This study was approved by the Ethical Committee of the Rio de Janeiro State University (number 2.281.471). It is a retrospective cephalometric study designed to compare the dentoskeletal effects produced by two treatment modalities for the correction of Class II malocclusion: the maxillary splint headgear and the cervical headgear appliances. Sample size calculation determined that for the independent t test, with a minimal detectable difference of 1.5 mm for the overjet, a standard deviation of 1.4 mm [10], an alpha level of 0.05, and a power of 0.80, 15 subjects were required for each group (BioEstat 5.0).
The inclusion criteria were as follows:
Class II dentoskeletal relationships (ANB angle greater than 4 degrees);
Overjet larger than 4 mm;
Class II molar relationship;
Absence of craniofacial anomalies (e.g., cleft lip and/or palate).
Lateral cephalograms for all subjects had to be available at the beginning of treatment (T1) and at the end of early treatment with an extraoral appliance (T2). Dental casts for the evaluation of the phase of the dentition had to be available at T1. Phase of the dentition was defined as follows [22]:
- 1.
Early mixed dentition: Shedding of the deciduous incisors, eruption of the first permanent molars and permanent incisors;
- 2.
Intermediate mixed dentition: Permanent incisors and first molars fully erupted, presence of all deciduous teeth in the buccal region (deciduous canine, first molar, and second molar);
- 3.
Late mixed dentition: Shedding of one or more deciduous canines and molars, eruption of the permanent canines and premolars;
- 4.
Early permanent dentition: Presence of all permanent teeth (possible presence of second molars; absence of third molars).
If the patient had some early loss of deciduous teeth, the stages of permanent teeth eruption were considered for classification.
All patients treated consecutively with maxillary splint headgear were derived from one private practice in Rio de Janeiro, while the cervical headgear group was from the records of the patients treated consecutively at the University of Florence. The maxillary splint headgear group (MSG) consisted of 28 patients (16 females and 12 males), and the cervical headgear group (CHG) included 28 patients (15 females and 13 males). During the observational period, no other appliances were used in both groups.
The maxillary splint headgear comprised a removable maxillary self-polymerizing acrylic resin splint extended laterally and occlusally, covering the cusps and approximately one third of the buccal surfaces of all teeth, and a high-pull orthopedic traction. The face bow was bent upwards at about 45° from the horizontal plane in relation to the intraoral arch and connected to an elastic (0.5 in. × 1.5 mm), which was attached to a head strap (Morelli Ortodontia Sorocaba, Brazil) (Fig. 1). The extraoral force was 400 g per side and it was calibrated with a dynamometer (Ohaus Corp., Florham Park, New Jersey, USA). Treatment with the maxillary splint headgear lasted 1.5 years on average. In the cervical headgear group, the upper first molars were banded, and a cervical orthopedic traction was applied. The face bow was bent upwards about 10° from the horizontal plane in relation to the intraoral arch. The cervical headgear delivered forces of 250 g per side that were calibrated with a dynamometer inserted in the spring modules (Leone Orthodontics Products, Sesto Fiorentino, Firenze). Treatment with cervical headgear lasted 1.7 years on average. Patients were instructed to wear both appliances for 14 h/day. During treatment, all patients were scheduled monthly so that the appliances could be adjusted, if necessary, and the forces could be verified. According to the patient’s records, if the patient had failed any appointment, this was re-scheduled as soon as possible and attendance was considered adequate for both groups. In the maxillary splint headgear, in cases of eruption of permanent canines and premolars, the acrylic resin was ground in order to obtain space for eruption.
Cephalometric analysis
All lateral cephalograms were digitized and a customized cephalometric analysis was provided by a specific cephalometric software (Viewbox, version 3.0, dHAL Software, Kifissia, Greece). Fourteen variables (3 angular and 11 linear) were assessed for each tracing, according to a modified Pancherz`s cephalometric analysis (Fig. 2) [23]. The established enlargement factor for all cephalograms was standardized to a magnification factor of 0%.
All lateral cephalograms were traced initially by the same operator (J.M.) and were checked by a second operator (L.F.) to verify anatomical outlines, landmark placement, and superimposition. Both operators were blinded with regard to the origin of the headfilms and the group to which the patient belonged. Any disagreements were resolved to the satisfaction of both operators.
Method error and statistical analysis
In order to determine the method error, the same operator redigitized twenty randomly selected cephalograms and recalculated the variables after 15 days. The systematic error was evaluated with the Wilcoxon signed-rank test. The random error was calculated with Springate’s method of moments’ estimator (MME) [24].
The Shapiro Wilk’s test showed normal distribution of cephalometric data at T1 and of T2-T1 changes. Therefore, statistical comparisons between the two groups for the cephalometric measurements at T1 (starting forms) and for the T2-T1 changes were performed by means of independent sample t tests. The difference in the distribution of the phases of the dentition in the 2 groups at T1 was assessed with Fisher Exact Probability Test (SPSS version 12.0, SPSS Inc., Chicago, IL).