Physical therapy and physical activity in patients with rheumatoid arthritis Hurkmans, E.J.

(1)

rheumatoid arthritis

Hurkmans, E.J.

Citation

Hurkmans, E. J. (2011, October 6). Physical therapy and physical activity in patients with rheumatoid arthritis. Department of Rheumatology , Faculty of Medicine, Leiden University Medical Center (LUMC), Leiden University.

Retrieved from https://hdl.handle.net/1887/17903

Version: Corrected Publisher’s Version

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/17903

Note: To cite this publication please use the final published version (if applicable).

(2)

••• Chapter 7 The effectiveness and safety of home-based exercise programs in patients with rheumatoid arthritis

E.J. Hurkmans F.J. van der Giesen T.P.M. Vliet Vlieland T.W. Schoones C.H.M. van den Ende

Cochrane Database of Systematic Reviews [Submitted]

(3)

• • • Abstract

Objective: To assess the effectiveness and safety of home-based exercise programs in patients with rheumatoid arthritis (RA).

Method: A literature search up to December 2009 within various databases was performed to identify relevant randomized controlled trials (RCTs) fulfilling the criteria: a) exercise program was performed by patients at home; b) instructions were given at least once; c) program was evaluated not more often than once in two weeks; d) the program included aerobic capacity, muscle strength, joint range of motion, and/or stretching exercises. Two blinded reviewers independently selected and categorized the studies, rated the risk of bias, and extracted data regarding outcome measures of effectiveness and safety. Where appropriate, quantitative and/or qualitative analyses were performed

Results: Eleven RCTs were included; 8 had a low risk of bias. The included studies were categorized as ‘home-based program versus control (no exercise) (2 trials)’,

‘home-based program versus supervised program’ (4 trials), and ‘home-based program versus another home-based program (5 trials)’. Home-based exercise programs (including aerobic capacity and/or muscle strength exercises) had in general a positive effect on functional ability and muscle strength immediately after the intervention compared to no intervention. There were no differences in effectiveness of home-based exercise programs (including aerobic capacity and/or muscle strength exercises) as compared to supervised exercise programs (including aerobic capacity and/or muscle strength exercises) with regard to functional ability. However, home-based exercise programs were less effective than supervised exercise programs with regard to aerobic capacity and muscle strength immediately after the intervention. When comparing different types of home-based exercise programs, a strengthening exercise program, whether or not combined with stretching exercises was more effective with regard to functional ability and self-reported pain than a stretching program. Moreover, an aerobic capacity and muscle strengthening exercise program that was goal directed was more effective with regard to disease activity compared to a similar, yet pain directed program. No differences in effectiveness were seen between an individualized program including aerobic capacity exercises and a general program

• • • • • • • •

(4)

including similar exercises, and between a program including aerobic capacity en strengthening exercising and a relaxation program. In the studies measuring disease activity, self-reported pain and radiological damage, no deleterious effects of any of the home-based exercise programs were reported.

Conclusion: Home based exercise programs comprising aerobic and muscle strengthening exercises have a positive effect on functional ability and muscle strength. Home based exercise programs with muscle strengthening exercises and goal-directed programs are more effective than programs with stretching exercises or pain directed programs, respectively. Supervised exercise programs comprising aerobic and muscle strengthening exercises are superior to similar home based exercise programs with respect to the improvement of aerobic capacity and muscle strength and are therefore to be preferred. These results suggest that after an initial period of supervision, goal-directed home based programs with stretching and muscle strengthening exercises should be considered.

• • • Background

Rheumatoid arthritis (RA) is a chronic systemic inflammatory disease mainly affecting the musculoskeletal system [1, 2]. The cornerstones of treatment include control of inflammation, relieve of pain and stiffness, and restoration of functioning [1]. Although there is no cure for RA, the current drug treatments are able to control the disease and/or significantly postpone joint damage in many patients [2]. Despite these advancements, many patients with RA experience a decreased ability to perform daily tasks and limitations in participation.

Exercise programs are of crucial importance for the maintenance and restoration of functioning in patients with RA [3]. Exercise programs usually comprise a combination of muscle strengthening exercises, exercise to increase cardio respiratory fitness (aerobic capacity exercises) and flexibility exercises, in combination with training of specific movement patterns or daily activities.

Exercise programs for RA patients are usually provided by a trained (health) professional, and can either be supervised or home based. Supervision implies that a (health) professional is present while the exercises are performed, allowing for continuous instruction, modification and evaluation of the program. With

Home-based exercise programs in patients with RA

• • • • • • • •

(5)

home based programs the exercises are done by patients on their own after initial instructions, with sparse evaluations. So far, reviews on the effectiveness of exercise programs in patients with RA have not made a distinction between supervised and home-based exercise programs [4, 5] or focused only on supervised exercise programs [3]. Research on the effectiveness and safety of home-based exercise programs has not yet been systematically summarized.

The aim of the present study was therefore to systematically review all available evidence with regard to the effectiveness and safety of home-based exercise programs in patients with RA.

Objectives

To determine the effectiveness and safety of home-based exercise programs in patients with RA.

• • • Methods

Criteria for considering studies for this review

Studies: Any randomized controlled trial comparing home-based exercise programs in patients with RA with another form of exercise or with a non-exercising control group.

Types of participants: Trials in which patients (males and females > 18 years old) with classical or definite RA according to the 1958 ARA criteria [6] or the 1987 ARA criteria [7] were included were selected. Trials pooling data of patients with RA with other diseases were excluded.

Types of interventions: Trials with the following interventions were selected:

• The intervention concerned an exercise program (including aerobic capacity, muscle strength, range of motion, stability/coordination and/or functional exercises) mainly performed by the patient at home with no (health) professional present.

• Instructions (written or oral) were given once or twice (maximum) by a (health) professional, i.e. a physical therapist or a person specially trained to provide an

• • • • • • • •

(6)

exercise program.

• The exercise program was not evaluated more frequently than once in two weeks by a (health) professional, although not face to face.

Types of outcome measures: Any trial reporting data on one or more of the primary and secondary outcome measures listed below fulfilled this inclusion criterion.

The outcome measures selected were those which have confirmed properties in terms of reliability, validity and sensitivity to change as described in the literature or outcome measures which are generally accepted in the literature. The following primary (P) and secondary (S) outcome measures were considered relevant:

Effectiveness:

1. Functional ability as assessed by validated measures preferably by the McMaster Toronto Arthritis Patient Preference Interview (MACTAR), the Health Assessment Questionnaire (HAQ) [8], the dimension Physical function of the Arthritis Impact Measurement Scales (AIMS) [9] or any other validated health related quality of life (QoL) measure with an incorporated functional ability dimension like the Short Form-36 (SF-36), the Fries Index [10] or the Functional Status Index (FSI) [11] (P) or a standardized functional performance test like the Timed Up and Go.

2. Aerobic capacity (VO2max, in ml/kg/min) determined by a maximal or submaximal ergometer test (S).

3. Muscle strength of the knee extensors in Nm, preferably assessed on an (isokinetic) dynamometer (S).

Safety:

1. Self-reported pain, preferably scored on a visual analog scale (VAS) (S).

2. Adverse effects (exercise induced injuries or substantially increased pain levels as a direct effect of treatment) (S).

3. Disease activity preferably assessed by the Westergren Erythrocytes Sedimentation Rate (ESR) in mm/h, or C-reactive protein (CRP) level in mg/l, a swollen or tender joint count, patient/physician global assessment as measured

• • • • • • • •

(7)

on a visual analog scale or rating scale, or the Disease Activity Score (DAS) (S).

4. Radiological joint damage preferably assessed by a Larsen score (S) of radiographs.

Self reported pain, disease activity and radiological damage were considered important outcome measures for safety. In case of improvement, those outcome measures could also be indicators for effectiveness.

Search methods for identification of studies

Articles in English, Dutch, French or German language were considered for inclusion in this review.

Electronic searches

The Cochrane Central Register of Controlled Trials (CENTRAL), Pubmed, EMBASE, Web of Science, CINAHL, Pedro, the Database of Abstracts of Reviews of Effectiveness (DARE) and Health Technology Assessment database (HTA) were used initially to identify all randomised controlled trials up to December 2009 relating to an exercise program in RA. The following strategy was used for identification of controlled trials, all performed by review author EJH:

(a) The Pubmed (up to December 2009) database was searched using an optimal sensitive Pubmed search strategy:

• search string for RA: rheumatoid arthritis [MeSH Term] OR arthritis[tiab] OR arthritic[ti] OR rheumatic disease[ti] OR rheumatoid[ti] AND;

• search string for intervention: exercise therapy [MeSH Term] OR exercis*[tiab]

OR physical education and training [MeSH Term] OR physical education[tiab]

OR training[tiab] OR gymnastics [MeSH Term] OR gymnast*[tiab] OR physical fitness [MeSH Term] OR physical fitness[tiab] OR physical activity [MeSH Term]) OR physical activit*[tiab] AND

• search string for identifying randomised controlled trials [12, 13]: randomized controlled trial OR controlled clinical trial OR randomized controlled trials OR random allocation OR double-blind method OR single-blind method OR clinical trial OR clinical trials OR “clinical trial” OR ((singl*

• • • • • • • •

(8)

OR doubl* OR trebl* OR tripl* ) AND (mask* OR blind* )) OR “latin square”

OR placebos OR placebo* OR random* OR research design [mh:noexp] OR comparative study OR evaluation studies OR follow-up studies OR prospective studies OR cross-over studies OR prospective* OR volunteer* OR randomised controlled trial OR randomised controlled trials.

(b) The same search strategy was translated by a trained medical librarian to make it applicable for the EMBASE database (up to 2009), the Web of Science database (up to 2009), the CINAHL database (up to 2009), the Pedro database (up to 2009) and The Cochrane Database (CENTRAL, DARE and HTA) (up to 2009).

Searching other resources

(c) The reference lists of the identified randomized controlled trials on the effect of home-based exercise programs were scanned.

(d) In case of abstracts with no published full-length article, the authors were contacted by mail and asked for a written report.

(e) The table of content of the major journals in the domain of rheumatology (n=16) were searched (from 2000 up to 2009).

(f) The controlled trial register was checked for ongoing trials (http://www.

controlled-trials.com). Authors of an ongoing trial were contacted for provisional results.

Data collection and analysis

Selection for inclusion in the review, data extraction and risk of bias in included studies were performed in three separate steps. The articles were independently assessed by review authors EJH and FG. The review authors were blinded in all three steps to the journal, for the authors and institution, by providing them with edited copies of the articles. In the case of persistent disagreement between the two review authors, a third review author (TVV) decided.

Step I: Selection of studies

First, all titles and/or abstracts were screened for the following criteria: 1) the study included patients with RA; and 2) and the intervention concerned

Home-based exercise programs in patients with RA Home-based exercise programs in patients with RA

• • • • • • • •

(9)

a type of a home-based exercise program. Second, the full length articles of the selected titles and/or abstracts were gathered and screened for the inclusion criteria as stated before. Only full length articles or full length unpublished reports were considered for inclusion in this systematic review.

Step II: Data extraction

Two authors independently assessed the included RCTs. The included studies were categorized as ‘home-based program versus control (no exercise)’, ‘home-based program versus supervised program’, and ‘home-based program versus another home-based program’.

Study characteristics: Extracted data were recorded in a pre developed form. The following descriptive information was systematically extracted: the number of participating RA patients, sex, age (years), disease duration (years), inclusion and exclusion criteria (yes or no); a description of the home-based group and control group (type of exercises, frequency, duration and amount of supervision. For every continuous outcome measure, baseline values and measures of variability (preferably means and standard deviation) were extracted. Mean changes from baseline with standard deviations of outcome measures were obtained. Means and SD were converted from medians and interquartile ranges. For every dichotomous outcome, the number of patients treated in the different intervention groups and/

or control group and the number incurring the event of interest in each group were extracted.

Measures of treatment effect: In case of statistically significant differences between groups, a standardized mean difference (effect size) with corresponding 95% confidence intervals was calculated for continuous variables. For dichotomous variables, risk ratios with corresponding 95% confidence intervals were computed.

If continuous as well as dichotomous data was present, the dichotomous data (risk ratio as well as odds ratio) was re-expressed as a standardized mean difference [14].

• • • • • • • •

(10)

Quality of selected studies (risk of bias): To assess the risk of bias, the following 6 key domains were reported by two authors (EJ and FG) (Cochrane handbook):

• Adequate sequence generation? The method used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should produce comparable groups.

• Allocation concealment? The method used to conceal the allocation sequence in sufficient detail to determine whether intervention allocations could have been foreseen in advance of, or during, enrollment.

• Blinding? (= yes when the assessor was blinded) All measures used, if any, to blind study participants and personnel from knowledge of which intervention a participant received. Any information relating to whether the intended blinding was effective.

• Incomplete outcome data? Completeness of outcome data for each main outcome, including attrition and exclusions from the analysis.

• Free of selective reporting? The possibility of selective outcome reporting.

• Free of other bias? Important concerns about bias not addressed in the other domains in the tool.

In case of disagreement between the review authors (EJ and FG), the decision was made by consensus. Studies fulfilling 4 or more key domains were regarded as ‘low risk of bias studies’, while studies fulfilling three or less key domains were considered to be ‘high risk of bias studies’.

Step III: Qualitative analyses

The quality of the body of the evidence for each outcome within each category was assessed with the GRADE approach (Grades of Recommendation, Assessment, Development and Evaluation) by two independent reviewers (EJ, FG) (Cochrane Handbook). The quality of the evidence was assessed based on six domains 1) the study design, 2) risk of bias (see quality of selected studies), 3) consistency of results, 4) directness (generalizability), 5) precision (sufficient data), and 6) reporting of the results for studies that measure one particular outcome (See Appendix 1 for definitions of these domains).

• • • • • • • •

(11)

High quality of evidence: Further research is very unlikely to change our confidence in the estimate of effect. There are consistent findings among 75% of RCTs with low risk of bias as to be accepted in the studied population. There are sufficient data, with narrow confidence intervals. All domains were met.

Moderate quality of evidence: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.

One of the domains is not met.

Low quality of evidence: 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.

Two of the domains are not met.

Very low quality of evidence: The estimate is very uncertain. Three of the domains are not met.

Step IV: Quantitative analyses: A quantitative approach was considered appropriate if the included studies were not clinically diverse and/or statistically heterogeneous. Clinical diversity among studies was assessed by two authors (EJH and FG) taking into account the classification of patients (severity of the disease), intervention (intensity and nature of exercise forms and duration, frequency and setting of exercise sessions) and outcome measures (dimensions of outcome measures). Disagreement was resolved by consensus. Statistical heterogeneity was tested by applying a chi-squared test and an I² statistic test (Cochrane Handbook). When the P value of this test was lower than 0.25 or when the I² statistic test showed a value greater than 50%, we considered this to indicate substantial heterogeneity.

In case of meta-analysis, the pooled standardized mean difference was calculated with a random effects model (Cochrane handbook). An effect size of 0.20 was considered small, 0.50 medium and 0.80 large.

Step V: Sensitivity analyses: Qualitative and/or quantitative analyses were again performed, taking now only ‘low risk of bias studies’ or ‘long term exercise programs’ (> 3 months) into account. These sensitivity analyses were performed because it was expected to influence the results (stronger evidence).

• • • • • • • •

(12)

• • • Description of studies Selection of included trials

The search strategy was applied up to December 2009. This resulted in a list of 2041 citations. Finally, based on the titles, abstracts, or both, a total of 84 citations were selected.

The full-length articles of these 84 citations were screened for the full set of inclusion criteria. Fifteen articles fulfilled the inclusion criteria [15-29]; one trial was described in five full-length articles [16-20], another trial was described in two full-length articles [26, 27]. The other 68 full length articles were excluded, mainly because: the exercise program was supervised, the trial was not randomized or the trial did not include RA patients or mixed various disorders.

After searching the references of the selected citations, no other trial that fulfilled the inclusion criteria was identified. No ongoing trials were found in the Current Controlled Trials register. In total, eleven trials described in 16 articles were included (see Table 1 “Characteristics of included studies”). The 68 articles that were screened but not included in the review are listed in the Table 2 “Characteristics of excluded studies”.

Description of included trials

The eleven trials included a total of 906 participants. In ten of the eleven included trials the majority of participants was female. In one trial the proportions of females and males were almost equal (14 males and 15 females) [26].

The average age of the participants varied from 48 and 55 years in most trials, except for two trials where the average age was higher, 61 and 62 years respectively [23, 25].

The average disease duration varied from 6 to 15 years, except for two trials. In one trial the average disease duration of the participants was 18 years [25], while in the other trial the average disease duration was 20 months [16].

In all trials patients with a stable disease activity and/or low to moderate limitations in daily functioning were included and in most trials patients with co-morbidity were excluded.

In all trials the exercise program was delivered by a trained professional.

• • • • • • • •

(13)

Table 1. Description of home-based physical activity and exercise programs (n RCTs = 11) in patients with RA. StudyMethodParticipantsInterventionOutcomesMethodolo- gical qualityTime- points Home-based programs versus no intervention Brodin 2008Prospective, randomized clinical trial, 2 groups 59 males / 169 females, Median (IQR) age: intervention group 54 (14) yrs, control group 56 (14) yrs Median (IQR) disease duration: intervention group 21 (5) months, control group 22 (4) months Inclusion criteria: Each RA patient > 18 years, able to communicate adequately in Swedish, able to perform body-function testing Exclusion criteria: no specific exclusion criteria were applied One physical activity program (promoting moderate intensity activities, 30 minutes/day, ≥ 4 days/week) versus no intervention Duration: 12 months Supervisor: physical therapists Amount of supervision: start session(s), continuous phone sup- port after 1 week and then once monthly Training supervisor: at least one coach from each participating clinic underwent a 1-day training program aimed at improving knowledge about the study . Setting: 10 rheumatology clinics at 7 university hospitals and 3 country hospitals Functional ability: HAQ Muscle strength: the Grippit Self-reported pain: VAS Disease activity: DAS28

Score: 4 (criteria 1, 3, 5, 6)

Baseline 12 months Ma- siero 2007

Prospective, randomized clinical trial, 2 groups

13 males / 57 females Mean (SD) age: intervention group 54 (10) yrs, control group 52 (12) yrs Mean (SD) disease duration: intervention group 15 (9) yrs, control group 16 (8) yrs Inclusion criteria: RA according to ARA criteria, aged 18-65 yrs, presented no variations in drug therapy in the 6 months before the trial, no present severe disability that seriously compromised independence in activities of daily living Exclusion criteria: previous participation in educational training, variations in drug therapy at any time during the trial, rehabilitation treatment or orthopedic surgery during the trial One educational-behavioral joint protection program (including various exercises) versus a control group (no intervention) Duration: 8 months Supervisor: persons trained to provide this program Amount of supervision: in total 4 meetings, every 3 weeks a 3-h session Training supervisor: unknown Setting: outpatient clinic Functional ability: HAQ Self-reported pain: VAS Disease activity: Ritchie Articular Index Score: 4 (criteria 3, 4, 5, 6)

Baseline 8 months Home-based programs versus supervised programs Van den Ende 1996

63 females/ 37 males Mean (SD) age: supervised high intensity program 51 (10) yrs, supervised group program 48 (14) yrs, supervised individual group 53 (12), home program 56 (11) Mean (SD) disease duration: supervised high intensity program 12 (8) yrs, supervised group program 8 (6) yrs, supervised individual group 9 (7), home program 11 (10) Inclusion criteria: age 20-70 yr, stable on medication, able to cycle. Exclusion criteria: arthroplasties of weight-bearing joints, co- morbidity One supervised high intensity program, one supervised low intensity group program, one supervised low intensity individual program and one home exercise program Duration 12 weeks and a 12 week follow-up Supervisor: physical therapists Amount of supervision (home-based program): advise once and written instructions Training supervisor: unknown Setting: outpatient clinic Functional ability: HAQ Aerobic capacity: VO2 max Muscle strength: isometric and isokinetic muscle strength Disease activity: ESR, no of tender and swollen joints Score: 3 (criteria 4, 5, 6)

Baseline 12 weeks 12 weeks follow- up Han- sen 1993

49 females / 26 males Median (25/75 percentiles) age: home-based program 55 (44/58), + individual program 52 (46/58), + group program 51 (42/56), + group program in water 54 (44/56), no training 51 (46/57) Median (25/75 percentiles) disease duration: home-based program 7 (5/9), + individual program 7 (3/16), + group program 7 (5/1), + group program in water 5 (4/9), no training 8 (5/11) Inclusion criteria: age 20-6- yr, RA according 1958 ARA criteria. Exclusion criteria: Steinbrocker III and IV, co-morbidity, presence of contra-indications for training, already training 3x per week One home-based training program, one home program + an individual supervised training program, one home program + a group supervised training program, one home program + a group supervised training program in water, one control group (no exercise) Duration of intervention: 24 months Supervisor: physical therapists Amount of supervision: (home-based program) Training supervisor: unknown Setting: hospital registry and physical therapy practices Functional ability: HAQ Aerobic capacity: aerobic fitness Muscle strength: isokinetic strength knee Self-reported pain: VAS Disease activity: ESR, swollen joint count Radiological damage: X-rays Score: 3 (criteria 3, 5, 6)

Baseline 24 months

• • • • • • • •

(14)

Table 1. Description of home-based physical activity and exercise programs (n RCTs = 11) in patients with RA. (continued) StudyMethodParticipantsInterventionOutcomesMethodolo- gical qualityTime- points Hsieh 2009Prospective, randomized clinical trial, 2 groups

30 females Mean (SD) age: supervised group 54 (8), home-based group 51 (12) yrs Mean (SD) disease duration: supervised group 8 (10) yrs, home-based group 8 (5) yrs Inclusion criteria: age range 20-65 yrs, RA according to ARA criteria, disease lasting at least 6 months, disease in well-controlled (stable) condition (no acute signs of inflammation) Exclusion criteria: arthroplasties or major operations in the knee or hip joints, presence of serious cardiac or pulmonary disease or any severe medical condition, severe arthritis or contracture of knee joints precluding bicycle exercise testing One supervised exercise group, one home-based exercise group Duration of intervention: 8 weeks Supervisor: physical therapists Amount of supervision: (home-based program) once and every two weeks a phone-call Training supervisor: unknown Setting: a private teaching hospital Functional ability: HAQ Aerobic capacity: bicycle ergometer Muscle strength: hand dynamometer Self-reported pain: VAS Disease activity: no of swollen and tender joints Score: 5 (criteria 1, 2, 3, 4, 5)

Baseline 8 weeks Lem- mey 2009

5 males / 23 females Mean (SD) age: supervised group 56 (8), home-based group 61 (11) Mean (SD) disease duration: supervised group 74 (76) months, home- based group 125 (101) months Inclusion/exclusion criteria: RA according to ARA criteria, age > 18 yrs, functional class I or II, not cognitively impaired, anti-inflammatory and/ or antirheumatic drug therapy stable in past 3 months, if on cortico- steroids, maintained on a dosage < 10mg/day, free of other cachectic diseases, free of medical conditions contraindicating regular high intensity exercise, not taking drugs of nutritional supplements known to be anabolic, not currently doing regular physical training, not pregnant One supervised exercise group, one home-based exercise (or control) group Duration of intervention: 24 weeks Supervisor: physiologists Amount of supervision: phone calls every two weeks Training supervisor: unknown Setting: rheumatology department Functional ability: HAQ Muscle strength: isokinetic dynamometer, knee extension Disease activity: DAS28 Score: 4 (criteria, 1, 4, 5, 6)

Baseline 24 weeks Home-based programs versus other home-based programs Van den Berg 2006

38 males / 122 females Median (IQR) age: individualized group 50 (13) yrs, general group 50 (14) yrs Median disease duration: individualized group 8 (9) yrs, general group 6 (11) yrs Inclusion criteria: RA according tot ACR 1987 criteria, not being physi- cally active for 30 minutes in succession at a moderate intensity level on at least 5 days a week, the availability of a computer with internet facilities, being able to cycle on a bicycle ergometer, being interested in a study on a physical activity program Exclusion criteria: having cardiopulmonary conditions that would not allow moderately intensive exercise Two physical activity programs, individualized and general (promoting moderate intensity activities, 30 minutes/day, ≥ 5 days/week) Duration: 12 months Supervisor: physical therapists Amount of supervision: individualized group once weekly by email, general group no personal contact with supervisor Training supervisor: unknown Setting: outpatient clinic of 3 hospitals Functional ability: HAQ Disease activity: DAS28Score: 5 (criteria 1, 2, 3, 4, 6)

Baseline 3 months 6 months 9 months 12 months Sten- ström 1994

Prospective, randomized clinical trial, 2 groups 15 females / 14 males Median (range) age: goal setting group 54 (26-68) yrs, pain attention group 58 (43-69) yrs Median (range) disease duration: goal setting group 10 (3-46) yrs, pain attention group 8 (3-35) yrs Inclusion criteria: RA according to the ARA criteria, age below 70, ARA functional class II, willingness to participate in the study Exclusion criteria: unable to perform the exercise program One home-based exercise program, goal contingent, one home- based exercise program, pain contingent Duration of intervention: 12 weeks and a 12 months follow-up Supervisor: physical therapists Amount of supervision: once and 3 phone calls after 1, 4 and 8 weeks Training supervisor: unknown Setting: rheumatological clinic Functional ability: HAQ Disease activity: Ritchie’s Articular Index

Score: 3 (criteria 3, 4, 5) Baseline 12 weeks 12 weeks follow- up

• • • • • • • •

(15)

Table 1. Description of home-based physical activity and exercise programs (n RCTs = 11) in patients with RA. (continued) StudyMethodParticipantsInterventionOutcomesMethodolo- gical qualityTime-points O’Brien 2005Pros- pective, randomized clinical trial, 3 groups

46 females / 21 males Mean (SD) age: joint protection program + strengthening + stretching exercises 62 (10) yrs, joint protection program + stretching exercises 57 (8) yrs, joint protection program 60 (13) yrs Mean (SD) disease duration: joint protection program + strengthening + stretching exercises 18 (14) yrs, joint protection program + stretching exercises 13 (12) yrs, joint protection program 10 (8) yrs Inclusion criteria: RA according to ARA criteria, age over 18 yrs Exclusion criteria: recent changes in drug regime, surgery of the wrist, hand, elbow or shoulder within the previous 6 months, sensory impair- ments to the hand, uncontrolled pain affecting the joints of the wrist or hand, and pregnancy One joint protection program + strengthening + stretching exercises, one joint protection program + stretching exercises, one joint protection program Duration of intervention: 6 months Supervisor: hand therapists (physical therapists and occupati- ons therapists) Amount of supervision: one information session Training supervisor: unknown Setting: rheumatology outpatient clinic Functional ability: AIMS Muscle strength: gross grip Disease activity: swollen and tender joints Score: 5 (criteria 1, 2, 3, 4, 5)

Baseline 1 month 3 months 6 months Sten- ström 1996, 1997

Pros- pective, randomized clinical trial, 2 groups 39 females / 15 males Median (IQR) age: high intensity program 54 (45-64) yrs, relaxation program 55 (48-64) yrs Median (IQR) disease duration: high intensity program 15 (7-24) yrs, relaxation program 7 (2-22) yrs Inclusion criteria: RA according to ARA criteria, aged 65 or below Exclusion criteria: in need of extensive treatment such as increase medication, inpatient rehabilitation, and/or surgery, already exercising within the health care system, severe psychosocial problems One home-based high intensity program, one home-based relaxation program Duration of intervention: 3 months Supervision: physical therapists Amount of supervision: one information session, phone calls after 1 week and every month Training supervisor: unknown Setting: rheumatology outpatient clinic Functional ability: AIMS Disease activity: Ritchie Articular Index

Score 4 (criteria 1, 3, 4, 5)

Baseline 3 months 12 months follow-up Häk- kinen 1998, 2001, 2003, 2004, 2004

Pros- pective, randomized clinical trial, 2 groups 40 females/25 males Mean (SD) age: intervention group 49 (10) yrs, control group 49 (11) yrs Mean (SD) disease duration: intervention group 10 (10) months, control group 8 (12) months Inclusion criteria: RA according to ARA criteria Exclusion criteria: previously taken glucocorticoids or disease-modi- fying anti-rheumatic drugs (DMARDs) One home-based strength training group, one control group (instruction for ROM and stretching exercises) Duration of intervention: 24 months Supervision: physical therapists Amount of supervision: 3 sessions and check-ups at 6, 12, 18 and 24 months Training supervisor: unknown Setting: hospital Functional ability: HAQ Muscle strength: dynamometer Disease activity: DAS28 Radiological damage: Larsen score Score 4 (criteria 1, 2, 4, 5) Baseline 6 months 12 months 18 months 24 months 36 months follow-up

• • • • • • • •

(16)

In eight trials instructions were provided by physical therapists [15, 16, 21, 22, 26-29]; in one trial by physical therapists and occupational therapists specialized in hands [25]; and in one trial by physiologists [23]. In one trial the professional background of the persons delivering the program was not described [24], although it was mentioned that they were trained to provide the program.

Risk of bias in included studies

Three trials fulfilled five out of the six key domains of risk of bias, five trials fulfilled four domains, while three trials met three domains. The following criteria were the least often fulfilled: allocation concealment (n = 7), adequate sequence generation (n = 4) and free of other bias (n = 4) (see table 1 ‘Characteristics of included studies’

for detailed information on risk of bias).

Table 2. Excluded studies

Callahan 2008 Self-reported arthritis Mengsoel 2005 Obsercational study Buljina 2001 Not home-based Bulthuis 2007 OA and RA mixed Bilberg 2005 Supervised program Lineker 2001 No RCT Goeppinger 2007 No RA patients de Jong Supervised program Barlow 2000 No RA patients

Callahan 2007 No RA patients Siu 2003 No RCT Scholten 1999 Not home-based

Nour 2007 OA and RA mixed Hakkinen 2003 Supervised program Lorig 1999 No RA patients Osborne 2007 Mixed population Suomi 2003 No RA according to criteria Lorig 1998 No RA patients Eversen 2007 Not home-based Moffet 2000 No control group Bostrom 1998 Not home-based

Nour 2006 OA and RA mixed de Jong 2003 Supervised program Kauppi 1998 No RCT

Lorig 2005 No RA according criteria Oh 2003 No randomised control group Rall 1996 No RCT Schoster 2005 Self-reported arthritis Belza 2002 OA population Noreau 1995 Supervised program

Marcora 2005 Supervised program Beame 2002 No RCT Baslund 1993 Supervised program

Ekdahl 1992 No RCT Hazes 1996 No RCT Perlman 1990 Observational study

Stenstrom 1996 Mixed population Lee 2006 Observational study Hoenig 1993 No relevant outcome Hakkinen 1994 Mixed population Munneke 2005 Supervised program Minor 1989 Not home-based

Ekdahl 1992 No RCT Neuberger 1997 No RCT Van Deusen

1987 No relevant outcome

Hawkes 1986 Not home-based Manoj Dash No RCT Danneskiold

1987 Observational study

Anandarajah 2004 No RCT McMeeken 1999 No home-based Byers 1985 No relevant ouscome

Martin 2002 No RA patients Bell 1998 Not really home-based Nordemar

1981 No relevant outcome Boutaugh 2003 No RA patients Komatireddy 1997 Not home-based Karten 1973 Not home-based

Barlow 1998 No RA patients Andersson 1996 Not home-based Stenstrom

1991 Supervised program Van den Ende

2000 Not home-based Lyngberg 1994 Observational study Ekblom 1995 Intervention not suf- ficiently described Hakkinen 1997 RA criteria unclear Brighton 1993 No relevant outcome

measures

• • • • • • • •

(17)

• • • Results

Home-based programs versus no intervention

Two trials compared a home-based exercise program (duration 8 and 12 months, respectively) to no intervention [15, 24]. Both home-based programs included aerobic capacity and/or muscle strength exercises. Neither of the trials had a follow-up period after the intervention. Both trials had a low risk of bias.

Concerning the effectiveness immediately after the home-based exercise program, in one study [24] a statistically significant effect was found on functional ability (p < 0.001) compared to no intervention. In the other study [15], a statistically significant effect was found on muscle strength (p < 0.05) compared to no intervention.

In neither of the two studies a greater increase of pain or disease activity was seen in the home-based exercise group than in the control group. In contrast, in the study by Maseiro et al [24], self-reported pain decreased significantly more in the intervention group (p=0.001) than in the control group. None of these trials included radiological damage as an outcome measure of safety.

Qualitative analyses: In Table 3 the GRADE score for each outcome measure is presented. Based on these results it can be concluded that there is low quality evidence that a home-based exercise program (including aerobic capacity and / or muscle strength exercises) is more effective with regard to functional ability compared to no intervention at all. Furthermore, there is moderate quality evidence that a home-based exercise program (including aerobic capacity and / or muscle strength exercises) is more effective with regard to muscle strength as compared to no intervention. There is low quality evidence that a home-based exercise program (including aerobic capacity and / or muscle strength exercises) is more effective with regard to self-reported pain compared to no intervention.

There is low quality evidence that home-based exercise programs (including aerobic capacity and / or muscle strength exercises) have no deleterious effects on disease activity. No conclusions can be drawn regarding the effects of the different interventions on the outcomes of aerobic capacity and radiological damage.

Quantitative analyses: Data could not be pooled due to clinical heterogeneity.

• • • • • • • •

(18)

Table 3. Results home-based programs versus no intervention StudyFunctional abilityAerobic capacityMuscle strengthSelf-reported painDisease activityRadiological damage Brodin 2008HAQ Intervention group Baseline: 0.5 (0, 2.25) 12 months: 0.5 (0, 2.0) Control group Baseline: 0.5 (0, 2.50) 12 months: 0.5 (0, 2.50) P between groups: ns -Grippet Intervention group Baseline: 370 (42, 1083) 12 months: 440 (32, 1236) Control group Baseline: 308 (24, 1100) 12 months: 340 (20, 1064) P between groups: < 0.05 ES 1.14 (95% CI: 0.83, 1.45) VAS Intervention group Baseline: 23 (0, 93) 12 months: 23 (0, 93) Control group Baseline: 22 (0, 98) 12 months: 26 (0, 98) P between groups: ns DAS28 Intervention group Mean difference 12 month - baseline: -0.12 (1.09) Control group Mean difference 12 months – baseline: 0.05 (1.03) P between groups: ns

- Masiero 2007HAQ Intervention group Baseline: 1.20 (0.56) 8 months: 0.93 (0.44) Control group Baseline: 1.17 (0.57) 8 months: 1.24 (0.54) P between groups: 0.000 ES -0.68 (95% CI: -1.17, -0.20)

--VAS Intervention group Baseline: 46 (21.6) 8 months: 37 (26.3) Control group Baseline: 39 (26.9) 8 months: 4.2 (16.4) P between groups: 0.001 ES -0.54 (95% CI: -1.02, -0.06) RAI Intervention group Baseline: 18.8 8 months: 17.9 Control group Baseline: 21.4 8 months: 22.1 P between groups: ns

- Qualitatice analy- sis (GRADE)Low quality evidence (inconsistency of results, imprecision of results)

Moderate (imprecision of results)Low (inconsistency of results, imprecision of results)Low (inconsistency of results, imprecision of results)- Quantitative analysis (pooled effect size)

n.a.-n.a.n.a. n.a.-

• • • • • • • •

(19)

Home-based programs versus supervised programs

In four trials a home-based exercise program (with a duration of 8 weeks, 12 weeks, 24 weeks and 24 months, respectively) was compared to a supervised program [21- 23, 29]. All of these four trials included home based exercise programs comprising aerobic capacity and/or muscle strength exercises and similar supervised programs.

One of the trials included a follow-up period of 12 weeks after the intervention [29]. Two of the four trials had a low risk of bias [22, 23].

Concerning the effectiveness immediately after the home-based exercise program, in none of the trials a statistically significant difference was found for functional ability compared to a supervised exercise program. In two of the studies [22, 29] the effect on aerobic capacity was significantly smaller after a home based program than after a supervised program (p < 0.05 and p = 0.001, respectively). In addition, in two of the trials a home-based exercise program was significantly less effective with respect to muscle strength (p = 0.02 and p = 0.006, respectively) than a supervised program [23, 29]. In the one trial including a follow-up period [29] the differences with respect to aerobic capacity and muscle strength did not remain after follow-up.

In none of the studies a statistically significant difference was found with regard to self-reported pain, and disease activity. In the one study [21] measuring radiological damage, no statistically significant difference was found between a home-based and a supervised exercise program.

Qualitative analyses: In Table 4 the GRADE score for each outcome measure is presented. Based on these results it can be concluded that there is low quality evidence that there is no difference in effectiveness between a home-based exercise program and a supervised exercise program (both including aerobic capacity and muscle strength exercises) with regard to functional ability. There is low quality evidence that a home-based exercise program is less effective with regard to aerobic capacity and muscle strength compared to a supervised exercise program (both including aerobic capacity and muscle strengthening exercises).

With regard to safety, there is low quality evidence that there is no difference in effect with regard to self-reported pain, disease activity and radiological damage

• • • • • • • •

(20)

Table 4. Results home-based programs versus supervised programs StudyFunctional abilityAerobic capacityMuscle strengthSelf-reported painDisease activityRadiological damage Van den Ende 1996

HAQ Supervised high intensity program Baseline: 0.83 (0.61) 12 weeks: -0.05 (-0.19, 0.09) 12 weeks follow-up: 0.05 (-0.11, 0.20) Supervised group program Baseline: 0.72 (0.53) 12 weeks: -0.05 (-0.20, 0.11) 12 weeks follow-up: -0.11 (-0.30, 0.07) Supervised individual program Baseline: 0.83 (0.65) 12 weeks: -0.03 (-0.13, 0.07) 12 weeks follow-up: 0.02 (-0.20, 0.23) Home program Baseline: 0.70 (0.61) 12 weeks: -0.03 (-0.13, 0.07) 12 weeks follow-up: 0.02 (-0.20, 0.23) P between groups: ns VO2 max Supervised high intensity program Baseline: 27.6 (7.1) 12 weeks: 4.7 (2.6, 6.8) 12 weeks follow-up: 1.1 (-1.6, 3.9) Supervised group program Baseline: 28.2 (9.5) 12 weeks: 0.9 (-1.2, 3.1) 12 weeks follow-up: 1.1 (-1.2, 3.4) Supervised individual program Baseline: 32.8 (9.8) 12 weeks: -1.2 (-2.8, 0.4) 12 weeks follow-up: -1.2 (-3.4, 1.0) Home program Baseline: 25.8 (6.1) 12 weeks: 0.3 (-1.6, 2.2) 12 weeks follow-up: 0.5 (-1.6, 2.7) P between groups: 0.001 (high intensity vs all other groups) Mean overall gain in muscle strength Supervised high intensity program Baseline: 12 weeks: 16.8% (8.4, 25.3) 12 weeks follow-up: 11.3% (3.4, 19.3) Supervised group program Baseline: 12 weeks: 7.7% (0, 15.4) 12 weeks follow-up: 17.0% (0.9, 33.1) Supervised individual program Baseline: 12 weeks: 7.4% (-2.3, 17.1) 12 weeks follow-up: 9.3% (-2.2, 20.9) Home program Baseline: 12 weeks: 0.9% (-6.0, 4.3) 12 weeks follow-up: -1.9% (-5.8, 2.0) P between groups: 0.02 (high intensity vs home program) -RAI Supervised high intensity program Baseline: 12.0 (7.9) 12 weeks: -0.5 (-2.2, 1.2) 12 weeks follow-up: -0.1 (-2.1, 1.9) Supervised group program Baseline: 10.7 (7.2) 12 weeks: -0.5 (-2.3, 1.9) 12 weeks follow-up: 0 (-2.8, 2.7) Supervised individual program Baseline: 12.5 (7.1) 12 weeks: 0 (-1.8, 1.8) 12 weeks follow-up: 0.3 (-2.1, 2.6) Home program Baseline: 12.4 (7.2) 12 weeks: 0.2 (-1.3, 1.7) 12 weeks follow-up: 0.1 (-1.5, 1.8) P between groups: ns

- Hansen 1993HAQ Home-based program Baseline: 0.63 (0.38/0.88) 24 months: 0.25 (0.12, 0.75) Home + sup. individual program Baseline: 0.57 (0.35/0.88) 24 months: 0.69 (0.44/1.15) Home + sup. group program Baseline: 0.50 (0.50/0.75) 24 months: 0.87 (0.11/1.25) Home + sup. group program in water Baseline: 0.75 (0.47/0.94) 24 months: 0.50 (0.18/1.00) No training Baseline: 0.50 (0.13, 0.88) 24 months: 0.62 (0.50, 1.06) P between groups: ns

Aerobic fitness Home-based program Baseline: 29 (26/37) 24 months: 26 (25/34) Home + sup. individual program Baseline: 32 (30/35) 24 months: 28 (24/31) Home + sup. group program Baseline: 41 (34/44) 24 months: 36 (32/42) Home + sup. group program in water Baseline: 32 (27/38) 24 months: 31 (25/35) No training Baseline: 34 (30/40) 24 months: 29 (27/35) P between groups: ns Isometric strength Home-based program Baseline: 29 (24/44) 24 months: 38 (32/52) Home + sup. individual program Baseline: 40 (24/55) 24 months: 63 (49/79) Home + sup. group program Baseline: 31 (26/38) 24 months: 57 (32/63) Home + sup. group program in water Baseline: 34 (22/42) 24 months: 54 (37/69) No training Baseline: 29 (19/35) 24 months: 45 (28/51) P between groups: ns VAS Home-based program Baseline: 1.6 (1.0/2.1) 24 months: 1.4 (0.9/2.0) Home + sup. individual program Baseline: 1.8 (1.1/2.6) 24 months: 1.9 (1.4/2.3) Home + sup. group program Baseline: 1.9 (1.2/2.6) 24 months: 2.1 (1.6/2.6) Home + sup. group program in water Baseline: 1.9 (1.5/2.2) 24 months: 1.4 (0.9/2.5) No training Baseline: 1.9 (1.5/2.3) 24 months: 1.9 (1.1/2.3) P between groups: ns No. of swollen joints Home-based program Baseline: 3.5 (1.3/5.8) 24 months: 3.5 (0.2/6.5) Home + sup. individual program Baseline: 3.2 (2.8/6.3) 24 months: 3.5 (0.4/5.8) Home + sup. group program Baseline: 2.3 (1.0/5.6) 24 months: (4.8 (0.4/8.0) Home + sup. group program in water Baseline: 3.6 (1.5/4.5) 24 months: 2.1 (1.1/5.9) No training Baseline: 3.8 (0.5/6.5) 24 months: 5.3 (0.3/12.5) P between groups: ns X ray scores Home-based program Baseline: 40 (13/79) 24 months: 50 (15/86) Home + sup. individual program Baseline: 49 (30/62) 24 months: 56 (35/84) Home + sup. group program Baseline: 37 (17/76) 24 months: 46 (32/89) Home in water Baseline: 33 (17/85) 24 months: 48 (22/88) No training Baseline: 68 (58/89) 24 months: 74 (65/97) P between groups: ns

• • • • • • • •

(21)

Table 4. Results home-based programs versus supervised programs (c0ntinued) StudyFunctional abilityAerobic capacityMuscle strengthSelf-reported painDisease activityRadiological damage Hsieh 2009HAQ Supervised program Baseline: 0.44 (0.42) 8 weeks: 0.36 (0.31) Home-based program Baseline: 0.41 (0.37) 8 weeks: 0.32 (0.27) P between groups: ns VO2 max Supervised program Baseline: 9.4 (1.8) 8 weeks: 10.7 (1.9) Home-based program Baseline: 9.5 (3.9) 8 weeks: 8.9 (2.8) P between groups: <0.05 ES 0.77 (95% CI: 0.03, 1.52) Grip strength Supervised program Baseline: 10.46 (2.66) 8 weeks: 12.00 (3.70) Home-based program Baseline: 12.27 (4.93) 8 weeks: 13.70 (5.37) P between groups: ns VAS Supervised program Baseline: 3.60 (1.88) 8 weeks: 2.70 (2.14) Home-based program Baseline: 2.70 (2.14) 8 weeks: 1.79 (2.42) Within group p< 0.05 P between groups: ns No. of swollen joints Supervised program Baseline: 9.07 (10.40) 8 weeks: 8.40 (9.93) Home-based program Baseline: 7.00 (4.42) 8 weeks: 6.13 (4.52) P between groups: ns

- Lem- mey 2009

HAQ Supervised program Baseline: 0.91 (0.68) 24 weeks: 0.82 (0.69) Home-based program Baseline: 0.58 (0.62) 24 weeks: 0.58 (0.59) P between groups: ns -Isometric knee extension Supervised program Baseline: 323 (79) 24 weeks: 404 (122) Home-based program Baseline: 308 (131) 24 weeks: 329 (123) P between groups: 0.006 ES: 0.34 -DAS28 Supervised program Baseline: 3.29 (1.27) 24 weeks: 3.12 (1.34) Home-based program Baseline: 3.28 (1.27) 24 weeks: 3.56 (0.71) P between groups: ns

- Quali- tative analysis (GRADE appro- ach

Low (inconsistency of results, high likelihood of bias)Low (inconsistency of results, high likelihood of bias)Low (inconsistency of results, high likelihood of bias)Low (inconsistency of results, high likelihood of bias)

Low (inconsistency of results, high likelihood of bias)Low (inconsistency of results, high likelihood of bias) Quan- titative analysis (effect size)

n.a.n.a.n.a.n.a.n.a.n.a.

• • • • • • • •