Articles and clinical guidelines with full-text availability, published in the English language between 01 January 2001 and 31 December 2021, that described therapeutic interventions and programmes designed for the prevention of MSK complications for children aged 0–18 years with severe CP (GMFCS levels III–V) were eligible. There was no limit on sample size or study design. Where more than one version of a clinical guideline was retrieved, only the most recent was included.

One researcher (S.v.A.) performed two independent extractions and critical appraisals for each article, at a minimum of 8 weeks apart to improve accuracy and consistency. Study details were then extracted onto a spreadsheet recording first author, year of publication, country or region of origin, sample size, age, diagnosis and GMFCS level, other participants, study design, main subject or concept, intervention, outcomes measured, main results or recommendations, adverse effects and identified gaps in knowledge (Van Aswegen et al. 2024:S2). The level of evidence of each record was noted, based on the work of Ackley et al. (2008), and methodological appraisal was applied using the AGREE-II assessment for clinical guidelines³ and Joanna Briggs Institute (JBI) checklists for the other studies.⁴ The JBI checklist and AGREE-II scores were reflected as very low, low, moderate and high, based on percentage of agreement with quality criteria for each record (Van Aswegen et al. 2024:S3, S4, S5).

As this review formed part of a larger study involving human research, ethical approval was obtained for the whole project from the University of Cape Town’s Human Research Ethics Committee with reference no. 024/2022.

Three CPGs came from national or regional health websites in HICs – the National Institute for Health and Care Excellence (NICE) in the United Kingdom (NICE 2012), the Waikato District Health Board (WDHB) in New Zealand (WDHB 2014) and the American Academy for Cerebral Palsy and Developmental Medicine (AACPDM) (Paleg et al. 2019). Two comprehensive SRs of reviews of therapeutic interventions aimed to guide clinical practice (Morgan et al. 2021; Novak et al. 2020), and one article was an SR of guidelines for the care of persons with CP, initiated by the WHO (Damiano et al. 2021). All guidelines adhered to AGREE-II and PRISMA criteria.

The NICE guideline (NICE 2012) and AACPDM Care Pathway (Paleg. et al., 2019) focussed on general medical management of CP and central hypotonia in children. No specific regimens or home programmes were given, but conditional recommendations for interventions targeting the body structures were included. The WDHB guided therapeutic management, providing detailed recommendations for assessment, interventions and equipment needs, all stratified into GMFCS levels and age bands.

One comprehensive SR addressed evidence-based medical and therapeutic interventions for children with CP, but recommendations for children with GMFCS levels IV and V were few and of questionable feasibility in RLSs (Novak et al. 2020).An SR on early intervention guidelines (Morgan et al. 2021) focussed on improving motor function and could make few recommendations in 0- to 2-year-olds for prevention of MSK complications. The WHO review (Damiano et al. 2021) only identified five official guidelines for CP of all ages, and relevant recommendations referenced guidelines already retrieved. Table 1 summarises the recommended non-medical modalities for prevention of MSK complications from each of the included sources.

From the results in Table 1, the conditional recommendations for the prevention or control of MSK complications for children with severe CP were:

The next sections discuss the evidence that informs the implementation and application of these recommendations as well as aspects to consider for successful home-based interventions.

At a Mac Keith Multidisciplinary meeting held in 2006, postural management was defined as ‘a planned approach encompassing all activities and interventions which impact on an individual’s posture and function’ (Gericke 2006). For this review, we developed a narrower definition, that is, ‘the supportive positioning of the whole body in lying, sitting and standing to preserve normal MSK alignment and to provide central support for functional activity’. This would include supported standing but exclude a focussed modality such as applying an orthosis.

Of the 27 articles included in this section, all were from HICs, except one from Malaysia (Htwe et al. 2016). Included articles were 18 cohort or quasi-experimental studies (Capati et al. 2020; Gibson et al. 2009; Holmes et al. 2003; Htwe et al. 2016; Kim et al. 2013; Macias-Merlo et al. 2015, 2016; Martinsson & Himmelmann 2011, 2021; Mol et al. 2012; Paleg et al. 2021; Porter et al. 2008; Picciolini et al. 2009, 2016; Pountney et al. 2002, 2009; Tornberg & Lauruschkus 2020; Vekerdy 2007), seven SRs (Chung et al. 2008; Gmelig Meyling et al. 2018; Humphreys et al. 2019; Paleg et al. 2013; Pérez-de la Cruz 2017; Pin 2007; Wynn & Wickham 2009), an expert consensus (Gericke 2006) and a narrative review (Kittelson-Aldred & Hoffman 2017).

Interventions in all three positions were represented in the studies, almost half reporting on standing programmes alone. Outcomes most often related to hip displacement, followed by lower limb ROM. Several sources reported on functional impact of postural management regimens and qualitative measures, for example, ease of performing daily activities, and sleep quality when using sleep systems. Table 2 and Table 3 provide detailed results of studies involving 24-h PM regimes and lying, seating and standing interventions.

For children in GMFCS levels IV and V, the expert consensus from 2006 recommended lying or sleeping with postural support at night, sitting from 6 months of age and supported standing from 12 months (Gericke 2006) The researchers found no revisions or updates for this regimen.

Unsupported holding, feeding and lying positions for children with GMFCS level V tended to result in spinal, pelvic and hip deformities in up to 95% of subjects (Porter et al. 2008). However, using a consistent 24-h PM approach in lying, seating with hip abduction for 5 to 6 h daily (Kim et al. 2013) and/or standing from the age of 18 months and upwards for 30–90 min per day significantly controlled and/or improved hip migration percentage (MP) and reduced frequency of spinal orthotic prescription, Botulinum toxin injections (BoNT) and hip surgery indications at age 5 (Gmelig Meyling et al. 2018; Pérez-de la Cruz 2017; Picciolini et al. 2016, 2009; Pountney et al. 2002, 2009).

Weak and inconclusive evidence from three studies reported the effects of using commercial sleep systems. Two reported significant reduction of hip MP after a minimum of 6 months and up to several years of consistent use, and hip ROM improved along with MSK comfort levels. Thermal comfort, however, was sometimes reduced, and parents required training to use these systems (Humphreys et al. 2019; Wynn & Wickham 2009). Caregivers reported that sleeping systems did not cause significant sleep disturbance, especially if used night and day (Mol et al. 2012).

Seating that incorporated a 3-point force system (Holmes et al. 2003), or modular or moulded contours with optional tilt, produced a decreased spinal Cobb angle and improved thoracic posture when used for at least 4 h per day (Chung et al. 2008). In addition, caregivers from two studies found trunk-supported seating eased everyday caregiving and improved the child’s social interaction (Chung et al. 2008; Vekerdy 2007).

Supported standing regimes were reported to significantly improve a range of outcomes, especially when commenced as early as 9 months of age (Paleg et al. 2013). Standing for 30 to 90 min per day for 3 to 7 days per week for young children from the age of 12 months up to 5 years significantly decreased hip MP at 5 years of age, the effects enhanced with additional total hip abduction up to 60° (Htwe et al. 2016; Macias-Merlo et al. 2016; Martinsson & Himmelmann 2011, 2021; Paleg et al. 2013). Standing for at least 30 min per day for at least 3 days per week also improved bone mineral density (BMD) significantly in the LLs and spine, especially when combined with whole body vibration (Paleg et al. 2013; Pin 2007) and reduced spasticity in the LLs for up to 35 min post intervention, which in turn led to a less scissored gait pattern (Macias-Merlo et al. 2015; Paleg et al. 2013; Pin 2007; Tornberg & Lauruschkus 2020). Standing for 45-60 min per day 3-5 days per week for 8-15 months was found to significantly increase LL ROM, which led to improvements in gait, social interaction and decreased burden of care (Capati et al. 2020; Gibson et al. 2009; Macias-Merlo et al. 2015; Martinsson & Himmelmann 2021; Paleg et al. 2013; Tornberg & Lauruschkus 2020).

Twelve articles, mostly from HICs, were included in this section, but only three were primary studies – a case-controlled time series from 2003 (Fragala et al. 2003) and two quasi-experimental trials (Laessker-Alkema & Eek 2016; Uzun Akkaya & Elbasan, 2021). The rest included six SRs (Autti-Rämö et al. 2006; Craig et al. 2016; Eldridge & Lavin 2016; Groppe et al. 2012; Wiart et al. 2008), two narrative reviews (Physiopedia 2016; Wilton 2003) and an expert consensus (Lannin et al. 2011). Within the five SRs that included manual stretching, there was considerable overlap in the sources, with several studies repeatedly cited (Craig et al. 2016; Eldridge & Lavin 2016; Groppe et al. 2012; Pin et al. 2006; Wiart et al. 2008). Overall, the evidence was scarce.

Although stretching is widely regarded as an essential part of conservative joint range maintenance in CP (Wiart et al. 2008), the researchers found no clear consensus on the scope of ‘stretching’ techniques nor any standardised stretching dosage regimes. All sources described manual (intermittent) stretching of a joint or muscle by a therapist, maintained for up to 60 s per repetition. Most sources also included at least one of the following: manual stretching with the addition of electrical stimulation (ES); active stretching with child participation; positional or sustained stretching, lasting 15 min or more and using equipment such as a standing frame or an orthosis; intramuscular stretching (similar to massage) and serial casting where a series of rigid casts were applied to a joint for progressive lengthening of shortened tissues. As the use of orthotics in children with severe CP is largely for joint range maintenance (Lannin et al. 2011), five studies reporting on orthotic use were included here. Outcomes included calf muscle spasticity, LL ROM and gross motor function. Characteristics of studies on stretching interventions are presented in Table 4.

There is overall weak and conflicting evidence for stretching interventions. Manual stretching interventions appear to be less effective for preventing deformities than sustained, positional stretching (Craig et al. 2016; Pin et al. 2006). Fragala et al. (2003) found that periods of non-intervention from a school-based manual stretch routine of 3 × 60 s per structure, taking place once to twice per week, produced inconsistent changes in LL ROM. The addition of ES to the antagonist during stretching significantly increased ROM and reduced spasticity (Groppe et al. 2012). Uzun Akkaya and Elbasan (2021) reported transient reduction in calf muscle spasticity after just 10 min of manual stretching, both with and without concurrent massage. No adverse effects were reported.

Wilton (2003), Autti-Rämö et al. (2006), Groppe et al. (2012) and Craig et al. (2016) all suggested that joint ROM may be maintained or increased in all limbs and spasticity reduced by an intensive serial casting programme if sustained for at least 3 weeks.

Positional stretching for 20 min using an orthotic in sitting or using a stander for 30 min at least three times per week for at least 8 weeks may be effective to increase LL ROM and may reduce calf muscle spasticity for up to 35 min after the stretch (Craig et al. 2016; Groppe et al. 2012; Laessker-Alkema & Eek 2016; Lannin et al. 2011; Pin et al. 2006). Wilton (2003) recommended wrist and hand orthotics be used at night or for 3–5 h per day to control deformities where spasticity is moderate to high although all other studies found inconclusive evidence of efficacy for ULs.

Although no stretching interventions resulted in improved motor function for the child, improved flexibility appeared to assist with symmetry in sitting (Physiopedia 2016) and ease of care when performing ADLs (Laessker-Alkema & Eek 2016). Adverse effects included hindrance of hand function while wearing UL orthotics (Autti-Rämö et al. 2006; Lannin et al. 2011; Wilton 2003) and bruising, skin breakdown, sleep disturbance, joint swelling and muscle cramps were reported in a few patients using LL orthoses (Craig et al. 2016; Laessker-Alkema & Eek 2016).

No studies were found evaluating the effectiveness of HBIPs to prevent MSK complications in severe CP. Twelve articles addressed the population of interest and aspects of HBIPs that influenced success.

Six were conducted in LMICs or involved RLSs. Three were reviews (Branjerdporn et al. 2021; Lord et al. 2018; Paleg 2005), and nine were primary studies, comprising eight qualitative studies (Bischof & Chirwa 2011; Colver et al. 2012; Halvarsson et al. 2010; Novak et al. 2012; Krüger & Sello 2008; Lillo-Navarro et al. 2015; Naidoo et al. 2019; Rezaie & Kendi 2020) and one quasi-experimental trial (Zuurmond et al. 2018). Studies explored essential programme components, training needs of caregivers and primary health workers, determinants of successful programme implementation and the impact on child development. Table 5 provides further details of each study.

Three South African studies reported on caregiver and CHWs’ views on their capacity for identifying and managing CP in the community. Outcomes included knowledge of the condition, practical handling skills and pain caused during activities of daily living (ADL). A facility-based study found a moderately high prevalence of pain in children and young adults with severe CP during daily care activities, which they observed to be directly influenced by their handling techniques (Bischof & Chirwa 2011). Community healthcare workers required specific training in CP screening and clear referral pathways and protocols for further management (Naidoo et al. 2019). Caregivers desired better knowledge of the condition, practical training for everyday routines and ongoing involvement in support groups with health professionals and respected members of their communities (Krüger & Sello 2008).

Limited evidence suggests that, in spite of poverty and low education levels of caregivers, parent-implemented interventions in LMICs positively influenced posture, positioning, self-care, feeding and social function in children while reducing parental stress (Branjerdporn et al. 2021; Zuurmond et al. 2018), and that a positive, equipped and informed environment improved overall participation of children with CP at home, in school and the community (Colver et al. 2012).

Factors affecting fidelity and adherence to HBIPs by caregivers were the presence of trusting relationships between child, caregiver and therapist, and establishing skills and coping strategies (Lord et al. 2018). Priorities from caregivers included evidence-based interventions, the element of fun, the incorporation of the programme into daily routines and written instructions with demonstration (Lillo-Navarro et al. 2015). Caregivers prioritised programme flexibility and mobility goals for the child, as well as ongoing support for monitoring and programme adaptation over time (Halvarsson et al. 2010). Priorities from therapists included effective communication with families, joint goal setting and equipping in CP clinical skills and resources (Rezaie & Kendi 2020). Zuurmond et al. (2018) found that monthly telephonic or face-to-face check-ins with caregivers and a group chat via social media for therapists and/or CHWs were effective support mechanisms for their community-based programme.

Existing studies of PM have focussed on the role of sitting and standing programmes in prevention and control of the hip and LL deformities. Neuromuscular scoliosis is another serious, common complication of severe CP (Hagglund et al. 2018) that 24-h PM may influence; yet, supporting research is lacking. Studies on sleeping systems have been limited to commercially available options. The efficacy of low-cost equipment for lying positions should be a research priority, particularly in children classified as GMFCS levels IV and V who spend a significant portion of the day in lying.

Evidenced-based stretching protocols should be developed, including dosage and stretching modalities for different treatment goals. In our review, there were few reported adverse effects; however, some MSK discomfort during stretching might be expected. The contribution of any stretching to relieve or exacerbate pain is an important consideration and is likely to affect programme adherence. Other questions should be considered about the psycho-sensory aspects of touch during manual stretches and if it can be effectively combined with massage. These questions warrant further research.

Our review found very little research into content, efficacy of and adherence to HBIPs in RLSs that include physical modalities designed to prevent MSK complications in severe CP. Further research could help identify structural and social determinants of successful programmes, establish effective protocols and provide ways to quantify participation and functional outcomes.

This review charts the existing literature over the past 20 years on manual interventions for limiting MSK complications in severe CP, including general CPGs, and efficacy of individual modalities and home programmes. The existing knowledge and evidence gaps were highlighted to help prioritise future research for populations living in RLSs.

The search was broad, using multiple sources including online platforms and included quality rating to support any recommendations made. Using a non-categorical diagnostic approach might have yielded useful studies involving other diagnoses and increased our evidence base. Our limit on language may have excluded some relevant studies from LMICs but was necessary owing to limited translation services. Several of the retrieved articles were, however, originally published in other languages. Initial screening, extraction and quality appraisal by a single researcher could have resulted in selection bias, especially with such a broad scoping purpose. However, this risk was mitigated by establishing clear and detailed exclusion criteria a priori and bringing in a second, independent reviewer for full-text review. Measures were taken to improve intra-rater reliability.