Self-management interventions are patient-centred and designed to foster active participation of patients in order to promote well-being and to manage symptoms. Over the past two decades, the role of self-management in chronic diseases has gained momentum. Self-management programmes are now acknowledged as a key element of quality care. New modes of delivery allow greater access to information and are tailored to address patient needs. This systematic review presents data from clinical studies of self-management over the past decade, summarises the evidence for programme effectiveness, and suggests future research directions.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Self-management is the ‘individual's ability to manage the symptoms, treatment, physical and psychosocial consequences and life style changes inherent in living with a chronic condition.’1 While often instigated by individuals, health professionals and patient organisations facilitate self-management in partnership with patients. Traditional patient education offers information and technical skills. In contrast, self-management interventions (SMIs) are problem focused, action-oriented and emphasise patient-generated care plans.2 SMIs include educational, behavioural and cognitive approaches to influencing health knowledge, attitudes, beliefs and behaviours to promote independence, maintain or adjust life roles, and address the psychological impact of disease. SMIs address five skills: problem solving, decision making, resource utilisation, collaborative patient/provider relationships and action taking.3 SMI success depends on needs assessments identifying the concerns of specific groups.
Health professionals view SMIs as structured education to develop patients' illness-management skills. However, patients view these differently.4 A qualitative study5 reported that patients with arthritis viewed SMIs as a way to bring order into their lives, helping them to recognise boundaries, mobilise resources, cope with change in self-identity, and plan, pace and prioritise. This study furthers the understanding of SMIs because it identifies and articulates the value of patient-centred approaches and has implications for SMI evaluation in terms of frequency and methodology. SMI evaluation occurs in stages—before, during and after intervention—representing a series of snapshots along a continuum and may miss the full impact on the patient. In addition, differences in outcome data may occur depending on who has asked questions and how questions were asked.
In the 1980s, Lorig and associates developed the Arthritis Self-Management Program (ASMP) to enhance well-being and quality of life. The original ASMP was not theory based, but considered that knowledge leads to behaviour change. During programme evaluation, participants linked programme satisfaction with an increased sense of disease control, thus Bandura's Social Cognitive Theory (SCT)6 was selected as a framework for change. As more SMIs developed, researchers recognised the value of peer-led as well as professional-led programmes.7 Lorig's ASMP is the most developed and studied of SMIs and is disseminated by patient organisations across the globe.
Most SMI research evaluates group programmes delivered face-to-face. Financial constraints and increasing access to technology have driven mailed and internet programme development. The ASMP group (Stanford University, USA) leads the way in developing new technologies for self-management. However, many SMIs are still delivered one-to-one by health professionals, the least evaluated delivery mode.
This review focuses on the most common conditions for which arthritis SMIs are provided: osteoarthritis (OA), rheumatoid arthritis (RA) and fibromyalgia (FM). Several systematic reviews of arthritis patient education and SMIs have evaluated studies of OA and RA6,–,11 published up to 2002 and for FM up to 1999.12 13 Reviews of RA SMIs concluded that only those using psychobehavioural approaches led to short-term (up to 9 months) significant improvements in function, but effects are rarely sustained.8 9 OA and RA reviews concluded that 40% of SMIs led to improved symptoms and disability.10 11 Effective programmes were more common in OA than RA, possibly because OA SMIs more often included exercises and recruited larger samples. Programmes based on cognitive-behavioural theory (CBT) or SCT yielded better outcomes. Few studies examined effectiveness for 12 months or more. Two systematic reviews of SMI trials in FM concluded that multimodal interventions yielded better outcomes than single-mode interventions, but improvements were modest and not maintained.12 13
Methodological limitations of SMI trials were highlighted, such as insufficient patient-relevant outcomes, lack of consensus for outcomes and patient attributes examined, and low statistical power. Recommendations for future SMI trials included: use OMERACT (outcome measures in rheumatoid arthritis clinical trials) core sets; assess fatigue and social participation; have sufficient power; be disease specific; identify which patient and programme characteristics yield beneficial outcomes; explicitly state theoretical basis; publish programme protocols/manuals; use leaders trained in group facilitation, problem-solving, goal-setting and CBT and/or SCT approaches; and use strategies for long-term behavioural change (eg, booster sessions).9 11
Materials and methods
For this review, Medline, PubMed, EMBASE, CINAHL, AMED, PsychInfo and the Cochrane Library were searched from 2000 for FM, from 2003 for OA and RA, and up to November 2009 using the key terms ‘arthritis’, ‘OA’, ‘rheum$’, ‘FM’, ‘self-management’, ‘self care’ and ‘patient education’. Studies were included if they: were in English; involved randomisation; stated that they were SMIs. In total, 30 articles that met the entry criteria were identified; some were economic evaluations and long-term follow-ups of earlier trials. Studies were excluded if they only reported observational data or long-term follow-up data from previous trials without comparison group data. This review aims to identify particularly whether longer-term (ie, ≥12 months), as well as short-term, benefits resulted, by diagnosis, and if any programme or patient characteristics are associated with improvement (table 1).
SMIs for patients with OA, RA or other arthritic conditions
Seven studies were identified: two were published within one article and three articles related to one study14,–,21 (table 2). Fifty to 75% of participants had OA, 15–35% had RA and 15–17% had other diagnoses. Two studies included 30–52% of people with FM (who may have other arthritis diagnoses).16 20 Six studies evaluated the ASMP or variations thereof, all including weekly exercise practice (stretch, strength and walking): two evaluated the ASMP,15 20 two a mailed, individualised version of the ASMP (SMART: provided over 12–18 months),14 one an internet-ASMP,16 and one a ‘one-time’ mailed ASMP version.21 The seventh evaluated the individually tailored ‘I'm Taking Charge of My Arthritis’ (ITCA) programme provided in individual home visits.17,–,19 All were community-based or home-based and all were either led by peer moderators or were self-completed (apart from the ITCA study, which was led by health professionals). All programmes involving contact lasted 6 weeks (in person or internet), and the other two were self-completed at a pace to suit participants. Two studies included booster sessions.14 17 18 Apart from the ITCA, all programmes included similar ASMP content and teaching strategies. Cognitive and behavioural strategies were taught by trained leaders, following the ASMP leader manual, using SCT approaches to facilitating change.
Methodologically, five studies used intent-to-treat (ITT) analysis14 15 21 (and only four included control groups14 16 17 21).Two conducted a priori sample size analyses and recruited sufficient participants,20 21 and a further four14,–,16 recruited 340–1090 patients, suggesting that these were large enough to detect differences.
All studies evaluated pain, function, mood, self-efficacy and health behaviours (table 1). Five reported follow-ups of 12 months or longer.14,–,16 20 At 12 months, all five showed improvements in self-efficacy: four in function,14,–,16 two in pain,15 16 and one in mood.16
The least effective ASMP recruited predominantly African–American participants.20 Most effective ASMP studies were delivered to predominantly non-Hispanic white participants, either face-to-face14 or via the internet.16 Most participants (64–75%) had OA. Only one study conducted subgroup analyses by diagnoses, showing that the internet ASMP was most effective in OA, next in RA, and not at all in FM.16 Follow-up longer than 12 months occurred in only one study with a control group comparison. At 2 years, the SMART programme, including booster materials, reduced perceived arthritis severity and doctor visits, but, by 3 years, there were no differences.14
SMIs for patients with OA
Eight studies were identified.22,–,29 Economic analyses were conducted for two studies30 31 (table 3). Three studies enrolled only patients with knee OA, and five studies enrolled patients with hip and/or knee OA. Three used a cluster randomised controlled trial approach with medical practices as the unit of analysis.22 23 27 Most participants (∼75%) were women. Three studies evaluated the group-format ASMP.24 28 29 Five evaluated other SMIs, comparing these with usual care or information booklets.22 23 25,–,27 Three studies were group interventions: two combining SMIs with supervised and home exercise practice23 26; and two evaluated individual SMIs.25 27 Most group interventions lasted 12 h over 6 weeks, apart from one of 4 h plus an individual home visit.22 None included booster sessions. Individual interventions lasted 1 h,25 or duration was not stated,27 with both including telephone follow-up. All but one study28 recruited patients from primary care. All but one24 were led by a health professional—general practitioner (GP), nurse and/or physical therapist—and group SMIs used SCT approaches. Methodologically, five studies used ITT analysis,24 26,–,29 and a priori power calculations were reported in four (table 3).
All studies evaluated pain and function. Four also examined health knowledge, behaviours and care use, three measured self-efficacy, but mood and fatigue were rarely assessed (table 1). Four studies only reported short-term outcomes (4–9 months), with three reporting improvements in pain and function,26,–,28 and one brief individual SMI reporting no differences.25 Two studies found no significant differences in outcomes at any follow-up, both having shorter duration interventions of 1 or 5 h.22 25 Two studies23 24 included economic analyses. Although both showed health benefits, only one demonstrated cost-effectiveness,31 a group format being better than individual interventions.26 Overall, of the four studies that evaluated long-term outcomes (ie, 12 months or longer), significant modest effects were found for pain,23 29 self-efficacy,24 29 function23 and mood.24 The most effective interventions longer term were protocolised group SMIs, including 6 weeks of supervised exercise, using either CBT or SCT approaches led by trained leaders.23 29
SMIs for patients diagnosed as having inflammatory arthritis
Five studies were identified32,–,36 (table 4). Four recruited people with RA only, and one, pragmatically, recruited people with RA (50%), early inflammatory arthritis (35%) and psoriatic arthritis (14%).36 Three studies recruited people with well-established disease (average 12–15 years), and two with early-to-established disease.33 36 The studies showed some similarities: all evaluated small group programmes (up to 10 participants), were hospital-based, and led by health professionals. All except one33 had control groups: usual care, information booklets or attention control (an information programme). All used ITT analysis. In other respects, studies were diverse. Sample sizes ranged from 59 to 218, with only three reporting sample size analyses and recruiting sufficient participants to detect differences.32 35 36
Programme duration varied from 12 to 52 h, over 6 weeks to 9 months, with three including booster sessions (3–9 months after the programme). Programmes also varied in content and delivery. All provided information about the disease and medication. Four explicitly used SCT and/or CBT approaches. Two were led by clinical psychologists using CBT combined with education from multidisciplinary team members using ‘more traditional didactic approaches’ for physical strategies.33 34 Two described use of leader manuals, staff training in SCT/CBT approaches, and applying these throughout programmes.32 36 One study did not explicitly describe its theoretical framework.35 Only two included regular exercise.35 36
All studies evaluated pain, function and mood. Disease status and self-efficacy were measured in four studies, and fatigue in three (table 1). Short-term benefits were found in four studies. Three had longer-term follow-up at 12 months, but only two showed benefits.32 36 Both used SCT/CBT approaches throughout the programmes, specially trained health professionals and longer-term booster sessions. Riemsma et al32 identified improved self-efficacy and fatigue in people attending a primarily cognitive strategy self-management programme without their partner, while those attending with partners showed no sustained benefits. Hammond et al36 showed significant improvements in pain, self-efficacy, perceived control and health behaviours. This was the only programme that consistently enabled both cognitive and behavioural strategies using SCT/CBT approaches with substantial skills practice. Patients also had shorter disease duration (average 7 compared with 12–15 years).
SMIs for patients with a diagnosis of FM
Seven studies were identified37,–,43 (table 5). Five of them recruited people from secondary care with FM diagnosed by rheumatologists using American College of Rheumatology criteria, and the other two recruited patients from primary care with FM diagnosed by GPs.37 41 Programmes were similar with respect to group size (up to 12 participants) and leadership (health professionals). In addition, an eighth study (internet-based ASMP) reported FM results separately16 (table 2). Most studies included a control group of usual care, wait list control, information leaflets or an attention control (relaxation) group. Only one had no control group, comparing exercise, SMI or combined exercise and SMI.43 Sample sizes ranged from 53 to 207, with all but two37 39 including sample size calculations and recruiting sufficient numbers to potentially demonstrate differences. Only four included ITT analysis.16 38 41 42 Programme duration varied between 18 and 50 h, with four being wholly or partially community based (eg, leisure centres). Four programmes lasted 2.5–6 weeks, and four 10–16 weeks. None included booster sessions.
All programmes included information about FM and its management. Nine different SMIs were evaluated, as one study evaluated two styles of SMI.37 Of these, six explicitly used SCT and/or CBT approaches, with four reporting use of trained leaders experienced in these (CBT group only37).16 42 43 All included cognitive components (eg, managing pain, relaxation), and six included supervised exercise practice. A systematic review identified that supervised moderate-intensity aerobic exercise (minimum 12 weeks, 3×/week) is effective in FM.44 However, only two programmes combined SMIs with this exercise intensity,38 43 with only one showing short-term improvements.43 Three studies included at least 10 sessions of supervised exercise spread over 6–10 weeks.39 41 42
All studies evaluated pain, function, perceived health and mood. Fatigue was measured in seven, self-efficacy and social role/support in four, health behaviours in three, and healthcare use in only two (table 1). Considering studies using either case-completer or ITT analysis, short-term benefits were found for 4–6 months in three studies for three or more outcomes.39 42 43 Three had controlled follow-ups of 8–12 months,16 40 42 but only two showed minimal continued benefits: in fitness40 and self-reported improvement.42 In summary, no FM studies showed that SMIs, with or without exercise, sustained benefits across a range of outcomes for more than 6 months.
Summary of the evidence
Previous reviews have identified that psychobehavioural SMIs lead to short-term benefits (ie, 6–9 months) but that longer-term benefits are not often evaluated and rarely sustained.11 32 In this updated review, we aimed to identify whether recent SMI trials identify any longer-term benefits and whether any programme or diagnostic or patient characteristics are associated with improved outcomes.
Of the 30 studies identified, only 14 had follow-ups of 12 months or longer, seven of which (two of the same SMART ASMP) led to sustained benefits in pain and/or function.14,–,16 23 29 36 All were methodologically sound—that is, met at least seven of the van Tulder criteria45: the method of randomisation was adequate and treatment allocation was concealed (apart from two that evaluated the ASMP without control groups,14 15 as an earlier trial had demonstrated that effectiveness46 groups were similar at baseline for the most important prognostic factors), the assessor (if relevant) was blinded to the allocation, timing of assessment in both groups was comparable, the drop-out rate was described and acceptable, and an ITT analysis was used. Little is known about benefits beyond 12 months as only two had longer-term follow-ups, although both showed some continuing benefits.14 23 Previous research has identified that the ASMP led to sustained benefits over 4 years and was cost-effective46—a multidisciplinary 20 h programme over 8 weeks led to significantly less pain in comparison with the control group at 5 years,47 and a 10 h educational-behavioural programme over 4 weeks led to significantly better function in comparison with the control group at 4 years48—but more research is needed in other SMIs. We recommend that SMI studies that demonstrate benefits at 6–12 months should seek funding to extend follow-up and evaluate cost-effectiveness, to provide data for service development decisions.
In terms of programme characteristics, the effective SMIs were the group ASMP, a variant (SMART or internet versions)14,–,16 29 or acknowledged ASMP influence on structure and format.23 36 Thus, common programme features were: duration of at least 6 weeks; explicit use of SCT and/or CBT approaches; individualised weekly action plans with progress review; highly protocolised with participant handbooks; and led by the same trained leaders. All face-to-face programmes included weekly supervised exercise and graded home aerobic exercise (eg, walking, tai chi). Three included aerobic exercise within programmes,24 29 36 and the mailed/internet ASMPs strongly promoted exercise action planning with clear details about exercise performance. We recommend these strategies be incorporated into clinical programmes and research trials. Evidence-based guidelines should include studies that meet SMI criteria3 and not treat all interventions equally. Although eight studies used boosters,14 17,–,19 27 32 34,–,36 only three showed significant benefits.14 27 36 Further evaluation of the timing, delivery (phone, mailing or group) and effectiveness of booster sessions is warranted.
In terms of patient characteristics, there is little research on who benefits most from SMIs. The majority of ASMP trials recruited community or online volunteers, who may be more motivated to change. In ASMP studies recruiting patients from primary or hospital settings (all with OA), there is conflicting evidence, as two studies (with enhanced exercise) were effective,28 29 but a third was not,24 supporting an earlier study's findings.49 Only one ASMP trial has evaluated whether different diagnostic groups fare better, concluding that the internet version is more effective in OA, less so in RA, and not effective in FM.16 Another study32 reported that patients with RA who attended without their partner fared better, as they could openly discuss problems without loved ones present. In addition, those not benefitting seemed less motivated to participate and change and had poorer health and more stressful life events.4 In an FM study, patients with initially higher self-efficacy reported greater improvements.42 Further research is needed to appropriately target interventions and/or modify SMIs to meet the needs of specific patient groups.
To best address the needs of patients with arthritis, we recommend that researchers who have demonstrated positive outcomes from SMIs with sufficient sample sizes should undertake secondary analyses to investigate whether any baseline characteristics predict better outcomes. We also recommend evaluating longer-term (12 months or more) benefits of SMIs, using protocolised SCT/CBT-based programmes of sufficient duration, incorporating exercise for six sessions or more (longer for FM, including twice-weekly supervised graded aerobic exercise), led by either specially trained health professionals or peer leaders and which incorporate assessment of patient-relevant outcome. Increased use of qualitative methods50 integrated into clinical trials (eg, patient diaries (written or audiorecorded), focus groups or interviews) are suggested to promote participants' expression of any perceived benefits on overall well-being, and to explore their perspectives on SMI effectiveness. Use of such qualitative approaches will help investigate the full impact of SMIs on individuals, determine the most patient-relevant outcomes for future trials, investigate the process of change, and shed light on what patient attributes influence benefits (or lack of) to better target and time SMIs (box 1).
Patient perspectives on self-management programmes and future directions
Undertake secondary analyses to investigate further whether any patient attributes are associated with better outcomes.
Examine longer-term (12 months and longer) benefits of self-management interventions (SMIs).
Standardise implementation using protocols and participant handbooks. Ensure programmes are Social Cognitive Theory (SCT)/cognitive-behavioural theory (CBT) based, use trained peer leaders or health professionals, are of sufficient duration, and incorporate exercise.
Patient diaries, creative writing or interviews are recommended to explore participants' overall well-being and processes of change and ascertain patient-relevant outcomes. For example, a woman attending a SMI course reported walking to her local shops for the first time in 12 years. Taking over 2 h to make the 200 yard trip was entirely irrelevant to her. Such change is not captured in outcome measures. Her ability to control her life was her relevant outcome.
Include patient-relevant outcomes. If using questionnaires, more frequent measures of outcomes may be useful in tracking the full impact of SMIs than longer intervals (baseline, 3 months and 1 year). Assess value of outcome from patient's perspective through utility weighting or other metric.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.