Do extractions influence the airway? A new study.
Despite the current lack of evidence, orthodontists are still promoting non-extraction treatment as part of “airway friendly orthodontics”. The authors of this new paper looked at the relationship between extraction treatment and airway changes.
Over the past ten years, there has been increasing interest in the relationship between orthodontics and disordered breathing. I have posted a lot about this before. We have also included an excellent summary of the AAO meeting on airway and orthodontics. In general, there is an absence of evidence to suggest that there is a relationship between orthodontics causing or treating breathing disorders.
A team from San Francisco did this new study. The Angle Orthodontist published their paper.
In their literature review, they mentioned that there is a belief that orthodontic extractions may result in a collapse of the airway. However, there is little evidence for this claim. So they did this study.
What did they ask?
They did the study to answer the following question;
“Is there a relationship between the extraction of premolars and changes in the airway dimensions in a sample of non-growing patients”?
What did they do?
They did a retrospective study.
In the first stage of the study, they searched the records of the patients who had been treated in their orthodontic department from January 2007 to June 2018. They used the following inclusion criteria to identify the patients in the study:
- The patients had to be at least 18 years old.
- They had pre and post-treatment CBCT images.
- At least two premolars were extracted as part of their treatment (extraction group)
- No extractions were done (non-extraction group).
At the end of this process, they had identified 42 patients in the extraction and the non-extraction groups. Within the extraction group, 6 patients had only had upper premolars extracted.
They analysed the CBCT scans by dividing the airway into three areas. These were the nasopharynx, retropalatal region of the oropharynx and the retroglossal region. They then calculated the airway volume for each part (mm3) and the most constricted area (MCA) of the airway in mm2. They also calculated a large number of other measurements.
Two judges made all the airway measurements. Finally, they did a large number of statistical tests across the variables. The extensive amount of data was presented in several complex tables.
What did they find?
I am sorry to be so honest here. But I do not know what they found. This is because the data presentation and statistical analysis were so complicated that I could not decipher it. I simply could not understand the statistical analysis and its description. However, I have extracted the data on MCA, as this was the most relevant information that I could find. I have put this in this table.
|Nasopharynx||289.30 (72.1)||266.7 (76.0)||256.8 (77.9)||256.8 (73.7)|
|Retropalatal||182.5 (86.1)||167.5 (110.9)||149.6 (75.9)||139.9 (83.0)|
|Retroglossal||172.0 (98.3)||174.4 (92.7)||134.3 (66.2)||126.0 (70.0)|
I could not work out if these differences were statistically significant. However, I did wonder if the effect sizes were small.
In their summary, the authors pointed out that there was no change in these measurements over time, there was a slight tendency for extraction cases to have larger measurements at the start and end of treatment and there were no differential effects with extraction.
Their overall conclusions were:
“There was no evidence that extractions in non-growing patients had negative consequences on the airway dimensions.”
What did I think?
Regular readers of this blog know that I do not usually review retrospective studies. However, I make an exception to this practice when the retrospective study provides us with interesting information that may be used to plan a randomised trial. I will also do this when a flawed retrospective study comes to firm conclusions that may add to the controversy. I do this to make sure that the flaws are understood and also to provide information for readers who are not sure how to interpret these papers. This paper falls into this latter group. I shall carefully outline the problems with this study.
Problems with the study design
Firstly, I would like us to think about the ability of the study to answer the question that they post. An excellent place to start would be to consider the ideal study design. This would be a trial in which we could take a group of patients and randomly allocate them to extraction or non-extraction treatment. However, there is some doubt that patients would enrol in such a study. This is because most would express a preference for non-extraction treatment.
An alternative would be a prospective study with well-matched participants with similar levels of crowding and other occlusal features who have treatment with or without extraction. These participants could have standard baseline records taken, including three-dimensional imaging as well as patient-focused assessment to evaluate any symptoms related to possible breathing disorders. These tests could then be repeated after treatment and indeed, even sometime later to assess whether an extraction-based treatment has a meaningful long-term effect.
Another method is a well-done retrospective study that is designed to minimise selection bias. If this is successful, then the groups would have minimal differences at the start of treatment. Unfortunately, in this study, 24 of the extraction group and only 5 of the non-extraction group had crowding greater than 7mm. As a result, the groups were very different at the start of treatment. This is why one group had extractions, and the other did not. I feel that this is a significant flaw in this study. There were also the well-recognised problems of measuring the airway using static measurements, possible errors in patient positioning and changes in airway volume due to breathing.
As a result, I am sorry to suggest that I disagree with the methodology that they used. Unfortunately, in my mind, this casts some doubt on the results of this study.