Classic Trials revisited:  The UNC Class II study.

Very few trials are being published at the moment. As a result, I have decided to look back at several classic orthodontic trials. To do this, I have asked an AI bot to identify the top classic orthodontic trials that may have changed practice. The list is interesting and is certainly well known to anyone familiar with the literature. Today I’m going to appraise the classic study carried out at the University of North Carolina on early treatment of Class II malocclusion. 

Most orthodontists are probably familiar with the results. However, I have reviewed them in light of our knowledge of how to conduct trials, which has expanded rapidly over the past 20 years. 

A team based in North Carolina, USA, conducted the trial. The American Journal of Orthodontics published the paper. 

I would like to declare an interest, as I provided input on using the PAR index as an outcome measure. I also knew the authors very well.

This group published several papers on the results of this trial. In this paper, they condense most of the results. 

What did they ask? 

However, at the start of this study in 1990, there was controversy over treatment timing for early Class II malocclusion.

1. That early Class II malocclusion could be intercepted with a course of orthodontic treatment in Phase I, followed by a comprehensive course of treatment in Phase II during adolescence.

2. That treatment should be deferred until the dentition was established, with a single course provided during adolescence.

As a result, they did this study to.

“Find out whether treatment started in the mixed dentition before adolescence, followed by a second phase of treatment in the early permanent dentition during adolescence, provides superior results to a single-phase treatment delayed until adolescence.” 

What did they do? 

They conducted a randomised trial with a 1:1:1 allocation of interventions. The PICO was 

Participants 

Children with an overjet of 7 mm or more in the mixed dentition, at least a year before their peak pubertal growth 

Intervention 1 

A Bionator appliance worn full-time. 

Intervention 2 

Cervical pull headgear.

Intervention 3. 

An untreated control group that was observed for 15 months. 

Outcomes 

The primary outcomes were selected cephalometric measurements. Secondary outcomes were Peer Assessment Rating (PAR) scores, complexity of treatment, and total treatment time. 

During the first phase of treatment, they randomly allocated patients to the three interventions. They concealed the allocation by using a pre-prepared randomisation, which was emailed to a research assistant, who informed the clinician. At the end of this phase of treatment, patients were re-randomised, again using standard randomisation techniques, to one of four orthodontists who completed their definitive treatment. 

They carried out the relevant univariate and multivariate statistical analyses. 

What did they find? 

170 participants started the trial, and 166 completed phase I treatment. After a waiting period, 145 returned for phase II treatment. Of these, 139 completed phase II. At the final point, they analysed data for 137 participants. 

At the end of phase I, they found modest skeletal changes. However, during phase II, the differences between the groups were lost. At the end of fixed appliance treatment, there were no significant differences among the three groups for any cephalometric measurements.  For example, the final ANB values were 4.3 (2.06) for the control, 3.79 (2.12) for the early functional and 4.0 (1.9) for the early headgear groups.

For occlusion, there were no differences in Peer Assessment Rating scores at the end of phase II. For the control group the mean PAR score was 9.3 points, for the early functional this was 8.8 and for the headgear group this was (7.2)

There were also no meaningful differences in the overall treatment time or the percentage of children who had teeth extracted as part of phase II treatment. 

The overall conclusion was 

“Early treatment should not be thought of as an efficient way to treat most Class II children.” 

Importantly, they pointed out that the trial illustrated the risk of relying on clinical impressions. This was because, early in the trial, they were very impressed with the progress of the children receiving early treatment and even considered stopping the trial then. 

What did I think? 

This is a classic paper outlining a very ambitious study for its time. It was well written, clear, and succinct in the usual style of the authors. I thought it was notable that the authors spent some time outlining the rationale for conducting a randomised controlled trial. This reflected a time when trial methodology was relatively new to orthodontics. 

As with all papers written by the UNC group, there were some great quotes. I was particularly struck by this one, originally made by David Sackett. 

“Enthusiastic treatment reports tend to have no controls, whereas well-controlled treatment reports tend to have no enthusiasm” 

We would do well to remember these words now, as we are suffering an onslaught of poor-quality studies and case reports promoting orthodontic fringe treatments.

Final thoughts. 

This paper was a valuable addition to our knowledge and set the scene for further, more complex trials that I and others could carry out.

We need to consider whether it withstands contemporary scrutiny. I think it is important to note that when this paper was included in Cochrane reviews, it was classified as having a high risk of bias. This was largely because the allocation concealment was not clearly reported, but more importantly, the cephalograms were digitised in a non-blinded manner because the headgear bands were left in place. As a result, the technician knew the treatment allocation. This does not mean the study is biased. It means the study is at high risk of bias. 

This is also similar to many studies carried out at the time and reflects the methodology we used and possible reporting deficiencies, as the CONSORT guidelines had not yet been produced.

Nevertheless, we need to consider the findings in light of these issues. I have thought about this very carefully. I still feel that this is a great piece of research that changed orthodontic practice. As a result, the findings are relevant. 

Finally, I feel that those who promote early treatment, expansion, or orthodontic-driven treatment for paediatric sleep disorders should revisit this paper. They will learn a lot about complex orthodontic research and how to interpret it. We must not forget our classic research as we attempt to move forward.