The objective of this study was to determine the degree of agreement between hand-wrist radiography and cervical vertebral maturation analysis in patients diagnosed with short stature.
A cross-sectional study was designed; 178 patients (90 girls, 88 boys) diagnosed with short stature and seeking treatment were selected. The patients were divided into 2 groups (76 with familial short stature, 102 with nonfamilial short stature). Hand-wrist and lateral cephalometric radiographs were obtained from the patients. The hand-wrist radiographs were analyzed using the Fishman method, and the lateral cephalometric views were categorized according to the method of Hassel and Farman. The degree of agreement between the 2 methods of predicting skeletal maturation was measured by calculating the contingency coefficient and the weighted kappa statistic.
A high degree of agreement was observed between the 2 methods of analyzing skeletal maturation. It was also observed that agreement was higher in girls in the familial short-stature group, whereas boys had higher agreement in the nonfamilial short-stature group.
Cervical vertebral maturation can be a valuable substitute for hand-wrist radiography in patients with short stature.
Cervical vertebral maturation is a reliable method for determining skeletal maturation in patients with short stature.
Short stature is a term attributed to any person who is significantly below the average height for a person of the same age and sex. If a child’s height is more than 2 SD below the average (ie, below the fifth percentile) of children of the same age and sex, the child is labeled as having short stature. There are various types of short stature, which can be differentiated by determining the pattern of skeletal maturity. Short stature can be categorized into 3 main types, each having its own characteristics. In patients with familial short stature, bone age is normal, corresponding to the patient’s chronologic age. Patients with constitutional delay of growth and puberty have delayed bone age, typically by 2 SD. Finally, in patients whose short stature is a result of a pathologic condition, bone age is severely delayed (usually more than 2 SD), and the delay worsens over time as the condition persists. Epidemiologic studies have shown that the prevalence of persons diagnosed with short stature ranges from 2.5% to 8.8% in the United States.
Orthodontic treatment in growing patients with jaw discrepancies consists of growth modification using orthopedic functional appliances. The timing of treatment becomes crucial with these appliances. It has been reported that the best time for initiating growth modification treatment is just before and during the peak velocity of mandibular growth. Generally, the pattern of growth in the mandible follows the overall growth of the body; thus, the timing of their corresponding peak velocities usually coincides.
The standard method for analyzing growth is to calculate skeletal maturation using hand-wrist radiography. However, a technique based on the calculation of cervical vertebral maturation (CVM) from routine lateral cephalometric views is rapidly gaining in popularity. It has been demonstrated that there is a significant correlation between CVM and hand-wrist radiography in determining skeletal maturation.
Although patients diagnosed with short stature constitute a significant portion of the population, no study has yet investigated the effectiveness of CVM in the analysis of skeletal maturation in these patients. Therefore, we aimed to correlate the results obtained from CVM with hand-wrist radiography to advocate its use in patients with short stature.
Material and methods
The samples were derived from 178 patients (90 girls, 88 boys) diagnosed with short stature seeking therapy at the Department of Endocrinology, Namazi Hospital, Faculty of Medicine, University of Medical Sciences, in Shiraz, Iran. The patients ranged in age from 8 years 2 months to 17 years 9 months. After paraclinical and hand-wrist evaluation, the patients were further divided into familial (76: 38 girls, 38 boys) and nonfamilial (102: 52 girls, 50 boys) groups. To have a sample that better represents the total population of familial short-stature patients, if 2 or more persons were related, only 1 was enrolled in the study. Our selection criteria included patients with height below the fifth percentile, no history of previous trauma or injury to the wrist and neck areas, and no malformations of the cervical vertebrae whether congenital or acquired.
Cephalometric views were obtained on the same day as the hand-wrist radiography. All radiographs were taken with the same equipment (cephalometer PM 2002 EC Proline, KV 85; Planmeca, Helsinki, Finland). Other parameters, including source-to-object and object-to-film distances and x-ray beam characteristics, were identical for all subjects. All assessments were performed by 2 observers, a board-certified oromaxillofacial radiologist (Sh.Sh.) and a resident of orthodontics (A.K.) trained by the radiologist before the study. The observations were made in a semidark room on a radiographic view box for best possible contrast of the views. A black cover was used for viewing the films, with a window showing only the area of cervical vertebrae C2 through C4. This cover omitted the light from the surrounding structures, hence enhancing the viewer’s focus on the desired field of view. Furthermore, using this cover helped to eliminate a possible source of bias from the surrounding structures in the craniofacial complex. The viewers performed a secondary assessment 6 months after the initial observation to determine the intraobserver agreement.
The hand-wrist radiograph was taken from the patient’s left hand according to the conventional method for hand-wrist radiography. Skeletal maturation was determined using the method of Fishman. The lateral cephalometric views were analyzed to determine the cervical vertebral stages according to the method of Hassel and Farman. The observers were blinded for both hand-wrist and lateral cephalometric radiography separately and were unaware of the patients’ chronologic ages.
Fishman’s method consists of 11 stages of development, and Hassel and Farman’s method has 6 stages. To assess the agreement of the 2 methods more easily, Fishman’s method was converted to 6 stages. Each 2 stages were merged into one, and the 11th stage was called stage 6.
We calculated descriptive statistics by determining the means and standard deviations of the chronologic ages for the 6 stages of the CVM. The degree of agreement between hand-wrist and CVM stages with each classifier was assessed by calculation of the contingency coefficient and the weighted kappa statistic. The weighted kappa statistic was used for interobserver and intraobserver agreement. According to Altman’s classification method, agreement ranges from 0 to 1 (poor: κ <0.2; fair: 0.2 < κ <0.4; moderate: 0.4 < κ <0.6; good: 0.6 < κ <0.8; very good: κ >0.8). Statistical analyses were performed with SPSS software (version 11; SPSS, Chicago, Ill).
We studied 76 persons with familial short stature and 102 persons with nonfamilial short stature. Table I shows the subjects’ distributions by sex and chronologic ages, grouped by CVM stage and short-stature state. As shown in Table I , in the familial short-stature group, stages 3, 4, and 5 appeared earlier in the boys, and stages 1, 2, and 6 appeared earlier in the girls. In the nonfamilial short-stature group, nearly every stage (except stage 1) appeared earlier in the girls than in the boys. The reproducibility of all the assessments was good. For interobserver agreement, the weighted kappa statistics for hand-wrist assessments and CVM assessments were 0.81 (95% confidence interval [CI], 0.79-0.83) and 0.89 (95% CI, 0.88-0.91), respectively. The weighted kappa statistics for intraobserver agreement were 0.94 (95% CI, 0.92-0.96) for hand-wrist assessments and 0.96 (95% CI, 0.95-0.97) for CVM assessments. The associations between hand-wrist stage and CVM stage were assessed for subjects according to their sex and short-stature state ( Tables II and III ). As shown in Tables II and III , lower stages of hand-wrist maturation were associated with lower CVM stages. Again, the higher the hand-wrist stage, the higher the CVM stage. Values of contingency coefficients and weighted kappa showed a highly significant association between familial short-stature and nonfamilial short-stature subjects ( Table IV ).
|Short-stature state||CVM stage||Sex||Subjects (n)||Chronologic age (y)|
|Sex||CVM stage||HW 1||HW 2||HW 3||HW 4||HW 5||HW 6||Total|
|Sex||CVM stage||HW 1||HW 2||HW 3||HW 4||HW 5||HW 6||Total|