17 Assessing Outcomes in Neuromuscular Scoliosis
“Would you tell me, please, which way I ought to go from here?”
“That depends a good deal on where you want to get to.”
“I don’t much care where —”
“Then it doesn’t matter which way you go.”
–Lewis Carroll, Alice in Wonderland
This chapter examines the outcomes of surgery for neuromuscular scoliosis and the methodologic problems inherent in defining a “good” result. An examination of retrospective case series highlights reasons for inconsistencies in the reported outcomes of scoliosis surgery and caregiver satisfaction rates. The concept of “cognitive dissonance” is introduced as a possible explanation for the dichotomy between published outcomes and the lack of substantial postsurgical functional gains. Improved cosmesis as an aspect of parental or caregiver perception are also discussed. As a counterpoint to the results of retrospective studies, a prospective report is highlighted. Finally, some suggestions for future research and outcomes assessment are made. Total-involvement spastic cerebral palsy is used as an exemplar for the group.
17.1 Background
Surgery for neuromuscular scoliosis has reflected the evolution of spinal surgery in general. Harrington first applied his rod to the treatment of deformity arising from paralytic poliomyelitis. However, it was the instrumentation developed by Eduardo Luque that ushered in the era of segmental correction, stable fixation, higher fusion rates, and the avoidance of prolonged postoperative immobilization or reliance on external orthosis (Fig. 17.1). When the importance of pelvic obliquity in contributing to pain and disability was recognized, further elaborations, such as the Galveston pelvic fixation technique and the unit rod, were introduced. The often-stated objectives of surgery for neuromuscular deformity can be summarized as follows:
Reduced pain and improved health-related quality of life;
Correction of pelvic obliquity and costopelvic impingement;
Spinal balance in the frontal and sagittal planes;
Solid arthrodesis;
Minimal complications.
17.1.1 High Rates of Caregiver Satisfaction
Many papers dating back to the early 1980s have documented high rates of caregiver satisfaction with the results of surgery. Comstock et al noted that 85% of parents or caregivers were very satisfied with the results of surgery and noted a beneficial impact on the patient’s ability to sit, physical appearance, ease of care, and comfort. 1
Using a questionnaire given to parents and caregivers, Tsirikos et al noted that caretakers did not recognize the effects of scoliotic deformity on patients’ head control, hand use, or feeding ability. 2 Both parents and caregivers reported a very positive impact of surgery on patients’ overall function, quality of life, and ease of care. Parents had more appreciation of the beneficial effects on their children’s appearance, whereas educators and therapists were more likely to acknowledge improvement in gross and oral motor function. Most parents (95.8%) and caretakers (84.3%) said they would recommend spine surgery.
Bohtz et al used a modified version of the Caregiver Priorities and Child Health Index of Life with Disabilities (CPCHILD) questionnaire to assess patients’ outcomes via their parents or caregivers. 3 A significant improvement in health-related quality of life was noted after the operation. The patients’ rate of satisfaction with the outcome of surgery was 91.7%.
17.1.2 Significant Complication Rates
The problem with spinal surgery in this patient group is the high complication rate. Lonstein et al reported 83 early complications in 54 patients, for an early complication rate of 58%. Reoperation during the initial hospitalization was required in two patients (1.1%): one for infection and one for proximal hook cutout and proximal junction kyphosis. There were 81 late complications in 44 patients (47%). Most of the complications were minor (i.e., the patient did not require additional care or surgery). Pseudarthrosis developed in seven patients (7.5%), presenting at an average of 30 months postoperatively. Late complications in eight patients required nine procedures: five for repair of a pseudarthrosis, three for removal of a prominent iliac screw, and one for superior junctional kyphosis. 4
In the series of Master et al, there were 46 major complications in 37 patients (28% prevalence), including two deaths. Non-walking status and a larger preoperative curve magnitude were associated with an increased prevalence of major complications. Nonambulatory patients were almost four times more likely than ambulatory patients to have a major complication. A preoperative major curve magnitude of more than 60 degrees was the most accurate indicator of an increased risk for a major complication. 5
Even in the era of pedicle screw instrumentation (in which there is theoretical advantage of less intraoperative blood loss than in the Luque–Galveston technique), complication rates are significant. Modi et al reported a 32% rate of major complications in their series, most of them pulmonary. There were two perioperative deaths, and one patient developed neurologic deficit due to screw impingement in the spinal canal, which resolved after removal. 6
17.1.3 Lack of Demonstrable Functional Benefit
A further problem is a dearth of papers that show significant functional improvement following surgery in neuromuscular scoliosis. Modi et al 6 retrospectively examined 32 patients at a mean of 3 years after surgery. Using a modified Ranchos Los Amigos Hospital functional rating system, in which grade 1 indicates an independent ambulator and grade 5 indicates a bed-bound patient, they demonstrated that 42% of the patients had gained functional improvement of one grade or more. For those who were non-walkers, the health gain was either improvement to independent sitting or conversion from a bed-bound state to some form of secured seating arrangement.
Watanabe et al 7 noted that functional improvements after surgery seemed limited, with improved sitting balance the most noteworthy (93%). Nonetheless, 8 to 40% of patients perceived their surgical results as an improvement, with an overall satisfaction rate of 92%.
The study of Bohtz et al 3 demonstrated no significant correlation between degree of scoliosis correction and parents’ or caregivers’ subjective rating of change in health-related quality of life. Of interest was the lack of correlation between the occurrence of complications or changes in health-related quality of life and the rate of satisfaction with the outcome of an operation.
Mention has already been made of the high satisfaction rate reported by Comstock et al. 1 It is therefore of note that there was a high rate of late complications. At a median follow-up of 4 years (range, 2–14), late progression of scoliosis, pelvic obliquity, and decompensation were noted in more than 30% of the patients. More than 75% of the patients with late progression were skeletally immature at the time of surgery and underwent a posterior procedure only. Disease progression necessitated a revision procedure in 21% of the patients.
In summary, despite the often high rates of complications after surgical treatment in patients with spastic neuromuscular deformity, the rates of parental and caregiver satisfaction are high. Moreover, there is limited evidence for significant functional gain in the majority of reports and no correlation between radiologic improvement (in scoliosis or pelvic obliquity) and reported satisfaction. Potential explanations for these apparent inconsistencies are discussed below.
17.2 Retrospective Studies
The literature on the outcomes of surgery in patients with neuromuscular deformity consists largely of retrospective series from single centers. In the majority, no validated outcome instruments or questionnaires are used. The inherent methodologic problems are illustrated by the study of Watanabe et al, 7 in which 84 patients with spastic cerebral palsy were evaluated at a mean of 6.2 years postoperatively (range, 2–16). In addition to radiologic evaluations, the investigators used a questionnaire that sought to determine aspects of pain, function, appearance, and other health-related quality-of-life issues specific to this patient group:
Expectations (1 question);
Cosmesis (2);
Function (6);
Patient care (3);
Quality of life (3);
Pulmonary issues (2);
Pain (1);
Comorbidities (1);
Self-image (1);
Satisfaction (3).
The authors demonstrated that after surgical treatment, the majority of patients (or more appropriately their parents and caregivers) were satisfied. Those patients who were less satisfied with their outcome had more postoperative complications and less correction of the major curve Cobb angle.
However, closer scrutiny of the paper reveals fundamental flaws. Because of severe learning difficulties, only four of the 84 patients (5%) answered the questionnaire; for the remaining patients, the caregivers or parents completed the questionnaire. Furthermore, the preoperative status and postoperative status for each functional domain were included in the same question. Thus, the validity of the results may be called into question because of recall bias (especially over a 16-year follow-up period in some cases) and third-party reporting (when a patient’s experience is given by proxy). Finally, the value of an outcomes instrument in which no explanations of internal consistency, construct validity, or test–retest reliability are included must be queried.