A Pilot Randomized Controlled Trial of Cognitive Restructuring for PTSD and Alcohol Misuse following Recent Sexual Assault: Initial Efficacy and Feasibility

Abstract

Sexual assault is a pervasive problem, particularly for US college women. Although many recover naturally, a significant minority develop posttraumatic stress disorder (PTSD) or alcohol misuse. Intervening acutely can prevent chronic psychopathology from developing. This study tested the efficacy of a newly developed one-session + four coaching call intervention (BRITE) adapted from Cognitive Processing Therapy (CPT), an evidence-based treatment for chronic PTSD. Individuals over 18, who identified as female, with symptoms of PTSD and alcohol misuse were recruited within 10 weeks of sexual assault for a RCT comparing BRITE to symptom monitoring. Participants (N = 57) were young (M = 21.63 years) and predominately White (61.4%). PTSD and alcohol use were assessed at baseline, weekly for 5 weeks, and at 3-month follow-up by masked evaluators. Participants assigned to BRITE reported significantly less PTSD symptoms (d = 2.69; 95% CI: 1.92, 3.45) than symptom monitoring (d = 1.19; 95% CI: 0.59, 1.79), when comparing baseline vs. 3-month follow-up. For alcohol misuse, participants reported fewer drinks per drinking day from baseline to 3-month follow-up in BRITE (d = 0.63 (95% CI: 0.05, 1.20) and symptom monitoring (d = 0.13; 95% CI: −0.42, 0.69) although the group difference was not significant. There was a similar pattern for other alcohol use outcomes (i.e., heavy episodic drinking frequency, alcohol use consequences). Pilot findings support this newly developed, brief, and accessible cognitive approach for promoting acute recovery with a vulnerable population.

Sexual assault, defined as non-consensual, unwanted sexual contact of any kind, is the most common form of trauma exposure in college settings and affects up to a quarter of college women (15–27%; Moylan et al., 2019; Moylan & Javorka, 2020; Papp & McClelland, 2021). When comparing across traumatic events sexual assault confers the highest risk of posttraumatic stress disorder (PTSD) of any trauma type (Jakob et al., 2017). Nearly half of those exposed to sexual assault meet PTSD criteria one year post assault (Dworkin et al., 2023), and during college approximately 9% of students meet PTSD criteria (Read et al., 2011).

Four primary PTSD temporal paths have been identified: 1) resilience (low or absent symptoms), 2) recovery (initial high symptoms that decrease over time), 3) delayed onset (low initial symptoms that increase over time), and 4) chronic (high symptoms that stay high; Galatzer-Levy et al., 2018). Following sexual assault, most PTSD symptom improvements occur over the first three months following assault; 81.42% of survivors report symptoms consistent with PTSD one week after assault, 53.60% 3-months post-assault, and 41.5% of survivors meet criteria for PTSD 12 months following assault (Dworkin et al., 2023). Given that sexual assault is especially associated with high rates of non-recovery and chronicity of PTSD (Goodman-Williams et al., 2023), there is a need for sexual assault specific early interventions.

Indicated preventative interventions, also called early or acute interventions, are those designed to be delivered to individuals at high risk for PTSD non-recovery, typically within the first three months post-trauma (Birur et al., 2017; Oosterbaan et al., 2019). These are individuals who present with elevated early symptoms, those that are not demonstrating a recovery trajectory, or those who are at risk for non-recovery for other reasons (e.g., trauma type, alcohol use). The goal is to prevent the development of chronic PTSD and related psychopathology. Although there are a few promising interventions, none to date have been established as a gold-standard intervention and there is a call within the field for more research to develop effective and accessible early interventions (Bisson et al., 2019). Furthermore, early interventions following sexual assault significantly reduce PTSD (Dworkin & Schumacher, 2018), but they do not typically target alcohol use.

Alcohol use is one predictor of non-recovery trajectories for sexual assault survivors (Gong et al., 2019; Peter-Hagene & Ullman, 2015). It is thought that using alcohol to avoid trauma-related thoughts, emotions, and reminders may impede emotional processing necessary for recovery (Kaysen et al., 2011). In addition, alcohol consumption increases the risk of revictimization (Cusack et al., 2021), and women who have been sexually assaulted, especially those with PTSD, are at risk of higher alcohol consumption and of drinking-related negative consequences (Blanco et al., 2013; Guinle & Sinha, 2020). These associations suggest that, post-assault, clinicians should assess both PTSD and alcohol use due to high rates of co-occurrence and evidence of poorer outcomes for survivors who drink, and when both are present and it is indicated, they should use interventions that address PTSD and alcohol misuse concurrently.

There are several explanatory models of PTSD and alcohol co-occurrence (for an overview, see (Simpson et al., 2014)), but self-medication or negative reinforcement theory (Stewart et al., 1998) is the most robustly supported (Hawn et al., 2020). In the context of sexual assault, this theory posits that alcohol consumption mitigates in-the-moment assault-related distress. The temporary reduction in distress in turn increases drinking to manage future symptoms. Over time, this can lead to both PTSD maintenance and escalation of alcohol use. There is ample cross-sectional (Moskal et al., 2024), longitudinal (Pedersen et al., 2024), and micro-longitudinal (Kaysen et al., 2014; Simpson et al., 2014) research suggesting increases in PTSD are associated with increases in urges to drink, actual consumption, and negative consequences from drinking.

Consistent with self-medication theory, efficacious therapies use cognitive-behavioral strategies to concurrently target co-occurrence of PTSD and alcohol use (Roberts et al., 2015). However, these approaches tend to be multi-session (e.g., 12–16 sessions) and are characterized by difficulties with patient engagement and high dropout (e.g., Kline et al., 2021). Less is known about targeting comorbidity acutely following assault when functional relationships between PTSD and alcohol misuse are not yet well-ingrained. Delivered acutely, lighter touch interventions (e.g., single session) may be more feasible and still efficacious.

Cognitive interventions, particularly cognitive processing therapy (CPT; Resick et al., 2016), have a strong evidence base for treating chronic PTSD. Based on social-cognitive theories, CPT helps survivors modify unhelpful beliefs about the traumatic event, teaching skills to change maladaptive beliefs about the event itself (assimilation) and over-generalized beliefs about self and the world (over-accommodation). Typically done in 12 sessions, CPT has strong support for reducing chronic PTSD following sexual assault (e.g., Asmundson et al., 2019) and has been demonstrated to reduce alcohol use more than control conditions (Elbarazi et al., 2022; Pearson et al., 2019; Simpson et al., 2022). Therapy providers have expressed interest in a CPT-based integrated intervention that addresses PTSD and substance misuse (Vujanovic et al., 2023)

Two studies (Nixon, 2012; Nixon et al., 2016) have examined a six session CPT protocol as an early intervention following assault. These studies found that the six session intervention did reduce PTSD more so than treatment as usual (Nixon et al., 2016), but not more so than supportive counseling (Nixon, 2012). Neither of these studies addressed alcohol use. In trying to understand the relatively modest findings, one potential reason was that the intervention focused too much on over-accommodated beliefs and failed to adequately address assimilated beliefs. More recent conceptualization of change in CPT posit that adequately addressing assimilation is a key mechanism in CPT (Farmer et al., 2017; Resick et al., 2024).

The current study (NCT02808468) is a pilot trial of a brief intervention (one 90-minute session + four 20-minute coaching calls) based on CPT (Brief Restructuring Intervention following Trauma Exposure; BRITE). An earlier study evaluated a single cognitive restructuring session + four coaching call intervention for people with co-occurring chronic PTSD and AUD and demonstrated promising results, particularly in reducing drinking (Stappenbeck et al., 2015). Participants in that study were recruited for chronic PTSD and alcohol use disorders and the intervention specifically targeted drinking-related cognitions. Thus, the intervention was not designed specifically for sexual assault survivors, was not designed to promote early recovery following trauma, and did not explicitly address beliefs maintaining PTSD, especially assimilated beliefs. Addressing these behaviors during the critical window when symptoms are not yet entrenched may mean they are easier to shift with a low dose approach and may prevent the development of related comorbidities and life disruptions associated with PTSD. The current intervention builds on this prior intervention and other previously tested interventions. It is tailored to be used in the acute aftermath of trauma exposure, it addresses PTSD and alcohol use in an integrated way, and it is briefer than the six-session CPT protocol previously tested. Thus, a small pilot study was warranted to determine proof of concept, safety and tolerability. To explore preliminary efficacy and feasibility, this study compares BRITE to a symptom monitoring control condition from baseline to 3-month follow-up on PTSD symptom severity and alcohol use and related consequences. This study used weekly self-report measures to model symptom change for PTSD and alcohol use for BRITE and the control condition to better understand both intervention-related and natural patterns of recovery for PTSD and alcohol use following sexual assault. We hypothesized that participants randomly assigned to the BRITE intervention would show greater decreases in PTSD and alcohol use than those assigned to symptom monitoring at both post and follow-up, and those assigned to BRITE would show greater reductions week by week in self-reported PTSD and alcohol use compared to symptom monitoring.

Method

Participants

Participants enrolled in this pilot trial were 57 female-identifying individuals who reported an unwanted sexual experience, defined as any experience that was unwanted and distressing and involved some form of unwanted sexual contact, including touching, kissing, and penetration, within the last 10 weeks. Participants were recruited using university registrar emails, referrals from university and community organizations, and flyers. Figure 1 depicts study participants’ flow through recruitment, screening, intervention, and assessments. Other inclusion criteria were (1) meeting criteria for at least three of four DSM-5 PTSD symptom clusters (i.e., intrusions, persistent avoidance, negative alterations in cognitions/mood, arousal and reactivity; American Psychiatric Association, 2013), and (2) reporting hazardous drinking, defined as at least one incident of heavy episodic drinking (four drinks or more on one occasion; NIAAA, 2004) plus at least two negative consequences of drinking in the past month. Exclusion criteria were limited to those deemed necessary to reflect appropriate clinical care and feasibility of delivering the intervention and included: (1) being acutely suicidal (i.e., current active suicidal intent or plan; suicide attempt in the past 3 months), (2) exhibiting current psychosis (i.e., meets criteria for a current psychotic disorder as assessed by structured clinical interview), (3) not able to understand, speak, and/or write English, (4) planned absences that would interfere with 5 weeks of participation, and (5) lacking access to a telephone.

Figure 1: Consort Diagram of Recruitment and Enrollment.

Demographics for enrolled participants are presented in Table 1. Of the 57 participants enrolled in the trial, all but one were enrolled at a local university. Additionally, 31 (54.4%) reported working at least part-time, and 71.4% of the sample reported an annual personal income of less than $10,000. Participants ranged in age from 18–38 years (M = 21.63, SD = 3.73). At baseline, participants had moderate interview-rated PTSD symptoms (M = 32.84, SD = 8.33; possible range 0–80, observed range 11–54). All met DSM-5 diagnostic criteria for PTSD (i.e., had a minimum of one reexperiencing, one avoidance, two mood/cognition, and two hyperarousal symptoms; APA, 2013) excepting the requirement that symptoms have persisted for a month. Participants reported drinking at potentially hazardous levels (average number of drinks consumed per week in the past month [M = 10.2, SD = 6.1], past month number of heavy episodic drinking episodes [M = 5.60, SD = 3.95]). Over a third of the sample (38.5%) reported at least one incident of high intensity drinking in the previous month, defined as drinking at twice the level of a binge episode which is eight or more drinks per occasion for women (Hingson et al., 2017).

Table 1.

Participant demographics and descriptive statistics of the total sample and comparison by condition.

		Condition		Test of difference
Characteristic	Total sample (n = 57)	Control (n = 29)	Intervention (n = 28)	t	χ 2	p
Age in years: M (SD)	21.6 (3.7)	22.1 (3.4)	21.1 (4.0)	1.0		.15
Sexual orientation: n (%)					3.3	.35
Heterosexual	41 (71.9)	18 (62.1)	23 (82.1)
Bisexual	12 (21.1)	8 (27.6)	4 (14.3)
Lesbian	1 (13.5)	1 (3.5)	0 (0.0)
Questioning	3 (5.3)	2 (6.9)	1 (3.6)
Ethnicity: n (%)					3.7	.34
Caucasian/White	35 (61.4)	21 (72.4)	14 (50.0)
Black/African American	2 (3.5)	1 (3.5)	1 (3.6)
Asian/Pacific Islander	13 (22.8)	5 (17.2)	8 (28.6)
Other	7 (3.0)	2 (6.9)	5 (17.9)
Hispanic: n (%)	7 (12.3)	3 (10.3)	4 (14.3)		0.2	.65
Treatment history (y/n): n (%)	38 (66.7)	20 (69.0)	18 (64.3)		0.1	.71
PSS-I: M (SD)
Baseline	32.8 (8.3)	33.2 (8.4)	32.4 (8.4)	0.37		.36
3-month follow-up	16.8 (10.3)	22.1 (10.4)	11.4 (7.0)	4.04		.0001
TLFB drinks per drinking day: M (SD)
Baseline	4.1 (1.4)	4.0 (1.3)	4.2 (1.6)	−0.51		.69
3-month follow-up	3.2 (1.6)	3.1 (1.7)	3.2 (1.6)	−0.21		.58
TLFB heavy episodic drinking: M (SD)
Baseline	5.6 (3.9)	5.8 (3.5)	5.4 (4.4)	0.45		.33
3-month follow-up	2.9 (3.3)	3.7 (3.7)	2.2 (2.7)	1.53		.07
Alcohol use consequences: M (SD)
Baseline	18.74 (20.1)	19.8 (15.3)	17.7 (24.3)	0.39		.35
3-month follow-up	9.1 (11.7)	11.0 (8.8)	7.4 (14.0)	1.00		.16
PCL symptom severity: M (SD)
Baseline	44.2 (11.4)	43.1 (12.0)	45.4 (11.4)	−0.77		.78
Week 0	38.0 (13.6)	38.9 (14.9)	37.0 (12.4)	0.51		.31
Week 1	32.9 (12.7)	36.8 (11.1)	29.2 (13.2)	2.23		.02
Week 2	28.1 (1.8)	33.8 (11.4)	22.5 (10.8)	3.51		.001
Week 3	25.4 (13.9)	30.8 (12.9)	19.8 (12.8)	2.97		.002
Week 4	23.4 (1.92	28.4 (11.8)	18.4 (13.6)	2.78		.004
3-month follow-up	17.6 (12.4)	23.3 (13.2)	12.2 (8.8)	3.28		.001
DDQ drinks per drinking day: M (SD)
Baseline	3.8 (1.8)	3.6 (2.0)	3.9 (1.6)	−0.57		.71
Week 0	3.7 (2.7)	4.1 (2.6)	3.4 (2.7)	1.00		.16
Week 1	3.3 (2.3)	3.2 (1.9)	3.3 (2.6)	−0.07		.53
Week 2	3.1 (2.8)	3.5 (2.4)	2.8 (3.2)	0.83		.21
Week 3	3.1 (2.6)	3.2 (2.3)	3.0 (2.9)	0.34		.37
Week 4	3.0 (2.6)	3.2 (2.4)	2.8 (2.9)	0.52		.30
3-month follow-up	3.0 (2.0)	3.3 (1.9)	2.6 (2.0)	1.14		.13
DDQ heavy episodic drinking: M (SD)
Baseline	1.6 (1.4)	1.5 (1.1)	1.8 (1.6)	−0.84		.80
Week 0	1.3 (1.3)	1.7 (1.4)	1.0 (1.1)	2.13		.02
Week 1	1.2 (1.2)	1.4 (1.3)	1.0 (1.1)	0.94		.18
Week 2	1.0 (1.2)	1.3 (1.3)	0.7 (1.0)	1.89		.03
Week 3	1.0 (1.20	1.3 (1.3)	0.6 (0.9)	2.05		.02
Week 4	1.1 (1.3)	1.2 (1.2)	1.0 (1.4)	0.53		.30
3-month follow-up	0.9 (1.2)	1.2 (1.2)	0.5 (1.0)	2.03		.02

Measures

Eligibility Interview Measures.

Index sexual assault was assessed by trained interviewer who queried on the date of the event, the degree of threat, and the perpetrator, and obtained a qualitative narrative description of what occurred to ensure it met DSM-5 criterion A (APA, 2013). The mood, psychosis, alcohol, and substance use modules of the Mini International Neuropsychiatric Interview for DSM-5 (MINI v7; Sheehan, 1998) were administered to rule out exclusion criteria of active psychosis and active suicidality and confirm inclusion criteria of maladaptive alcohol use (i.e., negative consequences of use).

Primary Outcome Interview Measures.

Interview measures were administered at baseline and 3-month follow-up. The Posttraumatic Symptom Scale Interview Version for DSM-5 (PSS-I-5; Foa & Capaldi, 2013) was used to assess past two-week PTSD symptoms anchored to the index assault (i.e., the assault in the past 10 weeks for which they were seeking treatment). The PSS-I has excellent psychometric properties including high test-retest reliability, strong convergent validity with other PTSD assessments, and excellent discriminant validity when compared with measures of anxiety and depression (Foa et al., 2016). The Timeline Followback (TLFB; Sobell & Sobell, 1992), an interview-guided calendar-based method of assessing alcohol use quantity and frequency, was administered to confirm alcohol use patterns in the last 30 days, including at least one day of heavy episodic drinking (i.e., four+ drinks on one occasion). The TLFB is a gold standard measure of alcohol use with high reliability in clinical samples and high test-retest reliability (Sobell et al., 1988).

Secondary Outcome Self-Report Measures.

Participants completed self-report measures at all assessment time points except alcohol use consequences, which was collected only at baseline and 3-month follow-up. The PTSD Checklist for DSM-5, civilian version (PCL-5; Weathers et al., 2003) was used to assess self-reported PTSD symptoms specifically related to their index sexual assault over the past month (intake, follow-up) or past week (weekly surveys). The PCL-5 contains 20 items corresponding to the DSM-5 criteria for PTSD and each item is rated 0–4 (range: 0–80). The Daily Drinking Questionnaire (DDQ; Collins et al., 1985) assessed drinking quantity and frequency. The DDQ asks participants to rate the number of drinks and hours spent drinking on each day of a typical week in the last three months (intake, follow-up) or on each day of the last week (weekly surveys). The Drinkers Inventory of Consequences (DrInC; Miller et al., 1995) was used to collect self-reported negative consequences of drinking in the past month. The DrInC negative consequences scale assesses negative consequences such as injuries, accidents, educational impairments, and interpersonal problems. The DrInC sums 45 consequences items rated from 0 never to 3 daily or almost daily (score range 0–135).

Satisfaction and Acceptability.

Participants randomized to the intervention were asked several questions to assess perceptions of the intervention. All questions used a 1 – 7 Likert scale. Items asked about interest, distress and helpfulness for the in-person intervention session, coaching calls and therapy skills. Additional items asked about whether their needs were met and whether they would refer a friend with a similar problem to this program.

Procedures

All procedures were approved by the IRB at the first authors’ university and were conducted on the telephone and in person at the institution. Study procedures took place between October 2017 and April 2020.

Recruitment.

Recruitment efforts included flyers, online advertisements, and registrar emails sent to female-identifying students enrolled at the university. Advertisements described the study as providing an intervention to reduce distress associated with sexual assault. Interested participants contacted through registrar emails could contact the study by phone or email or complete an online screening survey to be contacted directly by study staff. All participants provided verbal consent and were screened for preliminary eligibility over the phone. Although recruitment strategies targeted both the general local community as well as the authors’ university, most participants were recruited into the study using university-specific advertising techniques.

Baseline.

Participants deemed potentially eligible on the phone screen were scheduled for an in-person 90-minute intake. In this interview, a trained evaluator obtained informed consent. Participants then completed the PSS-I, TLFB, and MINI interviews and self-report measures (e.g., DDQ, DrInC). Evaluators also presented participants with a handout on common reactions to trauma and psychoeducation about natural recovery of symptoms.

Intakes were discussed weekly at a treatment team meeting consisting of study evaluators, study clinicians (including the study PI), and the research coordinator. The team made final entry decisions based on group discussion and consensus regarding eligibility criteria. Participants were then contacted and informed of final eligibility decision along with randomization condition (BRITE intervention, symptom monitoring) by the study coordinator.

Randomization.

An Excel random number generator was used to assign participants randomly to condition (BRITE intervention vs. symptom monitoring). Given the pilot nature of the study, no stratification was used for randomization. The research coordinator was responsible for randomization of each participant. Evaluators were masked to intervention assignment as well as to engagement in intervention (e.g., completer vs not).

BRITE Intervention.

Prior to completing the intervention, participants completed weekly assessments (Week 0 assessment point; e.g., PSS-SR, DDQ). The intervention included one 90-minute in-person session and four consecutive weekly 20-minute coaching calls. Participants also completed online self-report measures of PTSD and alcohol use each week (Week 1–4 assessment points). Therapy was delivered by PhD level therapists, faculty and fellows, with the same interventionist providing the in-person therapy and the coaching calls. The intervention provided psychoeducation on a cognitive model of PTSD, and the rationale for using cognitive strategies to process trauma-related beliefs and emotions. It also included psychoeducation on common reactions, natural recovery, and factors that predict the development of PTSD, including the role of using alcohol to avoid or manage trauma-related thoughts and feelings (e.g., self-medication). Participants were taught basic skills around identifying thoughts and feelings and using cognitive restructuring to challenge thoughts related to the recent sexual assault and alcohol use (see Bedard-Gilligan et al., 2020 for detailed intervention description).

Participants completed worksheets with the interventionist and took others with them as “homework.” Worksheets consisted of an ABC worksheet (i.e., a thought record) that guided the participant through connecting events, thoughts, and feelings related to the assault, drinking behavior, or both. The ABC worksheet consisted of an activating event column (A), a belief column with space to record unhelpful beliefs (B), and a third column to record emotions, behaviors, and consequences related to having the belief (C). It also included a challenging questions column with a subset of challenging questions (e.g., “what is the evidence for this thought?” “Is it all or none thinking?”) to use to challenge that belief, and a final column with space to generate a new more balanced belief and to rerate emotions. Participants also received supplemental worksheets that were intended to act as “how to” and problem-solving guides including tips for identifying unhelpful beliefs, additional questions that could be used to challenge thoughts, and problem-solving strategies for completing ABC sheets. These sheets were reviewed in session and participants were sent home with several copies and encouraged to use them to guide their at-home practice. Participants were also sent home with a recording of the intervention, not their personalized session but a script of intervention main points, to use as needed at home. Thus, the intervention retained the “core” skills of CPT (and other cognitive therapies) - identifying thoughts and feelings and learning to question those thoughts to build more helpful and accurate beliefs - but did not include other skills included in CPT (e.g., incremental building of identifying and challenging beliefs, identifying patterns of thinking). In addition, the intervention prioritized assimilated beliefs (e.g., self-blame for the sexual assault).

After the in-person session, participants completed four weekly 15–20-minute coaching calls with the interventionist. Prior to each call participants completed weekly assessment measures. Coaching calls reinforced participants’ use of restructuring skills around sexual assault and alcohol-related beliefs, problem-solved difficulties with skill-building, discussed trauma symptoms and alcohol use, and provided support.

Symptom Monitoring.

Participants assigned to the symptom monitoring condition completed online quantitative self-report measures of trauma symptoms and alcohol use each week for five weeks to correspond to the BRITE intervention (Week 0–4 assessment points). They did not have additional contact with study staff unless specifically requested by them or if they reported symptoms that were concerning or needed immediate intervention (e.g., alcohol use at potentially lethal levels). Those in symptom monitoring were offered the option to receive BRITE following the 3-month follow-up but outcome data was not collected for these participants treatment phase.

Follow up.

Participants came in-person for 3-month follow-up interviews (PSS-I, TLFB, MINI) and self-report measures. Participants met with the same evaluator from intake.

Treatment Fidelity.

Weekly supervision meetings were held with the therapist team and audio recordings of sessions were reviewed and discussed to minimize therapist drift. Fifty-seven core intervention elements (e.g., teaching cognitive restructuring skills, guiding participant through examples, assigning practice) were coded for 20% of completed sessions. The intervention coder was a PhD level clinician, not affiliated with the study, who had expertise in working with survivors of sexual assault and in providing both trauma-focused and alcohol-focused interventions. Fidelity to these coded elements was high, with coded sessions containing a mean of 51.60 (SD = 5.41; Range 43–56) or 91% of the core elements. There was not a clear pattern to session missed components with the exception that three sessions were coded as missing the review of strategies at the end of the session.

Evaluator Reliability.

Evaluators presented cases weekly to the study team. Discussion centered around ratings and study inclusion/exclusion to help ensure reliability among evaluators. Following data collection, assessment interrater reliability was calculated for 10% of the sample. The agreement between the evaluators was good, with a high overlap in PTSD severity ratings (PSS-I; ICC = .99). In addition, raters had 100% agreement on diagnoses of mood disorders, alcohol use and substance use disorders, and psychotic disorders on the MINI.

Results, study protocol, and analysis plan are registered on clinicaltrials.gov (NCT02808468). Data, intervention materials, and statistical code are available from the first author. There were no adverse events reported during the study.

Analytic Approach

Preliminary Analyses

Stata 15.1 (StataCorp, 2017) was used for all analyses. Potential differences by treatment conditions in demographics and background characteristics including age, race, ethnicity, sexual orientation, and treatment history were examined with t-tests or chi-square tests.

Primary Analyses

First, we examined intervention satisfaction and acceptability. To determine satisfaction and acceptability of BRITE, we examined means and standard deviations of each item. Next, to evaluate differences between BRITE and the symptom monitoring condition in PTSD and alcohol use outcomes between baseline and 3-month follow-up, we conducted separate multilevel mixed-effects models for PTSD symptom severity, average drinks per drinking day, heavy episodic drinking frequency, and alcohol consequences. We fit two-level random-intercept models with an independent covariance structure between the random effects. Each model controlled for treatment history (0 = none; 1 = any) as this was a low-dose intervention, and it was thought that prior experience with mental health treatment could affect response. Of primary interest was a condition (0 = control; 1 = intervention) x time (0 = baseline, 1 = 3-month follow-up) interaction. We examined pairwise comparisons to elucidate all intervention x time effects. Effect sizes for each outcome were computed by comparing the marginal means at baseline and 3-month follow-up for the intervention group versus control (Lipsey & Wilson, 2001).

Finally, we conducted similar multilevel mixed-effects models examining differences between BRITE and the symptom monitoring control condition in PTSD symptom severity, average drinks per drinking day, and heavy episodic drinking frequency using self-reports at baseline, intervention session/week 0, each of the four weeks of coaching calls (weeks 1–4), and the 3-month follow-up. Each of these models also controlled for treatment history (0 = none; 1 = any). Condition and time were considered categorical predictors. We were primarily interested in a condition (0 = control; 1 = intervention) x time (0 = baseline, 1 = week 0, 2 = week 1, 3 = week 2, 4 = week 3, 5 = week 4, 6 = 3-month follow-up) interaction. Pairwise comparisons for intervention x time effects were examined as described above. Effect sizes for each outcome were computed by comparing the marginal means at baseline and 3-month follow-up for the intervention group versus control (Lipsey & Wilson, 2001).

Power and Sample Size.

Power was calculated a priori for the hypothesis comparing the intervention to symptom monitoring for PTSD and drinking. We expected a medium-to-large effect size for outcomes based on effect sizes in past trials using CPT for PTSD. Using G*Power 3.1.7 with an ANOVA with 3 repeated measures, a sample size of 60 (30 per condition) was needed to identify medium-to-large effects with power of .80. Our final sample was smaller than anticipated due to recruitment challenges and the COVID-19 pandemic.

Missing Data and Attrition

Of the 57 participants, 51 (89.5%) completed the Week 1 assessment, 48 (84.2%) completed Week 2, 49 (86.0%) completed Week 3, 50 (87.7%) completed Week 4, and 44 (77.2%) completed the 3-month follow-up. There were no differences in demographic characteristics and baseline reports of PTSD severity, alcohol use, or alcohol consequences between those who completed all possible surveys (68.4%, n = 39) and those who were missing at least one survey (31.6%, n = 18), all p’s > .13. Thus, missing data appears primarily due to attrition. Rates of completed follow up are consistent with other studies on early intervention for symptoms following sexual assault (Gilmore et al., 2021). Intent-to-treat analyses were conducted and therefore included all available data.

Results

Preliminary Analyses

Descriptive statistics are in Table 1. There were no significant demographic differences or differences in baseline PTSD or drinking (i.e., heavy episodic drinking, drinks per drinking day, drinking consequences) between intervention versus control. Of the 28 participants randomized to the intervention condition, all (100%) completed the in-person intervention session, 22 (78.6%) completed all four coaching calls, three (10.7%) completed three calls, two (7.1%) completed only one call, and one (3.6%) did not complete any coaching calls.

Primary Analyses

Satisfaction and Acceptability of BRITE

Data on satisfaction with BRITE was collected from 24 (of 28) participants. In general, they reported overall high ratings of satisfaction with the intervention. Participants reported that the intervention session (M = 5.0, SD = 1.5), coaching calls (M = 5.0, SD = 1.8), and intervention skills (M = 5.2, SD = 1.8) were on average helpful (possible range 1 = not helpful to 7 = helpful). Distress ratings for each portion, intervention session (M = 3.9, SD = 1.8), coaching calls (M = 3.4, SD = 1.7), and intervention skills (M = 3.2, SD = 1.8), indicated low to moderate distress during the intervention (possible range 1 = not distressing to 7 = very distressing). Interest ratings for the intervention session (M = 2.8, SD = 1.4), coaching calls (M = 3.2, SD = 1.8), and intervention skills (M = 2.7, SD = 1.6) suggested interest in all aspects of the program (possible range 1 = very interesting to 7 = very boring). Additionally, on average participants tended to agree (1 = agree strongly; 7 = disagree strongly) with the statement that the program met their needs overall (M = 3.1, SD = 1.7) and were likely to recommend this program to a friend with similar problems (M = 2.5, SD = 2.0; possible range 1 = very likely; 7 = not likely).

Symptom Change from Baseline to 3-Month Follow-up (Table 2)

Table 2.

Results of the multilevel mixed-effects models and planned pairwise comparisons examining change from baseline to 3-month follow-up by condition.

	PSS-I	Average drinks per drinking day	Heavy episodic drinking frequency	Alcohol use consequences
Model parameters
Constant	37.11 (2.03)***	4.15 (0.38)***	5.59 (1.64)***	21.09 (4.34)***
Treatment History (none/any)	−5.62 (2.04)**	−0.21 (0.37)	0.35 (0.90)	−1.94 (4.38)
Intervention (vs. Control)	−1.08 (2.10)	0.19 (0.40)	−0.45 (0.96)	−2.17 (4.45)
3-Month FU (vs. Baseline)	−11.43 (1.45)***	−0.95 (0.32)**	−2.22 (0.76)**	−8.64 (3.00)**
Intervention x 3-Month FU	−9.60 (2.06)***	0.07 (0.45)	−0.85 (1.07)	−2.77 (4.21)
Contrasts
Baseline vs. 3m FU: Control	−11.43 (1.45)***	−0.95 (0.32)**	−2.22 (0.76)**	−8.64 (3.00)**
Baseline vs. 3m FU: Intervention	−21.04 (1.46)***	−0.88 (0.32)**	−3.07 (0.76)***	−11.41 (2.95)***
Intervention vs. Control: Baseline	−1.08 (2.10)	0.19 (0.40)	−0.45 (0.96)	−2.17 (4.45)
Intervention vs. Control: 3m FU	−10.68 (2.28)***	0.25 (0.44)	−1.30 (1.06)	−4.94(4.85)

Note. Unstandardized estimate (standard error) are represented.

^***p < .001,

^**p < .01.

After accounting for treatment history, there was a significant main effect of time on PSS-I scores such that PTSD severity was lower at 3-month follow-up compared to baseline regardless of condition. There was also a significant time x condition interaction. The intervention condition baseline vs. 3-month follow-up effect size was d = 2.69 (95% CI: 1.92, 3.45) and for control was d = 1.19 (95% CI: 0.59, 1.79). Pairwise comparisons showed significant decreases from baseline to 3-month follow-up in both conditions. However, intervention participants had significantly lower PTSD severity at the 3-month follow-up than those in control (see Figure 2).

Figure 2. Interview Rated PTSD from Baseline to 3-Month Follow-up by Intervention Condition.

For average drinks per drinking day, there was a significant main effect of time. After accounting for treatment history, participants in both conditions reported fewer drinks per drinking day on average at follow-up compared to baseline. The effect size for intervention condition baseline vs. 3-month follow-up was d = 0.63 (95% CI: 0.05, 1.20) and for control was d = 0.13 (95% CI: −0.42, 0.69). There was a similar pattern for both heavy episodic drinking frequency and alcohol use consequences, which decreased from baseline to 3-month follow-up for participants regardless of condition. For heavy episodic drinking frequency, the intervention condition baseline vs. 3-month follow-up d = 0.85 (95% CI: 0.27, 1.44) and for control was d = 0.59 (95% CI: 0.02, 1.15). For consequences, the intervention baseline vs. 3-month follow-up d = 0.50 (95% CI: −0.06, 1.07) and for control was d = 0.68 (95% CI: 0.11, 1.25). Pairwise comparisons showed significant decreases from baseline to 3-month follow-up for participants in both conditions and there were no statistically significant differences by condition.

Weekly Symptom Change (Table 3)

Table 3.

Results of the multilevel mixed-effects models and planned pairwise comparisons examining weekly change and differences between condition at eachs time point.

	PCL	Average drinks per drinking day	Heavy episodic drinking
Model parameters
Constant	46.37 (3.03)***	3.83 (0.54)***	1.45 (0.28)***
Treatment History (none/any)	−4.73 (2.93)	−0.30 (0.43)	0.05 (0.24)
Intervention (vs. Control)	2.10 (3.23)	0.26 (0.64)	0.31 (0.32)
Time
Week 0 (vs. Baseline)	−3.77 (1.86)*	0.45 (0.55)	0.18 (0.25)
Week 1 (vs. Baseline)	−5.39 (1.93)**	−0.22 (0.57)	−0.12 (0.26)
Week 2 (vs. Baseline)	−7.83 (1.95)***	−0.11 (0.58)	−0.21 (0.27)
Week 3 (vs. Baseline)	−10.63 (1.93)***	−0.34 (0.57)	−0.23 (0.26)
Week 4 (vs. Baseline)	−12.90 (1.93)***	−0.34 (0.57)	−0.33 (0.26)
3-Month FU (vs. Baseline)	−17.90 (2.04)***	−0.21 (0.60)	−0.31 (0.28)
Time x Condition
Intervention x Week 0	−4.62 (2.63)	−0.97 (0.78)	−1.00 (0.36)**
Intervention x Week 1	−10.98 (2.71)***	−0.49 (0.80)	−0.67 (0.37)
Intervention x Week 2	−14.65 (2.77)***	−1.00 (0.82)	−0.89 (0.38)*
Intervention x Week 3	−14.99 (2.75)***	−0.49 (0.81)	−0.93 (0.37)*
Intervention x Week 4	−14.21 (2.73)***	−0.78 (0.81)	−0.49 (0.37)
Intervention x 3-Month FU	−14.26 (2.87)***	−0.96 (0.85)	−0.91 (0.39)*
Contrasts
Baseline vs. Week 0: Control	−3.77 (1.86)*	0.45 (0.55)	0.18 (0.25)
Baseline vs. Week 0: Intervention	−8.39 (1.87)***	−0.51 (0.55)	−0.82 (0.25)**
Baseline vs. Week 1: Control	−5.39 (1.93)**	−0.22 (0.57)	−0.12 (0.26)
Baseline vs. Week 1: Intervention	−16.37 (1.91)***	−0.70 (0.57)	−0.79 (0.26)**
Baseline vs. Week 2: Control	−7.83 (1.95)***	−0.11 (0.58)	−0.21 (0.27)
Baseline vs. Week 2: Intervention	−22.48 (1.96)***	−1.12 (0.58)	−1.10 (0.27)***
Baseline vs. Week 3: Control	−10.63 (1.93)***	−0.34 (0.57)	−0.23 (0.26)
Baseline vs. Week 3: Intervention	−25.62 (1.96)***	−0.83 (0.58)	−1.16 (0.27)***
Baseline vs. Week 4: Control	−12.90 (1.93)***	−0.34 (0.57)	−0.33 (0.26)
Baseline vs. Week 4: Intervention	−27.11 (1.94)***	−1.12 (0.57)	−0.82 (0.26)**
Baseline vs. 3m FU: Control	−17.90 (2.04)***	−0.21 (0.60)	−0.31 (0.28)
Baseline vs. 3m FU: Intervention	−32.16 (2.04)***	−1.17 (0.60)	−1.22 (0.27)***
Week 0 vs. Week 1: Control	−1.62 (1.95)	−0.67 (0.57)	−0.29 (0.26)
Week 0 vs. Week 1: Intervention	−7.98 (1.91)***	−0.19 (0.57)	0.03 (0.26)
Week 0 vs. Week 2: Control	−4.06 (1.96)*	−0.56 (0.58)	−0.39 (0.27)
Week 0 vs. Week 2: Intervention	−14.08 (1.96)***	−0.60 (0.58)	−0.28 (0.27)
Week 0 vs. Week 3: Control	−6.86 (1.94)**	−0.79 (0.57)	−0.41 (0.26)
Week 0 vs. Week 3: Intervention	−17.22 (1.96)***	−0.31 (0.58)	−0.34 (0.27)
Week 0 vs. Week 4: Control	−9.13 (1.94)***	−0.79 (0.57)	−0.51 (0.26)
Week 0 vs. Week 4: Intervention	−18.72 (1.94)***	−0.60 (0.57)	−0.001 (0.26)
Week 0 vs. 3m FU: Control	−14.13 (2.05)***	−0.66 (0.60)	−0.49 (0.28)
Week 0 vs 3m FU: Intervention	−23.77 (2.02)***	−0.65 (0.60)	−0.40 (0.27)
Week 1 vs. Week 2: Control	−2.44 (2.02)	0.10 (0.60)	−0.09 (0.27)
Week 1 vs. Week 2: Intervention	−6.11 (1.99)**	−0.41 (0.59)	−0.31 (0.27)
Week 1 vs. Week 3: Control	−5.24 (1.99)**	−0.12 (0.59)	−0.11 (0.27)
Week 1 vs. Week 3: Intervention	−9.25 (1.99)***	−0.12 (0.59)	−0.37 (0.27)
Week 1 vs. Week 4: Control	−7.51 (1.99)***	−0.12 (0.59)	−0.21 (0.27)
Week 1 vs. Week 4: Intervention	−10.74 (1.96)***	−0.41 (0.59)	−0.03 (0.27)
Week 1 vs. 3m FU: Control	−12.51 (2.09)***	0.01 (0.62)	−0.19 (0.28)
Week 1 vs. 3m FU: Intervention	−15.79 (2.04)***	−0.47 (0.60)	0.43 (0.28)
Week 2 vs. Week 3: Control	−2.79 (2.01)	−0.22 (0.59)	−0.02 (0.27)
Week 2 vs. Week 3: Intervention	−3.14 (2.03)	0.29 (0.60)	−0.06 (0.28)
Week 2 vs. Week 4: Control	−5.07 (2.01)*	−0.22 (0.59)	−0.12 (0.27)
Week 2 vs. Week 4: Intervention	−4.63 (2.00)*	−0.001 (0.59)	0.28 (0.27)
Week 2 vs. 3m FU: Control	−10.07 (2.11)***	−0.09 (0.62)	−0.10 (0.29)
Week 2 vs. 3m FU: Intervention	−9.68 (2.07)***	−0.05 (0.61)	−0.12 (0.28)
Week 3 vs. Week 4: Control	−2.27 (1.98)	0.002 (0.59)	−0.10 (0.27)
Week 3 vs. Week 4: Intervention	−1.49 (2.01)	−0.29 (0.59)	0.34 (0.27)
Week 3 vs. 3m FU: Control	−7.28 (2.08)***	0.13 (0.62)	−0.08 (0.28)
Week 3 vs. 3m FU: Intervention	−6.55 (2.08)**	−0.34 (0.61)	−0.06 (0.28)
Week 4 vs. 3m FU: Control	−5.00 (2.09)*	0.13 (0.62)	−0.02 (0.28)
Week 4 vs. 3m FU: Intervention	−5.05 (2.05)*	−0.05 (0.61)	−0.40 (0.28)
Intervention vs. Control: Baseline	2.10 (3.23)	0.26 (0.64)	0.31 (0.32)
Intervention vs. Control: Week 0	−2.52 (3.24)	−0.70 (0.65)	−0.69 (0.32)*
Intervention vs. Control: Week 1	−8.88 (3.31)**	−0.22 (0.67)	−0.37 (0.33)
Intervention vs. Control: Week 2	−12.54 (3.36)***	−0.74 (0.69)	−0.59 (0.34)
Intervention vs. Control: Week 3	−12.89 (3.34)***	−0.23 (0.68)	−0.63 (0.34)
Intervention vs. Control: Week 4	−12.11 (3.33)***	−0.52 (0.68)	−0.19 (0.34)
Intervention vs. Control: 3m FU	−12.16 (3.44)***	−0.70 (0.72)	−0.60 (0.36)

Note. Unstandardized estimates (standard error) are represented. The reference group for the intervention condition is the control condition. The reference group for each of the time effects is baseline.

^***p < .001,

^**p < .01,

^*p < .05.

After accounting for the effects of participants’ treatment history, there was a significant main effect of time on PTSD symptom severity. PTSD symptom severity was significantly lower across all participants at intervention/week 0 and all subsequent time points compared to baseline. In general, although the week-to-week (e.g., Week 3 relative to Week 2) decreases in PTSD symptoms during the intervention period were not significant, with the exception of a significant decrease Week 2 relative to Week 1 for those who received the intervention, there were significant decreases in self-reported PTSD severity at the 3-month follow-up compared to all other timepoints for all participants. Moreover, there was a significant time x condition interaction on PTSD severity (Figure 3). The intervention condition baseline vs. 3-month follow-up was d = 3.21 (95% CI: 2.37, 4.05) and for control was d = 1.58 (95% CI: 0.95, 2.21). Post hoc pairwise comparisons revealed that women who received the intervention had significantly lower PTSD symptoms relative to those in the monitoring condition at Week 1 (after the in-person intervention session) and all subsequent assessments through 3-month follow-up. Those in intervention versus control did not differ on PTSD severity at baseline or intervention/week 0.

Figure 3. Self-reported PTSD Symptoms from Baseline to 3-Month Follow-up including Weekly Assessments.

Examining alcohol use outcomes across the weekly assessments through the 3-month follow-up, there were no significant effects of time, condition, or the time x condition interaction on average drinks per drinking day. However, a significant time x condition interaction emerged with regards to heavy episodic drinking frequency. For average drinks per drinking day, the intervention condition baseline vs. 3-month follow-up d = 0.73 (95% CI: 0.15, 1.30) and for control was d = 0.15 (95% CI: −0.40, 0.71). Individuals who received the intervention reported less frequent heavy episodic drinking episodes at intervention delivery/week 0 and all subsequent assessments compared to baseline though no other significant decreases following the initial decrease from baseline were observed. Individuals who were randomized to control had no significant changes in weekly self-reported heavy episodic drinking episodes. Additionally, those who received the intervention reported less frequent heavy episodic drinking episodes at intervention/week 0 compared to control, though the two conditions did not differ from each other at any other assessment point. For heavy episodic drinking frequency, the intervention condition baseline vs. 3-month follow-up effect size was d = 0.95 (95% CI: 0.36, 1.54) and for control was d = 0.26 (95% CI: −0.29, 0.82).

Discussion

Results of this small, pilot RCT are encouraging for a single session + four coaching call intervention for reduction of PTSD and alcohol misuse following sexual assault. PTSD and alcohol misuse decreased across both conditions over the 3-month follow-up which is consistent with natural recovery trajectories observed in epidemiological studies (Dworkin et al., 2023), but also notable given that this was a sample of individuals seeking treatment for assault related distress and impairment. For PTSD symptoms, statistically significant and clinically meaningful reductions were observed compared to symptom monitoring (i.e., natural recovery). Reductions in drinking were not statistically significantly different for BRITE compared to symptom monitoring. These findings show promise for an early intervention that has the potential to dramatically change PTSD trajectories following high risk events like sexual assault and does not appear to increase the risk of harmful alcohol use. Accessible, brief interventions are key to scalability and reaching a vulnerable group of women quickly following trauma exposure.

This pilot trial sought to examine feasibility, acceptability, and satisfaction with a newly designed intervention, although it is based on an established PTSD treatment (CPT; Resick et al., 2016). Of note, the intervention had a 100% completion rate; all individuals randomized to BRITE completed the in-person intervention. We also had high engagement in the coaching calls, with almost 80% completing all four coaching calls. Interventions for PTSD and alcohol misuse are characterized by high dropout, both non-initiation (attending no sessions after randomization) and non-completion (attending fewer sessions than intended) of treatment (Roberts et al., 2015). The observed high rates of both initiating and completing this brief intervention speaks to the accessibility of an approach using higher-dose initial contact with skills provided right away, followed by lower-dose support. In addition, participants rated the intervention as interesting, not distressing, and helpful. Positive perceptions predict engagement and outcomes in early interventions following sexual assault (Dworkin & Schumacher, 2018). Regarding feasibility, the study began recruiting widely across the community and university setting. However, the focus on college students was more successful and may be an indicator of the ideal target population and setting for the intervention. Dissemination and outreach in other settings should be explored in future studies.

Consistent with hypotheses, BRITE intervention led to consistent and robust decreases in PTSD compared to symptom monitoring. The PTSD field has several well-tested and highly efficacious treatments for chronic PTSD but lacks gold-standard early (i.e., within the first 3 months of trauma exposure) intervention options (Bisson et al., 2019). There are options, albeit relatively few, for early treatment of PTSD, with EMDR, CBT with a trauma focus, and cognitive therapy having research support (Roberts et al., 2019). However, indicated preventative interventions are at the early stages of establishing efficacy (Oosterbaan et al., 2019; Roberts & Nixon, 2023). BRITE may be an option as a low-dose, low-burden prevention strategy for PTSD. In particular, it may be a good fit in clinics serving college students (e.g., campus health clinics) where providers have limited number of sessions to provide services to students, although given the small sample this trial requires replication.

Contrary to expectations, BRITE was not superior to symptom monitoring of alcohol misuse: both conditions improved on this outcome. Monitoring alcohol can be an effective intervention (Ball et al., 2007), even among those with chronic PTSD and alcohol misuse (Simpson et al., 2022). In addition, individuals may change their drinking when they seek treatment prior to an intervention beginning (Stasiewicz et al., 2019), which was consistent with weekly changes in our data.

Another possible explanation for the differential findings of the intervention for PTSD vs drinking relates to self-medication. BRITE directly addresses the functional relationship where alcohol use increases as an attempt to control assault-related distress (Hawn et al., 2020). However, it is unknown how long these relationships take to be established. Our follow-up period may have been too short to see associated changes. In addition, young adults, including those with sexual assault histories, drink for many potential reasons, not solely to reduce negative affect or decrease PTSD symptoms (Huh et al., 2015). Anecdotally, clinicians heard BRITE patients make statements that their motives for drinking had changed (e.g., “My anxiety is less, so I am attending more parties and drinking with my friends, but I am no longer drinking alone to cope”). BRITE targets coping-related alcohol use specifically, but drinking for other motives (e.g., enhancement motives) may not have been affected by the intervention.

Limitations include the pilot nature of the trial, although even with a relatively small sample, we observed clinically meaningful reductions in PTSD. Recruitment was challenging, and we had the greatest success enrolling college students specifically. Still, we did not meet our anticipated sample size, partly due to COVID-19 disruptions at the end of the funding period, and the study is underpowered. The focus on female-identifying individuals is consistent with higher rates of sexual assault among women (Mellins et al., 2017), but future studies should considering expanding to individuals who identify as other genders. Another limitation is a relatively weak comparison condition (Leon & Davis, 2009). Future directions should include a larger RCT with a credible control to act as a valid therapy placebo to better understand the efficacy of the BRITE intervention. In addition, adaptations of BRITE for delivery via telehealth, web platforms, or an app could help increase accessibility and scalability. The definition of sexual assault used in this study was broad and included all types of unwanted sexual contact from unwanted touching to completed rape. This is helpful regarding generalizability but may mean we included participants with less severe events. Of note, all participants were treatment-seeking and reporting significant symptoms and impairment. Future tests of the BRITE intervention should also include factors such as drinking motivations (i.e., coping, enhancement) to understand mechanisms of change in the intervention. Lastly, the follow-up period for this study was brief and possibly not sufficient to see sustained effects for PTSD prevention and/or delayed effects on drinking.

The combination of PTSD and alcohol misuse is associated with high personal and public health costs which are exacerbated when sexual assault survivors wait months to years before starting treatment. We have effective interventions for PTSD, but when PTSD is accompanied by alcohol misuse, treatments are less effective with higher drop out (Bedard-Gilligan et al., 2018). There is a strong need for feasible, effective interventions in the acute aftermath of trauma. BRITE, as a brief intervention, with strong effects in reducing PTSD, and with high treatment engagement, demonstrates promise as a preventative intervention.

Highlights.

• One session + four coaching calls reduces PTSD when delivered acutely after sexual assault

• Effects were smaller for alcohol misuse, though reductions still occurred

• College students may be a particularly appropriate population for this intervention