories as a potential mechanism underlying PPMT’s impacts on PTSD symptoms.

Lastly, despite PPMT’s short-term modality, we found that five participants showed reliable changes in therapeutic alliance; of those, four participants reported increases in therapeutic alliance from pre-PPMT to post-PMT. For these four participants, increasing therapeutic alliance across sessions may help to lower posttrauma symptoms (Ormhaug et al., 2014) and to promote positive treatment outcomes (Flückiger et al., 2018). Future research needs to examine whether greater therapeutic alliance influences PPMT outcomes or whether improved therapeutic alliance is a response to improved outcomes associated with PPMT. Further, factors such as number of PPMT sessions, and client diagnostic profiles may have influenced the study findings.

Implications

Our data showed that about 42–67% of individuals benefitted from PPMT on one or more of the posttrauma indicators such as PTSD severity, depression severity, negative affect levels, and posttrauma cognitions. Results are consistent with existing statistics on how many individuals respond positively to trauma interventions (Cusack et al., 2016; Schnurr, 2017). An area of needed investigation is why certain individuals did not respond well to PPMT. As seen from the results, participants varied in improvement and deterioration levels for the different outcomes, and the same participant did not consistently show non-/poor response to PPMT across the different outcomes. Thus, PPMT may work differently for individuals across the different outcomes. Drawing on clinical observations (Fondren et al., 2021) and the existing literature, some potential factors could account for these idiographic variations: uncertainty about optimal dosage of sessions (Contractor, Banducci, et al., 2021a), comorbid conditions (e.g., substance use) that may influence positive memory retrieval (Banducci et al., 2020); and characteristics of traumas (e.g., interpersonal traumas; Contractor, Kearns, et al., 2021).

Study results also support and add to the Positive Memory-PTSD Model (Contractor et al., 2018) supporting direct effects of PPMT on affect, cognitions, and posttrauma symptoms. This being said, we note that perhaps, PPMT may better target improving the ability to regulate positive affect via exposure/habituation rather than solely improving levels of positive affect. This consideration has been outlined in the updated Positive Memory-PTSD Model (Contractor, Banducci, et al., 2021a). Positive emotion regulation strategies incorporated into PPMT may help to modify emotional processes, to promote cognitive reappraisals of emotional information, and to alter responses to positive emotions (Carl et al., 2013).

In addition, PPMT improved the ability to retrieve positive memories only among 25% of participants. Whether PPMT helps to improve phenomenological characteristics of retrieved positive memories beyond count, or whether certain therapist/client factors should be considered to maximize PPMT’s impacts on the ability for positive memory retrieval is an avenue for future research. Also, in the process of conducting this study, we developed and refined clinical practices integral to the ethical and effective conduction of this study such as preparation of fidelity checklists, informed consent documents, and data collection/monitoring procedures. Relatedly, we demonstrated acceptable indices and/or procedures of therapist training, treatment delivery, adherence to PPMT, as well as PPMT fidelity ratings. These factors are foundational to conducting larger and randomized clinical trials of PPMT.

We also note information on dropout rates for this study. The dropout rate of the current study is 30%; this estimate aligns with findings on dropout rates for trauma-focused interventions (Goetter et al., 2015; Lewis et al., 2020). For example, meta-analyses indicate that the pooled dropout rate from evidence-based, trauma-focused interventions ranges from 22% to 36% (Goetter et al., 2015; Lewis et al., 2020). Both external factors and internal factors may have contributed to dropouts in the current study. For example, external factors such as the COVID-19 pandemic may have posed challenges for participants (e.g., difficulties finding a confidential space for teletherapy, conflicts in schedule; Moring et al., 2020). In the current study, one participant discontinued the study after completing two sessions due to COVD-19-related closures. Further, internal factors such as avoidance/ambivalence, lack of readiness for change, and perceived stigma for mental health treatment may have contributed to the dropout rates (Moring et al., 2020). In the current study, one participant withdrew due to subjective discomfort with memory tasks. Thereby, exploring both external and internal factors in the pre-PPMT stage and explicitly addressing potential barriers to treatment retention is needed in future studies.

Some limitations need to be considered when interpreting study findings. First, this study did not have a control group, which means that we cannot rule out the influence of time or of non-specific/common therapy factors on PPMT outcomes. Relatedly, we are not able to discern whether common factors of therapy such as empathetic support and positive regard (Wampold, 2015) or PPMT’s active ingredients influenced study results. Second, we emphasize the pilot nature of this study. This pilot study was not powered for hypothesis testing, does not provide effect size estimates to guide future studies, and results are limited to this study’s sample characteristics and design (Leon et al., 2011). However, such pilot studies are important because they are foundational to examining impacts of and informing future implementations of novel interventions (Leon et al., 2011). Also, there is precedence for using small sample sizes (e.g., ranging from 7 to 11) in open label and pilot studies using statistical approaches similar or complementary to those applied in the current study (e.g., Sloan et al., 2013). Third, we acknowledge the effects of the COVID-19 pandemic on the current study in terms of transitioning the study to a telehealth platform mid-way and a considerable slowdown of participant recruitment and retention. Fourth, we did not use “real-world” variables (e.g., functional status) to determine CS (Wise, 2003); and RCI and CS estimates are fairly conservative and stringent indicators of treatment improvement (Wise, 2003). Perhaps, an individual may have demonstrated reliable decreases in symptoms but their post-PPMT score on that measure may not reflect a clinically significant change (Wise, 2003, 2004). Fifth, due to limited sample size as well as the uncontrolled nature of this pilot study, we were unable to do a component analyses of PPMT as well as examine specific mechanisms that potentially underlie PPMT’s effects on PTSD (e.g., improved affect/cognitions, elicitation of values and strengths and other VAST elements, completion of homework assignments such as engaging in a meaningful behavioral activity). These questions need further empirical investigation.

Lastly, there are some factors to consider given that this sample was selected from a psychology training clinic located at a university. There could have been a self-selection bias among participants who either chose PPMT due to its immediate availability (especially participants who perceived the need for immediate therapy and may be experiencing distress) or due to monetary compensation. Relatedly, participants could have perceived the need to report symptom improvement due to receiving monetary compensation. There could have also been a self-selection bias among participants who experienced more symptoms/distress and preferred to wait to receive more established therapeutic interventions by doctoral students in training who were supervised by licensed professionals. Also, the alternative treatment to PPMT as well as the treatment offered after PPMT to participants was contingent on the supervising faculty for the doctoral student therapist in training. These treatments could have ranged from psychodynamic therapy to cognitive-behavioral therapy, and the length of treatment varied contingent on participant’s needs as well as the modality.

In summary, PPMT may potentially impact certain posttrauma targets more effectively such as PTSD severity, depression severity, negative affect levels, and posttrauma cognitions. Further, PPMT may be more effective in improving regulation of experienced positive affect rather than primarily levels of experienced positive affect. Also, research is needed to examine what can improve PPMT’s impacts on the ability to retrieve positive memories, and whether PPMT better impacts quality vs. quantity of retrieved positive memories. Results of this study provide pilot data for a larger efficacy trial for PPMT.

Public Health Significance Statement.

The present study suggests that Processing of Positive Memories Technique (PPMT) may be a promising and feasible intervention for posttraumatic stress disorder (PTSD), pending further empirical investigation. Clinicians may use the technique of asking clients to retrieve and process positive memories repeatedly to alleviate PTSD/depression severity and improve affect (including ability to regulate positive emotions) and cognitions.

Data Availability:

The data that support the findings of this study are available from the corresponding author, AAC, upon reasonable request. We had not sought permission from study participants via the informed consent document to deposit de-identified data into an open source and public data repository.