Abstract
Purpose
-
Up to 90% of nonambulatory patients with cerebral palsy (CP) experience hip displacement during their lifetime. Reconstructive surgery is recommended. Redisplacement rate is an outcome parameter.
Methods
-
In a systematic literature review (MEDLINE, Embase and CENTRAL databases) until January 2023 we searched for reports with redisplacement rates after bony hip reconstructive surgery in nonambulatory patients. Quantitative data synthesis, subgroup analysis and meta-regression with moderators were carried out.
Results
-
The pooled mean redisplacement rate was 16% (95% CI: 12–21%) with a prediction interval of 3–51% (Q: 149; df: 32; P < 0.001; I2: 78%; τ2: 0.67 and τ: 0.82) in 28 studies (1540 hips). Varus derotation osteotomy (VDRO) alone showed a higher redisplacement rate than VDRO + pelvic osteotomy (30% vs 12%, P < .0001). Mean age in the VDRO-alone subgroup was 7.1 years and in the combined group 9.5 years (P = .004). In meta-regression, lower redisplacement rates were observed with higher preoperative migration index (MI) (correlation coefficient: −0.0279; P = .0137), where comprehensive surgery was performed. Variance in true effects are explained by type of bone surgery (57%), preoperative MI (11%), age (5%) and MI for definition of failure (20%). No significant reduction in the redisplacement rate could be observed over the mid-years of studies (1977-2015).
Conclusion
-
Our pooled data support the more extensive surgical approach in patients with high preoperative MI and emphasize the superiority of combined surgery. Studies should report a coordinated set of parameters and outcome classifications according to internationally accepted gradings to reduce redisplacement in future.
Introduction
The prevalence of cerebral palsy (CP) in industrialized countries where the disease is documented in appropriate registers is 1.2–3.6/1000 births, without major fluctuations in the last 50 years. All patients have motor impairments; 77–93% are affected by spasticity; and up to 35% are nonambulatory (10–15% Gross Motor Function Classification System (GMFCS) level IV and 12–19% GMFCS level V) (1). The incidence rate of hip displacement in CP is 35% and correlates with GMFCS level: 0% GMFCS level I, 90% GMFCS level V (2). Marcström et al. reported in a cross-sectional retrospective register study, which included >95% of children with CP in Sweden (952 children), that hip pain prevalence increased with increasing migration index (MI) (3) in hip displacement: 12.9% (MI: 31–40%) vs 25% (MI: 41–100%) (4). To benefit health-related quality of life (HRQoL) (5), hip reconstructive surgery is recommended to parents for their affected children. Due to the often significant comorbidities and aggravating social circumstances, parents understandably find it difficult to give their consent to these operations. In our experience, a question that is often asked during these information talks is about redislocation, i.e. ‘to what extent the effort is worth it’. This makes more sense if one considers the retrospective matched cohort study of DiFazio et al. (6) showing that CP patients experienced almost twice as many complications as non-CP patients. Ruzbarsky et al. reported a cumulative complication rate per patient, including failures to cure, of 47.6% in reconstructive hip procedures for nonambulatory patients with CP (7). Data from the Cerebral Palsy Integrated Pathway Scotland (CPIPS) surveillance programme indicated hip redisplacement in 15 children from 138 operated hips (11%) with 121 patients with GMFCS IV or V included (8).
Summarizing reviews in the sense of systematic reviews or meta-analyses have already been carried out. Agarwal et al. showed in a meta-analysis with 14 included studies (literature search up to January 2017) that combined pelvic/femoral interventions were superior to isolated femoral osteotomy and the latter, in turn, to soft tissue interventions with regard to the odds ratio for redisplacement. However, this was a comparative meta-analysis, so only studies with both a varus derotation osteotomy (VDRO) group and a combined group were included. Furthermore, the focus was not only on nonambulatory patients (9). The same observation applies to a systematic review by El-Sobky et al. (36 studies included), which gave an overview of the study situation up to July 2016 with regard to redisplacement rates for both pelvi-femoral and isolated varus derotation osteotomies in GMFCS III–V. A summarizing quantitative data synthesis was not carried out in this work (10).
The main goal of the present systematic review with meta-analysis was to be able to fully answer the frequently asked question about redisplacement rate, based on the basis of the numerous individual observational studies published to date for nonambulatory patients according to GMFCS IV or V, i.e. patients with higher comorbidities and higher hip displacement rates. Furthermore, relationships between patient- and surgery-related variables, like age at surgery, preoperative MI and type of bony surgery (i.e. VDRO vs VDRO + pelvic osteotomy), should also be computed, as well as study-related moderators like follow-up time, definition of failure and study period.
Methods
This systematic review was prepared in accordance with the PRISMA guidelines (11). There was no advance registration in PROSPERO.
Search strategy
Based on the Cochrane Handbook for Systematic Reviews of Interventions, version 6.3, 2022, MEDLINE (PubMed), Embase and CENTRAL were used as databases. (12) The database search was performed on March 18, 2022. The search was last updated on January 1, 2023. The study objective was transferred to the PICO (population, intervention, comparison, outcomes) scheme (Table 1).
PICO criteria.
| Criteria | Description |
|---|---|
| Population | Nonambulatory patients with cerebral palsy and hip displacement/dislocation |
| Intervention | Surgical bone hip reconstruction |
| Comparison | No |
| Outcomes | Percentage of hip redisplacement at follow-up |
Search words and combinations with Boolean operators are listed in Table 2.
Search words and combinations with Boolean operators.
| Health condition of interest | Intervention | |
|---|---|---|
| I (#1) | II (#2) | (#3) |
| Cerebral palsy | Hip dysplasia | Reconstruction |
| Hip subluxation | Osteotomy | |
| Hip luxation | Surgery | |
| Hip dislocation | ||
| Hip displacement | ||
| Hip instability | ||
| Hip migration | ||
| All terms combined with OR | All terms combined with OR | |
Combining hits from #1 AND #2 AND #3 is the final number of reports for title/abstract screening.
All three databases were searched using this strategy, and the records found were exported to a shared EndNoteTM library (version X9.2; Fa. Clarivate) that could be viewed by all reviewers involved. Duplicates were eliminated, and titles/abstracts were screened according to the inclusion and exclusion criteria listed in Table 3. Full texts were obtained and attached for the remaining reports. Furthermore, bibliographic data were transferred to a shared database in Microsoft® Excel® for Office 365 MSO (version 2210 Build 16.0.15726.20070). Decisions of reviewers in full-text screens and, later, extracted data were summarized and saved in it. The screening processes were performed by HD and YG. Inclusion and exclusion criteria for full-text screening are represented in Table 4. Disagreement was resolved by discussion of all authors.
Inclusion and exclusion criteria for title/abstract screening.
| Inclusion criteria | Exclusion criteria |
|---|---|
| • Patients with cerebral palsy and hip displacement | • Case report, review, conference abstract, expert opinion, letter |
| • Bony hip reconstruction (osteotomy of pelvis and/or femur and with or without possibly additional soft tissue release and with or without possibly open hip reduction) • Original study with outcome results • Full text in English available |
• Only soft tissue release • Salvage surgeries • Neurosurgical interventions • Hip arthroplasty • Disease other than cerebral palsy |
Inclusion criteria for full-text screening.
| Inclusion criteria |
|---|
| Specification of: |
| • Number of patients and operated hips |
| • Period in which the surgeries were performed |
| • Patients age |
| • GMFCS IV or V or ‘unable to walk’, ‘wheelchair’, ‘sitter’, ‘nonambulatory’, ‘nonwalker’ |
| • Follow-up period |
| • Description of surgical procedure |
| • Migration percentage according to Reimers (3) preoperative |
| • Failure rate (redisplacement/redislocation) |
| • Definition failure |
Data extraction
From the included studies, the following data were extracted: first author, year of publication, number of total hip surgeries, number of failed hip reconstructions in the sense of redisplacement, age of treated patients at surgery (mean ± s.d.), follow-up time (mean ± s.d.), value of MI set as failure, preoperative MI (mean ± s.d.) and type of bony reconstruction (VDRO alone, combined pelvic osteotomy + VDRO, mix of VDRO alone and combined procedure within the study). For case series in which data were presented for each patient, this was recalculated according to the desired parameters for the nonambulatory patients only (13, 14, 15, 16, 17, 18, 19, 20, 21, 22). When no definition of redisplacement was provided in these studies, we considered a MI of >30 as failure. Mean and s.d. were calculated, where necessary, in order to combine the results, from median, minimum and maximum values and/or the first and third quartiles according to the method of Wan et al. (23).
Risk of bias assessment
According to the manual ‘“Assessment of bias risk in interventional studies’, version 2.0 of May 10, 2021, from Cochrane Germany (24), the literature does not (yet) provide any clear criteria according to which the risk of bias in noncomparative intervention studies should be assessed. However, the features listed there, which increase confidence in the study results, were taken into account in the final decision for inclusion in the study. Additionally, the ‘Checklist for case series’ as a critical appraisal tool for use in systematic reviews of Joanna Briggs Institute, Adelaide, Australia was applied (26).
Data synthesis and statistics
The extracted data were read into the software comprehensive meta-analysis (CMA) version 4 (Borenstein, Hedges, Higgins, Rothstein; Biostat, Englewood, NJ, 2022). The data entry format was ‘one group/dichotomous (number of events)/raw data/events and sample size’. The random-effects model was employed for the analysis. For the calculation of the pooled event rate, the subgroups of five studies (21, 26, 27, 28, 29) with clearly separated subgroups ‘VDRO alone’ and ‘VDRO + pelvic osteotomy (= combined)’ were considered as independent studies. For subgroup analysis, ‘VDRO alone vs VDRO + pelvic osteotomy’ studies that included both surgical groups were assigned according to their majority (16, 18, 19, 30, 31, 32, 33, 34). In studies that looked at both groups in detail (21, 26, 27, 28, 29), the two study arms were assigned to the respective subgroup. This constellation was applied also for meta-regression with patient-related moderators ‘age at surgery’ and ‘preoperative MI’ while meta-regression with study-related moderators was performed with the study constellation for pooled redisplacement rate. In the subgroup analysis, the fixed-effects model was used at the subgroup level. τ2 was computed within subgroups and pooled across subgroups. IBM SPSS Statistics, version 29.0.0.0, was used to compare mean differences of age, MI and mid-year in the subgroup analysis as well as the distribution of the threshold MI for definition of failure in both groups. Controlling of normal distribution was performed with the Shapiro–Wilk test. The assumption of a nonnormal distribution was made at P < 0.05, and the nonparametric Mann–Whitney U-test was then used to test significant mean differences. In meta-regression with single covariates, P-values of <0.05 were considered to indicate statistically significant correlations.
Results
Study selection
The PRISMA flow diagram for literature selection is shown in Fig. 1. In all, 28 studies that met the inclusion criteria were found. Only those studies were included in which the preoperative MI and failure rate could be clearly assigned to a nonambulatory status comparable to GMFCS IV or V. Included studies are shown in Table 5. In some case series, there were data on resubluxation and redislocation, but data on the other parameters required by the inclusion criteria were missing (Table 6; (35, 36, 37, 38, 39, 40)). Reported outcomes of contralateral hips operated solely for reasons of symmetry or balancing were not included in the overall calculations (27, 32, 41) or could be disregarded because of their minor proportion in the study (30, 42).
PRISMA flow diagram for study election.
Citation: EFORT Open Reviews 9, 8; 10.1530/EOR-23-0043
Included reports in this study and in previous reviews.
| Study | Country | Previous inclusions‡ | Patients, n | Hips, n | Failures. n⁋ | Mean age (years) | Mean FU (years) | Bony surgery | PMI (%) | DOF (%)§ | Study period | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COMB | VDRO | ATB | Duration | Mid-year | ||||||||||
| Alassaf et al. (46) | Canada | no | 50 | 60 | 4 | 8.4 | 6.6 | ✓ | 72.8 | 50 | 1995–2009 | 2002 | ||
| Al-Ghadir et al. (26) | Canada | 1, 3 | 28 | 36 | 0 | 9.4 | 4.4* | ✓ | 61.6 | 50 | 1997–2007 | 2002 | ||
| 11 | 16 | 4 | 5.1 | 6.6* | ✓ | 63.3 | ||||||||
| Braatz et al. (47) | Germany | 3 | 67 | 91 | 6 | 10.9 | 7.7 | ✓ | 100 | 100 | 1990–2000 | 1995 | ||
| Canavese et al. (44) | France, Switzerland | no | 54 | 64 | 2 | 9.1 | 3.7 | ✓ | 66.8 | 30 | 2002–2017 | 2009 | ||
| Carroll et al. (30) | USA | no | 183 | 305 | 116 | 7.3 | 6.5 | 134 | 171 | 0 | 50.5 | 25 | 1992–2015 | 2004 |
| Debnath et al. (13)** | UK | 1, 2 | 7 | 7 | 1 | 14.6 | 13.3 | ✓ | 81.0 | 30 | 1986–1993 | 1990 | ||
| Dhawale et al. (41) | USA | 1, 3 | 19 | 22 | 2 | 7.5 | 11.7 | ✓ | 79.4 | 40 | 1989–1995 | 1992 | ||
| Iwase et al. (43) | Japan | no | 15 | 22 | 8 | 8.4 | 7.3 | ✓ | 92.0 | 40 | 2004–2014 | 2009 | ||
| Kim et al. (14)** | South Korea | 1, 3 | 22 | 31 | 0 | 8.6 | 2.4 | ✓ | 75.2 | 30 | 2006–2010 | 2008 | ||
| Koch et al. (48) | Poland | 1 | 81 | 115 | 8 | 9 | 5.5 | ✓ | 98.3 | 50 | 1998–2008 | 2003 | ||
| Mallet et al. (15)** | France | 3 | 19 | 19 | 2 | 7.8 | 9.1 | ✓ | 60.6 | 30 | 1994–2006 | 2000 | ||
| McCartney et al. (16)** | USA | no | 13 | 18 | 6 | 12.1 | 1.6 | 12 | 6 | 0 | 69.6 | 30 | 1986–1990 | 1988 |
| Mubarak et al. (17)** | USA | 1, 2 | 11 | 18 | 1 | 8.4 | 6.8 | ✓ | 78 | 30 | 1982–1984 | 1983 | ||
| Oetgen et al. (31) | USA | 1 | 20 | 20 | 7 | 15.1 | 2.2 | 11 | 7 | 2 | 57 | 25 | 2005–2011 | 2008 |
| Oh et al. (27) | South Korea, USA*** | 1, 2, 3 | 31 | 36 | 7 | 8.3 | 10.8 | All*** | 72.9 | 60 | 1989–1991 | 1990 | ||
| 24 | 4 | 8.9 | 10.8 | ✓*** | 76.1 | |||||||||
| 12 | 3 | 7.8 | 10.8 | ✓ | 66.8 | |||||||||
| Park et al. (45) | South Korea | no | 72 | 144 | 31 | 6.2 | 7.0 | ✓ | 62 | 30 | 2000–2015 | 2008 | ||
| Persiani et al. (18)** | Italy | 1, 2, 3 | 8 | 8 | 3 | 9.7 | 4.7 | ✓† | 75.6 | 30 | 1990–2005 | 1998 | ||
| Reidy et al. (42) | Switzerland | 3 | 29 | 44 | 2 | 8.8 | 5.5 | ✓ | 68.3 | 33 | 2005–2010 | 2008 | ||
| Robb et al. (49) | UK, Switzerland | 2 | 47 | 52 | 5 | 14 | 4 | ✓ | 70 | 25 | n.m. | |||
| Sankar et al. (22)** | USA | 1, 2, 3 | 9 | 11 | 0 | 9 | 16.6 | ✓ | 100 | 30 | 1973–1981 | 1977 | ||
| Shea et al. (19)** | USA | 1, 3 | 4 | 6 | 0 | 8.7 | 9.9 | 5 | 1 | 0 | 58.2 | 30 | 1976–1990 | 1983 |
| Shukla et al. (32) | USA | 1 | 35 | 35 | 12 | 9.2 | 5.0 | 29 | 2 | 4 | 60 | 40 | 1995–2008 | 2002 |
| Song et al. (21)** | USA | 3 | 31 | 22 | 3 | 8.3 | 3.5 | ✓ | 64.4 | 30 | 1969–1992 | 1981 | ||
| 24 | 7 | 8 | 5.9 | ✓ | 59.7 | 30 | ||||||||
| Terjesen (28) | Norway | no | 31 | 19 | 3 | 6.7 | 7.1 | ✓ | 75 | 50 | 2007–2014 | 2011 | ||
| 20 | 6 | 5.2 | 7.1 | ✓ | 63 | |||||||||
| Vallim et al. (33) | Brazil | no | 42 | 61 | 5 | 8.7 | 3.4 | 31 | 30 | 0 | 69.9 | 100 | 2013–2015 | 2014 |
| Zakrzewski et al. (34) | USA | no | 66**** | 141 | 13 | 7.3 | 4.1 | 110 | 87 | 0 | 51.5 | 30 | 2010–2019 | 2015 |
| Zenios et al. (20)** | UK | no | 12 | 14 | 6 | 7.9 | 11.3 | ✓ | 59.4 | 30 | 1985–1994 | 1990 | ||
| Zhang et al. (29) | New Zealand | 1 | 34 | 19 | 2 | 8* | 5.9* | ✓ | 69* | 60 | 1997–2009 | 2003 | ||
| 39 | 13 | 6* | 6.5* | ✓ | 51* | |||||||||
†Chiari group (five patients) not considered; ‡Included in previous reviews: 1 = Agarwal et al. (9), 2 = Bouwhuis et al. (53), 3 = El-Sobky et al. (10); ⁋Number of hips; §Threshold MI; *Estimated mean (23); **Recalculated; ***Probably same patients in combined group like in Dhawale et al., therefore only-VDRO group; ****Not clearly mentioned, 108 patients including 56 GMFCS I–III patients (66 ≙ mean between 52 and 80).
ATB, acetabular; COMB, combined; DOF, definition of failure; FU, follow-up; PMI, pre-operative MI.
Reports with information for failure rate but missing data for other inclusion parameters.
| Study | Previous inclusion* | Failure rate | Definition of failure | Reason for exclusion | Bony surgery |
|---|---|---|---|---|---|
| Axt et al. (39) | No | 13/28 (46.4%; hips, VDRO), 9/46 (19.6%; hips, VDRO + Dega) | MI > 30 | GMFCS III patients included, statistically no outcome difference between GMFCS III, IV, V | 2 groups: VDRO and VDRO + Dega |
| Bean et al. (38) | No | 24/150 (16%) | MI > 50 or revision | Probably included in Carroll et al. (30) | Mixed |
| Herndon et al. (37) | 2, 3 | 5/48 (10.4%; hips) | No MI (pre-, post-operative and at follow-up) with standard deviation for the entire patient group | VDRO | |
| Minaie et al. (40) | No | 13/121 (10.7% patients), in GMFCS IV and V | MI > 50 | Reported MI values included patients with GMFCS < IV and V | 51% femoral osteotomy, 41% combined, 8% isolated acetabular osteotomy (values include all GMFCS) |
| Osterkamp et al. (36) | No | 2/13 (15.4%; hips), | Dislocation | No MI, indication of CE | Chiari osteotomy |
| Shore et al. (35) | 3 | 24% resp. 42% failure (5- resp. 10-year FU) in 205 patients | MI > 50 or subsequent surgical procedures | No MI | Mixed |
Redisplacement rate
The pooled mean redisplacement rate was 16% with a 95% CI of 12–21% and a prediction interval from 3% to 51% (Q: 149; df: 32; P < 0.001; I2: 78%; τ2: 0.67 and τ: 0.82; 28 studies; 1540 hips; 1051 patients). The forest plot is shown in Fig. 2 and the distribution in Fig. 3. In the sensitivity analysis, the relative weights of contributing studies ranged from 1.13 (19) to 4.65 (30). This means that the number of operated hips in each study was taken into account in the overall result. Using the tool ‘one study removed’, the pooled mean ranged from 15.4% (20, 29, 32, 43) to 16.8% (44), indicating a robust result.
Distribution of true effects: redisplacement rate after bony reconstructive surgery in nonambulatory patients with CP (computed by CMA).
Citation: EFORT Open Reviews 9, 8; 10.1530/EOR-23-0043
Subgroup analysis: VDRO alone vs VDRO + pelvic osteotomy (combined)
The mean risk of redisplacement was 30% in studies that employed VDRO alone (95% CI: 22–41%; prediction interval: 12–59%; I2: 39%, 9 studies; 590 hips) (18, 21, 26, 27, 28, 29, 30, 43, 45). The mean risk of redisplacement in the combined group was 12% (95% CI 9–16%; prediction interval 4–30%; I2 63%, 24 studies; 950 hips) (13, 14, 15, 16, 17, 19, 20, 21, 22, 26, 27, 28, 29, 31, 32, 33, 34, 41, 42, 44, 46, 47, 48, 49). This difference was statistically significant (Q: 17.2; df: 1; P < 0.0001; τ2: 0.29; Fig. 4). Mean age in the VDRO-alone subgroup was 7.1 years and in the combined group was 9.5 years. This difference was statistically significant with P = 0.004. The difference in means of preoperative MI (64.9% vs 71.8%) was not significant (P = 0.142) nor were those of mid-year 2001 vs 1998 (P = 0.536). Furthermore, threshold MI for failure was distributed identically in both groups (P = 0.827).
Subgroup analysis of VDRO alone vs VDRO + pelvic osteotomy shows a significantly higher redisplacement rate in the VDRO-alone subgroup computed with CMA.
Citation: EFORT Open Reviews 9, 8; 10.1530/EOR-23-0043
Meta-regression with patient-related moderators ‘age at surgery’ and ‘preoperative MI’
Age
Age at surgery in the studies ranged from 5.1 years in the VDRO group of Al-Ghadir et al. (26) to 15.1 years (31). In the study population, there was no statistically significant correlation between age at surgery and redisplacement rate when testing age as a covariable alone (correlation coefficient −0.043; P = 0.565).
Preoperative MI
Preoperative MI ranged from 50.5% in the VDRO group of Carroll et al. (30) to 100% (22, 47). In meta-regression with the covariate ‘preoperative MI’ alone, there was a statistically significant negative correlation between this moderator and the redisplacement rate (correlation coefficient −0.0279; P = 0.0137; Fig. 5). Despite high preoperative MI in the studies of Sankar et al. (22), Koch et al. (48) and Braatz et al. (47) the event rate was only 4% respectively 7%.
Relationship between preoperative MI and redisplacement rate (logit), with confidence and prediction intervals computed with CMA.
Citation: EFORT Open Reviews 9, 8; 10.1530/EOR-23-0043
Meta-regression for relationship between type of bony surgery, age and preoperative MI
Running an analysis with bony surgery, age and preoperative MI as covariates in R 2 analogue statistics, 74% of the variance in true effects were explained by this model, with 57% from the type of bone surgery, 5% from age and 11% from preoperative MI (Fig. 6).
R2 for meta-regression model with surgery- and patient-related covariates: bony surgery, age and preoperative MI computed with CMA.
Citation: EFORT Open Reviews 9, 8; 10.1530/EOR-23-0043
Meta-regression with study-related moderators
Follow-up time
The mean values of the follow-up periods in the studies were between 1.6 (16) and 16.6 years (22). Within the framework of the meta-regression with the covariate ‘follow-up time’ alone, no significant correlation between the duration of the follow-up in the studies (mean value of each study) and the radiological failure rate could be found (correlation coefficient = 0.0136; P = 0.828).
Mid-year
The studies were conducted between 1977 (22) and 2015 (34) (mid-year of each study). In the meta-regression with the mid-year of study implementation as a covariate alone, there were no statistically significant changes in the radiological failure rate (correlation coefficient = −0.0072; P = 0.693).
Definition of failure (threshold MI)
The MI value beyond which redisplacement was defined was set differently in the studies (>25 (30, 31, 49) to ≥100 (33, 47)). With an alpha criterion P < 0.05, there was no statistically significant correlation between the definition of failure based on MI (correlation coefficient: −0.0114; P = 0.1741). However, in the R 2 analogue statistics, 20% of the variation in true effects (redisplacement rate) was explained by this variable.
Discussion
With the present meta-analysis including 28 studies and 1540 hips in 1051 patients, we can provide parents, caregivers and therapists with a pooled redisplacement rate after bony hip reconstruction in nonambulatory patients with CP based on the literature published to date. The mean effect size (redisplacement rate) is 16%, with a 95% CI of 12–22%. According to Borenstein (50), this means that in the universe of comparable populations, the true mean (redisplacement rate) probably falls within this interval. However, this statement deals with the mean and tells how precisely the mean has been estimated. The prediction interval indicating the expected event rate in 95% of all comparable studies ranges between 4% and 51%. If someone needed to decide whether to undergo this surgery, one can tell them that the risk of redisplacement is between 4% and 51%, and this would be correct for some 95% of all populations (50).
In Australia, parts of Europe and British Columbia, hip surveillance programmes for CP patients were established for early detection of the beginning of hip displacement to avoid more complex operations in the case of severe dislocations (51). The Scottish hip surveillance register was able to report redisplacement rates of 11% based on a larger number of patients (1646 registered patients, 138 operated children, but not only GMFCS IV or V). This rate lies close to the dimensions that were computed by us, although the number of operated patients in the present work is seven times larger. With the constantly increasing number of individual studies, pooling under defined criteria seems obvious and necessary to get solid, robust data. Furthermore, it seemed particularly important to us to focus on the more frequently and more severely affected patients, who are often mixed with patients with lower GMFCS levels in reports, as in the aforementioned data of the Scottish hip surveillance register. We were unable to include quite a few studies because of this fact. However, it can be assumed that the finding of a 16% redisplacement rate will deter few decision makers from an operation, especially against the background of the natural course of a progressive hip dislocation with salvage operations that could become necessary as a later alternative (52).
Nevertheless, due to the high level of heterogeneity and the broad prediction interval, subgroup analysis for the only categorical variable collected, namely ‘type of bony surgery: VDRO alone vs VDRO + pelvic osteotomy’, was added. It was observed that the mean redisplacement rate in the group of studies where only femoral osteotomy was carried out (30%) was significantly higher than in the group with additional pelvic osteotomy (12%). Heterogeneity decreased within these groups. Agarwal et al. (9) found in their comparative meta-analysis the same results: In their comparison of combined vs isolated femoral treatment, 45 of 323 hips (13.9%) treated with combined osteotomies were resubluxated/reoperated vs 63 of 303 hips (20.8%) treated with femoral osteotomy only. Since theirs was a comparative meta-analysis, only studies that compared both surgical procedures (VDRO vs VDRO + pelvic osteotomy) were included. Therefore, not as many patients were included as in our meta-analysis. A further difference is that they also included studies with patients who had GMFCS lower than IV or where the ambulatory status was not mentioned. This could be a reason for underestimation of the redisplacement rate in VDRO alone because one would expect that less extensive surgery would be performed on less severely affected patients who were still able to walk and whose displacement rates per se are not that high. El-Sobky et al. (10) concluded in their systematic review that there is limited evidence that would support isolated femoral reconstruction. They included patients with GMFCS III–V and did not perform a quantitative synthesis. In the review by Bouwhuis et al. (53), 189 patients with GMFCS IV or V (289 operated hips) in nine studies were summarized. They found 49 of the 289 hips (17%) were subluxated at follow-up and 9 were dislocated. This result is similar to ours, although only six of the studies in their review met our inclusion criteria. These authors also supported the fact that the redisplacement rate is lower with combined osteotomies.
Searching for reasons for lower redisplacement rates in combined osteotomies, we compared the means of age at the time of surgery and those of preoperative MIs of both subgroups, and patients in the VDRO-alone group were significantly younger. There were no differences in the means of preoperative MI and those of threshold MI for failure definition. It is known that children with CP are usually born with normal hips and that femoral head migration occurs due to muscle imbalance (54). The acetabular dysplasia, in turn, appears again later with progressing migration. For this reason, the patients with a combined intervention were probably older. Minaie et al. reported a 13% failure rate in their study of 291 hips/179 patients: age below 6 years at the time of surgery significantly increased the failure rate (26% vs 6.3%), as did a preoperative MP over 70% (28.9% vs 9.9%). Receiving an acetabular osteotomy was protective against failure (9.1% vs 16.9%), particularly in patients with a preoperative acetabular index over 25° (40). We excluded this study because it did not only contain patients with GMFCS IV or V, but still, the results reflected ours.
In order to examine the relationship between failure rate and age, preoperative MI and type of surgery, we carried out a meta-regression with these variables, both for each variable individually and for all three combined. While only a subordinate and no significant role as a moderator was observed for age, this was the case for type of bony surgery but also preoperative MI. As in the subgroup analysis, it was observed in the meta-regression that the decision for combined femoral and pelvic osteotomies significantly reduced the risk of failure of the surgical procedure with regard to redisplacement. Interestingly, however, preoperative MI was negatively correlated with the redisplacement rate. This initially seemed confusing in relation to the results of some other primary studies (40, 45, 55), where the rate of redislocation increased with increasing preoperative MI. However, this is a meta-regression regarding the results of studies, not individuals. Considering the two studies that show low redisplacement rates with the highest preoperative MI (100% in Braatz et al. (47) and 98% in Koch et al. (48)), the surgical technique may be relevant. In the study by Koch et al. (48) with 81 patients and 115 hips, the following were reported: bilateral release of the adductors, psoas and hamstrings distal lengthening; neurectomy of the anterior branch of the obturator nerve on the side of the dislocation; exposing the hip by using the Watson–Jones lateral approach; detaching of gluteal muscles and capsular attachment of the rectus femoris; opening of the hip capsule; removing of the ligamentum teres, the transverse ligament and connective tissue within the joint; defining the shortening of the femur by the distance between the upper edge of the acetabulum and the most proximally positioned part of the femoral head; performing the VDRO with femoral shortening; reducing the hip; Dega osteotomy and immobilization in a spica cast for 4–6 weeks. This procedure seemed to be relatively safe in terms of the redisplacement rate in nonambulatory patients with CP with high preoperative MI. It is a comprehensive approach and nothing more can be added, so we assume that this is the reason for the negative correlation between preoperative MI and redisplacement rate. If the hips are initially better, compromises in the procedure are possible. The data from the study by Braatz et al. (47) also support this thesis: of the 91 operated hips with an initial MI of 100%, all patients underwent pelvic surgery; 72 patients underwent open reduction, and 247 soft tissue procedures were performed. Neurectomy of the anterior branch of the obturator nerve on the side of the dislocation was not performed in these patients. In their study, Oetgen et al. concluded as follows: when all radiographic abnormalities were addressed during surgery (VDRO for neck-shaft angle > 135°, open reduction for MI >50%, acetabular osteotomy for acetabular index >25°), successful radiographic outcome at final follow-up is much more likely than when an intervention was less comprehensive (31).
Considering the extracted study-related parameters ‘follow-up time’, ‘definition of failure’ and ‘mid-year’, the regression coefficients behaved as expected but without statistical significance: the failure rate increased with increasing follow-up time, while it decreased if failure was defined at a higher MI value. It was not surprising that this last fact was included in the R2 analogous statistic (at 20%) and apparently explained the variance of the effect size to this extent. In further research efforts to optimize the rate of displacement after surgery, the limiting MI value should be defined by convention. This is directly linked to a recommendation regarding at which MI value surgery should be indicated. In children registered in the CPIPS Surveillance Programme, Wordie et al. found an MI of 46% to be a “point of no return”; however, GMFCS, CP subtype, or other patient-related factors were not considered (56). The implementation of more recent studies has led to slightly lower failure rates, which could be gratifying, as it shows that dealing with the topic in a variety of approaches has led to improvements in outcomes. However, none of these observations met the significance level.
We would like to emphasize again that we selected only studies where only non-ambulatory patients equivalent to GMFCS IV or V were included. Ruzbarsky et al. themed also noted that it is difficult to compare studies with mixed populations of ambulators and non-ambulators, especially since GMFCS level has been shown to be an important determinant for both the incidence of hip displacement and the risk of complications (7). They referred to a study by Stasikelis et al. (57) for this argument. Furthermore, they summarized the results, complication rates and Sink grades (58) of related studies and considered the proportion of ambulatory patients.
Our study has some limitations. First, we only pooled redisplacement rates for nonambulatory patients. It would be interesting to see a similar analysis for ambulatory patients. Second, when assigning to the subgroup ‘type of bony surgery’, studies were assigned to the subgroup depending on which surgical method was used in the majority of patients, as demonstrated by Borenstein (50). Third, we could only extract the variables: type of bony surgery, age at surgery, preoperative MI, period of study implementation, follow-up time and MI for failure definition as moderators. These were the parameters that were given in almost all studies and therefore could be made synthesizable. Fourth, acetabular- and neck-shaft-angles (NSA) were not extracted because they were not consistently reported. Furthermore, from our point of view, NSA is less valid due to its susceptibility to projection-related measurement errors. An X-ray of the pelvis, where the real NSA is projected, cannot be safely obtained in this patient population. Possible alternatives would be exact measurement intraoperatively with fluoroscopy or preoperatively in a reconstructed computer tomogram. Unfortunately, such a procedure is not part of the clinical routine until now. Since the NSA has a crucial correction function in the VDRO and overcorrection of it could possibly achieve a better outcome (36), comparable measurement techniques should be discussed for future studies. Fifth, one can imagine that the type of CP, degree of spasticity, situation of the opposite hip, presence of scoliosis, pelvic obliquity, as well as possibly direction of displacement also played a considerable role in this connection, but these data were not provided in all studies either. Sixth, the soft tissue interventions performed could not be recorded consistently due to their diversity. Seventh, postoperative follow-up treatment should also be taken into account. Finally, we considered only radiological failure rates. For success of this type of surgery, other outcome parameters must of course be taken into account, e.g. pain and complication rate.
Conclusion
With this meta-analysis, we present a possible way to pool data from single-centre studies including moderating variables, although, of course, multicentre studies and evaluation of these data in the context of hip surveillance programmes are also desirable. We computed a pooled redisplacement rate of 16% for nonambulatory patients with CP after bony hip reconstruction. As Terjesen (59) demanded, certain requirements for future studies should be defined. He stated that, in particular, age and GMFCS level have to be evaluated and that postoperative follow-up should be satisfactory, i.e. ≥5 years. Classification of the outcome should be in accordance with internationally accepted gradings in order to compare results (59). From our point of view, the parameters mentioned above (real NSA, type of CP, degree of spasticity, situation of the opposite hip, presence of scoliosis, pelvic obliquity, direction of displacement, soft tissue procedure) should be documented in future observational studies. Last but not least, the threshold MI, above which a failure is defined, should also be standardized. We can add further evidence for combined surgery for nonambulatory patients based on a larger, systematically created patient group. Furthermore, we have observed benefits for the more extensive surgical approach in patients with a high preoperative MI. Hopefully, taking these findings into account will lead to a reduction in the redisplacement rate over the years, which we have not been able to show significantly up to now based on the included studies.
ICMJE Conflict of Interest Statement
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the study reported.
Funding Statement
This work did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.
Author contribution statement
HD: conceptualization, methodology, software, literature search, study selection, data extraction, risk of bias assessment, data synthesis, statistics and meta-analysis, writing manuscript, visualization; YG: literature search, study selection, data extraction, risk of bias assessment; RB: conceptualization, literature review, supervision, manuscript revision; FH: supervision, manuscript revision. All the authors read and approved the final version of the manuscript.
References
- 1↑
Blair E. Epidemiology of the cerebral palsies. Orthopedic Clinics of North America 2010 41 441–455. (https://doi.org/10.1016/j.ocl.2010.06.004)
- 2↑
Soo B, Howard JJ, Boyd RN, Reid SM, Lanigan A, Wolfe R, Reddihough D, & Graham HK. Hip displacement in cerebral palsy. Journal of Bone and Joint Surgery 2006 88 121–129. (https://doi.org/10.2106/JBJS.E.00071)
- 3↑
Reimers J. The stability of the hip in children. A radiological study of the results of muscle surgery in cerebral palsy. Acta Orthopaedica Scandinavica. Supplementum 1980 184 1–100. (https://doi.org/10.3109/ort.1980.51.suppl-184.01)
- 4↑
Marcström A, Hägglund G, & Alriksson-Schmidt AI. Hip pain in children with cerebral palsy: a population-based registry study of risk factors. BMC Musculoskeletal Disorders 2019 20 62. (https://doi.org/10.1186/s12891-019-2449-8)
- 5↑
DiFazio RL, Vessey JA, Miller PE, Snyder BD, & Shore BJ. Health-related quality of life and caregiver burden after hip reconstruction and spinal fusion in children with spastic cerebral palsy. Developmental Medicine and Child Neurology 2022 64 80–87. (https://doi.org/10.1111/dmcn.14994)
- 6↑
DiFazio R, Vessey JA, Miller P, van Nostrand K, & Snyder B. Postoperative complications after hip surgery in patients with cerebral palsy: a retrospective matched cohort study. Journal of Pediatric Orthopedics 2016 36 56–62. (https://doi.org/10.1097/BPO.0000000000000404)
- 7↑
Ruzbarsky JJ, Beck NA, Baldwin KD, Sankar WN, Flynn JM, & Spiegel DA. Risk factors and complications in hip reconstruction for nonambulatory patients with cerebral palsy. Journal of Children’s Orthopaedics 2013 7 487–500. (https://doi.org/10.1007/s11832-013-0536-1)
- 8↑
Wordie SJ, Robb JE, Hägglund G, Bugler KE, & Gaston MS. Hip displacement and dislocation in a total population of children with cerebral palsy in Scotland. Bone and Joint Journal 2020 102–B 383–387. (https://doi.org/10.1302/0301-620X.102B3.BJJ-2019-1203.R1)
- 9↑
Agarwal KN, Chen C, Scher DM, & Dodwell ER. Migration percentage and odds of recurrence/subsequent surgery after treatment for hip subluxation in pediatric cerebral palsy: a meta-analysis and systematic review. Journal of Children’s Orthopaedics 2019 13 582–592. (https://doi.org/10.1302/1863-2548.13.190064)
- 10↑
El-Sobky TA, Fayyad TA, Kotb AM, & Kaldas B. Bony reconstruction of hip in cerebral palsy children Gross Motor Function Classification System levels III to V: a systematic review. Journal of Pediatric Orthopedics. Part B 2018 27 221–230. (https://doi.org/10.1097/BPB.0000000000000503)
- 11↑
Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, Shamseer L, Tetzlaff JM, Akl EA, Brennan SE, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021. (https://doi.org/10.1136/bmj.n71)
- 12↑
Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, & Welch VA, Eds. Cochrane Handbook for Systematic Reviews of Interventions version 6.3 updated 2022. Available at: www.training.cochrane.org/handbook.
- 13↑
Debnath UK, Guha AR, Karlakki S, Varghese J, & Evans GA. Combined femoral and Chiari osteotomies for reconstruction of the painful subluxation or dislocation of the hip in cerebral palsy. A long-term outcome study. Journal of Bone and Joint Surgery 2006 88 1373–1378. (https://doi.org/10.1302/0301-620X.88B10.17742)
- 14↑
Kim HT, Jang JH, Ahn JM, Lee JS, & Kang DJ. Early results of one-stage correction for hip instability in cerebral palsy. Clinics in Orthopedic Surgery 2012 4 139–148. (https://doi.org/10.4055/cios.2012.4.2.139)
- 15↑
Mallet C, Ilharreborde B, Presedo A, Khairouni A, Mazda K, & Penneçot GF. One-stage hip reconstruction in children with cerebral palsy: long-term results at skeletal maturity. Journal of Children’s Orthopaedics 2014 8 221–228. (https://doi.org/10.1007/s11832-014-0589-9)
- 16↑
McCartney DK, & Frankovitch KF. Proximal femoral diaphysectomy in cerebral palsy. Contemporary Orthopaedics 1994 29 52–58.
- 17↑
Mubarak SJ, Valencia FG, & Wenger DR. One-stage correction of the spastic dislocated hip. Use of pericapsular acetabuloplasty to improve coverage. Journal of Bone and Joint Surgery 1992 74 1347–1357. (https://doi.org/10.2106/00004623-199274090-00008)
- 18↑
Persiani P, Molayem I, Calistri A, Rosi S, Bove M, & Villani C. Hip subluxation and dislocation in cerebral palsy: outcome of bone surgery in 21 hips. Acta Orthopaedica Belgica 2008 74 609–614.
- 19↑
Shea KG, Coleman SS, Carroll K, Stevens P, & van Boerum DH. Pemberton pericapsular osteotomy to treat a dysplastic hip in cerebral palsy. Journal of Bone and Joint Surgery 1997 79 1342–1351. (https://doi.org/10.2106/00004623-199709000-00008)
- 20↑
Zenios M, Hannan M, Zafar S, Henry A, Galasko CSB, & Khan T. Clinical and radiological outcome of combined femoral and Chiari osteotomies for subluxed or dislocated hips secondary to neuromuscular conditions: a minimum of 10-year follow-up. Musculoskeletal Surgery 2012 96 101–106. (https://doi.org/10.1007/s12306-012-0201-8)
- 21↑
Song HR, & Carroll NC. Femoral varus derotation osteotomy with or without acetabuloplasty for unstable hips in cerebral palsy. Journal of Pediatric Orthopedics 1998 18 62–68. (https://doi.org/10.1097/00004694-199801000-00013)
- 22↑
Sankar WN, Spiegel DA, Gregg JR, & Sennett BJ. Long-term follow-up after one-stage reconstruction of dislocated hips in patients with cerebral palsy. Journal of Pediatric Orthopedics 2006 26 1–7. (https://doi.org/10.1097/01.bpo.0000190842.77036.d0)
- 23↑
Wan X, Wang W, Liu J, & Tong T. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Medical Research Methodology 2014 14 135. (https://doi.org/10.1186/1471-2288-14-135)
- 24↑
Braun C, Schmucker C, Nothacker M, Nitschke K, Schaefer C, Bollig C, Muche-Borowski C, Kopp I & & Meerpohl JJ Manual Bewertung des Biasrisi kos in Interventionsstudien. Albert-Ludwigs-Universität Freiburg, 2021.
- 25
Munn Z, Barker TH, Moola S, Tufanaru C, Stern C, McArthur A, Stephenson M & & Aromataris E. Methodological quality of case series studies: an introduction to the JBI critical appraisal tool. JBI Evidence Synthesis 2020 18 2127–2133. (https://doi.org/10.11124/JBISRIR-D-19-00099)
- 26↑
Al-Ghadir M, Masquijo JJ, Guerra LA, & Willis B. Combined femoral and pelvic osteotomies versus femoral osteotomy alone in the treatment of hip dysplasia in children with cerebral palsy. Journal of Pediatric Orthopedics 2009 29 779–783. (https://doi.org/10.1097/BPO.0b013e3181b76968)
- 27↑
Oh C-W, Presedo A, Dabney KW, & Miller F. Factors affecting femoral varus osteotomy in cerebral palsy: a long-term result over 10 years. Journal of Pediatric Orthopedics. Part B 2007 16 23–30. (https://doi.org/10.1097/01.bpb.0000228393.70302.ce)
- 28↑
Terjesen T. Femoral and pelvic osteotomies for severe hip displacement in nonambulatory children with cerebral palsy: a prospective population-based study of 31 patients with 7 years' follow-up. Acta Orthopaedica 2019 90 614–621. (https://doi.org/10.1080/17453674.2019.1675928)
- 29↑
Zhang S, Wilson NC, Mackey AH, & Stott NS. Radiological outcome of reconstructive hip surgery in children with gross motor function classification system IV and V cerebral palsy. Journal of Pediatric Orthopedics. Part B 2014 23 430–434. (https://doi.org/10.1097/BPB.0000000000000075)
- 30↑
Carroll KL, Stotts AK, Baird GO, Thorman AL, Talmage M, Moss WD, McMulkin ML, & MacWilliams BA. Factors influencing outcomes of the dysplastic hip in nonambulatory children with cerebral palsy. Journal of Pediatric Orthopedics 2021 41 221–226. (https://doi.org/10.1097/BPO.0000000000001760)
- 31↑
Oetgen ME, Ayyala H, & Martin BD. Treatment of hip subluxation in skeletally mature patients with cerebral palsy. Orthopedics 2015 38 e248–e252. (https://doi.org/10.3928/01477447-20150402-50)
- 32↑
Shukla PY, Mann S, Braun SV, & Gholve PA. Unilateral hip reconstruction in children with cerebral palsy: predictors for failure. Journal of Pediatric Orthopedics 2013 33 175–181. (https://doi.org/10.1097/BPO.0b013e31827d0b73)
- 33↑
Vallim FCdM, Da Cruz HAD, Rodrigues RC, Abreu de CSG, Godoy EDP, & Cunha MG. The use of pediatric locked plates in the paralytic hip: preliminary results of 61 cases. Revista Brasileira de Ortopedia 2018 53 674–680. (https://doi.org/10.1016/j.rboe.2017.09.009)
- 34↑
Zakrzewski AM, Bryant AJ, & McCarthy JJ. Can over-containment prevent recurrence in children with cerebral palsy and hip dysplasia undergoing hip reconstruction? Journal of Pediatric Orthopedics 2022 42 300–306. (https://doi.org/10.1097/BPO.0000000000002160)
- 35↑
Shore BJ, Zurakowski D, Dufreny C, Powell D, Matheney TH, & Snyder BD. Proximal femoral Varus derotation osteotomy in children with cerebral palsy: the effect of age, gross motor function classification system level, and surgeon volume on surgical success. Journal of Bone and Joint Surgery 2015 97 2024–2031. (https://doi.org/10.2106/JBJS.O.00505)
- 36↑
Osterkamp J, Caillouette JT, & Hoffer MM. Chiari osteotomy in cerebral palsy. Journal of Pediatric Orthopedics 1988 8 274–277. (https://doi.org/10.1097/01241398-198805000-00004)
- 37↑
Herndon WA, Bolano L, & Sullivan JA. Hip stabilization in severely involved cerebral palsy patients. Journal of Pediatric Orthopedics 1992 12 68–73. (https://doi.org/10.1097/01241398-199201000-00011)
- 38↑
Bean BK, Baird GO, Caskey PM, Bronson WB, McMulkin ML, & Tompkins BJ. Early bony hip reconstructive surgery for hip subluxation in children with severe cerebral palsy. Orthopedics 2021 44 e294–e300. (https://doi.org/10.3928/01477447-20201210-06)
- 39↑
Axt MW, & Wadley DL. The unstable hip in children with cerebral palsy: does an acetabuloplasty add midterm stability? Journal of Children’s Orthopaedics 2021 15 564–570. (https://doi.org/10.1302/1863-2548.15.210154)
- 40↑
Minaie A, Gordon JE, Schoenecker P, & Hosseinzadeh P. Failure of hip reconstruction in children with cerebral palsy: what are the risk factors? Journal of Pediatric Orthopedics 2022 42 e78–e82. (https://doi.org/10.1097/BPO.0000000000001989)
- 41↑
Dhawale AA, Karatas AF, Holmes L, Rogers KJ, Dabney KW, & Miller F. Long-term outcome of reconstruction of the hip in young children with cerebral palsy. Bone and Joint Journal 2013 95–B 259–265. (https://doi.org/10.1302/0301-620X.95B2.30374)
- 42↑
Reidy K, Heidt C, Dierauer S, & Huber H. A balanced approach for stable hips in children with cerebral palsy: a combination of moderate VDRO and pelvic osteotomy. Journal of Children’s Orthopaedics 2016 10 281–288. (https://doi.org/10.1007/s11832-016-0753-5)
- 43↑
Iwase D, Fukushima K, Kusumoto Y, Metoki Y, Aikawa J, Kenmoku T, Minato S, Matsuo A, & Takaso M. Femoral varus derotational osteotomy without pelvic osteotomy in nonambulatory children with cerebral palsy: minimum 5 years follow-up. Medicine 2022 101 e28604. (https://doi.org/10.1097/MD.0000000000028604)
- 44↑
Canavese F, Marengo L, & Coulon de G. Results and complications of percutaneous pelvic osteotomy and intertrochanteric varus shortening osteotomy in 54 consecutively operated GMFCS level IV and V cerebral palsy patients. European Journal of Orthopaedic Surgery & Traumatology 2017 27 513–519. (https://doi.org/10.1007/s00590-017-1902-3)
- 45↑
Park H, Abdel-Baki SW, Park K-B, Park BK, Rhee I, Hong S-P, & Kim HW. Outcome of femoral Varus derotational osteotomy for the spastic hip displacement: implication for the indication of concomitant pelvic osteotomy. Journal of Clinical Medicine 2020 9. (https://doi.org/10.3390/jcm9010256)
- 46↑
Alassaf N, Saran N, Benaroch T, & Hamdy RC. Combined pelvic and femoral reconstruction in children with cerebral palsy. Journal of International Medical Research 2018 46 475–484. (https://doi.org/10.1177/0300060517723797)
- 47↑
Braatz F, Eidemüller A, Klotz MC, Beckmann NA, Wolf SI, & Dreher T. Hip reconstruction surgery is successful in restoring joint congruity in patients with cerebral palsy: long-term outcome. International Orthopaedics 2014 38 2237–2243. (https://doi.org/10.1007/s00264-014-2379-x)
- 48↑
Koch A, Jozwiak M, Idzior M, Molinska-Glura M, & Szulc A. Avascular necrosis as a complication of the treatment of dislocation of the hip in children with cerebral palsy. Bone and Joint Journal 2015 97–B 270–276. (https://doi.org/10.1302/0301-620X.97B2.34280)
- 49↑
Robb JE, & Brunner R. A Dega-type osteotomy after closure of the triradiate cartilage in non-walking patients with severe cerebral palsy. Journal of Bone and Joint Surgery 2006 88 933–937. (https://doi.org/10.1302/0301-620X.88B7.17506)
- 50↑
Video | Comprehensive meta-analysis, 2023. Available at: https://www.meta-analysis.com/pages/video_cma4.php?VideoName=eventrate. Accessed on 4 Feb 2023.
- 51↑
Shrader MW, Wimberly L, & Thompson R. Hip surveillance in children with cerebral palsy. Journal of the American Academy of Orthopaedic Surgeons 2019 27 760–768. (https://doi.org/10.5435/JAAOS-D-18-00184)
- 52↑
Shaw KA, Hire JM, & Cearley DM. Salvage treatment options for painful hip dislocations in nonambulatory cerebral palsy patients. Journal of the American Academy of Orthopaedic Surgeons 2020 28 363–375. (https://doi.org/10.5435/JAAOS-D-19-00349)
- 53↑
Bouwhuis CB, van der Heijden-Maessen HC, Boldingh EJK, Bos CFA, & Lankhorst GJ. Effectiveness of preventive and corrective surgical intervention on hip disorders in severe cerebral palsy: a systematic review. Disability and Rehabilitation 2015 37 97–105. (https://doi.org/10.3109/09638288.2014.908961)
- 54↑
Lins LAB, Watkins CJ, & Shore BJ. Natural history of spastic hip disease. Journal of Pediatric Orthopedics 2019 39 S33–S37. (https://doi.org/10.1097/BPO.0000000000001347)
- 55↑
Rutz E, Vavken P, Camathias C, Haase C, Jünemann S, & Brunner R. Long-term results and outcome predictors in one-stage hip reconstruction in children with cerebral palsy. Journal of Bone and Joint Surgery 2015 97 500–506. (https://doi.org/10.2106/JBJS.N.00676)
- 56↑
Wordie SJ, Bugler KE, Bessell PR, Robb JE, & Gaston MS. Hip displacement in children with cerebral palsy. Bone and Joint Journal 2021 103–B 411–414. (https://doi.org/10.1302/0301-620X.103B2.BJJ-2020-1528.R1)
- 57↑
Stasikelis PJ, Lee DD, & Sullivan CM. Complications of osteotomies in severe cerebral palsy. Journal of Pediatric Orthopedics 1999 19 207–210. (https://doi.org/10.1097/00004694-199903000-00014)
- 58↑
Sink EL, Beaulé PE, Sucato D, Kim Y-J, Millis MB, Dayton M, Trousdale RT, Sierra RJ, Zaltz I, Schoenecker P, et al. Multicenter study of complications following surgical dislocation of the hip. Journal of Bone and Joint Surgery 2011 93 1132–1136. (https://doi.org/10.2106/JBJS.J.00794)
- 59↑
Terjesen T, Wagner P, & Hägglund G. Letter to the Editor and reply concerning: hip development after surgery to prevent hip dislocation in cerebral palsy: a longitudinal register study of 252 children. Acta Orthopaedica 2022 93 294–295. (https://doi.org/10.2340/17453674.2022.2030)
