Cerebral palsy (CP) is one of the most common causes of physical disability in childhood, with a worldwide prevalence of 2.11 per 1000 live births . Despite being considered a static encephalopathy, joint contractures often progress and require careful evaluation and management . The clinical assessment of children with CP routinely includes passive range of motion (ROM) measurements, which not only effectively estimate the contractures of major joints in the lower limbs but also serve as guidance for surgical treatment. However, the reliability of passive ROM examination in conscious CP patients is insufficient [3–5], which may confound treatment decisions and the response to surgical interventions.
Before surgery, the passive ROM of target joints is commonly reassessed postanesthetically because ROM examinations under general anesthesia (GA) are more reliable than those performed while the patient is conscious . It is useful to differentiate between a fixed contracture requiring surgical release and a dynamic tightness induced by spasticity and patient resistance. Theoretically, the passive ROM of target joints will increase after GA because of eliminated influences of spasticity and patient resistance and optimized patient compliance. Previous researches into the impact of GA on the measurement of lower limb contractures have shown limited and differing results. McNee et al.  reported no significant differences in the popliteal angle of 30 limbs in 15 children with CP. McMulkin et al.  found a significant difference in the Silfverskiöld test only in CP patients younger than 11 years old. To the best of our knowledge on the basis of literature review, the influences of GA on measurements of the hip abduction angle and Thomas test have not been addressed. The aim of this study was to evaluate the post-GA increase in ROM and associated predicting factors for children with CP using a retrospective review of prospectively collected data among children with spastic cerebral palsy (SCP) receiving corrective surgeries at our institute. The null hypothesis was that the hip abduction angle, the Thomas test, the popliteal angle, and the Silfverskiöld test would not change significantly after GA.
Patients and methods
The institutional review board approved the present study. Informed consents from participants were unnecessary because of the retrospective design and lack of identifying materials in data processing. The inclusion criteria were as follows: (a) SCP cases aged younger than 18 years, and (b) admission to our institute for corrective surgeries of lower limb deformities. From November 2001 to January 2011, 91 pediatric SCP cases were admitted for surgical soft tissue release, with or without contemporary bony procedures. In all cases, the diagnosis of SCP was confirmed by senior pediatric neurologists. Seven patients were excluded because of previous corrective surgeries for lower limb deformity. A total of 84 cases were enrolled in this study (Table 1). The average age of the patients at the time of surgery was 8.4 years (ranging from 4 to 18 years). Most cases had been diagnosed with spastic diplegia (39 cases) or quadriplegia (35 cases). All enrolled participants completed a preoperative assessment of the modified Ashworth scale (MAS), the Gross Motor Function Classification System (GMFCS) and Gross Motor Function Measure – Dimension D (GMFM-D). In total, 435 soft tissue procedures had been carried out on 168 lower limbs (an average of 2.59 procedures per limb).
Measurements of joint range of motion
The ROM of hip, knee, and ankle joints of all the participants were measured just 1 day before the operation and immediately after the completion of surgical anesthesia in the operation room. All measurements were performed by the same examiners using the same protocols and equipments. The examiners included one senior orthopedic surgeon and one physiotherapist. The physiotherapist acted as the holder who stretched the joint to the endpoint of ROM and the surgeon was the principal assessor who checked the end ROM performed by the holder and measured the angles using a goniometer. Five physical examinations were performed in uniform sequence: hip abduction angle, Thomas test angle, popliteal angle, and ankle dorsiflexion angle with knee flexion and extension. The examination procedure and the goniometer alignment are detailed in Table 2. The endpoint of ROM was reached as slowly as possible, without triggering of spasticity. Each physical examination involved three repeated measurements and the mean was used as the final record. It should be noted that the popliteal angle is the angle left before full extension of the knee joint, not the angle between the thigh and the calf. A negative value in ankle dorsiflexion measurement represented the endpoint in plantar flexion. Before corrective procedures were performed, the same measurement protocols were performed again by the same examiners when the patient had reached the third stage of GA (surgical anesthesia) in the operating room .
Evaluation of alternation of surgical plans
The pre-GA surgical plans were obtained from the surgical consents for comparison with the post-GA surgical plans from the operative reports. All of the surgical plans were prepared by the same orthopedic surgeon who was also responsible for measurement of joint ROM, interpretation of instrumented gait analysis, and operations. At our institute, all of the CP cases were subjected to a three-dimensional instrumented gait analysis before corrective surgeries. Surgical release of each tendomuscular structure was indicated according to not only the corresponding measurement of passive ROM, but data of kinematics, kinetics, and surface electromyography of key muscles (Table 3) . The ROM measurements (before and after GA), interpretation of gait analysis, surgical planning, and corrective surgeries were all performed by one senior surgeon (the corresponding author).
The statistical analysis of this study consisted of two parts: (a) A paired Student’s t-test was used to evaluate the post-GA change in the five ROM examinations; (b) Multivariate linear regression was used to evaluate the impacts of the pre-GA angle, age, body weight, sex, functional scores (MAS, GMFCS, and GMFM-D), and geographic classification on post-GA changes of the five ROM examinations; (c) McNemar’s test was used to evaluate the difference in surgical planning between conscious and post-GA states; (d) Individual t-test was used to compare pre-GA measurements and post-GA differences between patients with and without altered surgical plans. All statistical analyses were carried out using SPSS statistics (version 17.0; IBM, New York, New York, USA). All P values less than 0.05 were considered to be statistically significant.
The null hypothesis of this study was rejected. GA significantly impacted the values of the hip abduction angle, the Thomas test, the popliteal angle, and the ankle dorsiflexion angle with flexion and extension of the knee joint (Fig. 1). The assessed tightness reduced after GA in all of the five physical examinations. Because greater contracture of the iliopsoas and hamstring muscles generated larger values in the Thomas test and the popliteal angle, respectively, negative post-GA variances in these two measurements also represented a reduction in contractures. The greatest post-GA variance occurred in ankle dorsiflexion with knee extension (7.0° and 109% increased), followed by ankle dorsiflexion with knee flexion (6.6° and 39.8% increased), the hip abduction angle (11.1° and 39.5% increased), the popliteal angle (15.0° and 19.1% decreased), and the Thomas test (3.7° and 18.0% decreased).
Several clinical factors significantly influenced the post-GA variance of lower limb contractures (Table 4). Among these, the pre-GA angle had the most significant effect on all of the five measurements (P = 0.000). Age and body weight also impacted the post-GA variance in the hip abduction angle and the Thomas test, with less post-GA relaxation reported in older and lighter patients. However, sex, geographic classification, and functional scores, including GMFCS, GMFM-D, and MAS, did not affect the post-GA ROM, except greater post-GA angle of the Silfverskiöd test with knee flexion found in patients with higher MAS of triceps surae muscle.
More importantly, GA also significantly impacted the surgical planning for lower limb contractures (Table 5). The overall post-GA cancelation rate of soft tissue releasing procedures reached 9.5%. According to measurements when patients were anesthetized, up to 42.3% of tenoachilles-lengthening procedures were changed to gastrocnemius recession (34.6%) or canceled (7.7%) after GA. The individual cancelation rate was statistically significant for all of the soft tissue procedures used in this study, except over-the-brim psoas lengthening. Patients with altered surgical plans sustained lower pre-GA contracture and higher post-GA increase in ROM, despite the absence of statistical significance (Table 6).
Contracture measurement through physical examination provides instant and significant information to treat children with CP. Despite its irreplaceable role in clinical practice, insufficient reliability of contracture measurements may complicate treatment decisions. Previous studies have reported inconsistent results of intraobserver and interobserver reliability in the measurements of lower limb contractures [1,4,5,10]. However, in general, intraobserver variation has been proven to be lower than interobserver variation [1,10]. The single-observer design of the present study theoretically provided reliable measurements to analyze the post-GA change in lower limb contractures.
Limb contracture in SCP patients consists of dynamic and static components. The static component is generated by fibrosis of tendomuscular structures and shortened muscle fiber, and the dynamic one by spasticity and patient resistance . Few studies have addressed the direct effect of GA on spasticity in SCP patients. Fee and Miller  compared the pattern of the leg drop pendulum test between SCP and normal individuals before and after GA. The hypersensitive stretch reflex of quadriceps femoris in SCP patients is removed by GA and paralytic agents, which normalizes the phase plot of leg pendulum to a uniform ‘whirlpool’ pattern observed in normal children. Accordingly, GA could eliminate the dynamic component of contracture in SCP patients and lead to a post-GA decrease in lower limb contracture observed in the present study.
All the contractures measured in this study decreased significantly after GA (Fig. 1), which conflicted with the results of previous studies. McNee et al.  found no significant reductions in the popliteal angle in a small group of SCP patients (2.83° in 15 cases, P = 0.17). However, the severity of disease and surgical history were not available in the characteristics data. McMulkin et al.  researched post-GA change in the popliteal angle and ankle dorsiflexion in 35 SCP patients. They discovered significant increases in ankle dorsiflexion only in patients younger than 11 years old. Nevertheless, the selection biases including less severe disease (29 cases with diplegia, one case with quadriplegia) and previous surgeries on calf or hamstring muscles (35% of cases) probably confounded the results. A greater number of patients, more homogenous geographic classification of disease, and surgically naive patients enrolled in the present study could have produced more accurate results.
The pre-GA ROM had a significant impact on the post-GA ROM (Table 4). The greater the contracture in the conscious state, the more relaxed it was after GA. Because of the negative coefficient of the pre-GA angle in multivariate linear regression, greater pre-GA contracture leading to a smaller measurement angle (such as the hip abduction angle) led to more positive postanesthesia variance. Greater pre-GA contracture leading to a larger measurement angle (such as the popliteal angle) led to more negative post-GA variance. Both conditions meant reduced contractures in all five measurements after GA. It is commonly accepted that three major components lower ROM of joints in CP patients: dynamic tightness generated by spasticity, static contractures because of fibrosis of tendomuscular structures , and patient resistance. A positive correlation has been proven between level of spasticity and muscle contracture in SCP patients ; hence, we reasonably supposed that both components would increase in more severe cases and GA could eliminate the dynamic tightness produced by spasticity.
The post-GA increase in ROM was independent of the function level and geographic classification of SCP in most measurements (Table 4). This finding might conflict with the hypothesis that a greater post-GA increase in ROM occurred in patients with poorer function scale and higher spasticity. Nevertheless, the correlation research between MAS and GMFM-88 reported only a moderate correlation, which implied that other factors also played an important role in motor control and diluted the impact of function scale on the post-GA increase in ROM .
Age and body weight significantly impacted the post-GA increase in the hip abduction angle and the Thomas test only (Table 4), with less increase found in older and lighter patients. This might be explained by the fact that more fibrotic changes in musculotendinous structures were anticipated in older patients, which could not be eliminated by the muscle relaxation effect of GA. Greater amount of muscle mass in heavier patients, who were more susceptible to the muscle relaxation effect of GA, might have contributed towards the greater post-GA increase in ROM. However, the impacts of age and body weight were not found on the popliteal angle and the Silfverskiöld test. These findings deserve further investigations. The present study is the first investigation, to our knowledge, to evaluate the impact of GA on surgical planning for SCP patients. We found that GA significantly changed not only ROM measurements but also surgical plans. Overall, 9.5% of soft tissue procedures were canceled or downgraded (e.g. medial and lateral hamstring lengthening was changed to only medial hamstring lengthening, and tenoachilles lengthening was altered to gastrocnemius recession) according to the post-GA reassessment of joint ROM. Most notably, up to 42.3% (15 out of 26 cases) of planned tenoachilles-lengthening procedures were changed to gastrocnemius recession because of the 109% increase in ankle dorsiflexion with knee extension after GA. Because all of the factors, except ROM measurements, used for surgical planning were identical before and after GA, post-GA alternation of surgical planning could be attributed to ROM changes caused by surgical anesthesia. In terms of the negative effects of over-lengthened heel cord, which impairs the plantar flexion-knee extension couple and causes further deterioration in crouch gait, routine post-GA reassessment of ankle dorsiflexion is strongly recommended.
Patients with altered surgical plans sustained lower pre-GA contracture and higher post-GA increase in ROM (Table 6). However, there was no statistical significance, possibly because of the low case numbers in the canceled group. Although we found a high rate of canceled procedures after the post-GA reassessment of ROM, especially the tenoachilles lengthening, it was difficult to clarify a specific level of post-GA difference where the surgical plan was impacted. ROM measurements only play a role in decision-making. Kinematics and kinetics studies should also be integrated into surgical planning (Table 3).
The present study had several limitations. Despite the single-observer design of ROM measurement, the applied force without instrument control for both pre-GA and post-GA evaluation may have caused the intraobserver variation and bias. The documentation of three repetitive measurements and evaluation of intraobserver variance were also lacking. This might be compensated by the relatively large case numbers included in this study (1680 measurements in total). Furthermore, the agent and dosing of muscle relaxant and anesthetic inhalation were not controlled for each patient. The inhaled anesthetics used in this study included isoflurane (in 27.4% of patients) and sevoflurane (in 72.6% of patients), which have the same pharmacological effects on the central nervous system . Muscle relaxants used in the induction phase of GA included atracurium (in 88.1% of patients) and succinylcholine (in 11.9% of patients). Despite the difference in active duration between atracurium (20–50 min) and succinylcholine (9–13 min) , a statistical analysis using multivariate linear regression indicated an insignificant impact of muscle relaxants on post-GA ROM (P = 0.233–0.986 for all measurements). Therefore, the inconsistence in anesthetics used between patients is negligible in the present study.
GA significantly increased joint ROM of the lower limbs among SCP children, including the hip abduction angle, the Thomas test, the popliteal angle, and the Silfverskiöld test. The increase was universal and irrespective of function level, spasticity level, and geographic classification of SCP. Patients with lower pre-GA ROM showed a greater increase in ROM after being anesthetized. In terms of the severe weakness caused by overlengthening of musculotendinous structures, a routine post-GA reassessment of ROM should be considered before corrective surgery of lower limb contractures for SCP children.
The authors thank the faculty of the gait laboratory at our institute for providing detailed data of the patients included.
Conflicts of interest
There are no conflicts of interest.
1. Oskoui M, Coutinho F, Dykeman J, Jette N, Pringsheim T. An update on the prevalence of cerebral palsy
: a systematic review and meta-analysis. Dev Med Child Neurol. 2013; 55:509–519
2. Rosenbaum P, Paneth N, Leviton A, Goldstein M, Bax M, Damiano. A report: the definition and classification of cerebral palsy
April 2006. Dev Med Child Neurol Suppl. 2007; 109:8–14
3. Ten Berge SR, Halbertsma JP, Maathuis PG, Verheij NP, Dijkstra PU, Maathuis KG. Reliability of popliteal angle
measurement: a study in cerebral palsy
patients and healthy controls. J Pediatr Orthop. 2007; 27:648–652
4. McDowell BC, Hewitt V, Nurse A, Weston T, Baker R. The variability of goniometric measurements in ambulatory children with spastic cerebral palsy
. Gait Posture. 2000; 12:114–121
5. Kilgour G, McNair P, Stott NS. Intrarater reliability of lower limb sagittal range-of-motion measures in children with spastic diplegia. Dev Med Child Neurol. 2003; 45:391–399
6. McNee AE, Gough M, Shortland AP. The effect of anaesthesia on the assessment of hamstring length in spastic cerebral palsy
. J Pediatr Orthop B. 2007; 16:35–38
7. McMulkin ML, Gordon AB, Caskey PM, Ferguson RL, Baird GO. Range of motion measures under anesthesia compared with clinical measures for children with cerebral palsy
. J Pediatr Orthop B. 2008; 17
8. Hewer CL. The stages and signs of general anesthesia
. Br Med J. 1937; 2
9. Davids JR, Ounpuu S, DeLuca PA, Davis RB III. Optimization of walking ability of children with cerebral palsy
. J Bone Joint Surg Am. 2003; 85
10. Allington NJ, Leroy N, Doneux C. Ankle joint range of motion
measurements in spastic cerebral palsy
children: intraobserver and interobserver reliability and reproducibility of goniometry and visual estimation. J Pediatr Orthop B. 2002; 11
11. Zwick EB, Leistritz L, Milleit B, Saraph V, Zwick G, Galicki M, et alClassification of equinus in ambulatory children with cerebral palsy
-discrimination between dynamic tightness and fixed contracture. Gait Posture. 2004; 20
12. Fee JW Jr, Miller F. The Leg Drop Pendulum Test performed under general anesthesia
in spastic cerebral palsy
. Dev Med Child Neurol. 2004; 46
13. Hagglund G, Wagner P. Spasticity of the gastrosoleus muscle is related to the development of reduced passive dorsiflexion of the ankle in children with cerebral palsy
: a registry analysis of 2,796 examinations in 355 children. Acta Orthop. 2011; 82
14. Katusic A, Alimovic S. The relationship between spasticity and gross motor capability in nonambulatory children with spastic cerebral palsy
. Int J Rehabil Res. 2013; 36:205–210
15. Misha Perouansky RAP, Hugh CH. Miller RD. Inhaled anesthetics.. Miller’s anesthesia. 20148th edition, Philadelphia, PA: Elsevier614–637
16. Mohamed Naguib CAL, Claude M. Miller RD. Pharmacology of neuromuscular blocking drugs.. Miller’s anesthesia. 20148th edition, Philadelphia, PA: Elsevier958–994