Clinical functional outcome measures for children with cerebral palsy after gait corrective orthopaedic surgery: A scoping review

To identify the most frequently reported non‐instrumented measures of gait, activity, and participation in children with cerebral palsy (CP) after undergoing gait corrective orthopaedic surgery.

of orthopaedic surgical interventions must go beyond just assessing the changes in body structure and function associated with gait.
The introduction of the World Health Organization's International Classification of Functioning, Disability and Health (ICF) 16 encouraged a multidimensional perspective of measuring and documenting health outcomes.This framework created a shift in focus from just measuring gait-related outcomes ('impairment' of body function and structure) to also measuring 'activity' (action or task execution) and 'participation' (involvement in life situations) outcomes 16 to allow for more meaningful assessment of the child's function after gait corrective orthopaedic surgery.There is a long-held assumption that functional activity and participation will improve if gait impairments are treated. 12owever, due to limited reporting of activity and participation measures to reflect surgical outcome, 17 the correlation between gait impairment and activity and participation is poorly understood. 7Changes in gait may not necessarily translate into clinically significant or meaningful changes in activity and participation. 18espite the large body of literature evaluating the outcomes of gait corrective orthopaedic surgery for children with CP, 19 there is lack of consensus regarding which noninstrumented gait, activity, and participation outcome measures are most suitable to be collected before and after surgery. 17,20Therefore, the aim of this scoping review was to identify the most frequently reported non-instrumented measures of gait, activity, and participation in children with CP undergoing gait corrective orthopaedic surgery.By doing so, we may be able to help direct further work towards developing a core set of outcome measures for use in routine clinical practice.

Search strategy
A systematic search was conducted on the 9th December 2021 according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Scoping Review Consolidated Standards of Reporting Trials guidelines 22 of four databases from inception: MEDLINE, Embase, EMCare, and PubMed.The search term strategy, allowing for syntactical differences between databases, was as follows: 'cerebral palsy' AND 'orthopaedic surgery' OR 'orthopedic surgery' OR 'multi-level surgery' AND 'gait' OR 'gait analysis' OR 'assessment tool' OR 'functional mobility' OR 'walking capacity' OR 'endurance' OR 'fitness' OR 'energy expenditure'.All terms were designed to be searched within both titles and abstracts.

Eligibility criteria
Original full-text, randomized and non-randomized, peerreviewed publications were included if: (1) they evaluated the outcomes of gait corrective orthopaedic surgery as the primary intervention for children with CP; (2) participants were under the age of 18 years at the time of surgery; (3) outcome measures were reported before and after surgery; and (4) studies used an outcome measure of gait, activity, or participation (as categorized by the ICF). 16Studies that included the use of instrumented measures of gait such as 3DGA could only be included if they also reported on a non-instrumented measure of gait, activity, or participation.Articles that were unavailable in English were translated with Google Translate.
Studies were excluded if the only outcome measure was an assessment of an impairment of body structure (e.g.muscle strength, muscle tone, passive range of motion) or quality of life.Systematic reviews, meta-analyses, conference abstracts, and commentaries were also excluded (Table S1).

Reference review
Once duplications were removed, two authors (MF and LJ) screened titles and abstracts independently against the eligibility criteria.Any disagreement was resolved through discussion.The full texts of the studies included in this review were then reviewed independently against the inclusion and exclusion criteria, with any doubt over inclusion of an article resolved by a third investigator (NG).Targeted reference scanning and citation tracking of the included studies were also used to minimize the chance of missing key studies.

Data extraction
Data were extracted from each manuscript independently by two authors (MF, LJ), including study type, sample size, length of follow-up, and outcome measures.Extracted patient demographic information included sex, age, and Gross

What this paper adds
• Standardized outcome measures of gait, activity, and participation are needed after gait corrective orthopaedic surgery.

• The Edinburgh Visual Gait Scale and Functional
Mobility Scale should be included in a comprehensive suite of outcome measures.
Motor Function Classification System (GMFCS) level I and II (walking independently), III, or IV (walking with an assistive device) at the time of surgery.Authors were not contacted for further information.

Data analysis
Collected outcomes focused on before and after surgical measures of gait and function and were classified according to the ICF model under the gait (body and structure), activity, and participation domains. 16Data reporting participant characteristics from all studies were pooled to calculate descriptive information including sample size, GMFCS level, sex, mean and SD or median age at surgery, and follow-up time.The frequency of individual outcome measures and the frequency of measures within the ICF domains were calculated as percentages.

Search results
The search strategy yielded 547 unique records from which 442 were screened (title and abstract) after duplicates were removed.A total of 60 full-text records were screened for eligibility, from which 32 records were included.A manual screen of references identified 12 additional records, resulting in a total of 44 publications representing the final selection for data extraction (Figure S1).

Study type, design, and size
Of the 44 publications, 21 were prospective cohort studies, 17 retrospective cohort studies, five randomized controlled studies, and one phenomenographic study.Twenty-nine were non-comparative studies (Table 1) and 15 were comparative (Table 2).Sources of funding for each of the 44 studies are available in the Table S2

Length of follow-up
The duration of follow-up varied between 6 weeks and 20 years 6 months with a mean duration of 38 months and median duration of 24 months.Just over half of the studies (n = 26) assessed outcomes at multiple time points, with studies using different combinations of 6 weeks, 2 months, 3 months, 6 months, 9 months, 12 months, 18 months, 24 months, 36 months, 3 to 4 years, 5 years, 6 to 12 years, and 18 years; of these, the final median follow-up duration was 24 months.All these studies except two 36,49 assessed each outcome at the same time points.The outcome measures that used a different time point to other assessments were questionnaires (Functional Mobility Scale [FMS] and author-designed) reporting mobility at a much later time after surgery (at 5 and 8 years) (Table 3).

Outcome measures
For all reported measures, four studies used a noninstrumented measure of gait, 40 used a measure of activity, and 15 used a measure of participation.No studies evaluated outcomes using a combination of a non-instrumented measure of gait with both activity and participation measures (Tables 1 and 2).

Non-instrumented gait measures
The Edinburgh Visual Gait Scale (EVGS) was the only non-instrumented measure of gait and was used in four studies. 34,42,45,50The EVGS was measured after surgery at different time points for each of the four studies: 6 and 12 months; 34 17.1 months (median); 42 6 years 8 months (median); 45 and 2 years 7 months and 18 years (median). 50otably, 25 studies included 3DGA, with one study 42 reporting gait with both the EVGS and 3DGA.Sixteen studies did not report any measure of gait.

T A B L E 3 (Continued)
and participation.There is a lack of consistency and standardization of use of outcome measures, which was reflected in the reporting of a wide variety of activity and participation measures in the literature.Activity measures were the most frequently reported measures of outcome; the only non-instrumented tool for evaluating gait impairment was the EVGS. 34,42,45,506][67][68] It has also been identified as the most appropriate tool compared to other visual gait assessments used in a clinical setting 65 with good intra-and interrater reliability, 14,[28][29][30][31] having the highest correlations for concurrent and criterion validity of gait function 65 and being sensitive to orthopaedic surgical intervention. 34When gait impairment is the sole outcome measure, assessment of community function is incomplete.Therefore, gait impairment outcome should be used in combination with activity and participation outcome measures.
Measuring activity is important to both families and clinicians to assess if surgery has made a positive impact on how well the child performs or executes a task, thus leading to improvement in daily functioning.The FMS was the most reported activity measure identified in this review.It is administered as a patient-or observer-reported outcome measure.The FMS has been shown to discriminate between different GMFCS levels, 69 as well as having sound clinimetric properties and clinical utility. 29,69Clinicians tend to choose outcome measures that are quick to administer, easy to use, and do not require extensive training.The FMS complies with all these characteristics.However, a limitation of the FMS is that it only assesses walking ability over specific distances (home, school, and community distances) and lacks evaluation of functional activity in other contexts.Therefore, in determining activity outcomes, the FMS should be complemented by other objective assessments to quantify the child's postsurgical skills and capacity.
The 66-or 88-item GMFM, Gillette FAQ, and gait velocity were other common measures of activity identified by this review.Although the GMFM has good reliability and validity [32][33][34][35][36] and is considered the criterion standard tool for assessing gross motor function in children with CP, 34,36 it has ceiling effects for children classified in GMFCS level I, 34,37,38 accounting for approximately 30% of the study participants included (according to their presurgical GMFCS level).Similarly, the Gillette FAQ, a 10-point scale of parentreported walking ability, has good inter-and intra-reliability for community ambulators 77 but is not sensitive with floor effects for use with children classified in GMFCS level III who require assistive devices and mobility aids for daily functioning. 39Thus, assessment selection may need to be adjusted to children's functional abilities.For example, it may be more appropriate to assess children classified in GMFCS level III with the GMFM and use the Gillette FAQ for children classified in GMFCS level I or II.However, no studies in this review used different assessments for different GMFCS levels.
Many studies suggested that velocity of gait is an important measure of functional status in children with CP [40][41][42] and a useful measure of performance-based community walking activity. 80It is often used as an outcome measure when there are time and resource constraints because it only requires a stopwatch and measured distance to calculate gait velocity. 43Gait velocity was reported in 25% of studies in the absence of gait kinematics.This is comparable with previous systematic reviews examining orthopaedic gait surgeries in children with CP. 41,44 Gait velocity is not only a measure of walking ability but a crucial component in participating in life events. 80For example, a child with CP may want to walk at a velocity to keep up with peers or to meet environmental demands when crossing a road or using public transport.Hence, measuring velocity is only meaningful when evaluated in a life situation and should be assessed with a participation goal in mind.
Outcomes from this review revealed that less than half of the studies included participation measures and that these were an underused domain for measuring surgical outcome.In addition, measures that capture individualized child and family goals, a central tenet in paediatric rehabilitation, were lacking.Recent studies reported the importance of patient goals in determining the effectiveness of surgery. 83,84The PODCI was the most common participation measure used for children undergoing gait corrective orthopaedic surgery.The PODCI is a parent-and adolescent-reported outcome measure focusing on function and quality of life: upperextremity and physical functioning; transfer and basic mobility; sports and physical functioning; pain and comfort; happiness; satisfaction; and expectation.Despite the frequency of use of the PODCI, there is debate on its responsiveness when used with orthopaedic surgery in children with CP 19,36,45 due to large ceiling effects regarding reports of comfort and pain. 18,23,74Given that a reduction in pain was specifically identified as an important surgical outcome in a recent international multi-stakeholder consensus report, 85 the PODCI may be insufficient as a participation measure for gait corrective orthopaedic surgery.
For a comprehensive understanding of surgical outcome, a suite of outcome measures across all domains of the ICF is necessary to reflect outcomes that are meaningful to the patient, family, and surgeon. 19A combination of performancereflective questionnaires (to identify how a child participates in real-life situations), in addition to clinical objective measures (to quantify the child's underlying skills and capacity) are necessary to fully appreciate surgical outcome from the perspective of the child and family. 19The most common combination of measures identified in this review was the PODCI with velocity and the PODCI with the GMFM.This is comparable to the systematic review by Schiariti et al. 86 who identified that a combination of measures, including the GMFM or Pediatric Evaluation of Disability Inventory, plus the PODCI and Cerebral Palsy Quality of Life Questionnaire are required to capture the functional profile of the child with CP.
The selection of a suitable time frame to evaluate the postsurgical outcomes is an important consideration when assessing the impact of the intervention on functional change.The vast range in follow-up times reported in the literature (ranging from 6 months to 20 years after surgery) may be reflective of clinical reality but should also be considered when evaluating and comparing outcomes for children in different GMFCS levels, where postsurgical recovery rates may differ.For example, a child in GMFCS level III requiring surgical correction at multiple levels may require longer recovery time compared to a child in GMFCS level I who may only need correction at one level.Harmonization of timing of postoperative outcomes will assist in combining data for future research.However, our review indicated that in general, a 24-month follow-up time for children in GMFCS levels I to III will see functional change that is consistent with the study by Thomason et al., 13 who reported significant improvements in function for children in GMFCS levels II and III 24 months after multilevel surgery.
A limitation to the interpretation of this review is the heterogeneity of the participants and study interventions included.Although gait corrective orthopaedic surgery is only performed in children classified in GMFCS levels I to III, five non-comparative studies included participants in GMFCS level IV (although participants were few) who were non-independent walkers requiring an assistive device.The contradictory inclusion criteria may affect the choice of outcome measures, with authors leaning towards using activity measures responsive to lower functioning abilities.Furthermore, in three comparative studies, participants received selective dorsal rhizotomy as a stand-alone treatment or in addition to orthopaedic surgery.Because selective dorsal rhizotomy has a slower postoperative recovery rate with more intense rehabilitation requirements, there may be a biased influence of outcome measures chosen to accommodate this.There was also heterogeneity between interventions.For most comparative studies, lower-limb surgery was the intervention of interest, but two studies also evaluated different postsurgical rehabilitation styles as the intervention.This would create a bias in the selection of outcome measures to reflect rehabilitation outcome performed in a clinical setting.
Although achieving its purpose, this scoping review has been limited to identifying the most frequently reported outcome measures of gait, activity, and participation after gait corrective orthopaedic surgery.To provide robust recommendations in standardizing the use of measures in gait corrective orthopaedic surgery, further research is needed to report on the clinimetric properties of identified measures because frequency of use of a measure may not necessarily equate to being the most appropriate.In addition, future work is needed that considers other clinically relevant assessments for children with CP that have yet to be the focus in the orthopaedic literature, such as the Gait Outcomes Assessment List 87 (participation measure), High-Level Mobility Assessment Tool 88 (activity measure), or Canadian Occupational Performance Measure 89 (participation measure).Furthermore, as no studies in this review used a combination of non-instrumented measure of gait with both an activity and a participation measure as recommended by the ICF model, this should be a point for future research in the development of a comprehensive set of outcome measures.

Conclusion
This scoping review provides fundamental knowledge that can contribute to the development of a minimal core set of measures used in gait corrective orthopaedic surgery.Despite being the criterion standard, the lack of availability of 3DGA in most orthopaedic centres has led to this review identifying the most appropriate non-instrumented measures of gait and function in those settings.While there is consistency in the use of the EVGS and FMS to assess gait and functional activity in children classified in GMFCS levels I to III, there is uncertainty regarding the most suitable participation outcome measures.Also, to develop a comprehensive suite of outcome measures, the need to identify relevant activitybased clinical measures (especially for GMFCS levels I and II) and performance-reflective questionnaires is evident.This highlights the gap in knowledge of outcome measure selection that is meaningful and important to both the clinician and family and the need for further work to standardize the outcome measures used in gait corrective orthopaedic surgery in the clinical setting for children with CP.

•
Abbreviations: GMFCS, Gross Motor Function Classification System; ICF, International Classification of Functioning, Disability and Health; NR, not reported; PC, prospective cohort; Phenom, phenomenographic; RC, retrospective cohort.a Reported as decimal years as in original studies.b Median.
do not have access to the criterion standard 3DGA, there is a need to identify other relevant non-instrumented measures of gait and function for use in this setting.We found that no studies used outcomes assessing gait corrective orthopaedic surgery across all three domains of the ICF of impairment (via non-instrumented measures, i.e. not 3DGA), activity,