A single-center retrospective review of factors influencing surgical success in patients with cerebral palsy undergoing corrective hip surgery

1. Introduction

Progressive hip displacement is one of the most common orthopaedic problems in patients with cerebral palsy (CP), occurring in approximately one-third of all children with CP and in 90% of those with GMFCS level V [1]. Hip displacement can cause pain, limit function (including ambulation, sitting, and perineal care), and diminish quality of life [2]. The most frequently used measure of hip displacement is Reimer’s hip migration percentage (MP). Hips with a MP $>$ 30% are considered displaced, and those with a MP $>$ 90% are considered dislocated [3].

Surgical approaches for treating hip subluxation or dislocation are classified as preventive, reconstructive, or salvage [4]. Preventative procedures involve soft tissue releases (e.g., adductor or hamstring release) and are typically used to prevent subluxation. However, once a hip begins to undergo subluxation, preventing further displacement via soft tissue releases becomes difficult [5]. Reconstructive hip surgery typically involves proximal femoral varus derotational osteotomy (VDRO) to correct coxa valga and femoral anteversion and/or pelvic osteotomy to correct acetabular deficiency. Salvage operations are limited and include femoral head resection, hip arthroplasty, or hip arthrodesis. In general, a MP greater than 40% is an indication for surgical intervention to prevent hip dislocation [6].

Though multiple factors likely influence surgical outcomes for patients with CP and hip displacement, the timing of corrective hip surgery in the context of the patient’s pain (i.e., proactively (before onset of pain) or reactively (after pain is reported)) is of increasing clinical interest. The relationship between hip morphology and pain in patients with CP is unclear and remains controversial [7]. Furthermore, pain is often difficult to assess in this patient population, especially among severely affected patients with communication impairments. In such cases, physicians rely on caregiver reports of the patient’s pain to help determine when to proceed with hip surgery. Corrective bony hip surgery is an invasive procedure that requires a long recovery period; thus, patients and their caregivers may choose to delay surgery as long as possible. Because physical therapists often work closely with patients with CP and can assess pain levels, elucidating how preoperative pain influences surgical outcomes may help to inform physical therapists about when to refer patients for surgical evaluation.

To the authors’ knowledge, no study has reported on the associations of preoperative hip pain and MP with surgical outcomes in patients with CP undergoing corrective bony hip surgery. The purpose of the study was to investigate how patient age, sex, GMFCS level, type of surgery, type and extent of CP involvement, preoperative hip pain, and MP correlate with surgical outcomes in this patient population. This inquiry is important because it may influence the timing of corrective hip surgery, about which there is presently no consensus.

2. Methods

3. Results

4. Discussion

The results of this study suggest that patient sex, type or extent of CP involvement, type of surgery, and perioperative complications are factors that are independent of surgical success, as defined herein, among patients with CP who undergo corrective bony hip surgery. In contrast, the absence of preoperative hip pain ( p = 0.014), surgery after age 5 (0.041), and a milder preoperative MP ( 0.001) were significantly associated with surgical success. Although not a significant finding, GMFCS levels II or III were associated with higher rates of surgical success compared to GMFCS levels IV or V ( p = 0.058). However, these findings must be interpreted cautiously in the context of this single-center retrospective study.

The findings indicating that the absence of preoperative hip pain and a milder preoperative MP were associated with successful surgical outcomes may have especially important clinical implications because both factors are commonly considered in the surgical evaluation of hip displacement in this patient population. Hip pain and MP may be interrelated, but any relationship between the two is only ostensible; in the present study, patients may have had a completely dislocated hip and been free of hip pain or have had very little hip displacement and been bothered by hip pain. Hip pain itself may be assessed by caregivers or providers outside the clinic and may inform the referral of patients for surgical evaluation. Further, because hip pain and MP may develop and/or worsen over time, these findings suggest that early preventative surgical correction of hip displacement to prevent the onset of hip pain and minimize the preoperative MP may optimize clinical outcomes, though additional studies are needed to support this. Representative anteroposterior radiographs showing successful and unsuccessful surgical outcomes in two patients with CP undergoing corrective bony hip surgery are shown in Fig. 1.

OPEN IN VIEWER

The current findings align well with the relevant literature to support previous conclusions and suggest future directions for research. In one of the few reports of patient sex and surgical outcomes in patients with CP, Zwick et al. [10] reported in 2012 that although both boys and girls with spastic, diplegic CP initially benefited from single-event multilevel surgery (bone and soft-tissue), boys eventually deteriorated in walking ability at 10 years of follow-up. Although most patients in the current study were quadriplegic, the study by Zwick et al. suggests that additional follow-up is needed to evaluate surgical success, as we have defined, for our patients in the long term. In addition, Lee et al. [11] reported in 2009 that parental satisfaction after single-event multilevel surgery in ambulatory children with CP was greater in parents of younger patients. Clinically, this finding may need to be balanced by our subgroup analysis that suggests that surgical intervention before age 6 is associated with lower rates of surgical success. Further, multiple other reports suggest avoiding surgical hip correction on very young children with CP [12, 13], including one by Schmale et al. [14] that reported a reoperation rate of 74% at 5-year mean follow-up in children with spastic CP who underwent bilateral VDRO before age 6 years. Additionally, multiple reports indicate that lower GMFCS levels are strong predictors of surgical success, defined quantitatively in terms of MP or the need for revision surgery [15, 16]. These findings support the current study’s subgroup analysis by GMFCS level, which suggests that GMFCS levels II and III were associated with higher rates of surgical success compared with GMFCS levels IV and V, though this finding was not found to be statistically significant ( p = 0.058). In addition, previous studies have found that a smaller preoperative MP is associated with better long-term radiographic and clinical outcomes in patients with CP [17, 18, 19], in support of the current findings on preoperative MP and recommendations for early preventative surgical intervention. Further, similar to other studies [20, 21], the current study found overall improvements in hip pain after corrective bony hip surgery.

This retrospective study has several limitations. Forty-six patients were excluded because they had less than one-year follow-up, multiple/repeat bony hip surgeries, femoral head resections, and/or incomplete medical records. It is unknown what effect this has on the findings. Further, the relatively small sample of 38 patients (55 hips) limited the power of statistical analyses, including the ability to account for correlations between one patient’s bilateral hips. Most importantly, the assessment of pain is a limitation of our study, as standardized pain ratings and scales were not used in the clinical notes. In addition, retrospective chart reviews are inherently limited in the availability and reliability of data. Nonetheless, the authors believe that the presence of hip pain, regardless of the rating, should be considered in the surgical evaluation of these patients. Further, as with similar studies involving patients with severe CP, accurately assessing hip pain in this patient population can be challenging and often relies on caregivers or other providers. We agree that prospectively collecting data and evaluating pain over time with a reliable and reproducible grading system would strengthen our study. We present our results to add to the literature on when to intervene surgically to correct hip displacement in patients with CP, a challenging topic for which there is currently no consensus. Additional potential confounders include the relationship between hip morphology and pain, which was not investigated; the surgical algorithms followed by the treating physicians; and the limited follow-up time, especially in skeletally immature patients. Although the authors believe that one-year follow-up is sufficient to gain insight into the correlations of interest in this study, these patients will continue to be followed so that longer-term follow-up may be reported in the future.

In patients with CP and hip displacement, early preventative surgical correction of hip displacement before the onset of bony hip pain and worsening MP may lead to improved clinical outcomes, though these findings must be considered cautiously in the context of this single-center retrospective study. The authors believe that pain and MP should be considered along with patient and caregiver preferences to promote successful surgical outcomes. Future studies with larger patient populations, standardized pain rating systems, and longer-term follow-up are needed to provide more definitive clinical direction.