This randomized controlled dismantling trial has independent fidelity rating as the primary outcome and patient outcomes as the secondary outcomes. Clinicians are randomly assigned to six months of post-workshop support following an intensive two-day CPT workshop delivered by the last author. The conditions are: standard group tele-consultation, technology-enhanced group tele-consultation, or no consultation/fidelity monitoring only. The post-workshop consultation conditions vary as to whether audio-recorded therapy sessions and other work samples, such as CPT worksheets, are reviewed in the group. This design permits the evaluation of the additive value of each of the elements (i.e., expert consultation, technology-assisted expert consultation). Follow-up assessments of fidelity will be conducted at six months post-workshop for clinicians, and follow-up symptom assessments will be conducted at three months post-treatment for patients.
Participation in the study requires that all clinicians consent at least two clients to have all of their CPT sessions audio-recorded for use in consultation and/or fidelity rating. Four therapy sessions will be randomly selected across the six-month period for fidelity rating, and fidelity to treatment will be assessed by trained graduate students and post-doctoral fidelity assessors. Consistent with recommended clinical practice and research trials, patient outcomes will be assessed pre-treatment, at alternate therapy sessions, and at three months post-treatment. A secure web-based application is used to manage the continuous uploading of encrypted audio-recorded therapy sessions to the study server and to protect the privacy and confidentiality of patients and clinicians. Study-provided digital audio-recorders (USB-equipped) are used to record, store and upload CPT sessions to the secure server for fidelity ratings. Study staff retrieve the randomly-selected audio-recordings and assign them to the CPT fidelity raters, who are blinded to study condition. To check if rater blinding has been achieved, fidelity raters will be asked to guess the study condition for each session they rated. Patient self-report data is collected by the clinicians, who then upload the data directly to the server and clinician CPT activity outside of study-related procedures is also recorded.
Participants in all conditions attend the workshop, receive the CPT manual and related materials, and have access to resources available through the free CPT-web online training (https://cpt.musc.edu/index). All participants who upload the required number of session recordings and patient symptom measures will be eligible to receive ‘CPT Provider’ status. This status is not tied to any change in remuneration for services. Provider status will be awarded to clinicians who demonstrate an acceptable level of fidelity to CPT based on session recordings randomly selected from the last two weeks of the six-month post-workshop period. Workshop participants who provide informed consent are randomized to one of the three post-workshop consultation conditions after completing the workshop. Participants in the two tele-consultation conditions receive six months of weekly one-hour group tele-consultation with a CPT expert. Each call includes discussions of the CPT protocol, challenging cases, treatment obstacles, and specific issues raised by participants within each group. Both tele-consultation conditions include a maximum of six to eight clinicians per call. CPT experts provide consultation across both consultation conditions to control for any expert-related effects. The prescribed and proscribed activities for each consultation condition are described in manuals, and study team members (SWS, CMM, ML) review consultant calls on a monthly basis to ensure that fidelity to the consultation condition is maintained. Consultants are trained by study team members prior to beginning consultation calls and receive consultation and feedback every four to six weeks or as needed for the first two waves of consultation, and as needed during the third consultation wave.
Inclusion and exclusion criteria
Mental health clinicians from the Operational Stress Injury clinics, Canadian Forces mental health services, and the broader Canadian community are eligible to participate in the study if they: attend a CPT workshop; are licensed as a mental health clinician and are able to practice independently; currently provide psychotherapy to patients with PTSD; consent to be randomized to one of the study conditions; and are willing to solicit patient participation. Eligible patient participants are: diagnosed with PTSD; and consent to have their sessions audio-recorded and listened to by a fidelity rater and potentially other clinicians teaching and learning CPT. Patients are permitted to continue other psychotherapeutic interventions if they are not specifically focused on treating PTSD symptoms. Ineligible patients include those not eligible for CPT based on the state of research evidence, including those with: current uncontrolled psychotic or bipolar disorder; substance dependence; imminent suicidality or homicidality that requires acute care; and significant cognitive impairment (although mild to moderate traumatic brain injury is permitted).
Outcome measures and assessment procedures
Fidelity ratings and assessment of CPT modification from audio-recorded CPT sessions
This study uses a modified version of the CPT fidelity measure  that has been used in previous clinical trials . The CPT fidelity measure examines clinicians’ adherence to, and competence in, delivering specific CPT interventions prescribed in each session. Clinicians are rated on their adherence and competence in delivering these elements (rated on a 7-point, Likert-type scale). Two studies have assessed the reliability of the measure and found 97% agreement between two raters across all items for adherence in one study (agreement was not reported for competence scores; ) and 100% agreement for adherence and overall competence ratings in another . Pilot research has been conducted to assess the predictive validity of the unique and essential scales for a small sample of veterans. CPT competence was highly correlated across all sessions with competence in cognitive therapy as measured by a validated cognitive therapy fidelity-assessment instrument, r = 0.82, p = 0.001 [44, 57]. Therapist competence in the first session of CPT was positively correlated with self-reported symptom improvement among patients from baseline to post-treatment, r = 0.79, p = 0.02. Additionally, we use a framework of modifications to identify any deviations from or adaptations to CPT that occur in the sessions . At least 10% of the sessions will be rated by all raters to independently assess inter-rater agreement.
Patient outcome measures
Patients are asked to complete measures related to their PTSD symptoms, functional impairment and overall quality of life at baseline (pre-treatment), at every alternate CPT treatment session, and at post-treatment follow-up (three months post-treatment). The assessment measures include: ‘The Posttraumatic Stress Disorder Checklist’ (PCL-S; ), a 17-item questionnaire that reliably measures the severity of distress related to PTSD symptoms (Chronbach’s alpha = 0.94). Each item is measured on a 5-point Likert scale, which maps on to the DSM-IV criteria for PTSD; ‘The Outcome Questionnaire – 45’ (OQ- 45; ), a 45-item self-report measure of three dimensions of client functioning in the past week: symptom distress (e.g., anxiety, depression), interpersonal problems, and social role adjustment (Chronbach’s alpha = 0.71 to 0.93). The OQ-45 has established clinical cut-offs and a reliable change indices to signal meaningful improvements or deterioration; and ‘The Short-Form Quality of Life Health Survey’ (SF-12; ), a 12-item survey that measures mental components (vitality, social functioning, role-emotional, and mental health) and physical components (physical functioning, role-physical, bodily pain, and general health) of health related to quality of life. All questionnaires are available in English and French, and take clients around 15 to 20 minutes to complete. Therapists upload summary scores to a secure website at the same time that they upload their audio-recordings.
Clinician demographic characteristics and experience
A pre-workshop questionnaire is administered to assess relevant clinician demographic information (i.e., age, gender, education), clinician experience (years of licensed practice, hours of formal training in CBT, hours of supervised post-graduate CBT training), prior CPT training experience (workshop hours and supervision hours), experience treating clients with PTSD (number of clients), and volume of current clinical activity (caseload and hours devoted to psychotherapy).
Clinician monthly CPT activity reporting
During the six-month active study period, all clinicians complete a monthly report via a web-based survey on CPT activity. Questions include: the number of CPT sessions provided over the past month (irrespective of patients’ enrollment in the study); cumulative number of new CPT clients seen; frequency, duration, and satisfaction with consultation over the past month (as applicable); and a rating of clinician confidence in their ability to adhere to adherence to the CPT protocol.
Quantitative and qualitative measurement of potential barriers and facilitators to CPT skills uptake and implementation outcomes
The Organizational Readiness to Change Questionnaire  is a pre-workshop self-report questionnaire that assesses various elements of a clinician’s perceived support for training implementation. A total of 18 scales (Chronbach alphas = 0.56 to 0.92) are grouped within four dimensions of readiness (i.e., motivation for change, adequacy of resources, staff attributes, organizational climate). The Dimensions of Organizational Readiness-Revised (DOOR-R; ) scale assesses perspectives on intra- and extra-organizational implementation readiness. The Evidence-Based Practice Attitudes-50 (EBPAS-50; ) is a 50-item, self-report measure of clinician attitudes toward EBPs, consisting of 12 factors, with moderate to large factor loadings and fair to excellent internal consistency reliabilities.
We also interview clinicians at baseline, post-consultation, and three months post-consultation to further assess for barriers and facilitators to the use of CPT in their practice. The interviews query effective tactics and methods that may have been used outside of the consultation methods (i.e., positive deviance; ) that the clinicians deemed helpful in acquiring and maintaining CPT skills. We based our interview questions on those used to assess influences on training and implementation in previous research , which are based in part on the Consolidated Framework for Implementation Research (CFIR ).
Sample size and justification for power calculations
In conducting power calculations, we considered assessment attrition, within-clinician correlation, and within-setting effects. Based on a power curve calculation with the following assumptions: 30 clinicians per condition; 4 sessions rated for fidelity per clinician, randomly selected over time; with an alpha-level of 0.05, within provider correlation estimated at 0.38, 10% proportion of variance accounted for by the random effect associated with agency, based on a two-sided test, the detectable effect size for at least 80% power will be medium to large, per Cohen’s  classifications, ranging from d = 0.41 to 0.55.
Recruitment and randomization
We recruited OSI clinicians directly. Recruitment took place at CPT workshops in Canada from 2012 to 2013. Three months before the first workshop, the OSINN invited clinicians and clinic coordinators from the OSI clinic network, Canadian Forces health services, and approximately 800 community-based clinicians who provide services to veterans with PTSD. The clinicians include psychologists, social workers, psychiatric nurses, and psychiatrists. This invitation was accompanied by a letter soliciting participation in the study. Prior to enrolling clinicians, ethics approval for the study was obtained from the clinic or institution where the clinician worked. Community clinicians were informed that the CHIR-supported workshop was free, and 24 clinicians received up to $500 reimbursing travel and accommodation. Partial reimbursement for the community clinicians was intended to offset the barrier of cost associated with a two-day workshop. Following the initial CIHR-funded workshop, two provincial health authorities and one hospital hosted workshops. One of the authors (CMM) was the primary CPT trainer. Interested clinicians consented prior to, or within the week following the workshop. Consent forms outlined the purpose and requirements of the study, as well as potential risks. A random allotment of participants was generated by the study statistician for each condition. To avoid any biases in condition assignment, condition assignment was communicated by participant number to the study coordinator after participant numbers had been assigned to the consented participants. Clinicians received a notification e-mail from the study coordinator within one week after the workshop regarding their randomized condition. The e-mail also included consultant contact information and a schedule of consultation sessions (if applicable to their condition). As in other studies of this nature (e.g., ), the clinicians were asked to provide an information sheet about the study and obtain verbal consent from patients who agreed to participate in the study, then document patient consent in his/her medical records. Upon returning to their respective work places, participants began audio-recording CPT sessions with consented patients.
Proposed analytic strategy
Trial reporting will follow the CONSORT 2010 statement for reporting randomized controlled trials . Analyses will be preceded by careful examination of the descriptive statistics and distributional form of all variables. The effectiveness of randomization will be evaluated by comparing the different conditions on baseline measures, demographics, and key outcome variables using Chi-squared tests and t-tests. The primary analyses will be conducted using random effects regression modeling. These models provide more flexibility in modeling missing data than general linear models and are appropriate for nested data. ‘Intention-to-train’ principles will be used in the analyses of clinicians who are assigned to a consultation modality but either fail to engage or are unable to fulfill the post-workshop participation requirements (e.g., do not attend consultation sessions). Analyses of both the intention-to-train and completer samples will be conducted.
To explore the contribution of organization and clinician-level factors to the outcomes of interest, in addition to random effects regression modeling, we will conduct qualitative analyses on interview data. Qualitative and quantitative data will be integrated to facilitate a fine-grained understanding of processes and characteristics that influence training outcomes. Interviews will be analyzed in line with goals, objectives and key research questions. Using content analysis, we will identify analytical categories to describe and explain our observations. Our work will occur in five stages outlined in Pope, Ziebland, and Mays’ framework approach to qualitative analysis : Familiarization; Identifying a Thematic Framework; Indexing; Charting; and Mapping and Interpretation. In the second stage, codes will be derived deductively by identifying categories at the beginning of the research (e.g., elements of the CFIR), and inductively by identifying those that emerge gradually from the data. In subsequent phases, survey data on contextual factors will be examined in conjunction with fidelity scores for the purposes of validation, hypothesis generation, and expansion [71, 72] to understand ways in which contextual factors influence fidelity to CPT.
To date, six workshops have been completed, and, through efforts to enroll clinicians over three waves of recruitment, the recruitment goal of 90 clinicians was exceeded. The first and second waves of consultation have been completed, and the third wave is underway. Pre- and post-training interviews with providers from waves one and two are complete. Thus far, our retention rate and compliance with study procedures (e.g., uploading of sessions) has been high, and a high proportion of clinicians uploaded recordings at baseline and over the course of the consultation period. We anticipate that data collection will be complete in early 2014.