International Journal of Rheumatic Diseases 2010; 13: 324–334

REVIEW ARTICLE

Psychosocial interventions as adjunct therapy for patients

with rheumatoid arthritis: a systematic review
Rukmal Kumari DISSANAYAKE1* and James V. BERTOUCH2
1
Department of Medicine, Concord General Repatriation Hospital, Concord; and 2Department of Rheumatology, Prince of Wales
Hospital, Randwick, NSW, Australia

Abstract
Aim: To identify the psychological interventions for which there is consistent, high quality evidence of
efficacy in the treatment of patients with rheumatoid arthritis (RA).
Method: A computer-aided search and manual screening of identified papers was conducted. Randomised
controlled trials published in English in peer-reviewed journals, assessing the use of psychological interven-
tions in adult patients with RA were included.
Results: Thirty-four papers published between 1981 and 2009 encompassing 31 studies with 2021 patients
were included. There is consistent supportive evidence for the efficacy of disclosure therapy (four studies) and
cognitive behavioural therapy (CBT) with maintenance therapy (five studies). There is supportive evidence for
improvement with CBT of greater than 6 weeks duration (six studies) in the short-term but conflicting evi-
dence for its long-term efficacy. There is some evidence for improvement with biofeedback-based interven-
tions (two studies). There is conflicting evidence for the benefits of counselling (three studies), psychotherapy
(two studies) mindfulness and meditation (two studies), and CBT of less than 6 weeks duration (six studies).
There is limited evidence regarding relaxation therapy (two studies). Methodological limitations of the
reviewed literature included failure of allocation concealment, blinding and conduction of intention-to-treat
analysis, as well as the heterogeneity and choice of outcome measures.
Conclusions: This review shows consistent supportive evidence for the use of disclosure therapy, and CBT
with maintenance therapy as adjunct therapies in patients with RA. It also highlights methodological limita-
tions in the current literature and the need for future research in this area.
Key words: alternative medicine, complementary, rheumatoid arthritis.

biopsychosocial basis of the disease.1 For example,

INTRODUCTION
Rheumatoid arthritis (RA) is an autoimmune inflam-
matory process which causes chronic pain and dis-
ability. Anecdotally patients describe stressful life
events being exacerbating factors for joint symptoms,
which is in keeping with Engels’ 1977 paper on the
Correspondence: Dr Rukmal Kumari Dissanayake, BSc(Med),
MBBS, 45 Shoalhaven Rd, Sylvania Waters, NSW 2224,
Australia. Email: kumari_d_@hotmail.com
*Institution in which work completed: Prince of Wales Hos-
pital, Randwick.
McFarlane and Brooks2 considered the effect of psy-
chological status on flares of symptoms in patients
with RA and found that psychological factors consis-
tently predicted more of the variance in disability
than did disease activity.
Psychological interventions for chronic disorders
have been considered in two recent Cochrane reviews
for asthma and metastatic breast cancer.3,4 Both
reviews were however unable to draw a conclusion
due to lack of published data and the poor quality of
available trials.

ª 2010 The Authors

International Journal of Rheumatic Diseases
ª 2010 Asia Pacific League of Associations for Rheumatology and Blackwell Publishing Asia Pty Ltd

While there are similar problems in the quality of
data for RA, Astin et al.5 performed a meta-analysis of
psychological interventions for RA. They considered
25 trials and found a pooled positive effect at post-
intervention for pain, functional disability, psychologi-
cal status, coping and self-efficacy, and at follow-up,
significant improvement on joint tenderness, psycho-
logical status and coping. Psychological interventions
are a very heterogenous group, and thus the aim of
this review is to identify individual interventions for
which there is strong high-quality evidence.
METHODS
Randomised controlled trials considering psychologi-
cal interventions beyond education in adults with RA,
published in English in peer-reviewed journals were
considered for inclusion (Table 1).
A computer-aided search of MEDLINE, EMBASE and
the Cochrane Central Register of Controlled Trials util-
ising the search strategy in Table 2 was conducted.
This was followed by screening references from the
identified papers.
All clinical trials identified were reviewed indepen-
dently by the authors (JB and RD) to assess patient
selection and inclusion criteria. Consensus was used
to resolve differences.
Assessment of the methodological quality of the tri-
als included an assessment of selection bias, appropri-
ate randomization, performance bias, attrition bias
and detection bias and the application of a methodo-
logical score.
The score used by Ostello et al.6 in their Cochrane
review of behavioural interventions for chronic back
Table 1 Inclusion criteria
Types of studies
Randomised controlled trials published in English in a
peer-reviewed journal
Types of participants
Patients with confirmed rheumatoid arthritis
Control group
Appropriate control group included standard medical care;
wait list or attention control
Types of intervention
Any type of psychological intervention used in the
treatment of rheumatoid arthritis2
Outcome measures to be considered
Pain, biochemical and clinical markers of disease,
disability, mood and cognition, behaviour and patient
satisfaction
International Journal of Rheumatic Diseases 2010; 13: 324–334
Psychosocial interventions for rheumatoid arthritis
Table 2 Search strategy
Medline
1. Psychotherapy/or holistic health/or mind-body therapies/
or aromatherapy/or biofeedback, psychology/or breathing
exercises/or hypnosis/or ‘‘imagery (psychotherapy)’’/or
laughter therapy/or meditation/or mental healing/or
‘‘mind-body relations (metaphysics)’’/or psychodrama/
or psychophysiology/or relaxation therapy/or tai ji/or
therapeutic touch/or yoga/or psychology
2. Arthritis/or arthritis, rheumatoid/
3. 1 and 2
Embase
1. Psychotherapy/or meditation/or behavior therapy/or
cognitive therapy/or aromatherapy/or spiritual healing/or
patient counseling/
2. Rheumatoid arthritis/
3. 1 and 2
pain was applied. This assessment takes into consider-
ation the difficulty in blinding patients and caregivers
to these types of interventions. The score detailed in
Table 3 was assessed by JB and RD independently.
One point was given for a positive response, zero for
a negative or unknown response. Consensus was used
to resolve differences.
As per Ostello et al.6 consistent findings were
defined as 75% or more of the studies having statisti-
cally significant findings in the same direction. Sup-
portive evidence was defined as several high-quality
randomized controlled trials (RCTs) (score > 6) with
consistent findings. Limited evidence was defined as
one high-quality RCT or several low-quality RCTs
showing supportive evidence. Conflicting evidence was
defined as inconsistent findings among multiple RCTs.
No evidence meant that no RCT was available.
RESULTS
In this search 119 papers were identified {50 MED-
LINE + 22 COCHRANE + 47 EMBASE}. Of these 75
were clinical trials of psychological interventions in
patients with RA; 34 papers encompassing 31 studies
that met the criteria for inclusion (Fig. 1) were identi-
fied. The studies included 2021 patients in total and
were published between 1981 and 2009.
Twenty-two studies included by Astin et al.5 were
considered in this review, while three were excluded.
Two of the excluded articles were due to patient inclu-
sion criteria (one included patients with back pain
[Flor], and another with juvenile arthritis [Lavigne],
325
R. K. Dissanayake and J. V. Bertouch
Table 3 Methodological score
Was the method of randomization adequate?
A random (unpredictable) assignment sequence
Was the treatment allocation concealed?
Assignment generated by an independent person not
responsible for determining the eligibility of the patients.
This person has no information about the persons included
in the trial and has no influence on the assignment
sequence or on the decision about eligibility of the patient
Were the groups similar at baseline regarding the most important
prognostic indicators?
In order to receive a ‘‘yes,’’ groups have to be similar at
baseline regarding demographic factors, duration and
severity of complaints, ESR/CRP, joint count, and value of
main outcome measure(s)
Was the patient blind to the intervention?
Was the care provider blind to the intervention?
Was the outcome assessor blind to the intervention?
Was co-intervention avoided or similar?
Was the compliance acceptable in all groups?
50% attendance in 75% of the group was acceptable
Was the drop-out rate described and acceptable?
If the percentage of withdrawals and drop-outs does not
exceed 20% for short-term follow-up and 30% for
long-term follow-up and does not lead to substantial bias
a ‘‘yes’’ is scored
Was the timing of the outcome assessment in all groups similar?
Did the analysis include an intention to treat analysis?
Was their adequate follow-up?
4-month follow-up
ESR, erythrocyte sedimentation rate; CRP, C-reactive protein.
and the third [Lindroth] involved an educational
intervention. One additional paper (Poulson) that
was published within the Astin search period and 11
further papers published after the Astin paper were
included in the 34 studies that were analyzed.
As the types of interventions varied greatly, the
studies were divided based on their theoretical base.
The types of interventions included biofeedback (2),
relaxation training (2), disclosure (4), counselling
(3), psychotherapy (2), meditation and mindfulness
(2) and cognitive behavioural therapy (CBT) (16). As
more than half of the studies used CBT-based inter-
ventions, this group was further divided according to
the duration of treatment as short (< 6 weeks) (6),
long (> 6 weeks) (6) and CBT with maintenance
therapy throughout the follow-up period (5). The
studies, interventions and results are detailed in
Table 4.
Biofeedback involves providing patients with access
to physiological information in order to gain control
326
of processes under autonomic control. Two of the
34 papers, Appelbaum et al.7 and Bradley et al.8 exam-
ined a thermal biofeedback-based therapy. Bradley
et al.8 also included a social support arm. In thermal
biofeedback, patients were encouraged to relax in
order to increase skin temperature through vasodila-
tion. There was supportive evidence for these interven-
tions in the short term with improvement in pain or
pain behaviour noted, but no effect demonstrated at
follow-up. Both trials also used several treatment
modalities.
Relaxation therapy was considered in two papers.
Lundgren9 studied imagery techniques and Van Deu-
sen and Harlow10 examined a Range of Motion Dance
Program focusing on rest and exercise balance. Both
studies showed some significant improvement. How-
ever, as there were a large number of outcome mea-
sures with only a few showing significant effects, there
is limited evidence to assess the efficacy of relaxation
therapy.
Disclosure therapy was examined in four papers:
Keefe et al.11 Kelly et al.12 Middendorp et al.13 and
Smyth et al.14 All studies examined private disclosure
either verbally or written; Keefe et al.11 also consid-
ered physician-assisted disclosure. Three of the four
trials were of high quality. Immediately following the
intervention, three studies found increased negative
emotions. At follow-up three studies found significant
improvements. Kelly et al.12 performed a mixed-
model analysis of variance (ANOVA) and found signifi-
cant group linear effect on physical dysfunction
(F1,66 = 3.96, P = 0.05) and affective disturbance
(F1,66 = 4.99, P = 0.029) on the Arthritis Impact Mea-
surement Scale (AIMS-2). Middendorp et al.13 found
improved insight (P = 0.003), and positive language
through the course of the intervention. Smyth et al.14
found a 28% (P = 0.001) mean reduction in disease
severity (0–4) as assessed by a rheumatologist
blinded to the treatment group. However, Keefe
et al.11 performed a follow-up at 15 months and
found no effect at that time. Thus with three out of
four high-quality studies (methodological score
greater than size) showing some significant effects,
there is consistent supportive evidence for disclosure
therapy among the trials available in the short term.
Counselling was considered in three studies. Two of
these, Kaplan and Kozin15 and Maisaik et al.,16 used
non-directive counselling and one, Maisaik et al.,17 a
more formal structured approach. Interestingly Maisaik
et al.16 used counsellors with inflammatory arthritides.
Overall, with two supporting and one refuting study
International Journal of Rheumatic Diseases 2010; 13: 324–334
 Psychosocial interventions for rheumatoid arthritis

Identification

Medline n = 228

Embase n = 331 Reference review n = 7

Cochrane n = 22

Screening

340 Individual records screened 221 Records excluded

119 Abstracts obtained and 44 Records excluded

screened

Eligibility

75 Full text articles assessed 41 Articles excluded

for eligibility
n = 18 Not RCTs

n=3 Individual data for RA

n=6 Inappropriate contents

n = 13 Interventions not
Included
primarily psychological

n=1 Published in a language

other than English
34 Papers included in analysis

Figure 1 Summary of study selection.

at post-intervention, there is conflicting evidence for
efficacy in the short term. Two of the studies, Kaplan
and Kozin15 and Maisaik et al.,16 failed to perform
any follow-up and thus there is no available evidence
at follow-up for counselling.
Psychotherapy aims to alleviate difficulties through
the development of a therapeutic relationship to
encourage insight, and was examined in two papers,
Poulson18 and Strauss et al.19 Analytic psychotherapy
focuses on the unconscious mind and the importance
of past experiences in shaping current thinking. With
two low-quality trials, one showing some effect and
the other no effect, there is conflicting evidence for
the use of psychotherapy in RA.
Mindfulness and meditation were examined in two
trials, Pradan et al.20 and Zautra et al.,21 combining

International Journal of Rheumatic Diseases 2010; 13: 324–334 327

 Table 4 Studies examined

328
Intervention Studies (methodological No. Control Duration of Length of Evidence at Evidence at follow-up Level of
score); intervention patients group intervention follow-up post-treatment Statistically significant evidence
(months) Statistically significant findings if any
findings if any
Biofeedback Appelbaum et al.7 (4); relaxation 18 WL 6 weeks 18 Decrease in pain, affective No significant changes Some supportive
training, thermal biofeedback and response to pain, evidence but no
cognitive pain management strategies improved communication effect at follow-up
and ROM
Bradley et al.8 (7); thermal 68 SC ** 6 CBT: decreased pain CBT: decreased anxiety
biofeedback, group rlaxation training, behavior, rheumatoid only
behavioural training in goal setting activity score and anxiety.
and self reward. Active control with Attention control:
R. K. Dissanayake and J. V. Bertouch

social support decreased anxiety

Relaxation Lungren and Stenström9 (8); taped 68 SC 10 weeks 12 Improved self-care, 6 months: improves Limited evidence
therapy program of muscle relaxation and recreation and pastimes mobility, balance
imagery techniques for pain reduction co-ordination.;
12 months: no
significant difference
Van Deusen and Harlow10 (5); ROM 46 WL 8 weeks 4 Standard joint geniometry Improved ROM
Dance Program integrating principles performed and found shoulder, total upper
of Occupational Therapy and Tái Chi improved ankle plantar extremity combined
Chúan and discussion sessions regaring flexion and lower flexion and extension
compliance, enjoyment, coping, group extremity flexion.
interation, body awareness and Increased enjoyment of
experience of wellbeing exercise and rest
Disclosure Keefe et al.11 (3); private disclosure: 98 AC 3 weeks 15 Disclosure-produced 2 months: less pain Consistent
audio-recorded sessions discussing change in affect behavior in the supportive
stressful life events. clinician assisted; (negative), which was private disclosure evidence at
with a nurse trained in active listening reduced if nurse present group compared with follow-up
present the clinician. No
other group effect.
Long-term: no effect
Kelly et al.12 (7); personal disclosure to 79 AC 4 day 3 Increased dejection, Decreased physical
an audio recorder about a trauma or depression, anger, dysfunction, decreased
upheaval that has been experienced, hostility, fatigue, inertia affective
performed at home disturbance
Middendorp et al.13 (7); personal 72 AC 4 weeks 2 Higher immediate negative Optimism and positive
disclosure to an audio recorder affect, use of emotion, future directedness
performed at home insight in treatment
group
Smyth et al.14 (9); written disclosure 51 AC 3 day 4 No significant effect No changes at
about the most stressful experience 2 weeks/2 months,
they had ever undergone performed in but at 4 months
private rooms in the laboratory 47.1% of treatment
group had improved
disease activity
(0 = asymptomatic
to 4 = severe)

International Journal of Rheumatic Diseases 2010; 13: 324–334

 Table 4 (continued)
Intervention Studies (methodological No. Control Duration of Length of Evidence at Evidence at follow-up Level of
score); intervention patients group intervention follow-up post-treatment Statistically significant evidence
(months) Statistically significant findings if any
findings if any
Counselling Kaplan and Kozin15 (7); non-directive 34 WL 20 weeks – Improved realistic goals, No significant changes Conflicting
client centered group counselling improved expectations evidence with
and improved knowledge no effect at
Maisaik et al.16 (6); telephone 219 SC/AC 3 months 9 Improved affect, improved Not included in paper follow-up
conversation based counseling physical function. No
change in pain
Maisaik et al.17 (7); person centered, 60 SC 6 months – Non-significant effect Non-significant effect
non directive reflective telephone
based counselling
Psychotherapy Poulson18 (4); highly structured 28 SC 24 weeks 9 Improved psychic Improved psychic Conflicting
modified group analytic wellbeing and decreased wellbeing and decreased evidence
psychotherapy levels of alexithymia levels of alexithymia
Strauss et al.19 (5); conventional 57 SC/AC 6 months 6 No significant effect No significant effect
psychotherapy in group meetings
Mindfulness Pradan et al.20 (8); training and 63 WL 6 weeks 6 No significant differences Improvement in psycho- Conflicting
Meditation discussion about of mindfulness, logical distress and wellbe- evidence
experimental training in meditation ing, depressive symptoms

International Journal of Rheumatic Diseases 2010; 13: 324–334

and gentle yoga and mindfulness
Zautra et al.21 (8); CBT;coping skills, 144 AC 8 weeks 6 CBT: improved pain, and Not provided
alternate pain management IL-6 levels. Both active
approaches, memory/concentration, groups: improved coping
relapse prevention. Mindfulness; the efficacy. In patients with
bi-dimensional model of emotions, previous depression
acceptance, negative thought framing, improved affect and joint
positive emotions tenderness
CBT short DeVellis et al.22 (6); CBT to manage 111 SC 1 session 4 Treatment group improved ALL patients improved in Conflicting
problems raised in psychosocial in compliance and physical and evidence
interview and generic problem lifestyle problems psychological functioning Conflicting
solving skills evidence
Freeman et al.23 (6); Group based CBT 64 AC 4 weeks 6 Control: significantly less No significant changes
with samily members. The functional disability, less
intervention utilised the health belief helplessness. No change
model and the theory of self-efficacy with intervention
O’Leary et al.24 (6); group sessions 33 AC 5 weeks 1 Significant decrease in Not conducted
focussing on coping skills, several pain. Improved self-
cognitive and behavioural pain efficacy in management
management strategies, goal setting of pain, function and
general effects of RA
Radojevic et al.25 (6); Behavioural 65 SC/AC 4 weeks 2 Behavioral therapy Both behavioural therapy
therapy the subjects were taught pain containing programs groups improvements in
management skills, one arm added decreased swelling and swelling and joint count.
family support joint count. Family No difference with family
support yielded the best support
results
Psychosocial interventions for rheumatoid arthritis

329
330
Table 4 (continued)
Intervention Studies (methodological No. Control Duration of Length of Evidence at Evidence at follow-up Level of
score); intervention patients group intervention follow-up post-treatment Statistically significant evidence
(months) Statistically significant findings if any
findings if any
Scholten et al.26 (4); multidisciplinary 68 WL 2 weeks 12 Improved disability. increased Improved disability
education with coping strategies, coping, distraction, increased coping,
Jacobson stress management and knowledge and compliance. distraction, knowledge and
relaxation exercises with an emphasis Decreased depression compliance. Decreased
on control and efficacy depression
Taal et al.27 (5); The program included 140 SC 5 weeks 14 Improved functional disability, 4 months: positive effect on
R. K. Dissanayake and J. V. Bertouch

contracting, goal setting and performance of relaxation functional disability, joint

feedback; self management and exercises, physical exercise, tenderness, 4 months/
problem solving; communication self-management activities, 12 months: improvement
skills; coping with depression outcome expectations, performance of physical
self-efficacy and knowledge exercises, self-efficacy,
function and knowledge
CBT long Germond et al.28 (5); Stress 24 AC 8 weeks – No significant effect Not conducted Supportive
inoculation and Pain Management evidence in
Training the short term
Kraamiaat et al.29 (8); self- 77 WL/AC 10 weeks 6 Increased knowledge in CBT CBT: more distraction with But Conflicting
management of active coping and AC, CBT more pleasant activities. Disease evidence at
behaviour, training of coping distraction with pleasant progression in all groups follow-up
strategies such as progressive activities
relaxation, rational thinking, active
coping behaviours and goal setting
Leibing et al.30 (7); The CBT consisted 63 SC 12 weeks 9 CBT decreased depression, Not included in this paper
of information and education about helplessness and affective
the gate theory of pain, relaxation pain. Increased positive
and imagery, changing maladaptive reappraisal, positive
thinking and negative emotions, acceptance and resignation
pleasant activity scheduling
Sharpe et al.31 (8); Sharp et al.32; 53 SC 8 weeks 5 years Significantly decreased 6 months and 18 months:
Sharp et al.33 (8); The cognitive depression, reinterpretation decreased depression,
behavioural intervention was of pain, CRP anxiety and disability. Joint
developed from standard pain involvement maintained
management approaches and self BUT SC group also
help educational material improved 5 years: decreased
total health care use
Shearn and Fireman34 (5); identify 105 SC/AC 8 weeks 8 Both intervention groups did No significant effect
sources of stress as well as to learn better. Patients who scored
relaxation techniques and strategies highest tended to respond to
for coping. Attentional control; any treatment (regression
mutual support therapy to the mean)

International Journal of Rheumatic Diseases 2010; 13: 324–334

 Table 4 (continued)
Intervention Studies (methodological No. Control Duration of Length of Evidence at Evidence at follow-up Level of
score); intervention patients group intervention follow-up post-treatment Statistically significant evidence
(months) Statistically significant findings if any
findings if any
CBT with Carson et al.35 (8); Subjects taught 1) 167 SC/AC 10 weeks 18 Decreased joint pain, less Decreased joint pain and Consistent
maintenance attention diversion 2) activity negative mood, increased improved coping efficacy supportive
modification 3) recognition and problem-focused coping, MT made no change evidence in
changing pain related negative coping efficacy, positive the short and
cognitions; Maintenance: 3 sessions mood compared with long term
over 12 weeks standard care
Evers et al.39 (7); CBT focusing on 1) 278 SC 10 weeks 6 Decreased fatigue, Decreased fatigue,
pain and function 2) fatigue 3) depression, anxiety, and depression,
negative moods 4) social. The control helplessness + increased helplessness + increased
group received standard medical care. active coping perceived social support
Single booster 4 weeks post
intervention
Hammond et al.40 (6); behavioural 218 AC 9 months 12 Behavioural group had Better pain, self-efficacy
joint protection, ergonomic improvement in pain, and psychological status

International Journal of Rheumatic Diseases 2010; 13: 324–334

approaches to managing symptoms, fatigue, functional ability,
managing stress and coping with self-efficacy and greater
negative thoughts. Review at use of health behaviors
3–9 months
Parker et al.36 (7); Short hospital stay 83 SC 1 week 12 No data given 6 months/12 months:
addressing coping strategies, problem improved coping
solving techniques, relaxation strategies, decrease pain.
training, awareness of pain High adherence group:
behaviours, strategies of attention decreased pain, Arthritis
diversion followed by a support Helplessness Index
group meetings 1–3 monthly
Parker et al.37/Multon et al.38 (8); The 141 SC/AC 10 weeks 15 Decreased daily stressors, Changes maintained and
program included relaxation training helplessness and decrease in pain,
and instruction in CBT strategies to increased self-efficacy, improved AIMS-2, no
manage daily stresses. This was coping strategies, change in pain behavior
followed with 3 monthly confidence in the ability
maintenance program for 15 months to manage pain, no
change in pain behavior

WL, wait list; SC, standard medical care; AC, active control; **, not disclosed; CRP, C-reactive protein; ROM, range of movement; AIMS, Arthritis Impact Measurement Scale.
Psychosocial interventions for rheumatoid arthritis

331
R. K. Dissanayake and J. V. Bertouch
mindfulness and meditation techniques. Zautra et al.21
also included a CBT arm. At post-intervention, Pradan
et al.20 found no effect, while Zautra et al.21 found
both the CBT and meditation group had improved
coping efficacy, with the CBT group also showing
improvements in pain and in interleukin (IL)-6 levels.
At follow-up Pradan et al.20 showed some improve-
ments, while Zautra et al.21 showed no effect. Thus,
for meditation interventions there is conflicting
evidence at post-intervention and follow-up.
Cognitive behavioral therapy is based on the theo-
retical rationale that a patient’s mood and behaviour
are determined by the way in which they structure
the world, based on assumptions formed by past
experience.
Cognitive behavioral therapy uses cognitive tech-
niques of situation and assumption appraisal, in
which patients identify thoughts, assess their validity,
and try to replace them with more realistic and posi-
tive viewpoints. CBT also encourages altered cogni-
tions that influence behaviour and encourages more
mindful behavior, such as balancing pleasant and
unpleasant tasks and prioritising.
As 16 trials considered CBT these were further
divided into subgroups based on the length of inter-
vention. CBT-based intervention of shorter then
6 weeks duration were examined in six papers.22–27
Overall, there is conflicting evidence for short CBT
intervenions.
Five studies28–34 examined CBT interventions of
8 and 12 weeks duration. Overall there was
supportive evidence for these interventions in the
short term, but conflicting evidence for their long-term
efficacy.
Five studies considered the efficacy of CBT with a
maintenance program. Four interventions were long
courses of outpatient CBT, while one was an intensive
inpatient stay. Three were group34–40 and two were
individual interventions39,40. All were of high quality.
Evers et al.39 found at both post-intervention and
follow-up a significant improvement in fatigue
(P < 0.01), depression (P < 0.05), social functioning
(P = 0.05) on self-assessed questionnaires. Hammond
et al.40 compared CBT to education alone and found
that at 6 and 12 months the CBT with maintenance
group had lower levels of pain as measured on a
visual analog scale (P < 0.01) and self-efficacy on a
self-administered questionnaire (P < 0.01). Parker
et al.36 found that compared to standard care CBT
with maintenance resulted in improvement in the
Coping Strategies Questionnaire score (P = 0.0001);
332
this finding was particularly so in a small highly
adherent group. Parker et al.37 found that following
the active intervention, patients had improved pain
symptom scores, and lower extremity scores on the
Arthritis Impact Measurement Score, and reduction in
the daily Stress Index. Of these the lower extremity
scores on the AIMS was maintained. Carson et al.35
was the only study that compared CBT and CBT with
maintenance and found that the CBT group showed
improved pain (P = 0.0002), coping strategies
(P < 0.0001) and mood (P = 0.0001) as measured via
a daily questionnaire for 30 days at each endpoint,
compared with standard care, with no further
improvement in the maintenance group. Overall, there
appears to be supportive evidence compared with
standard care found consistently among the high-
quality trails available.
In addition to the methodological scores seen in
Table 4, an assessment of selection bias, performance
bias and attrition bias was conducted.
Aspects of selection bias assessed were allocation
concealment and randomization. Only eight studies
out of the 31 provided information of adequate allo-
cation concealment. This fact is reflected in the quality
scores. Of the studies only 20 stated the participation
rate and whether all the eligible patients were
approached. The method of randomization was fre-
quently not mentioned in the papers, with only 13
studies adequately discussing this.
Performance bias was assessed by considering blind-
ing of patients and caregivers. Sixteen studies utilized
attention controls, which involved education interven-
tions or symptom monitoring, and five wait-list con-
trols. As expected, none of the interventions were able
to blind treatment providers.
Attrition bias was assessed by the drop-out rate and
the handling of these losses during the study. Drop-
out rates ranged from 0% to 44% in these studies.
Only eight studies performed an intention-to-treat
analysis.
Detection bias was assessed through blinding of the
outcome assessor and choice of outcome measures.
Thirteen studies included blinding of the outcome
assessor.
The studies considered a wide variety of outcomes
and utilised numerous outcome measures. Each trial
considered between one and seven outcomes with 4.7
being the average. The outcomes included pain,
biochemical and clinical markers of disease severity,
functional disability, mood, cognition and behaviour
and patient satisfaction. Between one and twelve
International Journal of Rheumatic Diseases 2010; 13: 324–334

outcome measures were utilised, with 6.5 being the
average.
DISCUSSION
Based on the published literature, there is consistent
supportive evidence for two types of psychological
interventions in patients with RA. These were disclo-
sure therapy and CBT of > 6 weeks duration with
maintenance therapy. However, even in these areas
there are only small numbers of trials published. More
research is required to provide concrete answers about
the many other types of psychological interventions
available.
While it is possible to divide the psychological inter-
ventions considered according to their theoretical base,
even within these groups there is considerable hetero-
geneity. In studies that utilised interventions similar to
those already published, there was often little attempt
to utilise the trial methods of previous work, or pro-
vide comparisons between their study and previous
work. This factor limits the applicability of this review,
as each of the subgroups themselves had considerable
heterogeneity.
The lack of comparison between similar interven-
tions is particularly problematic in considering CBT
with maintenance therapy. All trials were of high qual-
ity and overall provided strong evidence for efficacy of
these interventions. However, only one study, Carson
et al.35 compared CBT with maintenance to CBT
alone. This study demonstrated that maintenance did
not seem to add to the efficacy of the intervention.
For this reason future trials examining CBT with main-
tenance need to ensure that two groups, one with and
one without maintenance, are included in addition to
standard medical care or attention controls.
Future studies in this area also need to address meth-
odological limitations evident in the current published
literature. Aspects that need to be considered include
the reporting of allocation concealment and randomi-
zation methods, as well as recruiting larger numbers of
patients to ensure that studies have sufficient power to
detect significant differences between interventions. It
is also imperative that intention-to-treat analysis is per-
formed on all data and that this is published. Where
possible the outcome assessors should be blinded to
interventions, particularly for assessment of clinical
indices of disease severity. Clearly this would not be
possible for psychological indices.
Overall the 34 papers included eight outcomes and
a total of 74 different outcome measures. Many of
International Journal of Rheumatic Diseases 2010; 13: 324–334
Psychosocial interventions for rheumatoid arthritis
these outcomes and outcome measures were not spe-
cifically reported, and most questionnaires, particularly
those measuring disability, were not reported upon as
a whole but as specific components. The use of such a
large number of outcome measures makes comparison
of results very difficult. Furthermore, the use of multi-
ple outcomes in each study, when reporting only
those outcomes showing positive results, leads to
potential reporting bias. Given the large number of
outcome measures, and varying statistical analyses
used by the papers, only a qualitative rather than
quantitative analysis was performed.
In conclusion, out of the numerous types of psycho-
logical interventions that can be utilised in clinical
practice, many areas need further study and the meth-
odological limitations identified in this paper need to
be appropriately addressed.
REFERENCES
1 Engel GL (1977; Apr 8) The need for a new medical
model: a challenge for biomedicine. Science 196 (4286),
129–36.
2 McFarlane AC, Brooks PM (1987) Determinants of dis-
ability in RA. Br J Rheumatol 26, 101–8.
3 Yorke J, Fleming SL, Shuldham CM (2006) Psychological
interventions for adults with asthma. Cochrane Database
Syst Rev (1), CD 002982.
4 Edwards AGK, Hulbert-Williams N, Neal RD (2004) Psy-
chological interventions for women with metastatic
breast cancer. Cochrane Database Syst Rev (2), CD
004253.
5 Astin JA, Beckner W, Soeken K, et al. (2002) Psychologi-
cal interventions for rheumatoid arthritis: a meta-analysis
of randomised controlled trials. Arthritis Rheum 47 (3),
291–302.
6 Ostelo RWJG, van Tulder MW, Vlaeyen JWS, et al. (2005)
Behavioural treatment for chronic low-back pain. Cochrane
Database Syst Rev (1), CD 002014.
7 Appelbaum KE, Blanchard EB, Hickling EJ, Alfonso M
(1988) Cognitive behavioural treatment of a vetran pop-
ulation with moderate to severe rheumatoid arthritis.
Behav Ther 19, 489–502.
8 Bradley LA, Young LD, Anderson KO, et al. (1987) Effects
of behavioural therapy on pain behaviour of rheumatoid
arthritis patients. Arthritis Rheum 30 (10), 1105–15.
9 Lundgren S, Stenström CH (1999) Muscle relaxation
training and quality of life in rheumatoid arthritis: a
randomised controlled clinical trial. Scand J Rheumatol
28, 47–53.
10 Van Deusen J, Harlow D (1987) The efficacy of the ROM
dance program for adults with rheumatoid arthritis. Am J
Occup Ther 41 (2), 90–5.
333
R. K. Dissanayake and J. V. Bertouch
11 Keefe F, Anderson T, Lumley M, et al. (2008) A rando-
mised controlled trial of emotional disclosure: can clini-
cian assistance enhance effect? Pain 37 (1), 164–72.
12 Kelley JE, Lumley MA, Leisen JCC (1997) Health effects
of emotional disclosure in rheumatoid arthritis patients.
Health Psychol 16 (4), 331–40.
13 Middendorp H, Sorbi M, van Doornen L, et al. (2007)
Feasibility and induced cognitive-emotional change of an
emotional disclosure intervention adapted for home
application. Patient Educ Couns 66, 177–87.
14 Smyth JN, Stone AA, Hurewitz A, Kaell A (1998) ) Affects
of writing about stressful experience on symptom reduc-
tion in patients with asthma or rheumatoid arthritis.
JAMA 281 (14), 1304–9.
15 Kaplan S, Kozin F (1981) A controlled study of group
counselling in rheumatoid arthritis. J Rheumatol S1, 91–9.
16 Maisiak R, Austin JS, West SG, Heck L (1996) The affect
of person centred counselling on the psychological status
of persons with systemic lupus erythematosus or rheuma-
toid arthritis. Arthritis Care Res 9 (1), 60–6.
17 Maisiak R, Austin J, Heck L (1996) Health outcomes of
two telephone interventions for patients with rheumatoid
arthritis or osteoarthritis. Arthritis Rheum 39 (8), 1391–9.
18 Poulsen A (1991) Psychodynamic, time-limited therapy
in rheumatic disease – a controlled study with special ref-
erence to Alexithymia. Psychother Psychosom 56, 12–23.
19 Strauss GD, Spiegel JS, Daniels M, et al. (1986) Group
therapies for rheumatoid arthritis: a controlled study of
two approaches. Arthritis Rheum 29 (10), 1203–9.
20 Pradan E, Baugarten M, Langenberg P, et al. (2007) Effect
of a mindfulness based stress reduction in rheumatoid
arthritis patients. Arthritis Rheum 57 (7), 1134–42.
21 Zautra A, Davis M, Reich J, et al. (2008) Comparison of
cognitive behavioural and mindfulness meditation inter-
ventions on adaptation to rheumatoid arthritis for
patients with and without history of recurrent depression.
J Consult Clin Psychol 76 (3), 408–12.
22 Devellis BE, Blalock SJ, Hanh PM, et al. (1988) Evalua-
tion of a problem solving intervention for patients with
rheumatoid arthritis. Patient Educ Couns 11 (29), 29–42.
23 Freeman K, Hammond A, Lincoln NB (2002) Use of cog-
nitive behavioural arthritis education programs in newly
diagnosed rheumatoid arthritis. Clin Rehabil 16, 828–36.
24 O’Leary A, Shoor S, Lorie K, Holman HR (1988) A cogni-
tive behavioural treatment for rheumatoid arthritis.
Health Psychol 7 (6), 527–44.
25 Radojevic V, Nicasso PM, Weisman MH (1992) Behavio-
ural intervention with and without family support for
rheumatoid arthritis. Behav Ther 23, 13–30.
26 Taal E, Riemsama RP, Brus HL, et al. (1993) Group edu-
cation for patients with rheumatoid arthritis. Patient Educ
Couns 20, 177–87.
334
27 Scholten C, Brodowicz T, Graninger W, et al. (1999)
Persistent functional and social benefits after a multidisci-
plinary arthritis training program. Arch Phys Med Rehabil
80, 1282–7.
28 Germond S, Schomer HH, Meyers OL, Weight L (1993)
Pain management in rheumatoid arthritis: a cognitive
behavioural intervention. S Afr J Psychol 23 (1), 1–8.
29 Kraamiaat FW, Brons MR, Geenen R, Bijlsma WJ (1995)
The effect of cognitive behaviour therapy in patients with
rheumatoid arthritis. Behav Res Ther 33 (5), 487–95.
30 Leibing ED, Pfingsten M, Bartmann UM, et al. (1999)
Cognitive-behavioural treatment in unselected rheuma-
toid arthritis outpatients. Clin J Pain 15 (1), 58–66.
31 Sharpe L, Sensky T, Timberlake N, et al. (2001) A blind,
randomised, controlled trial of cognitive-behavioural
intervention for patients with recent onset rheumatoid
arthritis: preventing psychological and physical morbid-
ity. Pain 89, 275–83.
32 Sharpe L, Sensky T, Timberlake N, et al. (2003) Long
term efficacy of a cognitive behavioural treatment from a
randomised controlled trial for patients recently diag-
nosed with rheumatoid arthritis. Rheumatology 42, 435–
41.
33 Sharpe L, Allard S, Sensky T (2008) Five year follow-up
of a cognitive-behavioural intervention for patients with
recently diagnosed rheumatoid arthrtiris: effects on health
care utilisation. Arthritis Rheum 59 (3), 311–6.
34 Shearn MA, Fireman BH (1985) Stress management and
mutual support groups in rheumatoid arthritis. Am J Med
78, 771–5.
35 Carson JW, Keefe FJ, Affleck G, et al. (2006) A compari-
son of conventional pain coping skills training and pain
coping skills training with a maintenance component: a
daily diary analysis of short and long term treatment
effects. J pain 7 (9), 615–25.
36 Parker JC, Frenk RG, Beck NC, et al. (1988) Pain man-
agement in rheumatoid arthritis patients: a cognitive
behavioural approach. Arthritis Rheum 31 (5), 593–601.
37 Parker JC, Smarr KL, Bukelew SP, et al. (1995) Affects of
stress management on clinical outcomes in rheumatoid
arthritis. Arthritis Rheum 38 (12), 1807–18.
38 Multon KD, Parker JC, Smarr KL, et al. (2001) Affects of
stress management on pain behaviour in rheumatoid
arthritis. Arthritis Care Res 44, 122–8.
39 Evers AWM, Kraaimaat FW, vanRiel PLCM, deJong ALJ
(2002) Tailored cognitive behavioural therapy in early
rheumatoid arthritis for patients at risk: a randomised
controlled trial. Pain 100, 141–53.
40 Hammond A, Bryan J, Hardy A (2008) Effects of a modu-
lar behavioural arthritis education programme: a prag-
matic parallel-group randomized controlled trial.
Rheumatology 47, 1712–8.
International Journal of Rheumatic Diseases 2010; 13: 324–334