Can patients change aligners at 7, 10 or 14 days? An RCT
This post is about a new RCT that looked at the effectiveness of differing aligner wear protocols. I thought that it was great to see a team carrying our a randomised trial into aligner treatment.
I have posted about aligners before and stated that for such a frequently carried out treatment; there is so little research. I was, therefore, very interested in this new paper. Before you read my opinion, I would like to point out that I have never done an aligner case. As a result, from a practical point of view, I know nothing. However, I am going to look at this from a scientific viewpoint.
This paper was really about finding out the effect of changing aligner at different time points. As a result, I thought that it was clinically relevant.
A team from Dubai and Virginia did this trial. The Angle Orthodontist published the paper.
What did they ask?
They did this study to:
“Evaluate and compare the efficacy of different orthodontic tooth movements with different aligner wear protocols (seven days, ten-day and fourteen days) by comparing the predicted outcome with the actual outcome”.
What did they do?
They did a three-arm parallel-group RCT with a 1:1:1 allocation. The PICO was
Participants: 80 patients with a malocclusion that could be treated with Invisalign aligners with a total sequence between 17 and 25 aligners.
Interventions: Changing aligners either every seven days, ten days or fourteen days
Outcome: The differences for each tooth between the actual and the predicted tooth positions.
A highly skilled experienced operator (Tier-level Diamond Plus Provider) who had treated more than 2500 cases, did the treatment.
They did remote randomisation, but they gave no information on allocation concealment. I was surprised that I could not find a sample size calculation. I will return to this later.
Data collection was blinded.
They did a complicated method of tooth analysis and superimposed the predicted and end of treatment scans on the best fit of the teeth. I was not sure about this, as the teeth are not stable structures.
What did they find?
They presented a large amount of tooth level data, and I have not space on this blog to insert the relevant tables etc. But, the paper is open access so that you can have a look at it yourself.
In summary, they found that there were no clinical or statistically significant differences between the wear protocols. They did find that there were five statistically significant differences between some of the tooth movement, but the effect sizes were very small. As they carried out 48 univariate statistical tests on the data, it should not surprise us that they found five statistically significant differences, as these probably occurred by chance. These spurious findings are what happens when you analyse data to kingdom come!
Their overall conclusions were
“Achieving a clinically similar accuracy between the seven and fourteen-day protocol in half the treatment time suggests a seven-day protocol is an acceptable treatment protocol”.
What did I think?
I have, previously, been accused of only being critical of papers that are supported by industry or done by KOLs. I have always thought that this was a little unfair. However, I have taken this comment on board and while I always try and be positive. I find it challenging to be very optimistic about this paper. Unfortunately, I thought that it was a poor paper and a missed opportunity. I wonder why the Angle referees did not identify the many problems with this paper?
Let’s start with the good points. It was great to see a study that was done in specialist practice. I want to congratulate Dr Kravitz for the large amount of work that he must have done in this study. I thought that the clinical component of the study was excellent.
I am also very heartened to see investigators researching aligner therapy. Many of us consider this to be some form of ‘dark art’; independent research will help to overcome this.
Problems with the method
Unfortunately, there are substantial problems with the research methods adopted by the academic researchers. These are:
I was surprised to see that there was no method of allocation concealment.
Most importantly, there was no sample size calculation. This problem means that we do not know if the study was sufficiently powered to detect a difference between the interventions. The absence of a sample size calculation is a fatal flaw in a randomised trial. It is usually a reason for the complete rejection of a publication. Nevertheless, the observed differences do appear to be minor and of limited if any clinical relevance.
I was not sure about the method of superimposition on the teeth. These are not stable and this may lead to errors in measurement.
I disagreed with the outcome measure. I do not understand why the researchers did not randomise a group of patients to the three groups and then measure the end of treatment result with an occlusal index. This information would have been more straightforward and had some meaning.
Finally, I am not 1oo% sure how a study had an ethical approval in Dubai, and yet the trial was carried out in the USA. This certainly is not common practice.
Are the results useful? I do not know.