Ghafari 1998.

Methods	Location: The University of Pennsylvania, USA Number of centres: 1 Recruitment period: not stated Funding source: this study was supported by grants RO1‐DE08722 and RR‐00040 (NIH) Trial design: parallel group RCT
Participants	Inclusion criteria: Class II, Division 1 malocclusion associated with bilateral distocclusion (unilateral Class I excluded) and a minimum ANB angle of 4.5°; between 7 and 12.5/13 years of age; no prior orthodontic treatment; and expected residential stability of 3 years Exclusion criteria: children with systemic, mental, behavioural, bleeding, and craniofacial disorders were excluded. If siblings presented with the same malocclusion, only 1 of them was recruited because they share in both the genetic background and environment Age at baseline: chronological age range 7 years 2 months to 13 years 4 months. Skeletal age range at baseline 5 years 9 months to 13 years 9 months and was basis of grouping participants into early (< 10 years for girls and < 10.5 years for boys) and late childhood Number randomised: 84 Number evaluated: 63
Interventions	Gp A (n = 35/41): headgear ‐ straight pull headgear inserted into the buccal tubes of bands cemented on permanent maxillary front molars Gp B (n = 28/43): Frankel function regulator type II to be worn at least 16 hours per day
Outcomes	Skeletal measurements from cephalograms, occlusal changes
Notes	Duration of randomised treatment: 2 years Sample size calculation: not specified
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomised. "Within each severity group, the children were assigned at random to treatment with either a headgear (n = 41) or a Frankel FR (n = 43)." Sequence generation method not described
Allocation concealment (selection bias)	Unclear risk	Not described
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not described
Incomplete outcome data (attrition bias) All outcomes	High risk	"Non cooperative children were those patients who, at some point in time, refused to receive treatment, despite all efforts to retain them. The largest percentage of these children were girls who wore the Fränkel regulator (42%); by contrast, the smallest number discontinued were girls in the headgear group (5%). The difference between these two groups of girls was statistically significant (P < 0.05). The percentages of boys lost to the study were similar in the headgear (24%) and FR (25%) groups." Dropouts in headgear 6/41 (15%), Frankel 15/43 (35%). This statistically significant difference between groups is likely to introduce bias.
Selective reporting (reporting bias)	Low risk	All variables reported
Other bias	Low risk	No other sources of bias identified Complete set of data not reported. Data for only 26/84 participants reported