At last, a trial on aligners v fixed appliances in complex cases!
Most readers of this blog will be familiar with me moaning about the lack of high-quality research into clear aligner treatment. Recently, I posted about several new studies but pointed out that there are problems with the papers. I was, therefore, interested to see this recent paper on the effectiveness of clear aligners and fixed appliances for patients with severe crowding.
I am generally under the impression that clear aligners provide orthodontic treatment for mild to moderate problems. However, there are currently no trials of using aligners to treat more severe malocclusions. As a result, I thought that this paper was very timely and exciting.
A team from Syria did this trial. Cureus Journal published the paper.
The Effectiveness of In-house Clear Aligners and Traditional Fixed Appliances in Achieving Good Occlusion in Complex Orthodontic Cases: A Randomized Control Clinical Trial.
Samer T. Jaber, Mohammad Y. Hajeer, Ahmad S. Burhan
Cureus 14(10): e30147. DOI 10.7759/cureus.30147
Now I wonder if you are thinking, “what is Cureus”? I was unsure, as I had not heard of this journal before I came across this paper. So, I started with Wikipedia, which describes it as follows;
“Cureus, also known as the Cureus Journal of Medical Science, is an open-access general medical journal and is among the growing number of journals using prepublication and post-publication peer review”.
They provide a rapid review and publication process. Furthermore, all the papers are open-access. I have looked very carefully to see if this is a predatory journal. I could not find any evidence that this was the case. Finally, I contacted the corresponding author to ask why this paper was not published in a mainstream orthodontic journal. He told me that they took this decision to enable rapid research publication.
So, let’s have a good critical look at this paper.
What did they ask?
They did the study to find out.
“What is the effectiveness and efficiency of in hour aligners compared to fixed appliances for premolar extraction complex orthodontic cases”?
What did they do?
They did a single centre 2 arm parallel group RCT. They registered the trial on ClinicalTrials.gov. The PICO was
Class I Orthodontic patients with severe dental crowding of more than 6mm of tooth size-arch length discrepancy.
In-house clear aligners
0.22 slot MBT brackets with TPA and lingual arch.
The primary outcomes were the change in the PAR index and Little’s Irregularity Index. The secondary outcome was treatment duration.
A single operator treated all the patients. They defined the endpoint of treatment when the patients had completed all stages of treatment, and two operators were satisfied with the end result of treatment.
They used pre-prepared remote randomisation. The team concealed the allocation by using sealed envelopes. As with most orthodontic trials, they could not blind the patients and operators to the intervention. However, they did data collection and analysis blind. They did a straightforward sample size calculation.
What did they find?
The team enrolled 36 participants. 18 (5 males and 13 females) were randomised to receive clear aligners, and 18 (7 males and 11 females) received fixed appliance treatment. No participants were lost to follow-up. The mean age of the participants was 21 years. There were no differences in Little’s index at the start of treatment.
They presented a large amount of PAR occlusal index data, and I will concentrate on this because it provides more information than Little’s Index.
|Group||PAR||% change PAR|
|Aligner||33.89 (8.01)||5.5 (2.85)||87.08|
|Fixed||30.44 (6.69)||4.04 (2.96)||83.5|
None of these differences were statistically or clinically significant.
I had a good look at the individual PAR score data to look for any differences in the components of the malocclusion. I did not find any clinically significant differences. The overall standard of care, according to the PAR index, was high.
The mean duration of treatment was 23.7 (5.2) months for the aligner group and 26.2(5.27) for the fixed group. This was not statistically different.
Their conclusion was;
“There were no significant differences between the clear aligner and fixed appliances groups for any of the component of the PAR index. In house clear aligners can effectively treat complex cases. There was no difference in treatment duration between the two interventions”.
What did I think?
This was a nicely done and written-up research project. The team followed the classical RCT methodology, and I could not find any significant sources of bias. The only one was possibly the assessment of the end of treatment. This appeared to be subjective, and the operators were not blinded to treatment allocation. However, this is a feature of many other orthodontic trials and reflects the “real world” assessment of the end of treatment.
My only other concern was that this was a single-operator study. This means that the findings may have limited generality. Nevertheless, this study is an excellent first step.
Good points were that the study was well done. The outcome measure (PAR) is well established and has been used in many trials. Notably, the index showed that the start dental malocclusions were relatively severe and that the overall treatment standard was high.
Finally, this study shows that orthodontic treatment with aligners is comparable regarding occlusal index scores with fixed appliances. It is a good and valuable study.
My only slight concern is that this paper would receive greater attention if the authors had published it in a mainstream orthodontic journal. However, I appreciate that they wanted a rapid review and publication of their work.
Emeritus Professor of Orthodontics, University of Manchester, UK.
Have your say!
This is encouraging but, I’d love to see the final panoramic X-ray for all participants. Root parallelism? Call me cautiously skeptical.
As in traditional treatments with fixed appliance, human factor (clinician expertise – experience with aligners), mechanics (how did they closed spaces, how much refinement) and materials (in house aligners materials, change frecuency etc) are in the way to consider to extrapolate this study to every day consideration in our practice
I studied at this prestigious university..and I fully know the passion of my professors for scientific research..we thank the efforts of Professor Muhammad Hujair and Professor Ahmed Burhan..we always wish you success…and indeed this study is very interesting, especially with the claim of manufacturers of transparent aligners.
An interesting study, which focused on clinical outcomes of the two appliances through established indices of PAR and Little’s Index. The authors have gone to significant lengths to perform a robust study to answer the research question in my opinion, using class 1 crowded cases with extractions of 4s.
There were a few points to note in the methodology. The aligners used were designed and planned ‘in house’, using a planned method described by the authors as ‘SAMMER’ essentially individual tooth retraction, including incisors. This is a novel approach to aligner tooth movement for extraction cases, as far as I am aware. Previous aligner research by Fan-Fan Dai 2018/2021 used invisalign protocols of staged retraction of canines followed by 2-2 retraction. It was unclear from the methods section how the fixed appliance case was managed for space closure. This may limit the findings to ‘conventionally planned’ aligner extraction cases. There was ‘0’ loss to follow up in both groups, which is exceptional for a 2.5 year study, however if I treated 36 patients i would expect at least one or two drop outs, and questions the generalisability of the study.
Attachment protocols were described, however the size, shape and indeed material were not mentioned. There is controversy in the effectiveness of attachments and design variation Costa 2020 and Momtaz 2016, and further information on the use of attachments would have been useful.
There as no mention of the use of elastics in the study, a key ‘protocol’ for invisalign space closure, and in my own use of aligners, in Best’s 2017 study he showed orthodontists use elastics 93% of the time with aligners.
Equivalence was demonstrated in the outcomes investigated, however from previous research by Fan-Fan Dai 2018 and 2021 on extraction and aligner use, one of the key findings was greater than expected anchorage loss, tipping of molars and the lack of apical movement of anterior teeth. Unfortunately PAR and LI do not consider these outcomes.
I congratulate the authors on the publishing this study, and it will work its way into my notes on aligners. Further outcomes to consider in the question of equivalence would be tipping Vs bodily movement into both extraction site as well as anterior and posterior movements. Anchorage loss between the appliances, inclination of the teeth at the end of treatment, Compliance from patients in wear over extended periods of treatment. One thing mentioned in Nada Haouile and Neal Kravitz’s repeat study on aligner predictability was that of users ‘overengineering’ planned movements, I wonder if that was used in this case. Refinement of case and if the case needed to be ‘replanned’ via the simulation outcome.
Interesting article, thanks Kevin. Also great to read all the excellent comments.
I just would like to point out, that, from a clinical point of view, I am surprised about the length of the fixed treatment. As far as I understand the cases, they seem mainly reciprocal anchorage cases, designed to solve the crowding primarily.
Now assuming an average tooth movement of 0.8mm/month I would expect a fixed treatment time well below 18 months without jeopardizing the outcome and the PAR improvement.
Probably it is my personal bias when I read aligner studies like this.
But I wonder about the treatment length.
I would absolutely disagree with the conclusions re: Treatment Outcome. The PAR index is a blunt tool when used to compare occlusal outcomes between two modalities, and will tend to show improvement across the board with orthodontic Tx. Thus, the use of the PAR index, instead of the ABO -OGS, invalidates the comparison between the fixed and removable appliances.
If the goal is to compare finished cases, it is imperative to assess 2nd order root angulation, marginal ridge leveling, BL inclination, occlusal contacts and relationships etc, none of which the PAR does (AFAIK).
Additionally, this flies in the face of two studies, the first of which showed the poor performance of aligners, as it pertains to first molar control and incisor retraction. The second study noted significant mesial tipping, buccal inclination, mesial displacement, and intrusion of the first molars, as well as distal tipping, lingual inclination, insufficient retraction, and intrusion of the canines and central incisors.
It is very likely that the results would have been very different, had the more valid ABO-OGS index been used.
1. Dai FF, Xu TM, Shu G. Comparison of achieved and predicted tooth movement of maxillary first molars and central incisors: First premolar extraction treatment with Invisalign. Angle Orthod. 2019 Sep;89(5):679-687.
2. Dai FF, Xu TM, Shu G. Comparison of achieved and predicted crown movement in adults after 4 first premolar extraction treatment with Invisalign. Am J Orthod Dentofacial Orthop. 2021 Dec;160(6):805-813
Yes Sir, exactly!
A very poor and weak article from a multitude of perspectives. It proves little and lends nothing to the evidence base. It’s hardly surprising there was no publication in a recognized journal. I would’ve expected it to be rejected.
Thank you for the article, and for the many thoughtful comments. The SAMMER protocol is quite interesting and I wonder how much influence it’s use had. A separate study comparing clear aligner techniques would be very interesting, even a taxonomy to help define or classify the techniques.
Reading the above raises many areas of contention:
– what are we measuring? This paper clearly states “a good occlusion”. In that case, their rulers were appropriate (not Little’s – shown time and again in the literature – but PAR). There are comments calling for radiographic and cephalometric post-treatment evaluation, all valid if we are extending our evaluation to the much broader and more difficult to assess “treatment outcome”. In that case we may argue for quality of life, patient satisfaction, aesthetic evaluation to compulsorily be included.
– There are still disagreements over the ruler to be utilized when assessing occlusion and alignment! PAR is truly validated thanks to Kevin, and out of the bunch, one of the best. ABO-OGS was introduced prior to any validation studies, apart from examiners of the ABO exam. It has a particularly low reliability and calibration has been suggested for any standardized use of the index. Clinical Use of the ABO-Scoring Index: Reliability and Subtraction Frequency William S. Lieber, DMD, MSD; Sean K. Carlson, DMD, MS; Sheldon Baumrind, DDS, MS; Donald R. Poulton, DDS Angle Orthod (2003) 73 (5): 556–564.
Subsequently, after its introduction it went through wider validation testing and appears to be by diffusion an accepted front runner for measuring occlusion and alignment. Perhaps the (calibrated) researchers may be able to use ABO on this data to satisfy some agitation? When we cant agree on what we need to measure, and then the ruler we need to apply, assuming it is actually valid and reliable, we are not capable of measuring anything in particular.
– Ernesto’ and Dr Ahmed’s comments above regarding specific protocols used are critical to any comparison of treatment modality
– Finally, the finish line – as pointed out by Kevin. If the examiner is not blinded to the fact that a study on outcome is being undertaken, we may assume a bias (in this case for both modalities) towards lengthier treatment duration in order to finish to a higher standard (may help to explain comment from Martin Baxmann). Real life fixed appliance practice has a subjective, moving finish line, whereas some researchers have inexplicably concluded that with computer programmed aligners, the finish line is the last aligner in the series. Neither of these scenarios are particularly helpful when measuring occlusal outcome – are we testing what is achievable, or regularly achieved? Computer programmed aligners suffer the indignation of having a visible and measurable end occlusal goal (not the same thing as end point) – and that “end goal ” of fixed appliance patients is sought after in a reactive manner , a figment created by clinicians that likely changes throughout treatment, and we may never know if its achieved. We would need to decide how to rationalize these differences prior to any attempt at measurement of treatment “outcome” versus “predictability”. Do we allow patients treated using CPA the same privilege of being treated until the clinician deems a good outcome is reached, or do we allow the computer default to set the finish line?
*VV Lectures on behalf of Align Technology
I agree with Vishnu’s analysis and Mark’s comments. Why this journal? Review was done in 24 hours and published four days later. I do not understand why authors did not want rigorous review, it would have only improved the paper.
I think one needs to define the outcome measure of interest appropriately in order to avoid confusion. Success is a rather broad, nebulous and non-specific term, when used to compare two very different modalities. Most, if not all, forms of orthodontic treatment provide some degree of improvement.
When evaluating finished results in extraction cases, parameters such as alignment, incisor torque, root positioning, arch form, occlusal contacts, 2nd molar positioning etc. have important ramifications as it pertains to esthetics, function and post-treatment stability. So, while the PAR may be a valid tool to measure within group changes (in terms of degree of success), it falls far short when comparing the performance of two different modalities.
That calls for a much more stringent analysis, like the ABO-OGS would allow for. Thus, it is very perplexing (or not, depending on one’s degree of cynicism) to see the less rigorous and robust PAR being boosted championed over the ABO-OGS.
Red herrings like the quality of life/patient satisfaction/aesthetic evaluation are simply not relevant when it comes to measuring the biomechanical ability of a modality to deliver the outcome. And the data on that, (Align KOL protestations notwithstanding) especially as it pertains to extraction cases, could not be more clear. Despite a multitude of attachments and other creative hijinks, aligners simply do not generate M:F ratios large enough to predictably allow translation, root correction, arch form and incisor control control at the level of precision inherent with fixed appliances.
Thanks Dr Vishnu for your comments, I am assuming they relate to mine above. You have touched on my 2 great passions, Occlusal Indices and Treatment Outcome Assessment and indirectly, Invisalign! *
Agree totally that we need to define the “measure of interest” appropriately; but the ruler still needs to be valid for that “outcome measure of interest”. The problem is that unless the index you have chosen is validated specifically for your “outcome measure of interest”, then it is of no scientific use. As you know, an index is only valid and reliable when used to measure what it was designed and validated to measure. The paper I quoted above suggests that ABO – OGS may only be reliable if examiners are calibrated. I have nothing against this index that may indeed be the best we have for measuring occlusion and alignment, and as you read, I suggested that the authors may choose to apply it to their data set – albeit it retrospectively. But to suggest that it holds more “validity” than the PAR – assuming both are used correctly and applied to what they were designed to measure – is untrue. Kevin, you may want to add a comment on validation process and misuse of indices?
In reference to the “red herring” comment, I respectfully bring to your attention the title of this research project ” The Effectiveness of In-house Clear Aligners and Traditional Fixed Appliances in Achieving Good Occlusion in Complex Orthodontic Cases: A Randomized Control Clinical Trial. If the purpose is indeed to assess “good occlusion” (for want of a better term), then the PAR or if you prefer, ABO-OGS may reasonably be applied.
When assessing the broader question : What is the effectiveness and efficiency of in house aligners compared to fixed appliances for premolar extraction complex orthodontic cases”?; this is a totally different assessment including both variables of treatment outcome as well as measuring efficacy and efficiency! I spoke to this (and this blog at the EAS 2021 meeting). When comprehensively measuring treatment outcome, I absolutely agree that additional significant variables as you mention- such as root parallelism, aesthetic evaluation, as well as periodontal indices, root resorption, patient satisfaction to name but a few should be assessed. These are not “herrings” of any color, but significant variables to be considered in treatment outcome assessment. The existence of a valid, comprehensive ruler with high inter and intra-examiner reliability is just one of the difficulties here, but that doesn’t mean as individuals we should not continually assess outcome to our best ability, albeit subjectively, if no objective yardstick is applicable.
Regarding Treatment efficacy/ predictability/ accuracy – similar but not interchangeable terminology and terribly abused in the literature- I am not aware of any study that can / has measured this with fixed appliances, so that a comparison with CPA is a lofty ambition. Unfortunately, for reasons I stated above, “predictability” or “accuracy” of traditional fixed appliance systems cannot be measured (unless digitally programmed). Even if our treatment goal could be pre-defined, maintained in 3D, then post-treatment models superimposed (on …rugae?) …do we count re-bonding, wire-bending, broken brackets, use of palatal arches, TADS , elastics , reverse curve wires, over-expanded archwires as failures / lack of efficacy or predictability of fixed appliance mechanics capabilities? May I only expand my archwire 2mm if I want 2mm of arch expansion, am I allowed to over-expand my wire, because I “know” that fixed appliances are neither 100% accurate in expressing “programmed” movement (or bracket prescription); nor are they predictable because we have a moving end point (that exists as a figment and that none of us even agree upon) , not to mention crown anatomy and oral function that throws out any possibility of predicting the exact expression of any particular fixed appliance system ; and like all mechanical systems, especially those used in vivo , utilized by clinicians of varying experience, are highly variable in effect?
Force systems, however, as we know, may be measured, although also a difficult if not impossible task with any complex appliance system attached to up to 32 moving parts, particularly in vivo. So, in reference to your statement “Despite a multitude of attachments and other creative hijinks, aligners simply do not generate M:F ratios large enough to predictably allow translation, root correction, arch form and incisor control at the level of precision inherent with fixed appliances.”, You are so correct- that there is/ are no paper/s that I may quote you, using (either fixed appliances or) CPA that will ever define the exact force system and thus the capacity of some CPA systems to distalize teeth, close extraction spaces “bodily” ( fixed appliances do not- its via series of tips that may result in parallel roots). There are of course many thousands of case reports of patients treated with the Invisalign system mechanics (sorry to mention “it” but as far as I know, only CPA with a significant body of “evidence”) and extraction of teeth, with corresponding radiographs pre and post treatment. The capacity for Invisalign aligners to create appropriate moment of force to close an extraction space, when managed (or sometimes not) via attachments and staging, is impossible to deny. Just search PubMed and you will find systematic reviews with meta analysis – (for what they are worth in the world of orthodontic literature) of papers measuring the range of forces capable with certain aligner materials, along with hundreds of case reports – lower quality evidence, still evidence; showing “parallel roots” resulting from biomechanical systems utilizing attachments and aligners. Show me 1 CPA case with tipped roots in an extraction space, I will show you 20 treated using fixed appliances- those patients that I re-treat every day using Invisalign, and gasp when I view the pre-treatment OPG. If you provide the references for the paper/ s that define the moment and force system in vivo , created by your particular individual bracket and wire brand of choice, or any for that matter, independently tested (because I assume that you will not accept data from Align Technology or “KOL”), we may simulate that study protocol using CPA. If only attachments were “creative hijinks”– we could treat patients without them! Unfortunately, as Professor Boyd and I pointed out in 2001, Australian Orthodontic Journal, we need them to extrude teeth. Since then we have learned that most movement except uncontrolled tipping will benefit from attachments, and not only those teeth but neighboring teeth (Melbourne University Thesis , Three Dimensional Radiographic Evaluation of the effect on Root Control of Increasing the Velocity of programmed Buccal Expansion with Invisalign®. Sequential aligners in Teenage Subjects 2016)
Finally, I do not agree that there is a (high), or any “level of precision inherent with fixed appliances”. We have all witnessed atrocious treatment outcomes in patients treated with fixed appliance systems, in all types of cases. If precision were “inherent”, there would be no debate about who in our profession should have the distinct privilege of treating patients with these appliances, there would be no need for outcome assessment, nor Board Exams ,or rebonding, wire bending, finishing positioners or whatever your favorite finishing methodology happens to be. There is nothing “inherently” magical about the ability of fixed appliances, (especially considering their 125 year evolution in their “modern era”,) in the experienced clinicians hands to close an extraction space via a series of small tipping movements. Its physics and experience. You may choose to disagree that some plastics and force-driven computer programmed aligner systems are unable to create ,via combination of geometry of attachments and material with appropriate modulus of elasticity, moments and force systems that are physiologically and mechanically compatible with expressing “bodily” tooth movement. You may choose to deny this, as you are correct that there is no single paper to your satisfaction proving this, and I would respectfully return the favour, with no corresponding existing “proof” for the multitude of individual fixed appliance prescriptions – in itself a statement on “inherent precision”.
The great news is that these digital systems have re-invigorated the (adult) orthodontic market (as Ricketts pointed out to me), spawned a flood of “copies”, some of which may be capable of producing satisfactory treatment outcomes in the right hands and if not, they are keeping others on their toes. We may not be comfortable with the corporations, the marketing or the palpable change, this I understand. The potential for quality of patient care to drop is of great concern if abused, as with any appliance when clinicians cease being clinicians, continue to delegate and to abrogate treatment responsibility to “laboratories” and computer defaults. But these digital systems have also escorted a largely stagnant and myopic specialty into the digital era with integration of scanners and 3D diagnostic soft-ware (that never existed prior to Align Technology) into teaching, traditional fixed appliance therapy, they have the potential to minimize iatrogenic damage and to increase patient quality of life during treatment- that “herring” – they have massive potential for AI learning and thus exponential improvement; they may actually be objectively measured in terms of degree of expression of movement ….and best of all: …those clinicians not wanting to learn how to use them or their associated digital innovations simply don’t have to!
*VV Lectures on Behalf of Align Technology
Thanks so much Dr. VV for your insight for evaluation and comments. I agree completely with all the points you made.
I would like to make only one point about the study from Syria. All of the aligners were started from silicone impressions, then poured into models which were trimed and then scanned separately, followed by the use of in-house software to set up the teeth, plan movement sequence, add appropriate attachments, and doing all of these steps by hand. This is followed by the creation of a series of perhaps as many as 30 to 50 models per arch for a total of as many as 100 appliances with any necessary refinements). Then the actual appliances made, trimmed and polished. All of this lab work is done in house.
I couldn’t help but think that if I told my residents that as a cost- saving measure, we would no longer be using the Invisalign company to make our aligners and that we would not use scanners but revert back to silicone impressions, I might find that my car would have the air removed from its tires in protest!