Hold that space! Which is the best space maintainer? A RCT

We all know that when a primary second molar is lost early we would like to hold the space with a space maintainer.  Surprisingly, there has been limited research into this common clinical problem. This new trial was fascinating.

When we see a patient with the early loss of a primary second molar, we need to decide whether to maintain the space or not. If we decide on space maintenance, we then choose between a fixed or a removable appliance.  The investigators of this new clinical trial looked at this dilemma.

What did they ask?

They did the study to find out:

“The efficacy of space maintenance of the primary second molars using a vacuum formed appliance or a banded space maintainer”.

What did they do?

They did a parallel-sided randomised trial with a 1:1 allocation.  The PICO was

Participants: 

Mixed dentition patients with the early loss of at least one primary second molar in the previous three weeks.

Intervention:

Vacuum-formed space maintainer. They asked the patients to wear this for at least 20 hours per day.

Control:

Banded fixed space maintainer.  When they had only lost one molar, they fitted a simple band and loop appliance. If they had a bilateral loss, then they provided a lingual arch.

Outcome:

Change in mesiodistal width of the extraction space. The authors recorded this at the start of treatment (T0), after three months (T2) and after six months (T3).

The operator fitted the space maintainers within one month of the extractions.

Secondary outcomes were several other dental measurements, for example, intermolar width and axial rotation of the molars. I will concentrate on the primary outcome in this post.

They did a sample size calculation based on a previous study looking at the closure of extraction spaces. They set the minimum difference to be detected at 0.5mm.  This calculation revealed that they needed to recruit 11 patients in each group. They increased the sample size to 30 in each group to take dropouts into account.

They used simple randomisation that one person performed, who informed the operator of the allocation.  As a result, randomisation and concealment were good.  They analysed the data blind.

What did they find?

Thirty patients took part in the trial. The mean age of the vacuum formed group was 7.2 (SD=1.0) years, and the fixed groups mean age was (8.2 (SD=0.8) years.

The baseline characteristics were similar for both groups. However, the fixed space maintainer group was one year older than the vacuum formed group.

I have included the data on the width of the extraction spaces for the groups in the table below. These are the width of the extraction space with 95% confidence intervals.

The mean change from baseline to 6 months was -0.33mm (95% CI=-0.6 to –0.03) for the vacuum formed and 0.01mm (95% CI= -0.15 to 0.16). These differences were statistically significant. However, I felt that these differences were not clinically significant, particularly when considering that the 95% confidence intervals included numbers that were very close to zero.

Their overall conclusion was:

“Both the retainers were clinically effective. Any differences between the retainers were not clinically significant”.

What did I think?

I thought that this was an interesting and well-done small study. It was great to see a study done into a common clinical problem.  It was also good to see a study reported simply with a statistical analysis that was relevant and concise.  The data was easy to interpret, and the authors addressed the issue of statistical and clinical significance very clearly.

However, few  studies are perfect, and I had two concerns about this study. Firstly, the authors reported that the data were collected blind. Unfortunately, they did not explain how they did this with the fixed space maintainer. I assume that they removed the appliances before they measured the models.

I was more concerned with the age differences between the two intervention groups. The fixed maintainer group were one year older than the vacuum formed retainer group.  One effect of randomisation is that the groups should be balanced for co-variables, for example, age.  As a result, this difference may have occurred entirely by chance. Alternatively, there could have been an unidentified bias that happened in randomisation.  The authors pointed this out and explained that they could have avoided this if they had done stratified randomisation.

Finally, we need to remember that this study reported short term outcomes.  I hope that they can extend this trial to evaluate long-term outcomes concerning the space maintainers’ effect on the need for orthodontic treatment when the children are older.

However, when we interpret this paper, we need to consider the clinical effect size. This difference is tiny. As a result, I can agree with their conclusions, albeit with a degree of uncertainty.

Final comments

This study provides us with handy information that suggests space maintainers are effective.  It was also interesting to see that they achieved high levels of co-operation with the removable maintainer.  I hope that they can extend the study to give us some long term effects.