Kevin, this is what you should know about Clear Aligner Treatment?
A few weeks ago, I posted my thoughts on clear aligner treatment and got many valuable comments. I was particularly struck by this detailed discussion by Farooq Ahmed. He makes many good points and does not necessarily agree with me!
Farooq is a UK-based orthodontic specialist and consultant. He works in a university teaching hospital and a private practice in London. He was fortunate to have undertaken his orthodontic training /residency in Manchester (misfortune has led him elsewhere). He also runs the very successful Orthodontics in Summary Podcast. This features lecture summaries, guest interviews as well as a clinical blog, which is excellent and worth looking at. www.orthoinsummary.com
This is what Farooq had to say about my post.
Introduction
For disclosure, I lead the education of aligners at the university teaching hospital, Guy’s and St Thomas’ NHS Foundation Trust. I am not a KOL and have no financial interest in aligner companies, products, or services. Furthermore, I enjoy the blog and consider it a staple of anyone’s orthodontic diet. I have learned and clinically used aligners for 7 years (I am a top 99% provider).
I will reply to the key topics mentioned in the blog. This is not for the sake of disagreement but to highlight the lack of reliable and reproducible research that has left uncertainty on the matter. This enforces a decision based on the clinician’s opinion, more so than on their powers of deduction from the research. I will direct readers to the phrase I believe Carl Sagan stated, “absence of evidence is not evidence of absence” and dismissing an item lacking evidence as “impatience with ambiguity”.
Accuracy of Aligners
Kevin’s statement
“The overall accuracy of treatment relative to predictions is only 50%”
Aligner research assessing accuracy varies, from 50% to 86%. There is a significant range in accuracy. I feel that the main challenge is with the method of assessment in research. These studies and most accuracy studies, base their evaluation on the ‘predicted Vs achieved’ method. This is a superimposition of the outcome simulator with the actual tooth position at the end of the series of aligners. The issue with this method is that the outcome simulation is the place where the aligner should move the teeth too. As opposed, to what the outcome simulator actually represents, which is a visual representation of the force delivery to the teeth.
The reason why (I think) this method is prevalent in aligner studies is the ease of assessment. Nevertheless, there is still value to this research as it aids users in planning overcorrection. For example, Al-Bahaa’s 2021 study showed 50% intrusion accuracy, and I now plan 200% correction. However, a more meaningful assessment would be ABO / PAR. Unfortunately, there are few studies with significantly varying outcomes. In Ke’s 2019 systematic review, “no statistically significant difference in CAT and fixed appliances with the ABO and PAR”.
Nevertheless, Yassir in his 2022 review, stated that although no difference was found in Ke’s paper, the non- statistically significant difference of 8 points on the ABO scale may be of clinical significance. We must bear in mind we do not assess directly bonded fixed appliances of their accuracy in this ‘prediction Vs achieved’ method. Consequently, our evaluation of the effectiveness of appliances has been based on occlusal outcomes (mainly), which appears unclear in terms of CAT.
Class 2 cases
Kevin’s statement “We don’t know how effective aligners are at correcting Class II problems”
Indeed we don’t know how effective they are. This is a sore topic as Professor Martyn Cobourne and I applied for funding to investigate this very question 2 years ago. After some initial enthusiasm and copious paperwork, Align Technology did not approve the funding. I, therefore, relent to the lowest form of research but significant component of evidence-based practice – clinical experience. Having treated some patients with Invisalign’s Mandibular advancement appliance, I had a (small) number of successes (enough to poke Martyn Cobourne into writing a research protocol). As a result, I continue to offer the appliance as an alternative to a removable functional appliance (mainly the Twinblock where I have had larger numbers of successes). This is a decision unsupported by research, in both directions.
CAT treatment duration.
Kevin’s statement
Eric Lin’s study found a 4.8-month difference in favour of fixed appliances. However, the fixed appliance group took, on average, 1.3 years for non-extraction mild crowding (less than 4mm). This is a long time for what appears to be a straightforward clinical case, and questions the reproducibility of the study.
These findings were at odds with other research; Ke’s systematic review found that mild-moderate cases when treated with Clear Aligner Treatment, were 6 months shorter in duration. Other treatment components should be considered, such as the number of appointments, breakages, and chairside time. Zheng’s 2017 systematic review showed 4 fewer appointments with Clear Aligner Treatment, 1 less emergency, and less chairside time by 7 minutes. However, this systematic review was from 2017 (several decades in aligner research years) and was based on a less reliable research design of cross-sectional studies. Thus, leaving the question of duration uncertain in terms of time and appointments.
Low predictability of movements apart from mild horizontal movements.
Kevin’s statement
It is interesting how a ‘predictability’ statistic is interpreted. Clinically it is as important (or even more important) to have the standard deviation of the predictability present. Unfortunately, this is not routine in studies. For example, if the intrusion of an incisor was 50% predictable with a standard deviation of the mean of 2%, I would be able to use this information to ‘predictably’ achieve 48-52% of my planned movements or overcorrect with a degree of predictability.
Root resorption.
Kevin’s statement
“We don’t know if aligners result in less root resorption than fixed braces”
I consider this topic less critical as I would expect some root resorption for any orthodontic appliance due to the physiology of orthodontic tooth movement. Nevertheless, a direct comparison is challenging to perform as an accurate comparison of Clear Aligner Treatment, and fixed appliances needs the roots of teeth to move the same distance in both appliances. Unfortunately, achieving such accuracy (I feel) is currently not possible to plan with either appliance, and this topic will remain uncertain.
What do I (Farooq) know about aligners?
There are known and unknown unknowns within the field of CAT. At this stage, decision-making is based largely on clinical experience and theoretical understanding. CAT is an evolving field, however, research has been less robust, there are three main reasons for this in my opinion:
- Digital technology has simultaneously entered orthodontics, resulting in significant variation in planning protocols. As clinicians trial, adapt and refine the use of CAT to the plethora of new processes available, the field remains in a constant state of change. Technology has also proved a challenge in research, as more complex analysis can direct answers away from clinical questions.
- Adults seeking orthodontics have gravitated to CAT, a different cohort to ‘conventional’ orthodontic patients. This group varies significantly in their treatment goals. They are more likely to require a ‘realistic’ orthodontic plan rather than an ‘idealistic’ plan. As result, this adds to variation in planning.
- Industry engagement, Kevin, I agree with you entirely on this point. Unfortunately, there has been little collaboration between industry and independent university institutions. I do view our colleagues in pharmacy with a degree of envy as innovation and research seem to be a collaborative endeavor between industry and universities in their field.
I do wish to congratulate the authors who have contributed to CAT research. Although my comments may appear critical, their efforts and work are adding to a growing body of knowledge. Research questions are becoming increasingly refined to produce more meaningful answers. I am optimistic that the answers to your questions Kevin will materialise as the aligner years pass.
Emeritus Professor of Orthodontics, University of Manchester, UK.
“Aligner research assessing accuracy varies, from 50% to86%. There is a significant range in accuracy. I feel that the main challenge is with the method of assessment in research. These studies and most accuracy studies, base their evaluation on the ‘predicted Vs achieved’ method.”
This posters premise is Double Speak.
Accuracy is not defined as a “range” and not a relative function. I have read Aligner results are more realistically in the 35%-50% level.
When deficiencies in Plastik Aligner tretments can be clearly defined and reported, such as but not limited to.
Inability to rotate cuspids, inability to open a deep bite, inability to close extraction spaces with root translation.
Of course, there are also the predictable Unintended reciprocal tooth movements that no one discusses, such as routine posterior open bites.
Blaming the methodology of the research is an attempt to mask adverse results.
I believe that “predicted results vs acheived results” is the definition of “evidenced based” treatments?
Plastik ortho has been a marketing success, but a failed orthodontic method mis represented by many as “equal” to or “better” than traditional orthodontics.
The longer the explanation, the shorter the Truth.
Dr. O’Brien, thank you for sponsoring an unbiased irregular platform for orthodontic discussions.
Thank you Tom for the comment. The findings of overall accuracy of 50% (Haouili 2020) and 86% (Bilello 2022 study) were taken from papers Neal Kravitz’s study as well as the latest paper on the topic. The statistics were used to represent the variability of predictable values from research. In both of the papers above the lowest predictable values given were 28% and 70%, however taking the ‘worst’ or ‘best’ value would not be representative of the appliance, hence the use of overall findings from these studies was used.
I agree there are deficiencies in the use of aligners, however less predictable movements are not the same as an inability to move, research shows some cuspid correction, overbite correction and root movement, the challenge with aligners is the low predictability or quantitive change of these movements. Indeed predictability values are evidence, but is it informative evidence of clinical outcomes, i would suggest they are not, an aligner outcome simulator is not a visual treatment objective.
The comments of methodology are to direct readers to more meaningful answers on aligner outcomes, which are uncertain currently in research, and to move away from predictability values, which are surrogate endpoints , clinical outcomes are more meaningful.
Thank you again for your comments
Great response and truthful. This comes from an orthodontist who loves braces but I love orthodontics more.
I don’t agree when you said about predictability.
Of course aligners should be evaluated using the method predicition Vs achieved. It is the way that we plan the case, the way that patients agree with the treatment outcome, and a powerful tool for marketing. However, It shows the deficiency in the treatment method.
For me, this is the real problem with aligners.
No treatment ever reaches the virtual prediction, and the company probably is aware of this. A complex case almost uses the 5 year warranty in the Comprehensive treatment.
To reach an ABO/ PAR outcome, probably more treatment time for aligners should be expected (more than predicted and more than clinicians and patients want).
Even though, new studies using this kind os assessment should be developed to clarify this question.
Hi Liana, thank you for your comment. Predictablity outcomes only hold value if the simulation was a visual treatment outcome, which I would argue it is it. The outcome simulator /Clincheck is incorrect in terminology, and should be a ‘visual representation of the force syatem’. Deep bite cases appear as AOBs in online set ups, however I can’t imagine anyone is aiming to achieve this outcome, so would be the use if assessing the predictability in this case?
I am not saying aligners are the panacea of orthodontics, there are limitations in movement and also in research, the post was to highlight the reasons for some of the latter.
Interesting discussion.
As a researcher, albeit honorary, with 3 recent publications, another 3 already accepted for this year plus another in review currently, just in AJODO, plus more in the Australian Ortho Journal I wish to explain the predicted/achieved rationale. I/we want to know how much the digital plan from a company produces the outcome so we can quantify any shortfall to inform our profession re over corrections – just as we know when to flip brackets or the torque loss from certain arch wires. Getting the numbers informs us of how to custom- make the “custom-made” appliance. We have access to a 14,000 patient database, so our numbers are a little different from Billlelo (10 patients) or Haouli (38 patients), from which those studies drew conclusions about every kind of movement for all kinds of teeth. One of the reasons aligners do not work is because we do not know how to make them work, we just believe the corporate Koolaid. Indeed, we also have increasingly identified movements that occur in the opposite direction to those planned, especially when moving roots through bone. This may be seen in 15-20 % of teeth performing those movements. I would argue we do need predicted vs achieved for the above reasons.
Hi Tony. Thank you for your comments and great work on the papers.
There are two questions here with confuses the topic, the first is the effectiveness of the appliance and the second is understanding of how the appliance works.
I entirely agree predicted vs achieved is one of best ways to understand how the appliance works, in my post I referenced Al Bahaa’s study for overbite correction and how useful this information is.
However the effectiveness of the appliance from predicted Vs achieved is not as useful as occlusal outcomes of treatment. Assessing the outcomes of a case ahould be at the end of treatment, not at the end of the 1st stage of treatment, which is what predicted Vs achieved literature tends to be. But as I say this is useful for understanding of the appliance but not for effectiveness of outcomes in treating patients