Received: 4 August 2020

| Revised: 29 December 2020

| Accepted: 16 January 2021

DOI: 10.1002/ejp.1734

ORIGINAL ARTICLE

Mindful Self-­Compassion program for chronic pain patients:

A randomized controlled trial

Marta Torrijos-­Zarcero1,2 | Roberto Mediavilla2 | Beatriz Rodríguez-­Vega1,2,3 |

María Del Río-­Diéguez3,* | Inés López-­Álvarez1,2 | Cristina Rocamora-­González1,2 |
Ángela Palao-­Tarrero1,2,3,*

1
Psychiatry, Clinical Psychology and
Mental Health Department, La Paz
Abstract
University Hospital, Madrid, Spain Background: Although evidence-­based psychological treatments for chronic pain
2
La Paz Hospital Institute for Health (CP) have been demonstrated to be effective for a variety of outcomes, modest ef-
Research (IdiPAZ), Madrid, Spain
fects observed in recent reviews indicate scope for improvement. Self-­compassion
3
Autonomous University of Madrid (UAM),
promotes a proactive attitude towards self-­care and actively seeking relief from suf-
Madrid, Spain
*APT and MDRD are certified teachers of
fering. Consequently, more compassionate people experience better physical, psy-
MSC program by UC San Diego. chological and interpersonal well-­being.
Methods: We conducted a single-­blind, randomized, controlled trial to examine the ef-
Correspondence
fects of a Mindful Self-­Compassion program (MSC) on relevant clinical outcomes in
Ángela Palao-­Tarrero, Department of
Psychiatry, Clinical Psychology and Mental patients with CP. Patients were randomly assigned to one of the two intervention arms:
Health, La Paz University Hospital, Paseo MSC or cognitive-­behavioural therapy (CBT). The protocols of both intervention arms
de la Castellana, 261, 28046 Madrid, Spain.
were standardized and consisted of a 150-­min session once a week during 8 weeks format-
Email: [email protected]
ted to groups of no more than 20 participants. The primary outcome was self-­compassion,
measured with the Self-­Compassion Scale (SCS). The secondary outcomes were other
pain-­related scores, quality-­of-­life measures, and anxiety and depression scores.
Results: In all, 62 and 61 patients were assigned to the MSC and CBT groups, respec-
tively. The MSC intervention was more effective than CBT for self-­compassion (average
treatment effect [ATE] = 0.126, p < 0.05). The secondary outcomes, pain acceptance
(ATE = 5.214, p < 0.01), pain interference (ATE = −0.393, p < 0.05), catastrophizing
(ATE = −2.139, p < 0.10) and anxiety (ATE = −0.902, p < 0.05), were also favoured in
the experimental arm (MSC). No serious adverse events were observed.
Conclusions: Mindful Self-­Compassion is an appropriate therapeutic approach for
CP patients and may result in greater benefits on self-­compassion and emotional
well-­being than CBT.
Significance: This randomized controlled trial compares the novel intervention
(MSC program) with the gold standard psychological intervention for CP (CBT).
MSC improves the levels of self-­compassion, a therapeutic target that is receiving
attention since the last two decades, and it also improves anxiety symptoms, pain
interference and pain acceptance more than what CBT does. These results provide
empirical support to guide clinical work towards the promotion of self-­compassion
in psychotherapeutic interventions for people with CP.

Eur J Pain. 2021;00:1–15. wileyonlinelibrary.com/journal/ejp © 2021 European Pain Federation -­EFIC® | 1

2
|    TORRIJOS-­ZARCERO et al.

1 | IN TRO D U C T ION helplessness) and aid in accepting pain-­related disabilities

(Smith & Osborn, 2007).
Evidence-­ based psychological theories and treatments to Compassion-­based interventions for CP promote positive
manage chronic pain (CP) have shifted from ‘first-­wave’ emotional outcomes (Montero-­Marín et al., 2018). A recent
behavioural approaches (Fordyce, 1976) to ‘second-­ study suggested that when emotional functioning is an im-
wave’ cognitive-­ behavioural approaches (McCracken & portant outcome besides daily functioning in CP, it may be
Turk, 2002) and then to ‘third-­wave’ –­mindfulness, com- beneficial to add self-­compassion, alone or as a component
passion (Gooding et al., 2020; Kabat-­ Zinn et al., 1986) in other therapies (Davey et al., 2020).
and contextual-­behavioural approaches (Hayes et al., 2006; However, compassion-­ based interventions have rarely
McCraken, 2005). been described well and barely been standardized; random-
Regardless of the wave, psychological interventions for ized controlled trials (RCTs) on these interventions are scarce
CP primarily target improvements in physical, emotional, so- (Kirby et al., 2017).
cial and occupational functioning rather than the resolution of The aim of the present RCT was to compare the effective-
pain (Sturgeon, 2014). Cognitive-­behavioural therapy (CBT), ness of the MSC (Germer & Neff, 2019) and CBT (Kovacs
which is considered the ‘gold-­standard’ psychological treat- & Moix, 2011; McCracken & Turk, 2002) programmes on
ment for CP (Häuser et al., 2010), tries to do so by reducing the basis of the primary outcome (self-­compassion) and sec-
distressing psychological symptoms, targeting maladaptive ondary outcomes (pain acceptance, pain interference, pain
behavioural and cognitive responses to pain, and address- intensity, catastrophizing, anxiety and depressive symptoms,
ing social contingencies that modify reactions to pain (Day and quality of life [QoL]) in a group of adult patients with CP.
et al., 2012). Third-­wave interventions, including acceptance-­
based and mindfulness-­based interventions (MBIs), focus on
promoting behaviours guided by important life values in- 2 | M ETHODS
stead of mitigating pain. They foster acceptance and change
the relationship between the person and his experiences. 2.1 | Design and participants
According to this framework, this relationship sustains psy-
chological distress more than the symptoms themselves do The main objective of this study was to compare the effec-
(Hayes et al., 2006). tiveness of the MSC program and CBT on the basis of the
Although all these approaches, especially the ones of self-­compassion outcome in a group of patients with CP.
the second and third waves, improved CP outcomes, such To that end, a parallel group, single-­blind (evaluator), ran-
as pain severity, disability and mood disturbance, the im- domized (1:1 ratio), controlled (vs. active comparator) trial
provements were only moderate. These modest effects, was implemented. We considered two intervention arms (the
ranging from small to medium in size, as observed in recent MSC course and CBT) and two assessment points (baseline
meta-­analyses, indicate scope for improvement (Harrison and post-­intervention).
et al., 2017; Hilton et al., 2017; Veehof et al., 2011; An active control was selected because recent sys-
Williams et al., 2012). tematic reviews have mentioned the lack of studies with
Mindfulness-­based interventions have evolved from the ‘head-­to-­head’ comparisons between MBIs and cognitive-­
Mindfulness-­Based Stress Reduction (MBSR) program de- behavioural therapies (Khoo et al., 2019); and this is an
veloped for people with chronic conditions, including CP important knowledge gap to be addressed. Moreover, inves-
(Kabat-­ Zinn, 1982), to more specific programmes based tigators working in the field of compassion have claimed
on mindfulness for CP, such as Mindfulness-­ Based Pain that compassion-­ based interventions and investigations
Management (Cusens et al., 2009). Furthermore, specific in the area are still in their infancy with only small-­scale
programmes that highlight the importance of the core com- RCTs (often with a non-­ active comparator) being per-
ponents of mindfulness, like the Mindful Self-­Compassion formed. Therefore, they recommend conducting RCTs that
(MSC) program, have also emerged (Germer & Neff, 2019). have adequately powered sample sizes and controls that are
Compassion is defined as ‘a sensitivity to the suffering not waitlists or treatments as usual, but active comparisons,
of self and others, with a deep commitment to alleviate it’ such as between MBI, ACT or CBT (Kirby et al., 2017).
(Neff, 2003a). Particularly, self-­ compassion promotes a CBT was chosen as the active control because it is a well-­
proactive attitude towards self-­care and seeking relief from established and prevalent psychological intervention for
suffering (Neff, 2003a). Consequently, more compassion- CP in the field (Khoo et al., 2019). When conducting such
ate people demonstrate better physical (Brion et al., 2014), active comparisons, it is important to clearly describe the
psychological (MacBeth & Gumley, 2012) and interper- protocols of the interventions to determine the differences
sonal well-­being. Self-­compassion might help confront the between the interventions and measurements that exam-
fear of pain, buffer difficult emotions (e.g. rage, shame and ine the process changes unique to the intervention. In this
TORRIJOS-­ZARCERO et al.    
| 3

regard, the MSC program, among all the compassion-­based on Methods, Measurement, and Pain Assessment in Clinical
interventions, is firmly standardized and easy to compare Trials (IMMPACT; Dworkin et al., 2005).
to CBT and to study differences and commonalities on
mechanisms of the changes.
The eligible participants were users of a Chronic Pain 2.2.1 | Baseline sociodemographic-­clinical
Liaison Program that was coordinated by the Mental Health questionnaire
Department (MHD) of a public general hospital in Madrid
(Spain). On average, over 450 patients enrolled in the pro- The following information was collected at baseline: gender,
gramme every year. Most of them were referred from the hos- age, marital status, education level and employment status. In
pital's CP unit and community mental health centres of the addition, relevant clinical variables, such as personal medical
catchment area. The recruitment period was from February history, years of CP and medical visits because of the pain in
2017 to October 2018. the last 3 months, were recorded.
Patients with the first appointment in the CP unit at least
3 months before the enrolment; ≥18 years of age; with a score
≥8 on the anxiety and/or depression subscales of the Hospital 2.2.2 | The primary outcome
Anxiety and Depression Scale (HADS); diagnosed with ad-
justment disorder, dysthymia or major depressive disorder Self-­compassion: The Self-­ Compassion Scale (SCS;
according to the Diagnostic and Statistical Manual of Mental Neff, 2003b) was a self-­reported instrument that consisted of
Disorders, Fifth Edition; and with signed informed consent 26 items including statements, such as ‘I am kind to myself
forms were included in the study. The following patients when I am experiencing suffering’, ‘when I see aspects of my-
were excluded: diagnosed with intellectual disability and/ self that I don't like I get down on myself’ and ‘when things
or any type of cognitive impairment, psychotic and/or manic are going badly for me, I see the difficulties as part of life that
symptoms and self-­harm or suicidal ideation at the time of everyone goes through’. The patients were asked to rate to
the study, and with previous formal training on mindfulness. what extent they experienced these feelings or situations on
Withdrawal criteria were as follows: participant's decision, a 5-­point Likert-­type scale, where 1 was ‘almost never’ and
hospitalization in a psychiatric unit or a worsening clinical 5 ‘almost always’. The final scores ranged from 1 to 5, with
condition identified by the researchers or the participant's at- higher values indicating greater self-­compassion. The higher
tending physician/s. scores were associated with lower negative effects, and lower
The convenience sample of care providers consisted of one disabilities and catastrophizing among people with CP (Costa
psychiatrist and one art therapist for MSC, and four clinical & Pinto-­Gouveia, 2013). The Spanish version of the SCS had
psychologists for CBT. The MSC therapists were trained and good internal consistency (α = 0.87) and test–­retest stability
certified by San Diego University (USA), and the CBT ther- (r = 0.92; Garcia-­Campayo et al., 2014).
apists were experienced clinical psychologists specifically
trained on CBT for CP. All therapists had wide experience
in the field of CP. Both interventions followed standardized 2.2.3 | Secondary outcomes
intervention manuals.
This study adhered to the tenets of the Declaration of Pain interference
Helsinki, SPIRIT 2013 (Chan et al., 2013) and CONSORT The Brief Pain Inventory (BPI) ‘interference’ subscale
2010 statements (Moher et al., 2012). Our institutional re- (Cleeland & Ryan, 1994) measured how much pain inter-
view board approved the trial (identifier 4,757). The study feres with daily aspects, such as mobility or social activi-
protocol was prospectively registered in December 2016 ties. It included 7-­point Likert-­type items that can be scored
(ClinicalTrials.gov identifier NCT03386422) and retrospec- from 0 (‘does not interfere’) to 10 (‘completely interferes’).
tively modified twice (March 2019 and May 2020). IMMPACT recommends its use as a measure of function-
ing in clinical trials (Dworkin et al., 2005). The Spanish ver-
sion of the BPI had good psychometric properties and the
2.2 | Measures interference subscale, in particular, presented high internal
consistency (α = 0.89) and acceptable test–­retest reliability
Participants completed a sociodemographic-­ clinical ques- (r = 0.77; Badia et al., 2003).
tionnaire (baseline) and a paper-­and-­pencil battery of instru-
ments (baseline and post-­intervention). Assessments included Pain intensity
measures of pain interference, pain intensity, emotional dis- The Pain Visual Analogue Scale (McCormack et al., 1988)
tress, QoL, self-­compassion, catastrophizing and pain accept- measured the intensity of the pain experienced. It consisted
ance. The instruments were selected in line with the Initiative of a horizontal line divided into 10 equal parts that range
4
|    TORRIJOS-­ZARCERO et al.

from ‘0 = no pain’ to ‘10 = greatest pain ever experienced’. attempts to control and/or avoid experiencing pain. Participants
Participants indicated the number in the horizontal line that rated each item on a 7-­point Likert scale, from 0 (‘never true’)
best reflects their level of pain. The scale had good test–­retest to 6 (‘always true’). The final scale score ranged from 0 to 48;
reliability (r = 0.87) for CP (Boonstra et al., 2008). higher scores indicated greater acceptance of pain. The Spanish
version of the scale showed good internal consistency (α = 0.82
Anxiety and Depression Symptoms for the activity engagement component and α = 0.78 for the
The Hospital Anxiety and Depression Scale (HADS; pain willingness component; Menéndez et al., 2010).
Zigmond & Snaith, 1983) was a 14-­item self-­reported instru-
ment designed for clinical populations with physical symp-
toms or conditions. It explored the symptoms of anxiety (7 2.3 | Interventions
items) and depression (7 items) experienced during the last
7 days with a 4-­point Likert scale. Two final scale scores The intervention programmes (MSC and CBT) were con-
were calculated, one for the anxiety subscale (from 0 to 21 ducted following two standardized treatment protocols (see
points) and one for the depression subscale (from 0 to 21 below). Both of them consisted of a 150-­min session (plus
points). Higher values indicated more severe symptoms. The homework between sessions) once a week during 8 weeks
HADS accurately discriminated between depressed and non-­ formatted to groups of no more than 20 participants. The two
depressed CP patients (Rusu et al., 2012). The Spanish ver- programmes were mutually exclusive; the MSC did not in-
sion was reliable, with α = 0.83 for the anxiety subscale and clude exercises that changed disruptive thoughts and beliefs
α = 0.87 for the depression subscale (Vallejo et al., 2012). and the CBT did not include mindfulness techniques or yoga
exercises. Table 1 presents an overview of the programmes.
QoL
The 36-­item Health Survey (SF-­36; Ware & Sherbourne, 1992)
was a self-­reported instrument that explored eight domains 2.3.1 | MSC
of QoL, namely physical functioning, physical role, body
pain, general health, vitality, social functioning, emotional Mindful Self-­Compassion is a protocol-­standardized interven-
role and mental health. IMMPACT recommended its use as a tion aimed at increasing mindfulness and self-­compassion and
measure of QoL in pain clinical trials (Dworkin et al., 2005). reducing the suffering associated with experiential avoidance.
Two summary components (physical and mental) have It was designed and protocolized by Neff and Germer (Germer
been developed from its eight original dimensions (Vilagut & Neff, 2019; Neff & Germer, 2013). It is not specific for CP.
et al., 2005). The Spanish version had an adequate internal Adherence to the standard MSC protocol was strict, without
consistency (α = 0.71–­0.94, except for the social functioning specific reference to pain as a source of suffering.
subscale whose α was 0.45) and acceptable test–­retest reli- The central components of the MSC were formal medita-
ability (r = 0.58–­0.99; Alonso, 1995). tion together with formal and informal self-­compassion prac-
tices aimed at developing cognitive, behavioural and physical
Pain catastrophizing abilities to soothe and comfort oneself when distressed. The
The Pain Catastrophizing Scale (PCS; Sullivan et al., 1995) outline of the programme was as follows: (a) general introduc-
was a 13-­item measure that explored pain-­related catastro- tion and a review of self-­compassion (what it is, and what it
phizing, including rumination, magnification and helpless- is not), (b) foundational knowledge and the practice of mind-
ness. Participants rated each item from 0 (‘not at all’) to 4 fulness, (c) application of self-­compassion in various aspects
(‘all the time’), based on how often they experienced a cer- of life and the practice of self-­kindness, (d) recognition of the
tain thought or feeling. The final scale score ranged from 0 to inner critic voice and development of a compassionate inner
52 points, where higher scores indicated greater catastrophic voice, (e) the importance of living in accordance with core
thinking in response to pain. The Spanish version of the scale values, (f) development of skills to deal with difficult emo-
had an adequate internal consistency (>0.7 for the total scale tions, like shame, (g) development of skills to deal with chal-
and subscales) and test–­retest reliability (>0.7 for the total lenging interpersonal relationships and (h) development of
scale and subscales; García-­Campayo et al., 2008). skills to relate to positive aspects of oneself and to one's life
with appreciation, including working on thankfulness.
Pain acceptance
The Chronic Pain Acceptance Questionnaire (CPAQ) was a
20-­item self-­reported measure assessing acceptance of pain 2.3.2 | CBT
(McCracken et al., 2004). The instrument explored (a) how
much a person engaged in life activities even when experienc- This was based on the Kovacs and Moix's protocol manual
ing pain and (b) how much a person disengaged from his or her (Kovacs & Moix, 2011) and focused on training participants to
TORRIJOS-­ZARCERO et al.    
| 5

TABLE 1 Outline of the MSC and CBT sessions

Session number MSC CBT

1 • Discovering MSC • Psychoeducation
• Soothing touch informal practice • Introducing CBT
• Self-­compassion break informal practice
2 • Practicing mindfulness • Understanding the vicious circle:
• Affectionate breathing strain–­pain–­strain
• ‘Soles of the feet’ informal practice • Relaxation:
• ‘Here-­and-­Now stones’ informal practice ∘ Progressive muscle relaxation
∘ Diaphragmatic breathing
3 • Practicing loving-­kindness • Attention techniques:
• Awakening our hearts exercise ∘ Distraction
• Compassion/loving-­kindness meditation for a loved one ∘ Visualization
• Finding loving-­kindness phrases informal practice • Presenting dysfunctional thoughts topic:
∘ ABC model
∘ About cognitive bias and dysfunctional
thoughts
4 • Discovering your compassionate voice • Working with dysfunctional thoughts:
• Self-­compassion/loving-­kindness Meditation for ourselves ∘ Working with diaries
• Motivating ourselves with compassion versus self-­criticism exercise ∘ Detecting cognitive bias and dysfunc-
• ‘Compassionate Letter to Myself’ informal practice tional thoughts
∘ Training strategies to change thoughts
cognitive discussion
5 • Living deeply • Psychoeducation on emotions
• Giving and receiving compassion meditation ∘ Explaining the relationship between dif-
• ‘Discovering Our Core Values’ exercise ficult emotions and pain
• ‘Living with a Vow’ informal practice • Emotion regulation exercises
• ‘Compassionate listening’ informal practice ∘ Identifying difficult emotions and its
Relation to pain
∘ Distance from unpleasant emotions
6 • Meeting difficult emotions • Interpersonal abilities
• Strategies for meeting difficult emotions • Working with interpersonal problems
• ‘Soften-­Soothe-­Allow’ informal practice • Assertiveness
• Topic on the emotion of shame
7 • Exploring challenging relationships • Working on life values and behavioural
• Compassionate friend meditation goals
• ‘Meeting Unmet Needs’ exercise • Time scheduling:
• ‘Self-­Compassion Break in Relationships’ informal practice ∘ including enjoyable activities
• ‘Compassion with Equanimity’ informal practice ∘ and self-­care time
8 • Embracing your life • Education about paced physical activity
• Compassion for self and others meditation • Education about body postures to prevent
• Cultivating happiness: ‘Savouring’ and ‘Gratitude’ informal exercises damage and pain
• ‘Self-­appreciation’ exercise • Relapse prevention
• Closing ‘What Would I Like to Remember?’
Abbreviations: CBT, cognitive-­behavioural therapy; MSC, Mindful Self-­Compassion.

manage their pain. Furthermore, we included CBT techniques psychoeducation on emotions and how to regulate them; (e)
most commonly practiced and studied for CP (Otis, 2007): interpersonal abilities; (f) attention techniques, such as dis-
(a) psychoeducation about CP and the relationship between traction and visualization exercises; (g) life values, behav-
thoughts, emotions and physical reactions; (b) relaxation and ioural goals, time scheduling (including pleasure activities)
breathing techniques (abdominal breathing and progressive and self-­care time and (h) paced physical activity and educa-
muscle relaxation); (c) cognitive restructuring, instruction tion about body postures to prevent pain.
and practice of changing dysfunctional thoughts (includ- Both interventions were offered as a supplement to usual
ing catastrophizing) and common beliefs among individuals care, which included primary care follow-­up, drug therapy,
with CP (e.g. inability to control pain, hurt equals harm); (d) physical therapy and/or surgical procedures.
6
|    TORRIJOS-­ZARCERO et al.

2.4 | Procedure To identify the baseline differences in relevant variables,

the Pearson's chi-­squared test was conducted for categorical
First, one of the researchers (APT) went over the records of the and ordinal variables, and independent samples t tests were
Chronic Pain Liaison Program to identify outpatients who had conducted for continuous variables (where the distribution
been visiting the department for more than 3 months. Then, she was not normal, the Mann–­Whitney U test was conducted).
excluded patients who were under 18 years. The remaining re- To estimate the comparative average treatment effect (ATE)
cords were of eligible participants. The team tried to contact all on the primary and secondary outcomes, generalized estimated
of them to inform them of the study. Participants who were in- equation (GEE) modelling was used (the post-­treatment score
terested were given an appointment where they were requested was the dependent variable and the treatment group was the in-
to fill in the informed consent form and the self-­administered dependent variable). The pre-­treatment score was introduced as
HADS and participate in a semi-­structured diagnostic interview. a covariate, even if the two groups were equivalent at baseline.
Participants who met all the inclusion criteria were asked to take The final ATE estimators (B) were obtained, along with their
a brief clinical interview. After considering the exclusion crite- 95% confidence intervals. An exchangeable working correlation
ria, a unique identifier was assigned along with completing the structure was introduced because it was assumed that the cor-
rest of the baseline assessment, either during this or the next relation between any two measurements was the same for each
scheduled visit. Once the baseline assessment was completed, individual. For these types of longitudinal data, GEE models are
the participants were informed about the date, time and location recommended because they allow for within-­subjects’ observa-
of the interventions. Neither the result of the allocation process tions to be correlated and for such correlation structures to be
nor the information regarding the study hypothesis was revealed introduced in the model (Zeger & Qaqish, 1988).
until the day before the first session. Additionally, between-­ group and within-­ group effect
The recruitment stopped either the day before the inter- sizes (standardized mean difference [SMD]) for all outcomes
vention started or once the maximum number of participants have been included.
per randomization (n = 40) was reached, whichever was ear- The main analyses were conducted on all randomized par-
lier. Then, one of the authors (RM), who neither took part in ticipants (intention-­to-­treat). If values were missing either at
the enrolment of the participants nor was present in the inter- random or completely at random, expectation-­maximization
vention group, randomly allocated each identifier to one of imputation techniques were used. Per-­protocol sub-­analyses
the treatment groups (ratio 1:1). The sequence was obtained were also carried out. All statistical analyses were performed
through the TeamMaker™ software (http://chir.ag/proje​cts/ using SPSS version 21.0 (IBM Inc.).
team-­maker/) and no restrictions (i.e. blocking) were applied.
Once randomized, the participants were contacted and in-
formed about the assigned treatment (they were not blinded 3 | RESULTS
to the type of intervention). Post-­treatment assessments were
conducted by a research assistant (CRG), who was blinded to A total of 251 eligible participants were evaluated over the
the treatment allocation for 7 days after the last session. This study period (18 months, 6 randomizations). Among them,
procedure was repeated at each randomization. 159 met the inclusion criteria, and 123 completed the base-
line assessment and were randomized. However, 120 re-
ceived the intervention (attended at least one session) and 89
2.5 | Sample size calculation were assessed for the primary outcome (SCS) at follow-­up
(MSC: n = 42, CBT: n = 47). Figure 1 depicts the flowchart
To detect the medium effect size (f2 = 0.15) on the primary of the participants.
outcome (SCS) with two predictors (treatment group and Participants’ characteristics and baseline outcomes are
baseline scores), 73 participants were required for α = 0.05 presented in Table 2. Most of the participants were women
(two-­tailed) and 1 − β = 0.90. Considering an attrition rate (87.8%). One-­third of the sample included highly educated
of 20%, we required a sample size of at least 88 participants. people (university degree), and only 22% of the participants
were working at the time of enrolment. The participants rated
the intensity of their pain with a mean of 7.5 points (out of
2.6 | Statistical analysis 10), and most of them were diagnosed with adjustment dis-
order (66.7%). Baseline characteristics and outcomes were
Appropriate descriptive statistics were calculated (percent- equivalent in both the groups.
ages, medians, means and standard deviations) depending Two statistical analyses were conducted. The first one was
on the variables’ distributions. Further analyses utilizing an intention-­to-­treat analysis with the imputation of missing
goodness-­of-­fit tests were conducted to check the probability data (expectation-­maximization method). The second one
distributions of continuous variables. was a per-­protocol analysis. Table 3 presents the main results
TORRIJOS-­ZARCERO et al.    
| 7

of the study. In both analyses, ATEs were higher in the ex- The per-­protocol analysis demonstrated that MSC pro-
perimental arm. duced greater effects compared to CBT on pain interference,
Between-­group and within-­group SMDs are provided in pain catastrophizing, anxiety and depressive symptoms, and
Table 4. pain acceptance.
No adaptations of MSC or CBT were required. No major
adverse events (i.e. hospitalization in a psychiatric unit or
3.1 | Primary outcome visits to the emergency department because of psychological
distress) were detected during the study period.
Average treatment effects on the primary outcome (self-­
compassion) in the intention-­to-­treat (ITT) analyses favoured
MSC with 95% confidence intervals. 4 | DISCUSSION
The per-­protocol analysis also favoured MSC over CBT
on the basis of the primary outcome. In accordance with our hypotheses, MSC was superior to
CBT in its effects on the primary (self-­compassion) and
most of the secondary outcomes (pain interference, anxiety
3.2 | Secondary outcomes and pain acceptance). Furthermore, the per-­protocol anal-
ysis showed that MSC had a greater effect on depression
Average treatment effects on the secondary outcomes, pain symptoms too.
interference, anxiety symptoms and pain acceptance, in the
ITT analyses, favoured MSC with 95% confidence intervals.
Results also suggest trends in that MSC might be more 4.1 | Primary outcome: self-­compassion
effective in reducing pain interference, pain catastrophiz-
ing and anxiety symptoms, and improving pain acceptance Regarding the primary outcome, in this study, it was found
(p < 0.10). that MSC was more effective than CBT for improving

Enrollment
Assessed for eligibility
(n=251)

Excluded (n=128)
• Not meeting inclusion criteria
(n=39)
• No contact (n=25)
• Declined to participate (n=28)
• No baseline assessment (n=36)

Randomized (n=123)

Allocation
Allocated to MSC (n=62) Allocated to CBT (n=61)
• Received allocated • Received allocated
intervention (n=60) intervention (n=60)
• Not received allocated • Not received allocated
intervention(n=2) intervention(n=1)

Post-treatment
• Lost to post-treatment assessment • Lost to post-treatment
assessment* (n=20) assessment* (n=14)
• Not completed (n=14) • Not completed (n=13)

Analysis
• Intention to treat (n=62) • Intention to treat (n=61)
• Per protocol (n=48) • Per protocol (n=48)

FIGURE 1 Participant flowchart. *Primary outcome. CBT, cognitive-­behavioural therapy; MSC, Mindful Self-­Compassion
8
|    TORRIJOS-­ZARCERO et al.

TABLE 2 Sociodemographic and clinical characteristics at baseline

Hypothesis testing (MSC vs.

MSC (n = 62) CBT (n = 61) Total sample (n = 123) CBT)
Age (years), M (SD) 48.29 (10.17) 49.25 (11.39) 48.76 (10.75) t(121) = –­0.491, p = 0.624a
Gender, n (%) χ2(1) = 0.740, p = 0.390b
Women 56 (90.3) 52 (85.2) 108 (87.8)
Men 6 (9.7) 9 (14.8) 15 (12.2)
Level of education, n (%) χ2(2) = 2.744, p = 0.433b
Primary degree 6 (9.7) 5 (8.2) 11 (8.9)
Secondary degree 37 (59.7) 34 (55.7) 71 (57.7)
University degree 19 (30.6) 22 (36.1) 41 (33.3)
Job status, n (%) χ2(4) = 0.609, p = 0.962b
Employed 13 (21) 14 (23.0) 27 (22)
Housework 7 (11.3) 6 (9.8) 13 (10.6)
Unemployed 7 (11.3) 5 (8.1) 12 (9.8)
Retired 9 (14.5) 9 (14.8) 18 (14.6)
Sick leave or disability 25 (41.9) 28 (45.9) 53 (43.1)
Medical visits in the last 3 months, χ2(3) = 0.751, p = 0.861b
n (%)
None 2 (4.5) 1 (2) 3 (3.2)
1 3 (6.8) 5 (9.8) 8 (8.4)
2–­5 12 (27.3) 14 (27.5) 26 (27.4)
>5 27 (61.4) 31 (60.8) 58 (61.1)
Duration of pain (in months), n (%) χ2(2) = 4.046, p = 0.132b
6–­12 months 5 (8.1) 1 (1.6) 6 (4.9)
1–­3 years 11 (17.7) 17 (27.9) 28 (22.8)
>3 years 46 (74.2) 43 (70.5) 89 (72.4)
Attrition, n (%) 13 (21.0) 14 (23.0) 27 (22.0) χ2(1) = 0.029, p = 0.865b
DSM-­5 diagnosis, n (%) χ2(2) = 0.306, p = 0.858b
Adjustment disorder 42 (67.7) 40 (65.6) 82 (66.7)
Major depressive disorder 8 (12.9) 10 (16.4) 18 (14.6)
Dysthymia 12 (19.4) 11 (18.0) 23 (18.7)
Primary outcome, M (SD)
Self-­compassion (SCS) (0–­5) 2.72 (0.58) 2.62 (0.43) 2.67 (0.51) t(113) = 1.068, p = 0.288a
Secondary outcomes, M (SD)
Pain intensity (PAVS) (0–­10) 7.52 (1.54) 7.52 (1.48) 7.52 (1.51) Z = –­0.385, p = 0.701c
Pain interference (BPI), (0–­10) 6.99 (1.29) 7.11 (1.82) 7.02 (1.57) t(108) = –­0.618, p = 0.788a
Pain acceptance (CPAQ), (0–­156) 39.99 (12.48) 37.64 (15.65) 38.82 (14.14) t(121) = 0.921, p = 0.359a
Catastrophizing (PCS), (0–­52) 33.36 (10.12) 35.60 (8.73) 34.47 (9.49) t(121) = –­1.314, p = 0.191a
Health, physical (SF-­36) (0–­100) 34.27 (7.57) 35.14 (9.02) 34.70 (8.30) t(121) = –­0.585, p = 0.560a
Health, mental (SF-­36) (0–­100) 23.05 (12.73) 22.39 (11.56) 22.72 (19.70) t(121) = 0.303, p = 0.763a
Depression (HADS) (0–­21) 11.51 (3.88) 11.56 (4.14) 11.53 (3.99) t(121) = –­0.065, p = 0.948a
Anxiety (HADS) (0–­21) 12.72 (3.21) 12.34 (3.67) 12.53 (3.44) t(121) = 0.604, p = 0.547a
Note: Italicized brackets show the score rank for each scale.
Abbreviations: BPI, Brief Pain Inventory; CBT, cognitive-­behavioural therapy; CPAQ, Chronic Pain Acceptance Questionnaire; HADS, Hospital Anxiety and Depression
Scale; MSC, Mindful Self-­Compassion; PAVS, Pain Visual Analogue Scale; PCS, Pain Catastrophizing Scale; SCS, Self-­Compassion Scale; SF-­36, SF-­36 Health Survey.
a
Independent samples t tests (degrees of freedom in brackets).
b
Chi-­squared test of independence (degrees of freedom in brackets).
c
Mann–­Whitney's U test.
TORRIJOS-­ZARCERO et al.    
| 9

self-­compassion (small-­to-­medium effect size difference). literature that has observed that MBIs increased pain ac-
This concurs with published research in various populations, ceptance (La Cour & Petersen, 2015; Turner et al., 2016).
other than in CP patients, that has found compassion-­based Traditionally, research is focused on pain, coping and cat-
interventions to have a significant effect on self-­compassion astrophizing as the typical action mechanisms of CBT.
when compared with an active control (Kirby et al., 2017). However, recent studies have projected acceptance also as
Prior to this study, no specific data existed on the changes an indicator of the benefits achieved with CBT (Åkerblom
in self-­compassion after interventions for CP since most et al., 2015; Baranoff et al., 2013; Turner et al., 2016).
recent studies in this field did not collect data on the self-­ Acceptance has been considered as one of the most relevant
compassion outcome (Carson et al., 2005; Montero-­Marín action mechanisms of third-­wave therapies, including MBIs
et al., 2018). Given that self-­compassion is an effective way (Day & Thorn, 2016; La Cour & Petersen, 2015). The MSC
to cope with life stressors, including CP (Wren et al., 2012), program proved effective in increasing acceptance and re-
this result seems relevant. Self-­compassionate individuals ducing avoidance in a couple of earlier studies (Edwards
ruminate less (Odou & Brinker, 2014), are usually not per- et al., 2019; Neff & Germer, 2013). It buffered the degree
fectionists, have less fear of failures (Killham et al., 2018) to which intolerable pain sensations were experienced and
and intrinsically motivate themselves with a compassion- immediately avoided (Shapiro et al., 2006). Despite the
ate voice to change their lives for the better (Zhang & level of pain, lesser avoidance led to better adjustment and
Chen, 2016). In contrast, self-­criticism, common among lesser pain interference (McCracken & Eccleston, 2005).
people with chronic medical conditions, results in poor self-­ Therefore, we hypothesized that improvement of pain ac-
care. Working on self-­ compassion may enhance health-­ ceptance may be a common result of different therapies,
promoting behaviours due in part to its link to adaptive even beyond CBT or MBIs.
emotions (Homan & Sirois, 2017; Sirois et al., 2015; Terry
et al., 2013), even in chronic medical populations (Brion
et al., 2014), including CP. 4.2.2 | Pain interference

In accordance with a recent meta-­ analysis (Veehof

4.2 | Secondary outcomes et al., 2016), MBIs reduced pain interference, as in this study,
where pain `reduction was higher in MSC than in CBT. We
4.2.1 | Anxiety and depression symptoms and hypothesized that MSC might regulate pain interference by
pain acceptance working on values and facilitating people with CP to focus
on and engage in valued or worthy aspects of their lives, in-
Among the secondary outcomes, the results on pain accept- stead of in pain and fear. This was achieved through less
ance and anxiety were remarkable; while both interventions avoidance and more proactive behaviour when coping with
were effective to some extent, MSC was superior to CBT in difficulties.
increasing pain acceptance (medium effect size difference)
and reducing anxiety (small effect size difference).
The first RCTs conducted to test the effectiveness of CBT 4.2.3 | QoL, pain intensity and pain
for CP found small-­to-­medium effect size changes in anxi- catastrophizing
ety. However, subsequent meta-­analyses (Morley et al., 1999;
Williams et al., 2012) concluded that CBT-­based programmes No treatment effects were found in this study for pain inten-
have no significant effect on mood and anxiety when com- sity and QoL general indexes. Studies that analysed the com-
pared with an active control. Acceptance-­based interventions ponents of QoL separately found improvements, particularly
and MBIs have shown medium effect size reductions in anxi- in vitality (de Jong et al., 2017; La Cour & Petersen, 2015)
ety symptoms after treatment (Luciano et al., 2014; Wicksell and physical functioning (Khoo et al., 2019). Besides, a sig-
et al., 2013; Wong et al., 2011). There have been few previ- nificant reduction in catastrophizing was found in the MSC
ous studies on self-­compassion-­based interventions for CP, arm; previous literature has stated that self-­ compassion
which are more relevant to our results. Montero-­Marín et al. was related to low levels of pain catastrophizing (Wren
found that a compassion-­based intervention (different from et al., 2012).
MSC) produced a large effect on anxiety (Montero-­Marín In our point of view, suffering in CP may be divided
et al., 2018). into five core components: (a) struggle with cognitive as-
Regarding pain acceptance, our study indicated that pects (high self-­criticism; [Smith & Osborn, 2007; Toye
MSC produced better results than CBT, although accep- et al., 2013], rumination about difficult aspects of the self,
tance had improved even after CBT (a small effect size over-­identification, worries about attaining personal goals
change after treatment); this is in accordance with previous and concerns about being able to fulfil one's personal and
 10
|   

TABLE 3 The estimated marginal means and average treatment effects (B) for primary and secondary outcomes at post-­intervention

Intention to treat (n = 123) Per-­protocol (n = 96)

MSC (n = 62) CBT (n = 61) Ba MSC (n = 48) CBT (n = 48) Ba

Primary outcome
Self-­compassion (SCS) (0–­5) 2.87 (2.64, 3.12) 2.75 (2.52, 2.98) 0.126 (–­0.000, 0.252)** 2.87 (2.59, 3.16) 2.72 (2.43, 3.01) 0.152 (–­0.002, 0.306)*
Secondary outcomes
Pain intensity (PAVS) (0–­10) 6.93 (5.04, 9.53) 7.10 (5.17, 9.76) –­0.024 (–­0.082, 0.034) 7.05 (4.91, 10.14) 7.21 (5.02, 10.36) –­0.022 (–­0.092, 0.048)
**
Pain interference (BPI), (0–­10) 6.59 (4.99, 8.19) 6.98 (5.45, 8.51) –­0.393 (–­0.760, –­0.029) 6.61 (4.61, 8.62) 7.05 (5.18, 8.91) –­0.433 (–­0.876, 0.010)*
Pain acceptance (CPAQ), (0–­156) 45.51 (–­24.49, 115.51) 40.29 (–­29.49, 110.12) 5.214 (1.870, 5.560)*** 45.92 (–­41.07, 132.92) 39.56 (–­47.66, 126.79) 6.361 (2.140, 10.580)***
*
Catastrophizing (PCS), (0–­52) 30.01 (–­18.86, 78.88) 32.15 (–­16.75, 81.05) –­2.139 (–­4.596, 0.317) 30.10 (–­29.97, 90.17) 38.82 (–­27.27, 92.91) –­2.791 (–­5.721, 0.283)*
Health, physical (SF-­36) (0–­100) 35.76 (0.85, 70.67) 34.25 (–­1.23, 69.72) 1.510 (–­0.347, 3.367) 36.03 (–­7.15, 79.21) 34.26 (–­9.56, 78.08) 1.772 (–­0.497, 4.040)
Health, mental (SF-­36) (0–­100) 25.77 (5.78, 114.84) 24.87 (5.74, 107.78) 0.035 (–­0.089, 0.160) 25.91 (4.36, 153.87) 23.99 (4.19, 137.42) 0.077 (–­0.075, 0.229)
Depression (HADS) (0–­21) 10.11 (4.51, 15.71) 10.74 (4.96, 16.52) –­0.628 (–­1.497, 0.241) 10.24 (3.39, 17.09) 11.23 (4.16, 18.31) –­0.993 (–­2.033, 0.047)*
Anxiety (HADS) (0–­21) 10.82 (4.67, 16.96) 11.72 (5.58, 17.86) –­0.902 (–­1.770, –­0.034)** 10.81 (2.92, 18.70) 11.99 (4.10, 19.89) –­1.183 (–­2.262, –­0.104)**
Note: Non-­italicized brackets show 95% confidence intervals; italicized brackets show the score range for each scale.
Abbreviations: BPI, Brief Pain Inventory; CBT, cognitive-­behavioural therapy; CPAQ, Chronic Pain Acceptance Questionnaire; HADS, Hospital Anxiety and Depression Scale; MSC, Mindful Self-­Compassion; PAVS, Pain
Visual Analogue Scale; PCS, Pain Catastrophizing Scale; SCS, Self-­Compassion Scale; SF-­36, SF-­36 Health Survey.
a
Dependent variables: SCS, PAVS, BPI, CPAQ, PCS and SF-­36 (2) and HADS (2) at post-­intervention; independent variable: treatment arm (MSC or CBT); covariate: SCS, PAVS, BPI, CPAQ, PCS and SF-­36 (2) and HADS
(2) at baseline.
*p < 0.10.
**p < 0.05.
***p < 0.01.
TORRIJOS-­ZARCERO et al.
TORRIJOS-­ZARCERO et al.    
| 11

TABLE 4 Between-­groups (post-­treatment) and within-­groups standardized mean differences (SMD)

Between-­groupsa Within-­groups CBT Within-­groups MSC

SMDb SMDb SMDb
Pain Interference (BPI) (0–­10) 0.33 0.003 0.29
Pain Intensity (PVAS) (0–­10) 0.07 0.24 0.35
Self-­Compassion (SCS) (0–­5) 0.39 0.24 0.35
Pain Acceptance (CPAQ) (0–­156) 0.49 0.11 0.50
Catastrophizing (PCS) (0–­52) 0.38 0.32 0.48
Physical Health (SF-­36) (0–­100) 0.10 0.06 0.15
Mental Health (SF-­36) (0–­100) 0.12 0.24 0.34
Depression (HADS) (0–­21) 0.18 0.20 0.43
Anxiety (HADS) (0–­21) 0.24 0.21 0.74
Note: Italicized brackets show the score range for each scale.
Abbreviations: BPI, Brief Pain Inventory; CBT, cognitive-­behavioural therapy; CPAQ, Chronic Pain Acceptance Questionnaire; HADS, Hospital Anxiety and
Depression Scale; MSC, Mindful Self-­Compassion; PAVS, Pain Visual Analogue Scale; PCS, Pain Catastrophizing Scale; SCS, Self-­Compassion Scale; SF-­36, SF-­36
Health Survey.
a
Every between-­groups SMD favoured MSC.
b
SMD = (M1 − M2)/Pooled SD.

work-­related responsibilities), (b) difficult emotions (fear 4.3 | Strengths and limitations
of pain, fear of being criticized or seen as a burden, shame,
guilt or helplessness; Purdie & Morley, 2016; Smith & Several methodological features of this study are note-
Osborn, 2007), (c) unpleasant or painful bodily sensations worthy. Since most participants had more than 3 years of
(related to physical pain itself or to physical sensations that pain (72.4%), high levels of emotional distress and psy-
correlates with difficult emotions), (d) behavioural aspects chopathology (66.7% adjustment disorder, 14.6% major
(pain avoidance and general experiential avoidance of ac- depressive disorder and 18.7% dysthymia), extensive his-
tivities and events that evoke difficult emotions, which, tories of unsuccessful treatments in specialized units with
in turn, increases disability) and (e) social disconnection, high rates of medical visits (61.1% had visited a doctor
isolation and loneliness derived from the previous points. more than 5 times in the last 3 months because of pain)
Self-­compassion helps people to cope with this suffering and short-­or long-­term disability to work (43.1%), the fact
through the following core mechanisms: (a) stimulating that psychological treatments were effective is encourag-
the soothing system related to attachment in mammals, ing. To protect external validity, we tried to minimize the
which is a natural regulator of the threat system (Stellar & selection bias (i.e. not rejecting people with pending dis-
Keltner, 2014), (b) regulating the influence of the achieve- ability claims or comorbidities). Other strengths included
ment system when evaluating oneself with respect to wor- IMMPACT-­ recommended outcomes, random allocation,
thiness (Depue & Morrone-­ Strupinsky, 2005; Purdie & blind outcome assessment, an active control group that has
Morley, 2016), (c) promoting active attitudes, reducing help- already widely demonstrated effectiveness, programmes
lessness and facilitating change providing encouragement conducted by certified MSC teachers and well-­trained CBT
through warm and supporting voices (Gardner-­Nix, 2009; therapists and high levels of therapy manualization to fa-
La Cour & Petersen, 2015), (d) facilitating the self-­efficacy cilitate replication.
perception when approaching and managing emotions and Limitations include moderate attrition rates (around
difficulties, thereby reducing experiential avoidance, (e) 20%, in accordance with most RCTs on psychotherapeutic
promoting non-­judgemental kindness, curiosity, openness, interventions for CP; Glombiewski et al., 2010; Luciano
moment to moment attitude towards the whole experience et al., 2014), absence of a third control non-­active group
(Kabat-­Zinn & Hanh, 2009), especially to the experience or usual care, absence of follow-­up and non-­systematic
of pain, suffering, and failures and understanding that these registration of adverse events, which would have been
experiences are unavoidable and part of the human condi- really valuable (Sharpe, 2020). All measures were based
tion, thus improving connectedness (Edwards et al., 2019) on patient-­ reported outcomes. Including objective out-
and (f) facilitating the engagement in value-­based activities comes, such as return to work and ecological momentary
and reducing the impact that CP has on important domains assessment method (Garcia-­ Palacios et al., 2014), may
of life, rather than reducing pain intensity itself (Edwards more clearly reflect wider impacts and improve ecological
et al., 2019). validity.
12
|    TORRIJOS-­ZARCERO et al.

5 | CO NC LU SION S CONFLICT OF INTEREST

The authors declare no conflicts of interest.
The results of this randomized, controlled trial comparing
two interventions in adult patients with CP conducted at the AUTHOR CONTRIBUTIONS
MHD of a tertiary hospital suggest that both MSC and CBT APT, BRV and MTZ conceived and designed the study and col-
have beneficial effects implemented together with standard lected the data. CRG and ILA collected the data. RM made ran-
medical management; however, MSC offers greater ben- domizations and data analysis. APT, MTZ and RM wrote the
efits to self-­compassion, pain interference, pain acceptance, paper and reviewed the successive versions of the manuscript.
pain catastrophizing and emotional well-­being than the CBT BRV reviewed the successive versions of the manuscript and
intervention. made relevant contributions to the manuscript. MDRD made
Our results were meaningful for a specific group of CP relevant contributions to the manuscript. All authors discussed
patients: women, highly educated people, patients with CP the results and commented on the manuscript, and all of them
and comorbid psychopathologies, undergoing treatment in read and approved the final version of the manuscript.
specialized units (non-­primary care units) who were referred
to the Chronic Pain Liaison Program of the MHD with a pro- R E F E R E NC E S
longed history of pain, medical visits and previous treatments Åkerblom, S., Perrin, S., Fischer, M. R., & McCracken, L. M. (2015).
(following the therapeutic ladder for pain management by the The mediating role of acceptance in multidisciplinary Cognitive-­
OMS). These kinds of patients have few therapeutic alterna- Behavioral Therapy for chronic pain. The Journal of Pain, 16, 606–­
615. https://doi.org/10.1016/j.jpain.2015.03.007
tives left; MSC seems to be a very valuable therapeutic alter-
Alonso, J. (1995). La versión española del SF-­ 36 Health Survey
native when there is a great level of suffering and previous
(Cuestionario de Salud SF-­36): Un instrumento para la medida de
treatments have failed. los resultados clínicos. Medicina Clínica, 104, 771–­776.
Future research may help in identifying the differences Badia, X., Muriel, C., Gracia, A., Manuel Núñez-­Olarte, J., Perulero,
and commonalities between MSC and CBT that may pro- N., Gálvez, R., Carulla, J., & Cleeland, C. S. (2003). Validación es-
mote pain-­related improvements and patient characteristics pañola del cuestionario Brief Pain Inventory en pacientes con dolor
that may predict better compatibility with specific treat- de causa neoplásica. Medicina Clínica, 120, 52–­ 59. https://doi.
ment approaches. Both aspects are essential to establish org/10.1016/s0025​-­7753(03)73601​-­x
Baranoff, J., Hanrahan, S. J., Kapur, D., & Connor, J. P. (2013).
clinical guidelines. Previous literature pointed out that
Acceptance as a process variable in relation to catastrophizing in
self-­
compassion alone improved functioning in CP pa-
multidisciplinary pain treatment. European Journal of Pain, 17,
tients (Edwards et al., 2019). Therefore, tailoring inter- 101–­110. https://doi.org/10.1002/j.1532-­2149.2012.00165.x
ventions that target self-­compassion more directly may be Boonstra, A. M., Schiphorst Preuper, H. R., Reneman, M. F., Posthumus,
warranted in the future, even if they are not compassion-­ J. B., & Stewart, R. E. (2008). Reliability and validity of the visual
based. Psychological treatments for CP, in any form (CBT, analogue scale for disability in patients with chronic musculoskele-
ACT, MBSR, etc.), may improve, in particular the emo- tal pain. International Journal of Rehabilitation Research, 31, 165–­
tional functioning outcomes (Davey et al., 2020), with 169. https://doi.org/10.1097/mrr.0b013​e3282​fc0f93
Brion, J. M., Leary, M. R., & Drabkin, A. S. (2014). Self-­compassion
the introduction of self-­compassion training and a self-­
and reactions to serious illness: The case of HIV. Journal of Health
compassionate attitude from the therapists. Psychology, 19, 218–­229. https://doi.org/10.1177/13591​05312​467391
Mindfulness-­ based intervention, in particular, Carson, J. W., Keefe, F. J., Lynch, T. R., Carson, K. M., Goli, V., Fras,
compassion-­ based interventions, help in recognizing a A. M., & Thorp, S. R. (2005). Loving-­Kindness meditation for
person as worthy of compassion, respect, dignity and for- chronic low back pain: Results from a pilot trial. Journal of Holistic
giveness, especially when facing failure, pain, discomfort, Nursing, 23, 287–­304. https://doi.org/10.1177/08980​10105​277651
physical and/or psychological suffering. Given the emo- Chan, A.-­W., Tetzlaff, J. M., Gotzsche, P. C., Altman, D. G., Mann,
H., Berlin, J. A., Dickersin, K., Hrobjartsson, A., Schulz, K. F.,
tional benefits of compassion, we would like to encourage
Parulekar, W. R., Krleza-­ Jeric, K., Laupacis, A., & Moher, D.
therapists to include this component and recognize the
(2013). SPIRIT 2013 explanation and elaboration: Guidance for pro-
importance of this human emotion in whatever practice or tocols of clinical trials. British Medical Journal, 346, e7586. https://
technique they adopt. doi.org/10.1136/bmj.e7586
Cleeland, C. S., & Ryan, K. M. (1994). Pain assessment: Global use of
ACKNOWLEDGEMENTS the Brief Pain Inventory. Annals, Academy of Medicine, Singapore,
We would like to thank the Chronic Pain Unit team of La Paz 23(2), 129–­138.
University Hospital and the professionals of the Community Costa, J., & Pinto-­Gouveia, J. (2013). Experiential avoidance and self-­
compassion in chronic pain. Journal of Applied Social Psychology,
Mental Health Centers of the catchment area for the collabo-
43, 1578–­1591. https://doi.org/10.1111/jasp.12107
rative work, and the biostatistical team of IdiPAZ and La Paz
Cusens, B., Duggan, G. B., Thorne, K., & Burch, V. (2009). Evaluation
University Hospital for their support along different phases of the breathworks mindfulness-­ based pain management
of this project.
TORRIJOS-­ZARCERO et al.    
| 13

programme: Effects on well-­being and multiple measures of mind- Glombiewski, J. A., Hartwich-­Tersek, J., & Rief, W. (2010). Attrition
fulness. Clinical Psychology & Psychotherapy, 17, 63–­78. https:// in cognitive-­behavioral treatment of chronic back pain. The Clinical
doi.org/10.1002/cpp.653 Journal of Pain, 26, 593–­ 601. https://doi.org/10.1097/ajp.0b013​
Davey, A., Chilcot, J., Driscoll, E., & McCracken, L. M. (2020). e3181​e37611
Psychological flexibility, self-­compassion and daily functioning in Gooding, H., Stedmon, J., & Crix, D. (2020). ‘All these things don’t take
chronic pain. Journal of Contextual Behavioral Science, 17, 79–­85. the pain away but they do help you to accept it’: Making the case for
https://doi.org/10.1016/j.jcbs.2020.06.005 compassion-­focused therapy in the management of persistent pain.
Day, M. A., & Thorn, B. E. (2016). The mediating role of pain accep- British Journal of Pain, 14, 31–­41. https://doi.org/10.1177/20494​
tance during mindfulness-­ based cognitive therapy for headache. 63719​857099
Complementary Therapies in Medicine, 25, 51–­ 54. https://doi. Harrison, A. M., Scott, W., Johns, L. C., Morris, E. M. J., & McCracken,
org/10.1016/j.ctim.2016.01.002 L. M. (2017). Are we speaking the same language? Finding theoret-
Day, M. A., Thorn, B. E., & Burns, J. W. (2012). The continuing evo- ical coherence and precision in “mindfulness-­based mechanisms” in
lution of biopsychosocial interventions for chronic pain. Journal of chronic pain. Pain Medicine, pnw310. https://doi.org/10.1093/pm/
Cognitive Psychotherapy, 26, 114–­129. https://doi.org/10.1891/088 pnw310
9-­8391.26.2.114 Häuser, W., Thieme, K., & Turk, D. C. (2010). Guidelines on the
de Jong, M., Peeters, F., Gard, T., Ashih, H., Doorley, J., Walker, R., management of fibromyalgia syndrome –­A systematic review.
Rhoades, L., Kulich, R. J., Kueppenbender, K. D., Alpert, J. E., European Journal of Pain, 14, 5–­ 10. https://doi.org/10.1016/j.
Hoge, E. A., Britton, W. B., Lazar, S. W., Fava, M., & Mischoulon, ejpain.2009.01.006
D. (2017). A Randomized controlled pilot study on mindfulness-­ Hayes, S. C., Luoma, J. B., Bond, F. W., Masuda, A., & Lillis, J. (2006).
based cognitive therapy for unipolar depression in patients with Acceptance and commitment therapy: Model, processes and out-
chronic pain. The Journal of Clinical Psychiatry, 79, 15m10160. comes. Behaviour Research and Therapy, 44, 1–­ 25. https://doi.
https://doi.org/10.4088/jcp.15m10160 org/10.1016/j.brat.2005.06.006
Depue, R. A., & Morrone-­Strupinsky, J. V. (2005). A neurobehavioral Hilton, L., Hempel, S., Ewing, B. A., Apaydin, E., Xenakis, L.,
model of affiliative bonding: Implications for conceptualizing a Newberry, S., Colaiaco, B., Maher, A. R., Shanman, R. M.,
human trait of affiliation. Behavioral and Brain Sciences, 28, 313–­ Sorbero, M. E., & Maglione, M. A. (2017). Mindfulness meditation
349. https://doi.org/10.1017/S0140​525X0​5000063 for chronic pain: Systematic review and meta-­analysis. Annals of
Dworkin, R. H., Turk, D. C., Farrar, J. T., Haythornthwaite, J. A., Behavioral Medicine, 51, 199–­213. https://doi.org/10.1007/s1216​
Jensen, M. P., Katz, N. P., Kerns, R. D., Stucki, G., Allen, R. R., 0-­016-­9844-­2
Bellamy, N., Carr, D. B., Chandler, J., Cowan, P., Dionne, R., Galer, Homan, K. J., & Sirois, F. M. (2017). Self-­compassion and physical
B. S., Hertz, S., Jadad, A. R., Kramer, L. D., Manning, D. C., … health: Exploring the roles of perceived stress and health-­promoting
Witter, J. (2005). Core outcome measures for chronic pain clinical behaviors. Health Psychology Open, 4, 205510291772954. https://
trials: IMMPACT recommendations. Pain, 113, 9–­19. https://doi. doi.org/10.1177/20551​02917​729542
org/10.1016/j.pain.2004.09.012 Kabat-­ Zinn, J. (1982). An outpatient program in behavioral med-
Edwards, K. A., Pielech, M., Hickman, J., Ashworth, J., Sowden, G., & icine for chronic pain patients based on the practice of mind-
Vowles, K. E. (2019). The relation of self-­compassion to function- fulness meditation: Theoretical considerations and prelimi-
ing among adults with chronic pain. European Journal of Pain, 23, nary results. General Hospital Psychiatry, 4, 33–­47. https://doi.
1538–­1547. https://doi.org/10.1002/ejp.1429 org/10.1016/0163-­8343(82)90026​-­3
Fordyce, W. E. (1976). Behavioral methods for chronic pain and illness. Kabat-­Zinn, J., & Hanh, T. N. (2009). Full catastrophe living: Using
Mosby. the wisdom of your body and mind to face stress, pain, and illness.
Garcia-­Campayo, J., Navarro-­ Gil, M., Andrés, E., Montero-­ Marin, Delta.
J., López-­Artal, L., & Demarzo, M. M. (2014). Validation of the Kabat-­Zinn, J., Lipworth, L., Burncy, R., & Sellers, W. (1986). Four-­
Spanish versions of the long (26 items) and short (12 items) forms year follow-­up of a meditation-­based program for the self-­regulation
of the Self-­Compassion Scale (SCS). Health and Quality of Life of chronic pain: Treatment outcomes and compliance. The Clinical
Outcomes, 12, 4. https://doi.org/10.1186/1477-­7525-­12-­4 Journal of Pain, 2, 159–­774. https://doi.org/10.1097/00002​508-­
García-­Campayo, J., Rodero, B., Alda, M., Sobradiel, N., Montero, J., 19860​2030-­00004
& Moreno, S. (2008). Validación de la versión española de la es- Khoo, E.-­L., Small, R., Cheng, W., Hatchard, T., Glynn, B., Rice, D.
cala de la catastrofización ante el dolor (Pain Catastrophizing Scale) B., Skidmore, B., Kenny, S., Hutton, B., & Poulin, P. A. (2019).
en la fibromialgia. Medicina Clínica, 131, 487–­492. https://doi. Comparative evaluation of group-­based mindfulness-­based stress
org/10.1157/13127277 reduction and cognitive behavioural therapy for the treatment and
Garcia-­ Palacios, A., Herrero, R., Belmonte, M. A., Castilla, management of chronic pain: A systematic review and network
D., Guixeres, J., Molinari, G., Baños, R. M., & Botella, C. meta-­analysis. Evidence Based Mental Health, 22, 26–­35. https://
(2014). Ecological momentary assessment for chronic pain doi.org/10.1136/ebmen​tal-­2018-­300062
in fibromyalgia using a smartphone: A randomized crossover Killham, M. E., Mosewich, A. D., Mack, D. E., Gunnell, K. E., & Ferguson,
study. European Journal of Pain, 18, 862–­ 872. https://doi. L. J. (2018). Women athletes’ self-­ compassion, self-­
criticism, and
org/10.1002/j.1532-­2149.2013.00425.x perceived sport performance. Sport, Exercise, and Performance
Gardner-­Nix, J. (2009). Mindfulness-­Based Stress Reduction for chronic Psychology, 7, 297. https://doi.org/10.1037/spy00​00127
pain management. In F. Didonna (Ed.), Clinical handbook of mind- Kirby, J. N., Tellegen, C. L., & Steindl, S. R. (2017). A meta-­analysis
fulness (pp. 369–­381). Springer. of compassion-­ based interventions: Current state of knowledge
Germer, C., & Neff, K. (2019). Teaching the Mindful Self-­Compassion and future directions. Behavior Therapy, 48, 778–­792. https://doi.
program: A guide for professionals. Guilford Publications. org/10.1016/j.beth.2017.06.003
14
|    TORRIJOS-­ZARCERO et al.

Kovacs, F., & Moix, J. (2011). Manual del dolor: Tratamiento cognitivo of Clinical Psychology, 69(1), 28–­ 44. https://doi.org/10.1002/
conductual del dolor crónico. Grupo Planeta. jclp.21923
La Cour, P., & Petersen, M. (2015). Effects of Mindfulness meditation Odou, N., & Brinker, J. (2014). Exploring the relationship between ru-
on chronic pain: A randomized controlled trial. Pain Medicine, 16, mination, self-­compassion, and mood. Self and Identity, 13, 449–­
641–­652. https://doi.org/10.1111/pme.12605 459. https://doi.org/10.1080/15298​868.2013.840332
Luciano, J. V., Guallar, J. A., Aguado, J., López-­del-­Hoyo, Y., Olivan, Otis, J. D. (2007). Managing chronic pain: A cognitive-­behavioral ther-
B., Magallón, R., Alda, M., Serrano-­ Blanco, A., Gili, M., & apy approach. Workbook. Oxford University Press.
­Garcia-­Campayo, J. (2014). Effectiveness of group acceptance and Purdie, F., & Morley, S. (2016). Compassion and chronic pain. Pain,
commitment therapy for fibromyalgia: A 6-­month randomized con- 157, 2625–­2627. https://doi.org/10.1097/j.pain.00000​00000​000638
trolled trial (EFFIGACT study). Pain, 155, 693–­702. https://doi. Rusu, A. C., Pincus, T., & Morley, S. (2012). Depressed pain patients
org/10.1016/j.pain.2013.12.029 differ from other depressed groups: Examination of cognitive con-
MacBeth, A., & Gumley, A. (2012). Exploring compassion: A meta-­ tent in a sentence completion task. Pain, 153, 1898–­1904. https://
analysis of the association between self-­compassion and psycho- doi.org/10.1016/j.pain.2012.05.034
pathology. Clinical Psychology Review, 32, 545–­552. https://doi. Shapiro, S. L., Carlson, L. E., Astin, J. A., & Freedman, B. (2006).
org/10.1016/j.cpr.2012.06.003 Mechanisms of mindfulness. Journal of Clinical Psychology, 62,
McCormack, H. M., Horne, D. J. L., & Sheather, S. (1988). Clinical ap- 373–­386. https://doi.org/10.1002/jclp.20237
plications of visual analogue scales: A critical review. Psychological Sharpe, L. (2020). “First do no harm”: Why donʼt we measure ad-
Medicine, 18, 1007–­1019. https://doi.org/10.1017/s0033​29170​ verse events routinely in psychological treatment trials for people
0009934 with chronic pain? Pain, 161, 666–­667. https://doi.org/10.1097/j.
McCracken, L. M., & Eccleston, C. (2005). A prospective study of ac- pain.00000​00000​001771
ceptance of pain and patient functioning with chronic pain. Pain, Sirois, F. M., Kitner, R., & Hirsch, J. K. (2015). Self-­compassion, affect,
118(1–­2), 164–­169. https://doi.org/10.1016/j.pain.2005.08.015 and health-­promoting behaviors. Health Psychology, 34(6), 661–­
McCracken, L. M., & Turk, D. C. (2002). Behavioral and cognitive–­ 669. https://doi.org/10.1037/hea00​00158
behavioral treatment for chronic pain: Outcome, predictors of out- Smith, J. A., & Osborn, M. (2007). Pain as an assault on the self: An in-
come, and treatment process. Spine, 27, 2564–­2573. https://doi. terpretative phenomenological analysis of the psychological impact
org/10.1097/00007​632-­20021​1150-­00033 of chronic benign low back pain. Psychology & Health, 22, 517–­
McCracken, L. M., Vowles, K. E., & Eccleston, C. (2004). Acceptance 534. https://doi.org/10.1080/14768​32060​0941756
of chronic pain: Component analysis and a revised assessment Stellar, J. E., & Keltner, D. (2014). Compassion. In M. M. Tugade, M.
method. Pain, 107(1–­2), 159–­166. https://doi.org/10.1016/j. N. Shiota, & L. D. Kirby (Eds.), Handbook of positive emotions (pp.
pain.2003.10.012 329–­341). Guilford Press.
McCraken, L. M. (2005). Contextual cognitive-­behavioral therapy for Sturgeon, J. A. (2014). Psychological therapies for the management of
chronic pain. Progress in pain research and management (Vol. 33). chronic pain. Psychology Research and Behavior Management, 7,
IASP Press. 115–­124. https://doi.org/10.2147/PRBM.S44762
Menéndez, A. G., García, P. F., & Viejo, I. T. (2010). Aceptación del Sullivan, M. J. L., Bishop, S. R., & Pivik, J. (1995). The pain catastroph-
dolor crónico en pacientes con fibromialgia: adaptación del Chronic izing scale: Development and validation. Psychological Assessment,
Pain Acceptance Questionnaire (CPAQ) a una muestra española. 7, 524–­532. https://doi.org/10.1037/1040-­3590.7.4.524
Psicothema, 22, 997–­1003. Terry, M. L., Leary, M. R., Mehta, S., & Henderson, K. (2013). Self-­
Moher, D., Hopewell, S., Schulz, K. F., Montori, V., Gøtzsche, P. C., compassionate reactions to health threats. Personality and Social
Devereaux, P. J., Elbourne, D., Egger, M., & Altman, D. G. (2012). Psychology Bulletin, 39, 911–­ 926. https://doi.org/10.1177/01461​
CONSORT 2010 explanation and elaboration: Updated guidelines 67213​488213
for reporting parallel group randomised trials. International Journal Toye, F., Seers, K., Allcock, N., Briggs, M., Carr, E., Andrews, J., &
of Surgery, 10, 28–­55. https://doi.org/10.1016/j.ijsu.2011.10.001 Barker, K. (2013). Patients’ experiences of chronic non-­malignant
Montero-­Marín, J., Navarro-­Gil, M., Puebla-­Guedea, M., Luciano, J. V., musculoskeletal pain: A qualitative systematic review. British
Van Gordon, W., Shonin, E., & García-­Campayo, J. (2018). Efficacy Journal of General Practice, 63(617), e829–­ e841. https://doi.
of “Attachment-­Based Compassion Therapy” in the treatment of fi- org/10.3399/bjgp1​3x675412
bromyalgia: A randomized controlled trial. Frontiers in Psychiatry, Turner, J. A., Anderson, M. L., Balderson, B. H., Cook, A. J., Sherman,
8, 307. https://doi.org/10.3389/fpsyt.2017.00307 K. J., & Cherkin, D. C. (2016). Mindfulness-­based stress reduc-
Morley, S., Eccleston, C., & Williams, A. (1999). Systematic review tion and cognitive-­behavioral therapy for chronic low back pain:
and meta-­analysis of randomized controlled trials of cognitive be- Similar effects on mindfulness, catastrophizing, self-­efficacy, and
haviour therapy and behaviour therapy for chronic pain in adults, acceptance in a randomized controlled trial. Pain, 157, 2434–­2444.
excluding headache. Pain, 80, 1–­13. https://doi.org/10.1016/S0304​ https://doi.org/10.1097/j.pain.00000​00000​000635
-­3959(98)00255​-­3 Vallejo, M. A., Rivera, J., Esteve-­Vives, J., & Rodríguez-­Muñoz, M. F.
Neff, K. D. (2003a). Self-­compassion: An alternative conceptualization (2012). Uso del cuestionario Hospital Anxiety and Depression Scale
of a healthy attitude toward oneself. Self and Identity, 2, 85–­101. (HADS) para evaluar la ansiedad y la depresión en pacientes con
https://doi.org/10.1080/15298​86030​9032 fibromialgia. Revista de Psiquiatría y Salud Mental, 5, 107–­114.
Neff, K. D. (2003b). The development and validation of a scale to mea- https://doi.org/10.1016/j.rpsm.2012.01.003
sure self-­compassion. Self and Identity, 2, 223–­ 250. https://doi. Veehof, M. M., Oskam, M.-­J., Schreurs, K. M. G., & Bohlmeijer, E. T.
org/10.1080/15298​86030​9027 (2011). Acceptance-­based interventions for the treatment of chronic
Neff, K. D., & Germer, C. K. (2013). A pilot study and randomized pain: A systematic review and meta-­analysis. Pain, 152, 533–­542.
controlled trial of the Mindful Self-­Compassion program. Journal https://doi.org/10.1016/j.pain.2010.11.002
TORRIJOS-­ZARCERO et al.    
| 15

Veehof, M. M., Trompetter, H. R., Bohlmeijer, E. T., & Schreurs, K. comparative trial. The Clinical Journal of Pain, 27, 724–­734. https://
M. G. (2016). Acceptance-­and mindfulness-­based interventions for doi.org/10.1097/ajp.0b013​e3182​183c6e
the treatment of chronic pain: A meta-­analytic review. Cognitive Wren, A. A., Somers, T. J., Wright, M. A., Goetz, M. C., Leary, M.
Behaviour Therapy, 45, 5–­ 31. https://doi.org/10.1080/16506​ R., Fras, A. M., Huh, B. K., Rogers, L. L., & Keefe, F. J. (2012).
073.2015.1098724 Self-­compassion in patients with persistent musculoskeletal pain:
Vilagut, G., Ferrer, M., Rajmil, L., Rebollo, P., Permanyer-­Miralda, G., Relationship of self-­compassion to adjustment to persistent pain.
Quintana, J. M., & Alonso, J. (2005). El Cuestionario de Salud SF-­ Journal of Pain and Symptom Management, 43, 759–­770. https://
36 español: Una década de experiencia y nuevos desarrollos. Gaceta doi.org/10.1016/j.jpain​symman.2011.04.014
Sanitaria, 19, 135–­150. https://doi.org/10.1590/S0213​-­91112​00500​ Zeger, S. L., & Qaqish, B. (1988). Markov regression models for time
0200007 series: A quasi-­Likelihood Approach. Biometrics, 44, 1019. https://
Ware Jr. J. E., & Sherbourne, C. D. (1992). The MOS 36-­item short-­ doi.org/10.2307/2531732
form health survey (SF-­36). I. Conceptual framework and item se- Zhang, J. W., & Chen, S. (2016). Self-­compassion promotes personal
lection. Medical Care, 30, 473–­483.Retrieved from https://www. improvement from regret experiences via acceptance. Personality
ncbi.nlm.nih.gov/pubme​d/1593914 and Social Psychology Bulletin, 42, 244–­ 258. https://doi.
Wicksell, R. K., Kemani, M., Jensen, K., Kosek, E., Kadetoff, D., org/10.1177/01461​67215​623271
Sorjonen, K., Ingvar, M., & Olsson, G. L. (2013). Acceptance Zigmond, A. S., & Snaith, R. P. (1983). The Hospital Anxiety and
and commitment therapy for fibromyalgia: A randomized con- Depression Scale. Acta Psychiatrica Scandinavica, 67, 361–­370.
trolled trial. European Journal of Pain, 17, 599–­611. https://doi. https://doi.org/10.1111/j.1600-­0447.1983.tb097​16.x
org/10.1002/j.1532-­2149.2012.00224.x
Williams, A. C. D. C., Eccleston, C., & Morley, S. (2012). Psychological
therapies for the management of chronic pain (excluding headache) How to cite this article: Torrijos-­Zarcero M,
in adults. Cochrane Database of Systematic Reviews, 11, CD007407. Mediavilla R, Rodríguez-­Vega B, et al. Mindful
https://doi.org/10.1002/14651​858.CD007​407.pub3 Self-­Compassion program for chronic pain patients: A
Wong, S.-­Y.-­S., Chan, F.-­W.-­K., Wong, R.-­L.-­P., Chu, M.-­C., Kitty
randomized controlled trial. Eur J Pain. 2021;00:1–­15.
Lam, Y.-­Y., Mercer, S. W., & Ma, S. H. (2011). Comparing the
effectiveness of Mindfulness-­ based Stress Reduction and multi-
https://doi.org/10.1002/ejp.1734
disciplinary intervention programs for chronic pain: A randomized