Self‐management interventions for children and young people with sickle cell disease: A systematic review

Increasing numbers of interventions are being developed to support self‐management for children and young people (CYP) with sickle cell disease (SCD), but no systematic review has systematically synthesized this evidence regarding their characteristics, effectiveness, acceptability and feasibility for all published intervention types.

Patient and Public Contribution: Three young people with SCD recruited acted as the review advisors. They were formally trained in the review process and involved in every aspect of the review: the design, conduct and interpretation of the findings. CYP involvement in the interventions' development and implementation was analysed as part of the review. This systematic review was conducted as part of a wider research project titled: Understanding fatigue experiences of CYP with SCD to guide the co-development of a fatigue self-management intervention. Two of the young advisors involved in the review were also involved in the development of the project funding application. States and the United Kingdom, respectively. 2 Alterations in the shape of red blood cells characterize SCD, leading to blood vessel occlusion and inflammation, infarction, organ damage, pain and profound chronic anaemia. 3 Consequently, children born with SCD experience comorbidities, have high healthcare needs and reduced life expectancy. 4,5 With medical advances, mainly in high-income countries, children born with SCD transition into adulthood. However, SCD is associated with significant medical and psychosocial challenges that often worsen when children and young people (CYP) reach an age to become responsible for managing their condition. As CYP with SCD mature, they have increased decision-making autonomy and self-care independence. Self-management is the purposeful performance of specific learned tasks, activities and behaviours to manage the medical, psychosocial and life impact of a chronic illness. 6-8 SCD selfmanagement strategies include symptom monitoring, following treatment plans, and ensuring health maintenance practices. Selfmanagement is crucial as CYP with SCD begin caring for themselves and their illness.
SCD-related management strategies can negatively impact CYP's perceived quality of life, disrupt school and work attendance and participation and influence their social interactions and relationships. [9][10][11][12][13][14][15] This may reduce commitment to treatment plans and health maintenance practices. 16 Effective selfmanagement support can help mitigate treatment challenges, engage CYP, improve health and social outcomes and reduce illness burden and health costs. 16,17 Consequently, recent efforts have focused on developing and implementing interventions to support and improve self-management.
Previous reviews have focused on specific psychological, 18 medication adherence, 19 transition programmes 20 and e-health interventions, 21 Table 1. The search was unrestricted by country or setting, language or publication year.

| Study selection
All citation records from the database and manual searches were exported into EndNote 20 ® and deduplicated. Records were independently screened in Covidence ® by two reviewers (B. A. P. and S. K.). Any disagreements were adjudicated by the third author (K. M. A.). Reasons for excluding papers were noted.

| Quality assessment
Included papers were appraised for methodological quality independently by B. A. P. and S. K. using design-specific standardized critical appraisal instruments from the JBI System for the Unified Management, Assessment and Review of Information (SUMARI). 24 For each of the appraisal criteria, the individual appraisers assessed whether Inclusion criteria Exclusion criteria Population CYP aged 8-25 years with SCD. The upper age limit is consistent with definitions of young people by the UN, WHO and UK NHS. The lower age limit is informed by the evidence that children with long-term health conditions may begin to engage (informally) in self-management around this age. 8 Papers including samples described as CYP without further specificity around age. Papers involving samples primarily encompassing CYP with a reported mean age between 8 and 25 years. Papers including children and/or young people with long-term conditions that reported findings for CYP with SCD.
Only included participants over 25 years old. Mean age > 25 years.

Intervention
We defined self-management as purposeful performance of specific learned tasks, activities and behaviours to manage the medical, psychosocial and life impact of a chronic illness. [6][7][8] And we defined a self-management intervention as any intervention, programme or approach designed to develop the ability of and/or support CYP with SCD to manage their long-term health condition through education, training and support to develop their knowledge, skills or psychosocial resources. All self-management interventions that target CYP with SCD. All types/forms of self-management intervention-health, social care and educational interventions-designed to support or facilitate CYP with SCD to take control of and manage their condition, promote their capacity for self-care and/or maintain and enhance their physical and/or mental health.
Interventions in which children and/or young people are not actively engaged and/or remain passive recipients of knowledge or instructions. Self-management interventions that target professionals, parents/carers/siblings or families as a whole, without a distinct component(s) for CYP. Papers only describing the development of an intervention. Papers where the effects of the self-management intervention cannot be distinguished from broader interventions for SCD. | 581 the article met each quality criterion ('yes'), if it failed to meet the criterion ('no'), if insufficient information was presented to assess the specific criterion adequately ('unclear') or if the criterion was not applicable due to the study type, design and/or intervention ('N/A') (Tables 3-7). The total number of appraisal criteria met determined the overall quality of a paper. Papers that met three-quarters or more of the appraisal criteria were considered to be of high quality, between half and three-quarters as moderate quality and less than half as poor quality (Tables 3-7). The quality appraisal did not impact study inclusion or exclusion but was considered in the synthesis.

| Data extraction
A typology from previous studies on self-management interventions guided data extraction on intervention characteristics. 25

| Data synthesis
Results from the included papers were synthesized narratively in line with the JBI approach. As only one qualitative research paper was included, it was impossible to follow the JBI convergent integrated approach for mixed-method review. The characteristics of the different interventions were examined and summarized in relation to their effectiveness, acceptability and feasibility for implementation in 'real life' settings.

| Patient and Public Involvement (PPI)
The review team worked with a PPI group of three young people with SCD recruited from a patient support group. They were trained and involved at every review stage to ensure that patient perspectives (and concerns) were reflected in the review. They commented on the search strategy and the data extraction dimensions developed by the reviewers, helping to identify additional search terms and quality criteria. In addition, they highlighted issues for discussion and recommendations after reviewing the extracted and synthesized data.

| Methodological quality of included studies
Overall, the body of evidence was of moderate methodological quality (Tables 3-7). Several of the randomised controlled trial (RCTs) (n = 7) were assessed as being of poor quality. Four were feasibility RCTs with poor reporting quality. Only four RCTs were rated as high quality, three of which were feasibility/pilot RCTs. The RCT studies lacked details of randomization methods, group allocations, blinding of assessors, group comparativeness and statistical analysis methods.
The limitations of the quasi-experimental studies were their short follow-up length, the multiplicity of outcome measurements and the reliability of the outcome measurement tools used. The cohort and cross-sectional studies failed to report on how confounding factors were managed in the analysis. Only two quantitative study article 46 While studies reported differences in the characteristics of those who completed the study and those who dropped out, they failed to report how these had been factored into their analysis. Some RCTs only reported data on the intervention groups. In some studies, not all outcomes were measured for the comparison groups or intervention delivery and the timing of outcome assessments was variable between and within the groups. Several studies assessed outcomes using nonvalidated tools, or there were missing outcome data. Eleven papers reported high follow-up attrition rates (Table 2). Many studies experienced recruitment challenges, achieving less than 50% of their target sample size. Commonly reported reasons for recruitment difficulties included people's lack of interest in the research, time constraints, inconvenience of intervention format and delivery approaches and perceived lack of intervention benefits.

| Description of self-management interventions
Thirty-two self-management interventions were reported in the 38 included articles ( Several interventions were referred to by names and/or acronyms that reflected their intent or objective(s) (see Table 2).

| Format and delivery methods
Twelve interventions were delivered both in-person and remotely, 29

T A B L E 3 Methodological quality of included RCTs
Quality criterion Green et al. 44 Barakat et al. 30 Broome et al. 31 Crosby et al. 36 Daniel et al. 37 Gil et al. 42 Gil et al. 43 Hood et al. 46 Kaslow et al. 47 Ketchen et al. 48 McClellan et al. 49 Palermo et al. 50 Schatz et al. 54 Schwartz et al. 55 Smaldone et al. 58 Was randomization used for assignment of participants to treatment groups? Were participants blind to treatment assignment?
Were those delivering treatment blind to treatment assignment?

T A B L E 3 (Continued)
Quality criterion Green et al. 44 Barakat et al. 30 Broome et al. 31 Crosby et al. 36 Daniel et al. 37 Gil et al. 42 Gil et al. 43 Hood et al. 46 Kaslow et al. 47 Ketchen et al. 48 McClellan et al. 49 Palermo et al. 50 Schatz et al. 54 Schwartz et al. 55 Smaldone et al. 58 Were participants analysed in the groups to which they were randomized? Rodgers-Melnick et al. 52 Sil et al. 56 Sil et al. 57 Smith et al. 59 Treadwell and Weissman 60 Viola et al. 61 Wihak et al. 62 Yoon and Rodgers-Melnick et al. 52 Sil et al. 56 Sil et al. 57 Smith et al. 59 Treadwell and Weissman 60 Viola et al. 61 Wihak et al. 62 Yoon and  29 Estepp et al. 40 Saulsberry et al. 53 Were the two groups similar and recruited from the same population?
Were the exposures measured similarly to assign people to both exposed and unexposed groups?

T A B L E 6 Methodological quality of included cross-sectional studies
Quality criterion Crosby et al. 33 Crosby et al. 34 Crosby et al. 35 Were In all the studies where race and ethnicity were identified, the SCD population were of African American background. There appeared to be an assumption that participants would be of low socioeconomic status and have strong family systems due to their racial and ethnic identity.
These assumptions influenced the intervention format, delivery and target population, with attempts to make interventions financially accessible and/or involve family relatives. Some studies also highlighted the black culture of the SCD population and included black music, sports celebrities, mentors, navigators, interventionists and content from black TV shows to make them 'culturally sensitive'. 37,47,49,55,61,62 The agency of CYP, however, appeared to be overlooked, with most interventions developed with minimal or no CYP involvement (see Table 8) to tailor them to the values, expectations and hopes of the CYP. All the interventions were individual-or family-focused, aimed at knowledge, skill acquisition, reinforcement and/or behaviour change, underpinned predominately by psychological models (see Table 8). There was little or no consideration for CYP's wider social contexts like school and/or workplace.

| Outcomes measurement
Thirty-one studies assessed the effectiveness of self-management interventions using multiple outcome measures (Table 9). Fifteen were feasibility/pilot studies that assessed and reported intervention effectiveness as primary outcomes. 28 Different validated and unvalidated generic and SCD-specific outcome measurement tools were used (Table 4). Most tools had not been formally evaluated for use with CYP with SCD, and different tools were used to measure the same outcomes across the papers (Table 9). Only three papers 44,58,60 reported the measurement of physiological parameters ( Table 9). Most of the studies reporting intervention effectiveness failed to report on the effect size and/or 95% confidence intervals of the results.
The evidence below must, therefore, be considered in the context of these limitations. Table 9 summarizes the effectiveness data.

Visits, Emergency Room (ER) Visits
State-Trait Anxiety Inventory for adults or children, Tennessee Self-Concept Scale for adults, Piers-Harris Children's Self-Concept Scale, Human Drawings Test

w:
Nonsignificant change in in-patient and emergency room visits and self-concept. Improvement in anxiety for children (p < .01) and adolescents (p < .05); reported headaches as crises symptoms (p < .01); reported pain medication days (p < .001); reported pain intensity (p < .001) and self-treated crises (p < .05).

Moderate
Daniel et al. 37

1-6 m:
Nonsignificant improvement in pharmacy refill (p = .33) and self-reported medication adherence. Less decrease in Personal best HbF in the intervention group compared to controls (p = .009). Controlling for intervention month and group assignment, the intervention group, progressed to Personal best by 2.3% per month during Months 0-4 (p = .30).

Moderate
Hazzard et al. 45 Knowledge, Perceived Social Support, Coping Strategies How Much Do I Know About SCD Scale, Perceived Social Support-Friends Scale, Kidcope At hospital discharge: Significant effect on all outcomes (p < .001).

Moderate
Hood et al. 46 Self-Efficacy, Self-Management (goal and confidence)
The overall strength of the evidence base for the psychobehavioural outcomes limits confidence in the results, particularly since 16 of the 20 articles were of moderate to poor methodological quality.
Only one high-quality study associated the self-management intervention (e-mentorship by medical students) with significant improvements in self-efficacy. 61 Furthermore, effectiveness appeared to reduce overtime for self-efficacy and coping. 30,31,47,52 Eight studies of mixed quality examined the impact of selfmanagement interventions on healthcare utilization which included emergency room visits, in-patient admission duration and/or hospitalization rate. 29,30,32,35,38,40,52,56 Only a 12-month CYP-personalized transition programme delivered by a designated social worker reported a statistically significant reduction in hospitalization rate at 12 months follow-up. 29 This study involved the largest number of participants (n = 110); however, outcome measurement was based on a retrospective record review, which would not have included hospitalizations outside the study centre.
Health-related quality of life was measured in eight studies. 28,[35][36][37]41,48,58,61 Only three studies of moderate quality reported statistically significant improvement in quality of life. 28,41,64 However, in two, improvement was only observed in the social domain. 28,64 The high-quality studies identified no significant improvement in quality of life. 37,61 Overall, our greatest confidence is in the evidence of improvements in SCD knowledge and social outcomes, where studies of predominantly moderate methodological quality reported positive outcomes (Table 9).
Sixteen studies of strong to moderate quality reported a statistically significant impact of self-management interventions on these outcomes. 27,29,36,38,40,41,44,45,47,48,52,[57][58][59][60][61]63 Eight studies found that educational programmes on SCD and pain improved knowledge and understanding. 27,40,47,52,57,60,63,64 Six studies found a positive association between improved knowledge and understanding and self-management confidence, behaviours and practices. 27,41,46,53,63,64 Similarly, six studies also found improvements in social functioning and perceived social support. 38,45,47,48,59,60 Although evidence on knowledge and social outcomes is based upon high-quality studies, many used different T A B L E 9 (Continued)  Six studies measured medical adherence as medication adherence, CYP responsibility for treatment, clinic attendance or transition success. 29,40,44,58,60,61 Four studies reported significant improvements in these outcomes. 29,44,58 The effective interventions were behavioural and skills training programmes delivered by a designated social care provider 29,44,58 and an educational programme delivered by lay counsellors. 60 None of the technological interventions using treatment reminders improved treatment adherence. 40,61 3.6 | Acceptability and engagement interventions was reported to be higher amongst parent participants than CYP for the interventions that involved CYP-parent dyads. 44,51,58 In addition, engagement levels were reported to be higher among younger CYP participants than in the older CYP cohort in the same studies. 34,43,46,47 Some studies reported on CYP participants' feedback regarding the negative aspects of the interventions and their preferences. 44,47,50,58,61,64 Negative aspects included the burden of home-based self-directed activities, 47 the lack of mobile phone-based technology interventions, 50,64 the frequency of reminder notifications and the surveillance features of some interventions. 44,50,58,61 Interestingly, participants recommended incorporating in-person sessions for interventions involving virtual platforms for social interaction and strengthening peer support. 61,64 Some studies described how the professionals delivering the interventions 31,61 reported improved communication and relationships with CYP and improved understanding of their challenges.

| Evidence related to self-management intervention implementation
The main challenges reported for intervention implementation related to study recruitment and retention and provider (interventionist) availability. Eleven studies reported high follow-up attrition rates, which ranged between 32% 40 and 53% 57 ( Training each multidisciplinary team member to deliver selfmanagement interventions was reportedly facilitative. Several studies made recommendations to facilitate future intervention acceptability and engagement (Table 10).

T A B L E 10 Recommendations for future self-management interventions
Self-management programmes/interventions need to: m-Health self-management interventions need to: Be personalized to CYP's goals and values. 34,45,46 Be family-oriented. 30 Use multiple formats and delivery approaches to enhance flexibility, convenience, acceptability, usefulness and engagement. 37,50 Include components to facilitate communication with health and social care professionals. 50 Develop the broad range of knowledge and skills for managing SCD. 50 Support older CYP's capacity to navigate psychological, relational, educational, vocational, economic and health and social care system challenges. 29,30,43,47,64 Be integrated within the wider healthcare system. 46 Involve schools. 47 Designate a lead for self-management in the multidisciplinary team to support CYP. 30 Only a third of the technological interventions (4/12) showed effectiveness. These findings are consistent with the findings of other reviews in which limited data was found to support the effectiveness of mobile or web-based applications for self-management of adolescents with a chronic illness 68 and for treatment adherence among CYP and adults with SCD. 19,21 There was limited evidence for interventions improving healthcare utilization or quality of life.
Conflicting and inconsistent evidence for the effectiveness of selfmanagement interventions on quality of life in CYP with chronic illnesses has been found and reported by others. 66 In addition to previous research on the effectiveness of self-management for CYP with chronic illnesses, 66  effectiveness. 69 Ignoring the agency of CYP, particularly from minority backgrounds like CYP with SCD, also has the potential to create and deepen health disparities and inequalities. This supports embedded coproduction to ensure self-management intervention design and implementation are consistent with CYP's needs, voices, experiences and expectations rather than prioritizing medical adherence.
Further, interventions focused on role management (i.e., social domains and social participation) were limited. Despite the importance of the social context on self-management, 70,71 the included studies overlooked home, school and/or work environments. This is despite several research studies reporting the difficulties of engaging in SCD self-management practices in these settings. 9,11,[13][14][15] The lack of accommodation for SCD self-management practices in the home, school and workplace has been highlighted, including how school policies and practices contradict self-management. 9 Overall, the highlighted limitations mean that there is limited evidence for any single self-management programme or programme components in relation to promoting health and social outcomes for CYP with SCD. However, we found effective and acceptable interventions appear to include those that provide support from peers and trusted adults (e.g., parents, teachers, relatives and care providers). While the ultimate goal for self-management is to increase CYP with SCD autonomy and independence in managing their illness, they depend on others for encouragement, communication and treatment support. Adults (e.g., parents, relatives, care providers, teachers and adults with SCD) can provide information on how CYP can independently handle health and social care needs, while peers can learn from and support one another by sharing personal experiences. A recent review highlights the pivotal role peers and adults play in supporting the adoption and promotion of selfmanagement strategies in CYP with chronic conditions. 69 Future interventions should, therefore, include components that facilitate peer and significant adult support. The involvement of peers, families, care providers and teachers is particularly important because CYP with SCD have been found to experience significant stigma in relation to their illness management, even from families, care providers and teachers due to their illness and race/ethnicity. 9,11,72,73

| STRENGTHS AND LIMITATIONS
This is the first review of self-management interventions for CYP with SCD that synthesizes the evidence from the range of programmes developed, the range of outcomes assessed and the diversity of evaluative study designs. Including young people with SCD in the review's conduct has enhanced the credibility of the process. However, there are some limitations. Only peer-reviewed articles published in English were included, posing a potential publication bias. It was impossible to synthesize study results using meta-analysis due to the heterogeneity of the interventions and outcome measures. As previously described, the robustness of the included studies limits the confidence that can be placed on the results. In addition, the cost-effectiveness or cost-benefits of interventions has not been evaluated, and there is a lack of qualitative research to inform understanding of intervention effectiveness and acceptability.

| CONCLUSION
Findings generated from this systematic review provide a critical summary of the characteristics, effectiveness, acceptability and feasibility of self-management interventions for CYP with SCD. This review found that interventions that promoted peer, parent/family and care provider support and combined technology and in-person group methods appear to be associated with the effectiveness and acceptability to CYP. However, there is limited evidence for any single self-management programme or programme components in relation to promoting health/social outcomes for CYP with SCD.
Future studies need to be robust in design, conducted across multiple and different settings and collect long-term outcome data to assess the effectiveness, acceptability, cost-benefits and generalizability of interventions across populations and settings. In addition, it is important that more validated health-related outcome tools are developed for CYP with SCD; CYP are involved in intervention development, study design and implementation; CYP participants from non-Black backgrounds are included and that attention is paid to the wider context of CYP's lives when designing and evaluating self-management interventions.