Self-compassion in Acceptance and Commitment Therapy for chronic pain: a pilot study

The COMP.ACT program

The development of the COMP.ACT program (Table 1) was based on the Psychological Flexibility Model (PFM) applied to CP [21, 22], with elements of compassion-based approaches [7, 23].

The COMP.ACT program has eight weekly sessions of 2 h and is delivered according to a standardized and previously written workbook (with sessions’ goals, therapist guidance, instructions to meditations, metaphors and overall exercises, guiding bullet-points for post-exercises inquiry, summary and indication of between sessions assignments and practices). Participants were provided with the following materials: complementary texts, handouts with activities worksheets, and recorded audio exercises (1. Mindful breathing, 2. Body-scan, 3. Soothing Rhythm Breathing, 4. Loving-kindness, and 5. Soften, Soothe and Allow).

Participants

The current study is part of a larger one that focused on the relationship between self-compassion, pain, psychopathology, and psychological flexibility processes in CP. Specifically, the sample in the present clinical study (n=16) partly overlaps with the sample (n=49) of a published cross-sectional study [24] (which is composed of the baseline scores of the 16 participants who are part of the current study, as well as participants who were not able or interested to participate in the current clinical study, but who agreed to participate in the cross-sectional study).

Recruitment occurred in the pain consultation of the anesthesiology service of Coimbra University Hospital (CHUC). Inclusion: a) non-malignant CP; b) female (the healthcare service has disproportionately more women with CP; this criterion was established a priori for attaining a homogeneous sample in terms of gender); c) adults (age 18–65); d) available to attend sessions. Exclusion: a) psychosis; b) severe depression; c) substance abuse; d) non-suicidal self-injury; e) suicidal ideation; f) attending another psychological intervention.

Participants had a mean age of 50.69 (SD=8.50). The majority were married (n=12; 75%), and some were single (n=2; 12.5%), divorced (n=1; 6.3%) and widowed (n=1; 6.3%). In terms of education, participants had 4th grade (n=3; 18.8%), 6th grade (n=3; 18.8%), 9th grade (n=4; 25%), high school (n=3; 18.8%), bachelors (n=2; 12.5%), and masters (n=1; 6.3%). In regards to duration of CP, participants reported having pain for less than 1 year (n=4; 25%), between 1 and 5 years (n=8; 50%), between 5 and 10 years (n=3; 18.8%), and for more than 10 years (n=1; 6.3%). CP diagnoses included fibromyalgia (n=12; 75%), osteoarthritis (n=1; 6.3%), rheumatoid arthritis (n=1; 6.3%), low back (n=2; 12.5%), neck pain (n=2; 12.5%), and other (n=6; 37.5%). Ten participants reported 1 CP diagnosis (62.5%), four reported having 2 CP diagnoses (25%), and two reported having 3 CP diagnoses (12.5%). Seven participants had other chronic illnesses (43.8%).

No significant differences were found between conditions in terms of age (t₍₁₄₎=1.216; p=0.244), marital status (χ²_{(3, n=16)}=2.12, p=0.549), education (χ²_{(5, n=16)}=2.79, p=0.732), number of CP diagnoses (t₍₁₄₎=−1.038; p=0.317), duration of CP (χ²_{(3, n=16)}=5.67, p=0.129), and presence of other chronic illnesses (χ²_{(1, n=16)}=0.91, p=0.341).

Procedure

This study is registered as a trial (NCT04210466) at ClinicalTrials.gov. Participants were informed about the study by their physician or nurse that participation was voluntary, and that personal information would be accessed exclusively by the research team. Those who provided informed consent were assessed for exclusion criteria (see Figure 1). Exclusion criteria were assessed by a clinical psychologist through a semi-structured 30-min clinical interview.

Figure 1:

Diagram of participants

Conditions differed in sessions 3 and 4, in which the COMP.ACT condition included two compassion-based sessions (see Table 1), and in the ACT-only condition these sessions consisted of questions and answers regarding previous sessions, willingness, and mindful meditation practice. The intervention was conducted in a co-therapy setting by two clinical psychologists with training in ACT and compassion approaches, and previous experience in delivering acceptance, mindfulness and compassion-based interventions.

No significant differences were found between participants who were lost to post-intervention assessment and participants who completed the program in terms of age (t₍₂₅₎=−0.98; p=0.338), marital status (χ²_{(3, n=27)}=2.70, p=0.440), education (χ²_{(5, n=27)}=4.05, p=0.541), number of CP diagnoses (t_(10.00)=1.19; p=0.262), duration of CP (χ²_{(3, n=24)}=1.68, p=0.641), and presence of other chronic illnesses (χ²_{(1, n=24)}=0.18, p=0.673).

It should be noted that the current study design present the following deviations from the registered clinical trial (NCT04210466): 1) the sample size is n=16, instead of the aimed n=20. The reason for this discrepancy relates to the dropout rate of participants throughout the study (see Figure 1). Also, this study was part of the PhD studies of the first author, who, due to time constraints and lack of human resources, was not able to continue recruitment; 2) as a result from these time constraints, the research team was unfortunately not able to conduct the 6-month follow-up assessment as was previously intended; 3) additionally, we added one more exclusion criterion that was not previously thought out: participants already attending psychological programs/care were not eligible, as this would confound results on the impact of COMP.ACT; 4) also, the study previously aimed to assess exclusively pain, psychopathological symptomatology, and quality of life. However, it has been pointed out that assessing the impact of contextual-behavioral interventions (e.g., ACT) should consider not exclusively symptom reduction, but more importantly changes in functioning and in the psychological processes targeted by these interventions [25]. In order to add measures of relevant processes (e.g., mindfulness, self-compassion, valued-living), and reduce the burden of an already extensive assessment battery, we ended up excluding the quality of life measure from the protocol.

Measures

In addition to socio-demographic and medical information, the protocol included self-report measures already validated for the Portuguese population. See Cronbach α for each scale in Table 2.

Numerical Pain Rating Scale (NPRS) [26] is a measure of pain intensity. Participants are asked to rate the average pain intensity in the last 24 h on an 11-point scale (0 = “No pain”; 10 = “Worst imaginable pain”). A pain intensity single score was created from ratings of: 1) current pain; 2) highest pain in last 24 h; 3) lowest pain in last 24 h. Higher scores indicate greater pain intensity.

Pain Disability Index (PDI) [27], is an 11-point scale (0 = “No disability”; 10 = “Worst disability”) that measures pain disability in seven daily life domains. Higher scores indicate higher pain disability.

Depression, Anxiety and Stress Scale-21 (DASS-21) [28] is a measure of depression, anxiety and stress symptoms with 21 items rated on a 4-point Likert scale (0 = “Did not apply to me at all”; 3 = “Applied to me very much or most of the time”).

Five Facet Mindfulness Questionnaire (FFMQ) [29] is a self-reported measure of mindfulness composed of 39-items. These are divided in five facets of mindfulness: Observing, Describing, Acting with awareness, Nonjudgmental, and Non-reactive, which can be computed in a total Mindfulness score. Each item is scored on a 5-point Likert scale (1 = “Never or very rarely true”; 5 = “Very often or always true”). Higher scores indicate higher levels of mindfulness.

Valuing Questionnaire (VQ) [30] is a 10-items measure of valued living. The questionnaire has a two-factor structure: 1) VQ-Obstruction (obstacles to valued living), and 2) VQ-Progress (progress in valued living). Higher scores in VQ-Obstruction indicate higher levels of experiencing obstructions to valued living, and higher scores in VQ-Progress indicate higher levels of experiencing progress in living according to values.

Self-Compassion Scale (SCS) [6] is a 26-items scale that measures self-compassion in two domains: compassionate self-responding (self-kindness, common humanity and mindfulness) and uncompassionate self-responding (self-judgment, isolation and over-identification). Items are scored in a 5-point Likert scale (1 = “Almost never”; 5 = “Almost always”). There is an ongoing debate around the factor structure of the SCS [31, 32]. In this study, we will follow the two-factor approach [31].

Data analysis

Preliminary analyses (descriptive analyses, t-test differences, and chi-square differences) were conducted in SPSS version 23 (IBM Corp., Armonk, N.Y., USA). Reliable and significant change examined the clinical meaningfulness of change (pre-intervention vs. post-intervention) according to Jacobson’s clinical significance analysis [33]. Reliable change index (RCI) tests whether individual change falls outside the range that could be due to measurement error. Reliable change occurs if it exceeds 1.96 times the standard error (see [34] for an in-depth description). Change is considered clinically significant (CSC) when a person’s score moves from the “dysfunctional population” to the “functional population” [35]. A cut-off is calculated considering the mean and standard deviation of the sample at pre-treatment, as well as from the comparison “functional” population. A person is considered “recovered” when the magnitude of change is above the level of the RCI and the post-treatment score is beyond the CSC cut-off (see [36]).