This clinical trial of 80 German inpatients with PTSD (Woud et al., 2021 – Psychotherapy and Psychosomatics), tested whether a computerized cognitive bias modification for appraisals (CBM-APP) can modify dysfunctional appraisals and reduce such appraisals of trauma reactions and PTSD symptoms. The CBM-APP schedule was 8 sessions over a 2-week period completed as an adjuvant to a multimodal inpatient treatment program. Relative to patients receiving sham training, patients receiving CBM-APP showed greater reductions in dysfunctional appraisals (d = 1.30), posttraumatic cognitions (d = 0.85), and PTSD symptoms (d = 0.68). Importantly, reductions in dysfunctional appraisals correlated with reductions in PTSD symptoms suggesting that proximal modification of appraisal biases may transfer to downstream reductions in symptoms. These findings could open new avenues for improving present therapeutic approaches to PTSD and are of particular importance given the limited effectiveness of current first line treatments for this disorder.
Woud, M. L., Blackwell, S. E., Shkreli, L., Würtz, F., Cwik, J. C., Margraf, J., … & Kessler, H. (2021). The effects of modifying dysfunctional appraisals in posttraumatic stress disorder using a form of cognitive bias modification: Results of a randomized controlled trial in an inpatient setting. Psychotherapy and Psychosomatics, 90:386–402.
Commentary by Courtney Beard
McLean Hospital and Harvard University – USA
Woud et al. (2021) present a pilot RCT comparing a trauma appraisal CBM to a control as an augmentation to an 8-week inpatient PTSD program. CBM successfully shifted appraisals, and change in appraisal mediated PTSD symptom improvement. In this brief commentary, I highlight notable aspects of the study, as well as considerations for future work in psychiatric settings.
The authors should be commended for such a well-designed and reported clinical trial. This study provides an excellent example for future CBM clinical trials. For example:
- The authors included many open science practices, including prospectively registering the trial, noting any deviations and rationale for those changes, and providing anonymized outcome data and analysis scripts. These practices should improve the replicability of CBM studies.
- The authors presented intention-to-treat analyses. Some CBM researchers from an experimental psychopathology background may be unfamiliar with the “once randomized, always analyzed” standard in RCT design. Intention-to-treat analyses are critical for avoiding overoptimistic estimates of CBM’s effects. Drop out and low adherence to a protocol are likely to occur even more in real-world clinics compared to a research study; thus, including all participants randomized provides a more realistic estimate of the intervention’s actual effect.
- Advances in the CBM field, and particularly attention bias modification (ABM), have been stymied by poor unreliability of assessments. I was delighted to see reporting of reliability estimates for all study measures, especially the cognitive bias assessments. As a frequent reviewer of CBM manuscripts, my most common critique is that authors say nothing about the psychometric properties of their primary cognitive bias outcome.
- I was also struck by the control condition. Having been involved in some of the first clinical trials for ABM and Interpretation Bias Modification, I understand the difficulty in designing an appropriate CBM control condition. The peripheral visual task control controlled for time, expectancy, completing a computer task, positive feedback about task performance, focusing on something (e.g., time away from intrusive thoughts), and attention from the staff. Importantly, it did not expose participants to the same type of stimuli as the appraisal training. This type of control is superior to a sham 50/50 training or neutral training because these are both diluted versions of CBM. The authors attempted to create a credible control and took care to present the study as training concentration (not appraisals). Despite these efforts, the CBM group reported higher credibility of their assigned intervention, underscoring the need to always assess this.
Considerations for psychiatric settings
- Unlike most CBM, the appraisal training became gradually more positive over the 8 sessions. This approach was suggested by patients in my hospital program as well. Typical benign interpretations may be perceived as unrealistic and laughable during acute crises and severe symptoms.
- Patients with suicidal ideation or substance use in past 6 months were excluded, although no rationale was provided. In acute psychiatric settings, excluding people with suicidal ideation would typically exclude most patients and those most in need of augmentation interventions that might accelerate improvement. It will be important for future work to include all patients and determine whether this intervention is effective for those with and without suicidal ideation.
- CBM is often assumed to be benign, but this may not be the case, especially in psychiatric settings. The authors systematically assessed for adverse effects, and a few participants reported worsening of symptoms due to the CBM triggering traumatic memories. It is crucial to present this potential risk to patients and clinic staff, as well as contextualize it as an expected risk of most PTSD treatments.
- Participants’ acceptability ratings hovered in the middle of a 9-point Likert scale. I encourage the authors and others to include stakeholders (patients, clinic staff) throughout the treatment development process using methods such as qualitative interviews or focus groups. The patient and clinician perspectives are crucial, and treatments will ultimately fail if they are ignored.
In sum, the results of this pilot RCT are quite encouraging, as it is no small feat to demonstrate improvement above and beyond such a powerful treatment as usual. I look forward to seeing the next step of the treatment development process and more options for people with PTSD.