Intervention Review

You have free access to this content

Non-pharmacological interventions for fatigue in rheumatoid arthritis

  1. Fiona Cramp1,*,
  2. Sarah Hewlett1,
  3. Celia Almeida1,
  4. John R Kirwan2,
  5. Ernest HS Choy3,
  6. Trudie Chalder4,
  7. Jon Pollock1,
  8. Robin Christensen5

Editorial Group: Cochrane Musculoskeletal Group

Published Online: 23 AUG 2013

Assessed as up-to-date: 1 OCT 2012

DOI: 10.1002/14651858.CD008322.pub2


How to Cite

Cramp F, Hewlett S, Almeida C, Kirwan JR, Choy EHS, Chalder T, Pollock J, Christensen R. Non-pharmacological interventions for fatigue in rheumatoid arthritis. Cochrane Database of Systematic Reviews 2013, Issue 8. Art. No.: CD008322. DOI: 10.1002/14651858.CD008322.pub2.

Author Information

  1. 1

    University of the West of England, Faculty of Health & Life Sciences, Bristol, UK

  2. 2

    University of Bristol, Bristol Royal Infirmary, Rheumatology Unit, Bristol, UK

  3. 3

    Cardiff University School of Medicine, Section of Rheumatology, Department of Medicine, Cardiff, UK

  4. 4

    Institute of Psychiatry, King's College London, Department of Psychological Medicine, London, UK

  5. 5

    Copenhagen University Hospital, Frederiksberg, Copenhagen, Denmark, Musculoskeletal Statistics Unit (MSU), The Parker Institute, Dept Rheumatology, Copenhagen, Denmark

*Fiona Cramp, Faculty of Health & Life Sciences, University of the West of England, Glenside campus, Blackberry Hill, Bristol, BS16 1DD, UK. fiona.cramp@uwe.ac.uk.

Publication History

  1. Publication Status: New
  2. Published Online: 23 AUG 2013

SEARCH

 
Characteristics of included studies [ordered by study ID]
Bilberg 2005

MethodsRCT: 2 arms


ParticipantsRA for 1-5 years, 3 months stable medication, functional class I, II or III, age 20-65


InterventionsExercise in temperate pool twice a week in groups of 8-9: moderate intensity, paced by music, 2 alternating physiotherapists 


OutcomesPrimary: aerobic capacity and SF-36 physical component

Relevant secondary: disability (HAQ)

Other: disease activity (DAS28), arthritis impact measurement scale (AIMS2), health related quality of life (SF-36), Upper and lower extremity muscle endurance, function, Quality of life, active forward and lateral rotation of shoulder, hand grip force


ExclusionsOther severe diseases or functional limitations with pool training


Fatigue outcomesSF36 Vitality, High=good


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskOptimal allocation with a computer program

Allocation concealment (selection bias)Unclear riskNot stated

Blinding (performance bias and detection bias)
All outcomes
Unclear riskNot stated for personnel or participants. A physiotherapist blinded to study arm membership conducted the examinations

Incomplete outcome data (attrition bias)
All outcomes
Low riskData analysis carried out by protocol

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasHigh riskSome missing data. Two control patients exercised in the pool group

Burgos 2009

MethodsRCT: 2 arms


Participants18-70 years, active RA (at least 1 swollen joint, ESR >20mm/h or CRP >20mg/dl), MTX partial responders


Interventions30 mg andrographolides tablet 3 times/day for 14 weeks


OutcomesPrimary outcomes: reduction of joint pain, stiffness and tiredness

Relevant secondary: disability (HAQ)

Other: immunological parameters, health-related quality of life (SF36), rheumatoid factor, ESR, CRP, complete blood counts, X-rays (chest, hands and feet), venous blood for safety evaluations


ExclusionsOther non-degenerative/other joint diseases that could interfere with RA evaluation (gout, chondrocalcinosis, psoriatic arthritis, infectious arthritis, reactive or spondylitic arthritis), women with child-bearing potential, pregnant or breast feeding and patients with severely limiting arthritis that rendered them subject to surgery or severely crippling or prostrated patients. Use of intra-articular steroids the month before enrolment, concomitant treatment with hydantoin, lithium, or anticoagulant drugs. History of peptic ulcer, or gastrointestinal bleeding in the preceding 6 months. History of hypersensitivity or adverse effects to NSAIDs. Renal failure, hepatic failure, severe heart failure, haematologic disease, history of alcohol or drug abuse, participation in other research in the preceding month, patients not willing to attend regular check ups.


Fatigue outcomesVAS tiredness, range 0-100mm, High = bad


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"random codes with a permuted randomisation scheme were generated by computer"

Allocation concealment (selection bias)Low risk"Two envelopes contained each individual's treatment assignment"

Blinding (performance bias and detection bias)
All outcomes
Unclear riskPatients and personnel not aware of assignment, no details re outcome assessors

Incomplete outcome data (attrition bias)
All outcomes
Low riskITT

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskNo evidence of other bias identified

Danoff-Burg 2006 - I

MethodsRCT: 3 arms


ParticipantsRA/Systemic Lupus Erythematosus (SLE), able to write for 20 minutes in English


InterventionsExpressive writing: encouraged participants to express their "deepest thoughts and feelings" about their experience with RA/SLE

Benefit finding: focus on "any positive thoughts and feelings" regarding their illness experience of RA/SLE

ALL arms had 3 week intervention, with 4 separate writing sessions within the 3-week period. Participants wrote alone in a private room in the laboratory for 20 minutes in accordance with their assigned instructions.


OutcomesPrimary: not stated

Relevant outcomes: Disability (HAQ), depression (CES-D), pain (VAS), anxiety (STAI)

Other: Profile of Mood States (POMS)


ExclusionsNone stated


Fatigue outcomesa) POMS Vigor 0-32, high=good
b) 100mm VAS: fatigue= no problem-major problem, High=bad  


NotesAuthors state positive mood measured by POMS vigor - it has 2 subscales: vigor/energy and fatigue. They measure mood so vigour is about feeling full of pep (energy)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear risk"randomly assigned" but no further details given

Allocation concealment (selection bias)Low risk"participants were given envelopes containing instructions that randomly assigned them to one of the three conditions"

Blinding (performance bias and detection bias)
All outcomes
High riskexperimenter/raters unaware of arm assignment, not possible to blind participants

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk2 participants did not complete due to time constraints: not clear how their results were handled

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskNo evidence of other bias identified

Danoff-Burg 2006 - II

MethodsRCT: 3 arms


ParticipantsRA/Systemic Lupus Erythematosus (SLE), able to write for 20 minutes in English


InterventionsExpressive writing: encouraged participants to express their "deepest thoughts and feelings" about their experience with RA/SLE

Benefit finding: focus on "any positive thoughts and feelings" regarding their illness experience of RA/SLE

ALL arms had 3 week intervention, with 4 separate writing sessions within the 3-week period. Participants wrote alone in a private room in the laboratory for 20 minutes in accordance with their assigned instructions.


OutcomesPrimary: not stated

Relevant outcomes: Disability (HAQ), depression (CES-D), pain (VAS), anxiety (STAI

Other: Profile of Mood States (POMS)


ExclusionsNone stated


Fatigue outcomesa) POMS Vigor 0-32, high=good
b) 100mm VAS: fatigue= no problem-major problem, High=bad  


NotesAuthors state positive mood measured by POMS vigor - it has 2 subscales: vigor/energy and fatigue. They measure mood so vigour is about feeling full of pep (energy)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear risk"randomly assigned" but no further details given

Allocation concealment (selection bias)Low risk"participants were given envelopes containing instructions that randomly assigned them to one of the three conditions"

Blinding (performance bias and detection bias)
All outcomes
High riskexperimenter/raters unaware of arm assignment, not possible to blind participants

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk2 participants did not complete due to time constraints: not clear how their results were handled

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskNo evidence of other bias identified

Das Gupta 2009

MethodsRCT: 2 arms


ParticipantsRA, disease duration >6 months, DAS 28 ≥3.2


InterventionsMain text indicates: 25mg indomethacin plus 3g omega-3 fatty acid. Abstract indicates: 75mg of indomethacin. Author contacted for clarification but no response received.


OutcomesPrimary outcome not identified

Relevant measures: disability (HAQ), tender joint count, swollen joint count, pain

Other: disease activity (DAS28), duration of morning stiffness, grip strength, ESR, health-related quality of life (SF-36)


ExclusionsMedical diseases including metabolic, endocrine, hepatic, renal, haematological, pulmonary, cardiovascular, or any other disease or condition that required active medical attention


Fatigue outcomesSF-36 vitality, range of scale 0-100, High = good


Notes1. Abstract indicates 75mg of indomethacin but main text suggests 25mg

2. No mention of SF-36 in methods but results reported

3. SDs for change scores not provided


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo details provided

Allocation concealment (selection bias)Unclear riskNo details provided

Blinding (performance bias and detection bias)
All outcomes
High riskParticipants in each arm received a different number of tablets

Incomplete outcome data (attrition bias)
All outcomes
High riskReasons for dropout not reported, no mention of how missing data were handled

Selective reporting (reporting bias)High riskSF-36 not listed in methods but results provided

Other biasHigh riskNo details of demographic information including disease duration and gender; possible errors in manuscript presentation, no information on adverse effects

Evans 2012

MethodsRandomised wait list control: 2 arms


ParticipantsRA for ≥ 6 months, age 16-35 years, stable medication for ≥ 4 weeks, ability to provide written informed consent, speak and understand English


InterventionsIvengar yoga with up to 7 people per class, lead by an experienced teacher and assisted by at least one junior teacher. 6 weeks (twice per week) for 1.5 hours per session (18 hours total)


OutcomesPrimary: not stated

Relevant outcomes: adverse events, Pain (pain disability index), Disability (HAQ), Anxiety and Depression (Brief Symptom Inventory)

Other: QoL (SF-36), disease activity (DAS), RA symptoms (Global Improvement scale), chronic pain acceptance, mindfulness, arthritis self-efficacy


ExclusionsPregnant, recently experienced an injury, history of drug or alcohol abuse, on any experimental medications in the previous 6 months


Fatigue outcomesSF-36 and FACIT-F


NotesStudy details obtained from unpublished report


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear risk'randomized in blocks' but no details of sequence generation

Allocation concealment (selection bias)Unclear risk‘randomised to either yoga or wait list control’ with no details of allocation concealment provided

Blinding (performance bias and detection bias)
All outcomes
High risk‘Principal investigators were blinded to participant randomization during the study process.’ No details provided for participant blinding or blinding of outcome assessors

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Selective reporting (reporting bias)Low risk

Other biasHigh riskWait list control design may have biased participants. Statistically significant difference between arms in time since diagnosis

Evers 2002

MethodsRCT: 2 arms


ParticipantsACR criteria for RA, ≥18 years, disease duration <8 years, screened as "at risk" i.e. high levels of negative mood/anxiety, helplessness and lower acceptance


InterventionsTailored Cognitive Behavioural Therapy (CBT): 2/4 possible treatment models chosen based on patient priorities (pain and functional disability; fatigue; negative mood; social relationships). Homework assignments 30 minutes/day


OutcomesPrimary outcomes: physical, psychological and social functioning

Relevant secondary: functional disability (Impact of Rheumatic Diseases on General Health and Lifestyle - IRLG), pain (IRGL-pain scale), anxiety (IRLG - anxiety scale), depression (Beck Depression Inventory)

Others: disease activity (DAS: ESR and clinical joint score); negative mood; social function; illness cognition questionnaire - helplessness and acceptance; Utrechtse Coping Lijst; Pain Coping inventory; problem focusing scale; avoidance scale; compliance with RA medication


ExclusionsCo-morbidities e.g. malignancy, cardiac, respiratory, hepatic and renal problems that may affect CBT


Fatigue outcomesChecklist Individual Strengths (CIS) Fatigue Scale, 8 items, 1=True -> 7=Not true measures level of fatigue over previous 2 weeks, High=high fatigue


NotesGender - Table 1 does not state if % refer to women, but it is confirmed in text


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskPreviously determined pattern of random numbers

Allocation concealment (selection bias)Unclear riskNot stated

Blinding (performance bias and detection bias)
All outcomes
High riskNo blinding of participants was reported and the nature of the intervention indicates that blinding was not undertaken

Incomplete outcome data (attrition bias)
All outcomes
Low risk"All analyses were conducted with completers as well as with intention-to-treat (ITT) analyses, using the last observation carried forward method"

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskITT and analysis of completers revealed same result. Differences between arms at baseline accounted for in analysis

Furst 1987

MethodsRCT: 2 arms


Participants18 years or over, RA diagnosis (ARA) for a minimum of 1 year, not hospitalised, permission to participate from general practitioner


InterventionsNew National Institutes of Health (NIH) Energy Conservation (EC) program: patient workbook: body position, rest, activity analysis and joint protection with didactic information, illustrations, practice assignments and self-evaluation.  Uses visual learning through text/illustrations, auditory learning through discussion with instructor, group and family. Homework assignments to reinforce learning


OutcomesPrimary outcome: not stated

Relevant outcomes: disability (HAQ), pain (HAQ VAS/Activity Record - ACTRE)

Other outcomes: psychosocial adjustment to illness (PAIS), knowledge (T/F questionnaire), disease activity (Richey-Camp Articular Index), Occupational Questionnaire adapted to measure behavioural objectives resulted in NIH Activity Record (ACTRE): log of activities for each 1/2hr of day over 2 days. Pain, fatigue, rest and PA measured as % of waking hours


ExclusionsInpatients, wheelchair-bound, temporary mobility limitations, currently receiving EC training


Fatigue outcomesa) VAS (added to HAQ), possible range 0-3, high=bad    b) ACTRE, During this time (activity) I felt fatigue 1=Not at all, 2=Very little, 3=Some, 4=A lot  per half hour, high=bad


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"pre-randomised treatment assignment lists for each centre were generated by computer"

Allocation concealment (selection bias)Low risk"randomly assigned to experimental or control groups". Patient ID numbers keyed to the centres

Blinding (performance bias and detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
High riskPost-test data not provided. No mention of participants who left the study early. No 9 month data available

Selective reporting (reporting bias)High riskFull results not reported for all measures

Other biasHigh riskNot clear who delivered the interventions or performed the outcome measures. Number of sessions attended and adherence not reported. ACTRE was not a validated measure

Giraudet-Le-Quintrec 2007

MethodsRCT: 2 arms


ParticipantsRA


InterventionsBoth arms received 2 information leaflets written by the research team

Interactive multi-disciplinary education program plus usual medical care: information on the disease, treatment, diet, coping strategies, relaxation, exercise program to be followed at home. Self-efficacy principles to reduce pain and stress at home, behaviour modification techniques to change behaviours/improve quality of life by modifying psychological and social contacts


OutcomesPrimary outcome: health and functional status (HAQ)

Relevant secondary: Anxiety and Depression (Hospital Anxiety and Depression Scale - HADS)

Other: disease activity (DAS28), Arthritis Helplessness Index (AHI) score for Coping, Quality of life (EMIR17: score for the short form of the Arthritis Impact Measurement Scale - AIMS2), physical activity (Baecke questionnaire), drug compliance, satisfaction with the program, exercise and leisure physical activity compliance


ExclusionsJuvenile chronic arthritis, Steinbrocker class 4, pregnancy, RA flare, patient unable to understand information 


Fatigue outcomesFACIT-F, high=bad


Notes Mayoux-Benhamou is further reference report


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"The allocation sequence was generated by randomly placing thoroughly shuffled marked cards into sequentially numbered sealed, opaque envelopes. This process was carried out by a statistics assistant not involved in the trial"

Allocation concealment (selection bias)Low risk"The allocation sequence was generated by randomly placing thoroughly shuffled marked cards into sequentially numbered sealed, opaque envelopes. This process was carried out by a statistics assistant not involved in the trial"

Blinding (performance bias and detection bias)
All outcomes
High risk12-month evaluation performed by 3 independent rheumatologists blind to arm allocation, not possible to blind participants

Incomplete outcome data (attrition bias)
All outcomes
Low risk"When possible (i.e. at least an available baseline value for the outcome) we handled missing data by the last observation carried forward method"

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskNo evidence of other bias identified

Hakkinen 2003

MethodsRCT: 2 arms


ParticipantsRecent-onset RA
(ACR 1987 criteria)


InterventionsIndividually designed home based dynamic strength training program. A physiotherapist initially guided the program during a 5 day inpatient period. Strength training included all the main muscle groups using dumb bells and elastic bands. Intensity was re-evaluated every 6 months


OutcomesPrimary outcome: functional capacity (Valpar whole body range of motion work sample and the HAQ)

Relevant secondary: Pain during transfers

Other: muscle strength, physical loading of work, disease activity (ESR, Ritchie index, Larsen score, pain), Transfer time


ExclusionsNone stated


Fatigue outcomes100mm VAS during Valpar transfers, 0-100mm, high=bad


NotesPossible contamination: control arm instructed to carry out exercises without any additional resistance and to continue recreational exercise with the exception of strength training. However, the only monitoring of this was self-report diaries and it is therefore possible that the control arm undertook resistance training. It is also possible that the intervention arm did not carry out the exercises at the intensity prescribed.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomisation performed using clusters of 4 stratified by age and sex but does not state how the sequence was generated

Allocation concealment (selection bias)Unclear riskNot stated

Blinding (performance bias and detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
Low riskIntention to treat

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskNo evidence of other bias identified

Hakkinen 2003-II

MethodsRCT: 2 arms


ParticipantsRecent-onset RA
(ACR 1987 criteria)


InterventionsIndividually designed home based dynamic strength training program. A physiotherapist initially guided the program during a 5 day inpatient period. Strength training included all the main muscle groups using dumb bells and elastic bands. Intensity was re-evaluated every 6 months


OutcomesPrimary outcome: functional capacity (Valpar whole body range of motion work sample and the HAQ)

Relevant secondary: Pain during transfers

Other: muscle strength, physical loading of work, disease activity (ESR, Ritchie index, Larsen score, pain), Transfer time


ExclusionsNone stated


Fatigue outcomes100mm VAS during Valpar transfers, 0-100mm, high=bad


NotesPossible contamination: control arm instructed to carry out exercises without any additional resistance and to continue recreational exercise with the exception of strength training. However, the only monitoring of this was self-report diaries and it is therefore possible that the control arm undertook resistance training. It is also possible that the intervention arm did not carry out the exercises at the intensity prescribed.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomisation performed using clusters of 4 stratified by age and sex but does not state how the sequence was generated

Allocation concealment (selection bias)Unclear riskNot stated

Blinding (performance bias and detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
Low riskIntention to treat

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskNo evidence of other bias identified

Hammond 2008

MethodsRCT: 2 arms


Participants≥18 years, RA or inflammatory arthritis. Psoriatic arthritis patients could also attend


InterventionsLifestyle Management for Arthritis Programme (LMAP): delivery by 3 therapists (2 occupational therapists and 1 physiotherapist) who had attended a training course and were observed with feedback to standardise delivery. Modules for patients = Looking after joints and keeping mobile and managing pain/mood. Modules involved self-monitoring, skills training, goal setting and action planning for home activity/exercise programmes


OutcomesPrimary outcome = Change in pain scores at 12 months

Relevant secondary: disability (modified HAQ)

Others: perceived health (very bad to very well), early morning stiffness, RA self-efficacy, arthritis self-efficacy scale, Arthritis Helplessness Index, self-management (Arthritis Stages of Change), visits to general practitioner over the past 6 months, medication 


ExclusionsSevere psychological difficulties/receiving mental health care, medical condition preventing safe exercise, previous education programme HAQ >2 indicating severe functional problems


Fatigue outcomesVAS: 100mm, No problem-major problem, high=bad


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"Randomised by a research administrator to the modular or standard programme using computer-generated random numbers"

Allocation concealment (selection bias)Low risk"Pre-prepared sealed, opaque, numbered envelopes"

Blinding (performance bias and detection bias)
All outcomes
High riskNot possible. Participants could not be blinded to programme allocation but programme differences were not discussed. Outcomes were self-report questionnaires

Incomplete outcome data (attrition bias)
All outcomes
Low riskIntention-to-treat analysis

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskNo evidence of other bias identified

Harkcom 1985

MethodsRCT: 4 arms


ParticipantsRA


InterventionsA: Bicycle ergometry at 70% HRmax. 5 minute bouts of exercise separated by 1 minute rest periods. Progressed to 15 minutes of exercise by week 6

B: Bicycle ergometry at 70% HRmax. 5 minute bouts of exercise separated by 1 minute rest periods. Progressed to 25 minutes of exercise by week 6

C: Bicycle ergometry at 70% HRmax. 5 minute bouts of exercise separated by 1 minute rest periods. Progressed to 35 minutes of exercise by week 6


OutcomesPrimary outcome: not stated

Relevant outcomes: patient perceptions of changes in amount of joint pain and tolerance of joint pain

Other: morning stiffness, sleep patterns, self care, ambulation, activities of daily living. Global assessment: patient perceptions of changes in strength, ability to do housework, participation in social activities, overall mood. Grip strength, 50 foot walk time, muscle strength (knee flexors and extensors, joint evaluation. Exercise tolerance, aerobic capacity


ExclusionsNone stated


Fatigue outcomesa) Perception of change in fatigue, Worse-same-better-much better

b) Time to fatigue during exercise - BUT this was physiological fatigue


NotesSample size too small for statistical test, possible arm contamination - control arm may have exercised


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot stated

Allocation concealment (selection bias)Low risk"Patients independently chose an exercise time slot convenient for them...without knowledge of the protocol to which they would be assigned. The 4 different protocols were, in turn, randomly allocated to these time slots"

Blinding (performance bias and detection bias)
All outcomes
High riskNot possible for participants

Incomplete outcome data (attrition bias)
All outcomes
High riskNo information on missing data for 3 participants who dropped out

Selective reporting (reporting bias)High riskNo control data for fatigue. Exercise arms are combined and data are not reported in sufficient detail

Other biasHigh riskSample size too small for the statistical tests performed. No control data. Possible contamination if patients in the control arm chose to exercise

Helliwell 1999

MethodsRCT: 2 arms


ParticipantsRA (1987 ARA criteria) for less than 5 years. Able to read and speak English


InterventionsEducation classes took place over 4 weeks in 4 afternoon sessions lasting 2 hours. Format was a talk from a non-medical health professional, a discussion period and distribution of supporting literature. Content: pathophysiology of RA, drug treatments, local treatments, mechanisms and control of pain, stress, exercise and rest, joint protection, task allocation, splinting and assistive equipment.


OutcomesPrimary: Radiological damage (Larsen score of hand and wrist)

Relevant Secondary: disability (HAQ)

Other: Quality of life (SF-36), patient knowledge questionnaire, compliance questionnaire, Ritchie articular index, plasma viscosity, admission and clinic attendance, drug consumption and changes


ExclusionsNone stated


Fatigue outcomesSF36 Vitality, High=good


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk‘Randomization codes were taken from a table of random numbers...’

Allocation concealment (selection bias)Low risk‘...concealed in consecutive, numbered, sealed envelopes.’

Blinding (performance bias and detection bias)
All outcomes
High riskx-rays were blinded and assessed by an independent person but there was no blinding of participants and blinding of personnel was not detailed

Incomplete outcome data (attrition bias)
All outcomes
Low riskIntention to treat

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasHigh riskNo attention control arm and no details of control arm in methods

Hewlett 2011

MethodsRCT: 2 arm


ParticipantsRA, scoring ≥6 on a VAS for fatigue during the past week


InterventionsCBT for fatigue self-management. Co-delivered by a clinical psychologist and OT. 6x2hour sessions during weeks 1-6 and a 1 hour consolidation session at week 14. Thoughts, feelings and behaviours related to fatigue were addressed using reflective questioning and guided discovery.  Problem solving, goal setting, self monitoring of rest/activity and activity management to improve well-being


OutcomesPrimary: Fatigue (MAF and VAS)

Relevant Secondary: pain (VAS), disability (HAQ), anxiety (HADS), depression (HADS)

Other: fatigue severity (VAS), fatigue coping (VAS), disease activity (VAS), personal impact (HAQ), QoL (RA QoL scale), helplessness (Arthritis Helplessness Index), self-efficacy (RA self-efficacy scale), sleep quality (binary scale)


ExclusionsChange in disease modifying drugs or biological agents in the previous 24 weeks or glucocorticoids in the previous 6 weeks


Fatigue outcomesMAF and VAS


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskComputer generated random numbers

Allocation concealment (selection bias)Low riskProject managers emailed participant codes to an epidemiologist who performed randomisation off site

Blinding (performance bias and detection bias)
All outcomes
High risk'It was not possible to blind participants and staff, but self report measures were analysed blind to arm allocation.'

Incomplete outcome data (attrition bias)
All outcomes
Low riskIntention-to-treat analysis

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskNo evidence of other bias identified

Huffstutter 2007

MethodsRCT: 2 arms


ParticipantsOver 18 years; capable of reading, understanding, and providing written informed consent forms; diagnosed with active RA; disease severe enough to warrant a visit to their rheumatologist at least every 3 months


InterventionsDuring the first 6 months of study participation (pre-randomisation), ALL enrolled patients and their physicians entered data into the Electronic Data Capture System at each visit; however, Health Tracker (HT) reports summarizing the patients’ and physicians’ assessments of the first 6 months of visits were not available to any patients or physicians.

For intervention arm only: Health Tracker (HT) reports integrated and summarized each patients data generated at specific study visits occurring 6 to 12 months after study entry, including graphical displays of the number of patient-defined painful joints, Health Assessment Questionnaire (HAQ) scores, Short Form (SF)-12 scores, physician and patient global assessment scores, and anti-rheumatic medications over time.


OutcomesPrimary outcomes: Patients overall satisfaction with 8 aspects of their care: VAS 0 cm (completely unsatisfied) to 10 cm (completely satisfied). Physicians: 6-item patient-physician interaction questionnaire using a VAS (lower scores = easier interactions)

Relevant secondary: pain (VAS), disability (HAQ), patients self-evaluated painful joints on a joint mannequin diagram in EDCS, and a trained assessor examined patients and separately entered the painful and swollen joint count in the EDCS.

Other: Global disease assessments by the patient and the physician. Health-related quality of life (SF-12)


ExclusionsCurrently receiving investigational drugs or participating in other clinical trials or safety registries for RA or RA-related therapies


Fatigue outcomesVAS, no fatigue to fatigue is a major problem/very severe, high=bad


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo details

Allocation concealment (selection bias)Unclear riskNo details

Blinding (performance bias and detection bias)
All outcomes
High riskParticipants were aware of the group allocation due to the nature of the intervention

Incomplete outcome data (attrition bias)
All outcomes
High riskNo mention of how missing participants' data were handled

Selective reporting (reporting bias)Unclear riskFatigue VAS data not reported in published paper

Other biasUnclear riskRandomised at 6 months but 0 months baseline measure used

Laforest 2008

MethodsRCT: 2 arms


ParticipantsHousebound (not leaving home independently more than twice a month), ≥50 years; reporting moderate to severe pain; suffering from OA or RA, reporting difficulties in performing domestic or daily living activities, able to speak English or French


InterventionsWeekly 1-hr individual home visits by a health care professional over 6 weeks. Participants explore themes of: (a) their life with arthritis, (b) physical and relaxation exercises, (c) managing pain and stiffness, (d) maintaining a positive attitude, (e) energy management, and (f) partnerships with health professionals and creative problem solving. Each session reviews previously discussed topics, explores a new theme, formulates a personal contract (i.e., defining an objective that is meaningful to the person and an action plan to meet it), and time for discussion and reflection. Session 1 planned to be 1.5 hours long, with subsequent sessions lasting 1 hour


OutcomesPrimary outcome: pain

Relevant secondary: depression (CES-D)

Others: Western Ontario and McMaster Universities Arthritis Index (functional limitations/stiffness), Arthritis Helplessness Index-brief, 4-item coping, arthritis self-efficacy scale, outcome expectations, frequency of physical/social activities, social life satisfaction 


ExclusionsPolymyalgia, recent health problems requiring rehabilitation services, cognitive problems, or previous participation in a similar intervention


Fatigue outcomesVAS: 100mm, No fatigue-fatigue as bad as it can be, high=bad


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"The randomization procedure used a list of numbers that predetermined group allocation according to the number of participants required in each group"

Allocation concealment (selection bias)Unclear riskNot stated

Blinding (performance bias and detection bias)
All outcomes
High risk"None of the interviewers were involved in the administration of the intervention and they were blind to participants' group assignment" BUT "some interviewers may have become aware of the assignment of participants to the experimental or control conditions during follow-up interviews"

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNot stated

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasHigh riskNo details of control arm

Lorig 2005

MethodsRCT: 2 arms


ParticipantsA diagnosis of any rheumatic condition supplied by physician, ≥18 years 


InterventionsPrograms held in community locations led by a pair of trained peer leaders, one/both have a rheumatic condition. Each program is attended by 10-15 individuals with arthritis as well as some participants' significant others


OutcomesPrimary outcome: Pain was used as the basis of the sample size calculation

Relevant secondary: disability (Health Assessment Instrument); Visual Numeric Pain

Other: Health Distress (worry/concern caused by chronic illness, 5 items); Self-Rated Global Health; Activity Limitations; number of times/week individuals practice mental stress management and relaxation techniques; self-reported exercise; self-efficacy for managing chronic disease; physician visits in past 4 month; hospitalizations in past 4 months


ExclusionsPrevious participation in the Arthritis Self-Management Program or the Chronic Disease Self-Management Program


Fatigue outcomesVNS for fatigue, probably no fatigue to worst fatigue imaginable (from Ref 25), high=bad


NotesCo-morbidities = 65.7% IV, 75.4% Ctrl


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"All participants were randomized into 2 groups using random number charts"

Allocation concealment (selection bias)Unclear riskNot stated

Blinding (performance bias and detection bias)
All outcomes
Low risk"Investigators were blinded to intervention assignment", self-report measures used

Incomplete outcome data (attrition bias)
All outcomes
Low riskAnalyses re-run with intention to treat and all missing data replaced by baseline values

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasHigh riskPossible contamination as recruited from Arthritis Foundation groups so patients randomized to different arms could have met regularly during the year

Lorig 2008

MethodsRCT: 2 arms


ParticipantsRA/osteoarthritis/fibromyalgia - self-report, ≥18 years, no previous Arthritis Self-Management Program participation, access to computer/email, agreement to 6 week commitment


Interventions25 patients per workshop, interactive web-based programme with 11 topics including bulletin board discussion, diaries, self-tests, arthritis handbook, online questionnaire. Aim=reduce pain and improve function


OutcomesPrimary outcome: Pain and function

Relevant secondary:disability (HQAQ-8 item)

Other: health distress, self-rated global health, disease impact (Activities Limitation Scale), health behaviour (stretching/strengthening, aerobic exercise, cognitive symptom management techniques, healthcare provider communication techniques, self-report outpatient visits), Arthritis Self-Efficacy Scale


ExclusionsActive cancer treatment in previous year


Fatigue outcomesVisual Numeric Scale 0-10, high=bad


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo details

Allocation concealment (selection bias)Unclear riskNo details

Blinding (performance bias and detection bias)
All outcomes
High riskParticipants were aware of arm allocation due to the nature of the intervention

Incomplete outcome data (attrition bias)
All outcomes
Low riskMissing data collected by phone, checked for differences in completers/non-completers at baseline

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasLow riskNo evidence of other bias identified

Lumley 2011-I

MethodsRCT: 2 x 2 factorial design


ParticipantsAdults meeting 1987 ACR criteria, reported pain or morning stiffness during the week prior to screening


InterventionsWritten emotional disclosure: one week duration, 4 writing sessions, blank journal for daily writing with specific instructions, intimal 20 minutes completed in clinic followed by 3 writing sessions at home for 20 minutes each. Participants identified a stressful life event to write about including thoughts and feelings about the event, meaning and effect upon beliefs or health and learning relating to coping that occurred as a result of the event.

 

Coping skills training: one session per week for 8 weeks individually with an advanced doctoral student or post doctoral fellow in clinical psychology for 75 minutes per session. Sessions were face to face where possible although some were carried out by phone. Therapist training consisted of a 2 day workshop. Training focused upon various cognitive behavioural skills.  Included education on gate control theory of pain, muscle relaxation, assertive communication skills, cognitive restructuring, activity-rest cycling, distraction, problem solving, relapse prevention.


OutcomesPrimary outcome: Disease activity (physician completed VAS) and pain (AIMS-2)

Relevant secondary: physical disability (AIMS-2), affective disturbance (AIMS-2)

Other: walking speed (timed 50ft walk), inflammation (C-reactive protein), intervention credibility (credibility scale)


ExclusionsOther autoimmune disorder, another disease that seriously affected function, illiterate or cognitively impaired, were in a formal behavioural pain management programme, had experienced a major stressful life change in the previous 6 months, unable to write or walk


Fatigue outcomesDaily diary NRS


NotesMajority of data were obtained directly from the author and were unpublished. Fatigue data were only available for the main effects RCT, that is: 1) all patients who undertook Written Emotional Disclosure versus Control Writing (irrespective of later training arms) and 2) all patients who undertook Coping Skills Training versus Education alone (irrespective of preceding writing arm)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk‘A person independent of the study staff used randomization software to develop the condition assignments’. Stratified by study site, medication class. Randomisation was done in blocks of 8, to one of 2 writing conditions followed by one of 2 training conditions

Allocation concealment (selection bias)Unclear risk‘assignments were sealed in envelopes’  does not state that they were opaque

Blinding (performance bias and detection bias)
All outcomes
High riskParticipants and study personnel were not blind to the interventions. There was blinded assessment of disease activity but the other measures were participant self-report

Incomplete outcome data (attrition bias)
All outcomes
Low risk'All analyses were conducted on the intent-to-treat sample of all 264 randomized participants. This is done by default in HLM, which estimates missing values; but for the ANCOVAs, we replaced those few missing values with the last available value (in most cases, using the baseline value for missing follow-up data).'

Selective reporting (reporting bias)Unclear riskNo daily outcome data were provided for fatigue, pain and positive and negative mood; states this will be presented in a separate paper

Other biasLow riskNo evidence of other bias identified

Lumley 2011-II

MethodsRCT: 2 x 2 factorial design


ParticipantsAdults meeting 1987 ACR criteria, reported pain or morning stiffness during the week prior to screening


InterventionsWritten emotional disclosure: one week duration, 4 writing sessions, blank journal for daily writing with specific instructions, intimal 20 minutes completed in clinic followed by 3 writing sessions at home for 20 minutes each. Participants identified a stressful life event to write about including thoughts and feelings about the event, meaning and effect upon beliefs or health and learning relating to coping that occurred as a result of the event.

 

Coping skills training: one session per week for 8 weeks individually with an advanced doctoral student or post doctoral fellow in clinical psychology for 75 minutes per session. Sessions were face to face where possible although some were carried out by phone. Therapist training consisted of a 2 day workshop. Training focused upon various cognitive behavioural skills. Included education on gate control theory of pain, muscle relaxation, assertive communication skills, cognitive restructuring, activity-rest cycling, distraction, problem solving, relapse prevention.


OutcomesPrimary outcome: Disease activity (physician completed VAS) and pain (AIMS-2)

Relevant secondary: physical disability (AIMS-2), affective disturbance (AIMS-2)

Other: walking speed (timed 50ft walk), inflammation (C-reactive protein), intervention credibility (credibility scale)


ExclusionsOther autoimmune disorder, another disease that seriously affected function, illiterate or cognitively impaired, were in a formal behavioural pain management programme, had experienced a major stressful life change in the previous 6 months, unable to write or walk


Fatigue outcomesdaily diary NRS


NotesMajority of data were obtained directly from the author and were unpublished. Fatigue data were only available for the main effects RCT, that is: 1) all patients who undertook Written Emotional Disclosure versus Control Writing (irrespective of later training arms) and 2) all patients who undertook Coping Skills Training versus Education alone (irrespective of preceding writing arm)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk‘A person independent of the study staff used randomization software to develop the condition assignments’. Stratified by study site, medication class.  Randomisation was done in blocks of 8, to one of 2 writing conditions followed by one of 2 training conditions

Allocation concealment (selection bias)Unclear risk‘assignments were sealed in envelopes’  does not state that they were opaque

Blinding (performance bias and detection bias)
All outcomes
High riskParticipants and study personnel were not blind to the interventions. There was blinded assessment of disease activity but the other measures were participant self-report

Incomplete outcome data (attrition bias)
All outcomes
Low risk'All analyses were conducted on the intent-to-treat sample of all 264 randomized participants. This is done by default in HLM, which estimates missing values; but for the ANCOVAs, we replaced those few missing values with the last available value (in most cases, using the baseline value for missing follow-up data).'

Selective reporting (reporting bias)Unclear riskNo daily outcome data were provided for fatigue, pain and positive and negative mood; states this will be presented in a separate paper

Other biasLow riskNo evidence of other bias identified

Neuberger 2007 - I

MethodsRCT: 3 arms


Participants40-70 years, RA, able to read/speak English, ambulatory, rheumatologist approval to participate


InterventionsLow impact aerobics; warm up/low-impact/strength/cool down set to individual heart rate targets
Class - at a fitness centre
Home - video at home


OutcomesPrimary outcome: not specified

Relevant outcomes: Pain (McGill Pain Questionnaire), depression (CES-D), total joint count.

Other: ESR/CRP, grip strength, walk time, amount of exercise, aerobic fitness level, optimism (Life orientation test), exercise benefits/barriers, self-efficacy for exercise, confidence to exercise


ExclusionsArryhtmias, recent myocardial infarction, acute infection, uncontrolled metabolic disease, known electrolyte abnormalities, systolic blood pressure >200mgHg or diastolic >115mmHg, doing ≥30 minutes of exercise ≥3 times/week, history of fibromyalgia or severe chronic obstructive pulmonary disease, taking beta blockers or digitalis.


Fatigue outcomesa) MAF, range 3-50, high=bad
b) POMS, range 0-4


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"a priori list of randomly generated permutations of 3 numbers: class, home, control"

Allocation concealment (selection bias)Unclear riskNo details

Blinding (performance bias and detection bias)
All outcomes
High riskNot possible to blind participants or instructors but "trained assessors were blinded to group assignment"

Incomplete outcome data (attrition bias)
All outcomes
Low risk"intention to treat used in all analyses"

Selective reporting (reporting bias)High riskSocial support data collected but not reported

Other biasHigh riskNo data obtained on 62 withdrawals prior to T1 assessment therefore chance that more highly motivated participants were analysed. Exercise self-efficacy and confidence to exercise measures not tested in all RA patients

Neuberger 2007 - II

MethodsRCT: 3 arms


Participants40-70 years, RA, able to read/speak English, ambulatory, rheumatologist approval to participate


InterventionsLow impact aerobics; warm up/low-impact/strength/cool down set to individual heart rate targets
Class - at a fitness centre
Home - video at home


OutcomesPrimary outcome: not specified

Relevant outcomes: Pain (McGill Pain Questionnaire), depression (CES-D), total joint count

Other: ESR/CRP, grip strength, walk time, amount of exercise, aerobic fitness level, optimism (Life orientation test), exercise benefits/barriers, self-efficacy for exercise, confidence to exercise


ExclusionsArrhythmias, recent myocardial infarction, acute infection, uncontrolled metabolic disease, known electrolyte abnormalities, systolic blood pressure >200mgHg or diastolic >115mmHg, doing ≥30 minutes of exercise ≥3 times/week, history of fibromyalgia or severe chronic obstructive pulmonary disease, taking beta blockers or digitalis


Fatigue outcomesa) MAF, range 3-50, high=bad
b) POMS, range 0-4


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"a priori list of randomly generated permutations of 3 numbers: class, home, control"

Allocation concealment (selection bias)Unclear riskNo details

Blinding (performance bias and detection bias)
All outcomes
High riskNot possible to blind participants or instructors but "trained assessors were blinded to group assignment"

Incomplete outcome data (attrition bias)
All outcomes
Low risk"intention to treat used in all analyses"

Selective reporting (reporting bias)High riskSocial support data collected but not reported

Other biasHigh riskNo data obtained on 62 withdrawals prior to T1 assessment therefore chance that more highly motivated participants were analysed. Exercise self-efficacy and confidence to exercise measures not tested in all RA patients

Otter 2010

MethodsSingle blind comparison trial: 2 arms


ParticipantsRA, >18 years


InterventionsReflexology delivered by one of four reflexologists using a standard protocol to stimulate various reflexology points on the feet. Once per week for 6 weeks


OutcomesPrimary outcome: Fatigue (MAF)

Relevant outcomes: Foot Pain (Manchester Foot Pain Disability Questionnaire)

Other: Sleep (Insomnia Severity Index)


ExclusionsChanges in drug regime during the trial or in the previous 3 months, sensory neuropathy, current foot ulceration and/or infection and previous experience of reflexology or foot massage treatment


Fatigue outcomesMAF


NotesMajority of the study details were obtained from the author and were unpublished


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear risk‘The principle investigator allocated participants in a 1:1 ratio to receive either reflexology or foot massage’. No further details provided

Allocation concealment (selection bias)Unclear risk‘The principle investigator allocated participants in a 1:1 ratio to receive either reflexology or foot massage’. No further details provided

Blinding (performance bias and detection bias)
All outcomes
Low riskParitcipants were blind to the intervention and outcomes were self-report but the study personnel were not blind

Incomplete outcome data (attrition bias)
All outcomes
Low riskAll data were obtained

Selective reporting (reporting bias)Low riskAll outcomes are presented

Other biasHigh riskGroups were not matched at baseline

Riemsma 2003 - I

MethodsRCT: 3 arms


ParticipantsMet at least 4 of the ACR criteria for RA. Significant other must be willing to participate


Interventions5 weekly 2 hour group sessions for 8 patients with significant other. 3 x 2 hour booster sessions after 3, 6 and 9 months. Each group had 2 trained leaders (nurses). Patients also received a programme book with information on the sessions, a self-help guide, brochures on RA and an audiotape with relaxation exercises. Programme included: contracting, goal setting and feedback; self-management and problem solving; information on RA and treatment; pain management and relaxation; physical exercises; coping with depression; and booster sessions


OutcomesPrimary outcomes: self-efficacy and health behaviour

Relevant secondary:

Other outcomes: disease activity (DAS28), health status (Dutch-AIMS2), social interactions, treatment credibility


ExclusionsResident in a nursing home, younger than 20 or older than 70 years


Fatigue outcomesVAS (tiredness on average during past wk due to arthritis), 0-100, high=bad


Notes42% of the group with significant others, 58% without significant others and 58% of the controls had no co-morbidities (53% total)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear risk"randomly assigned " but no details

Allocation concealment (selection bias)Unclear riskNot stated

Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNumbers at each time point unclear

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasUnclear riskStatistically significant differences between arms at baseline for coping with pain and perceived problematic support

Riemsma 2003 - II

MethodsRCT: 3 arms


ParticipantsMet at least 4 of the ACR criteria for RA. Significant other must be willing to participate


Interventions5 weekly 2 hour group sessions for 8 patients with significant other. 3 x 2 hour booster sessions after 3, 6 and 9 months. Each group had 2 trained leaders (nurses).  Patients also received a programme book with information on the sessions, a self-help guide, brochures on RA and an audiotape with relaxation exercises. Programme included: contracting, goal setting and feedback; self-management and problem solving; information on RA and treatment; pain management and relaxation; physical exercises; coping with depression; and booster sessions


OutcomesPrimary outcomes: self-efficacy and health behaviour

Relevant secondary:

Other outcomes: disease activity (DAS28), health status (Dutch-AIMS2), social interactions, treatment credibility


ExclusionsResident in a nursing home, younger than 20 or older than 70 years


Fatigue outcomesVAS (tiredness on average during past wk due to arthritis), 0-100, high=bad


Notes42% of the group with significant others, 58% without significant others and 58% of the controls had no co-morbidities (53% total)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear risk"randomly assigned " but no details

Allocation concealment (selection bias)Unclear riskNot stated

Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNumbers at each time point unclear

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasUnclear riskStatistically significant differences between arms at baseline for coping with pain and perceived problematic support

Skoldstam 2003

MethodsRCT: 2 arms


ParticipantsRA, diagnosed for at least 2 years, stable disease assessed by rheumatologist


InterventionsMediterranean Cretan diet (MD) including reduced fat, more green/black tea. Served by hospital for 3 weeks with 6 lessons about Mediterranean cooking from dietician. After rehabilitation same dietician was available weekly for telephone consultation and every third week for face to face consultation.
All MD patients got written instructions and recipes to facilitate the preparation of meals at home. The meals that both the MD and the control patients had to prepare at home were (a) every day breakfast and evening snacks; (b) meals for two weekends during rehabilitation; (c) every meal for the remaining nine weeks that followed rehabilitation. To enhance compliance, some food items were supplied free to the MD patients (olive oil, canola oil, liquid margarine, spreadable margarine, frozen vegetables, and tea).


OutcomesPrimary: disease activity (DAS 28), disability (Swedish HAQ), Health-related quality of life (SF-36 Swedish version), daily non-steroidal anti-inflammatory drugs (NSAIDs)

Relevant secondary:

Other: CRP, thrombocyte count, early morning stiffness, functional impairment, grip ability test


ExclusionsDisease-modifying anti-inflammatory drugs had to be unchanged for>3months, corticosteroids for >4 weeks, and NSAID for >10 days before beginning. Daily dose of oral corticosteroids could not exceed 12.5 mg of prednisolone. At baseline DAS28  >2.0 indicating active disease. Except for RA, the patients could have no other condition that demanded active medical attention. Patients who were vegetarians or who already lived on a Mediterranean-like diet were also excluded


Fatigue outcomesSF-36 Vitality, High=good


NotesCompliance data from Hagfors 2005a


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"Stratified for gender, block randomisation with 2-6 per block"

Allocation concealment (selection bias)Unclear riskNo details

Blinding (performance bias and detection bias)
All outcomes
High riskDue to the nature of the intervention participants were not blinded

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk5 patients excluded - no details of how any missing data handled

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasHigh riskPossible contamination of control arm, violations of study protocol - injections. Arms not balanced at baseline. Statistically significant differences in diet between intervention and control arm but not clear how closely they adhered to the intervention. Diet is also all self-report questionnaires

Wang 2008

MethodsRCT: 2 arms


ParticipantsClass I and II RA, 18 and over


InterventionsTai Chi, one hour group sessions. Based on classical Yang style


OutcomesPrimary:  the proportion of participants who attained an ACR 20 response criterion at 12 weeks

Relevant secondary: tender and swollen joint scores; disability (HAQ); pain (VAS); depression (CES-D)

Other: Ritchie articular index, patient and physician VAS for disease severity, ESR, CRP, physical function, health-related quality of life (SF-36 and ED5D).


ExclusionsPrior experience of Tai Chi or other similar types of alternative and complementary therapy and/or were unwilling to abstain from Tai Chi (other than that provided in the study) until study completion, cardiovascular or other severe disease limitations precluding full participation, mini-mental status below 24, pregnant or breastfeeding, non-English speaking, participation in any other clinical trial in the previous 30 days.


Fatigue outcomesa) VAS: 0-100mm high=bad b) SF-36: 0-100 high=good


Notes


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"Randomization assignments were made using computer-generated random numbers, in two groups of 10..."

Allocation concealment (selection bias)Low risk"...and were provided in sealed opaque envelopes upon agreement to participate"

Blinding (performance bias and detection bias)
All outcomes
Unclear riskNo details of participants or personnel, "physical examination and functional capacities were blindly assessed by the study rheumatologist and a physiologist"

Incomplete outcome data (attrition bias)
All outcomes
Low riskIntention to treat

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Other biasHigh riskAt baseline the Tai Chi arm had statistically significantly worse HAQ scores and CRP. Two patients in the Tai Chi arm changed their medication during the study period

Zangi 2012

MethodsRCT: 2 arms


ParticipantsInflammatory arthritis diagnosed for ≥1 year; age 20-70 years


InterventionsVitality training programme: groups of 8-12 people. 10 sessions over 15 weeks initially with a booster at 6 months. 4.5 hours per session (49.5 hours in total). Included mindfulness-based exercises, creative exercises, reflection. Between sessions participants listened to a compact disc with mindfulness based exercises. Group facilitators were health professionals trained in mindfulness-based exercises


OutcomesPrimary: Psychological distress (General Health Questionnaire-20), self-efficacy (Arthritis Self-Efficacy Scale) and emotion focused coping (Emotional Approach Coping Scale)

Relevant secondary: Pain (NRS)

Other: patient reported disease activity, self care ability and overall well being (all by NRS)


ExclusionsInability to understand Norwegian


Fatigue outcomesNRS


NotesMajority of the published results reported were for all participants with inflammatory arthritis. Specific data for RA participants were unpublished and obtained from the study authors.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk‘A statistician generated randomization lists’ using specialist computer software ‘with blocks of 10 and 15 for each department to ensure approximately equal sample sizes.

Allocation concealment (selection bias)Low risk‘Participants were given consecutive numbers and a person not involved in the data collection or intervention allocated each participant to the corresponding numbers on the randomization list’.

Blinding (performance bias and detection bias)
All outcomes
High riskStudy personnel and participants were not blinded. ‘The persons conducting telephone interviews were blinded for group assignments, and participants were instructed not to discuss their intervention with the interviewer.’ ‘A blinded statistician conducted statistical analysis. The randomization code was not opened until the primary analyses were completed.’

Incomplete outcome data (attrition bias)
All outcomes
Low riskMixed-effects model

Selective reporting (reporting bias)Low riskAll outcomes were reported

Other biasHigh riskControl arm were advised that they could access the intervention at the end of the study which is likely to have raised expectations and hence bias

 
Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

Barlow 1998Not an RCT

Barlow 2000RA data not reported separately for RA participants

Barlow 2009Not an RCT

Behandling og traeningUnobtainable

Bhat 2008RA data not reported separately for RA participants

Bhat 2010Not an RCT

Callahan 2008RA data not reported separately for RA participants

Daltroy 1995RA data not reported separately for RA participants

de Buck 2005RA data not reported separately for RA participants

Evers 2005Not an RCT - summary report of several studies

Furst 2000No fatigue outcome measure

Goldbach-Mansky 2009No fatigue outcome measure

Hagfors 2003Not an RCT

Hagfors 2005Not an RCT

He 2008No fatigue outcome measure

Heiberg 2007Not an RCT

Jiang 2012No fatigue outcome measure

Kremer 1987Not an RCT

Lee 2006Not an RCT

MacDonald 1994No fatigue outcome measure

Minnock 2008Not an RCT

Mur 2002No fatigue outcome measure

Neuberger 1997Not an RCT

Nordemar 1981Not an RCT

Noreau 1995Not an RCT

O'Leary 1986No fatigue outcome measure

Pan 2009Not an RCT

Poulsen 1991No fatigue outcome measure

Rall 1996Not an RCT

Rauma 1993No fatigue outcome measure

Solomon 2002RA data not reported separately for RA participants

Song 2007No fatigue outcome measure

Stenstrom 1993No fatigue outcome measure

Stenstrom 1994No fatigue outcome measure

Stenstrom 1996RA data not reported separately for RA participants

Taibi 2009No fatigue outcome measure

Toirov 2010Not an RCT

 
Characteristics of studies awaiting assessment [ordered by study ID]
Berbert 2005

MethodsData query - still trying to contact authors

Participants

Interventions

Outcomes

Notes

Wiesenauer 1991

MethodsAwaiting translation

Participants

Interventions

Outcomes

Notes

 
Comparison 1. Physical activity versus control

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Fatigue6371Std. Mean Difference (IV, Random, 95% CI)-0.36 [-0.62, -0.10]

 
Comparison 2. Psychosocial versus control

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Fatigue151556Std. Mean Difference (IV, Random, 95% CI)-0.24 [-0.40, -0.07]

 
Comparison 3. Mediterranean diet versus control

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Fatigue151Std. Mean Difference (IV, Random, 95% CI)0.37 [-0.18, 0.93]

 
Comparison 4. Andrographis paniculata versus control

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Fatigue158Std. Mean Difference (IV, Random, 95% CI)-0.25 [-0.77, 0.27]

 
Comparison 5. Omega-3 fatty acid versus control

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Fatigue181Std. Mean Difference (IV, Random, 95% CI)0.93 [0.47, 1.39]

 
Comparison 6. Data tracker versus control

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Fatigue1714Std. Mean Difference (IV, Random, 95% CI)0.06 [-0.10, 0.21]

 
Comparison 7. Reflexology versus control

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Fatigue111Std. Mean Difference (IV, Random, 95% CI)-1.24 [-2.59, 0.11]

 
Summary of findings for the main comparison. Non-pharmacological interventions for rheumatoid arthritis

Non-pharmacological interventions for rheumatoid arthritis

Patient or population: patients with rheumatoid arthritis
Settings: All
Intervention: non-pharmacological interventions

OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments

(95% CI)

Assumed riskCorresponding risk

ControlNon-Pharmacological Interventions

fatigue (physical activity interventions)
Various fatigue scales
The mean fatigue in the control groups was
6.3 VAS (score: 0-10)
The mean fatigue (physical activity interventions) in the intervention groups was
9% lower
(2% to 15% lower)1
371
(6 studies)
⊕⊕⊕⊝
moderate3
SMD -0.36 (-0.62 to -0.10)

Relative percent change -13.7% (-23.6 to -3.8)

NNTB 7 (4 to 26)

fatigue (psychosocial interventions)
Various fatigue scales
The mean fatigue in the control groups was
6.3 VAS (score: 0-10)
The mean fatigue (psychosocial interventions) in the intervention groups was
6% lower
(2% to 10% lower)1,2
1,556
(13 studies)
⊕⊕⊝⊝
low3,4
SMD -0.24 (-0.40 to -0.07)

Relative percent change -9.1% (-15.2 to -2.7)

NNTB 10 (6 to 33)

fatigue (Andrographis paniculata)The mean fatigue in the control groups was
6.3 VAS (score: 0-10)
The mean fatigue (Andrographis Paniculata) in the intervention groups was
6% lower
(18% lower to 6% higher)1,2
58
(1 study)
⊕⊕⊝⊝
low5
SMD -0.25 (-0.77 to 0.27)

Relative percent change -9.5% (-29.3 to 10.3)

NNTB not applicable

fatigue (data tracker)The mean fatigue in the control groups was
6.3 VAS (score: 0-10)
The mean fatigue (data tracker) in the intervention groups was
1% higher
(2% lower to 5% higher)1,2
714
(1 study)
⊕⊕⊝⊝
low5
SMD 0.06 (-0.10 to 0.21)

Relative percent change 2.3% (-3.8 to 8.0)

NNTB not applicable

fatigue (Omega-3 Fatty acids)The mean fatigue in the control groups was
6.3 VAS (score: 0-10)
The mean fatigue (Omega-3 Fatty acids) in the intervention groups was
22% higher
(11% to 33% higher)1,2
81
(1 study)
⊕⊕⊝⊝
low6
SMD 0.93 (0.47 to 1.39)

Relative percent change 35.4% (17.9 to 53.0)

NNTB not applicable

fatigue (Mediterranean diet)The mean fatigue in the control groups was
6.3 VAS (score: 0-10)
The mean fatigue (Mediterranean diet) in the intervention groups was
9% higher
(4% lower to 22% higher)1,2
51
(1 study)
⊕⊕⊕⊝
moderate5
SMD 0.37 (-0.18 to 0.93)

Relative percent change 14.1% (-6.9 to 35.4)

NNTB not applicable

fatigue (reflexology)The mean fatigue in the control group was

6.3 VAS (score: 0-10)
The mean fatigue (Reflexolgy) in the intervention groups was
30% lower
(62% lower to 3% higher)1,2
11

(1 study)
⊕⊕⊝⊝
low4
SMD -1.24 (-2.59 to 0.11)

Relative percent change -47.2% (-98.7 to 4.2)

NNTB not applicable

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; SMD: standardised mean difference; NNTB: number needed to treat for an additional beneficial outcome

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

 1 A lower score means less fatigue
2 Based on a conversion from Hewlett S et al (Mean: 6.3, SD: 2.4).
3 Limitations in the design as appropriate masking was not possible
4 Unexplained heterogeneity present between studies, represented by I2 of 55%
5 High probability of publication bias; only 1 study available; did not support a difference between groups
6 High probability of publication bias; only 1 study, differences between groups not assessed
 
Table 1. Interventions

Study IDIntervention categorySpecific intervention

Bilberg 2005Physical activitypool-based therapy

Evans 2012Physical activityYoga

Hakkinen 2003Physical activitystrength training

Harkcom 1985Physical activitystationary cycling

Neuberger 2007 - IPhysical activityaerobic exercise

Wang 2008Physical activityTai Chi

Danoff-Burg 2006 - IPsychosocialbenefit finding and expressive writing

Lumley 2011-IPsychosocialcognitive skills training & written emotional disclosure

Evers 2002Psychosocialcognitive behavioural therapy

Hewlett 2011Psychosocialcognitive behavioural therapy

Zangi 2012Psychosocialmindfulness based group intervention

Hammond 2008Psychosociallifestyle management

Furst 1987Psychosocialeducation incorporating energy conservation

Laforest 2008Psychosocialeducation incorporating self management

Lorig 2005Psychosocialeducation incorporating self management

Lorig 2008Psychosocialeducation incorporating self management

Riemsma 2003 - IPsychosocialgroup education

Helliwell 1999Psychosocialgroup education

Giraudet-Le-Quintrec 2007Psychosocialgroup education

Skoldstam 2003OtherMediterranean diet

Burgos 2009Otherherbal medicine

Das Gupta 2009Otheromega-3 fatty acid supplementation

Huffstutter 2007Otherelectronic data capture system

Otter 2010Otherreflexology