Braces are faster than Invisalign? An RCT
We need more research into clear aligner treatment. Some would suggest that this research is overdue, as this form of treatment is still to be scientifically evaluated. I was, therefore, pleased to see this new trial that compared the effectiveness of Invisalign and fixed braces.
Most research into clear aligners has been observational. There are very few clinical trials that have looked at this form of care. When we look back, it appears that clear aligners were widely adopted by dentists and orthodontists with little evidence of their effectiveness. The main drivers for this uptake were advertising, case reports, and key opinion leaders. In general terms, aligners are suitable for treating mild to moderate malocclusions. Although, skilled operators are using them for more complex care. This new RCT provides us with very useful information.
A team from Texas wrote this paper. The Angle Orthodontist published it.
Eric Lin et al. Angle Orthodontist: Online: DOI: 10.2319/032921-246.1
What did they ask?
They did the study to:
“Compare patients treated with traditional braces to patients treated with Invisalign”.
What did they do?
They did a randomised trial. The PICO was
Participants: Orthodontic patients with Class I molar and canine relationship, non-extraction treatment, mandibular crowding of 4 mm or less with no missing teeth.
Intervention: Invisalign treatment with SmartForce attachments.
Comparator: Fixed appliance treatment.
Outcome: ABO-Objective Grading System scores. Secondary outcome: Duration of treatment.
The authors carried out a sample size calculation that showed they needed to enroll 31 patients per group for sufficient power. They used pre-prepared randomisation. However, I could not find any information on allocation concealment.
One operator treated all the patients using standard protocols. The participants changed their aligners every two weeks. They collected data at the start and end of treatment and after 6 months of retention.
They recorded data from the cephalograms blind. Unfortunately, it was not clear whether they blinded the study model data collection.
What did they find?
They enrolled 40 patients into each intervention group. At the end of treatment, they analysed data for 31 braces and 32 Invisalign patients. It was relevant to see that 1 brace and 4 Invisalign patients did not complete their treatment. Other patients were excluded because of the absence of records. This is important, and I shall return to this later.
At the start of treatment, the mean age of the Invisalign patients was 26.7 years, and the braces patients were 25.9 years old.
The median DI scores for the Invisalign and braces groups at the start of treatment were 4.5 and 7.0, respectively. This was statistically significant, but this was a small difference.
At the end of treatment, there were no differences between the groups for the overall OGS scores. The median scores were 12.0 and 17.00 for the braces and Invisalign groups. This was not statistically significant.
Similarly, after 6 months of retention, there were no differences between the groups.
When they looked at treatment duration. This was 1.7 years for the Invisalign and 1.3 years for the braces group. So again, this was clinically and statistically significant.
Their overall conclusions were:
“There were no differences in the occlusal results of treatment between the two treatments”.
“Invisalign patients took 4.8 months longer to complete treatment than the braces patients”.
What did I think?
I felt that this was an ambitious trial to carry out. The good points were that the randomisation and sample size calculation was clear and logical. Importantly, all the patients were treated by the same experienced operator. This removed any operator effect. However, it does reduce the generality of the findings.
I was concerned with the lack of information on the concealment of the treatment allocation and the apparent lack of blinding of occlusal index scoring. Unfortunately, this means that the trial is at high risk of bias. We need to consider whether this had an effect. In this respect, I thought it was interesting to see that there were differences in the DI at the start of treatment. Again, this may reflect possible bias.
Finally, we need to consider that 4 of the Invisalign and 1 of the brace patients did not complete their treatment. This is important because in a trial we are measuring the effect of an intervention. This means that the outcome should be analysed regardless of whether the treatment is complete. This is called an Intention to Treat analysis. I wonder if this had been done whether this would have influenced the results?
Final comment
The most important take-home message is that the standard of treatment for mild malocclusions with braces and Invisalign is high. Nevertheless, the duration of treatment with Invisalign was much longer than with braces. Perhaps, this is the price to pay by our patients for having relatively invisible appliances? However, we do need to take the high risk of bias into consideration when we interpret these findings.
Emeritus Professor of Orthodontics, University of Manchester, UK.
I assume the Invisalign cases were 2/52 between changing aligners.
Many cases are now changing aligners at shorter intervals (assuming compliance and tracking following the Clincheck), thus the longer treatment time may not be generally true.
If aligners are changed weekly and they do not track in some patients, could this increase the average Invisalign treatment time? Or perhaps have them discontinue treatment, as there were already four times as many Invisalign patients who did not complete treatment.
Would be interesting to see a comparison with more difficult cases.
Standard protocol: aligner change every two weeks. It is now possible for more than half of patients to change aligners every week. This cuts treatment time in half, so that the outcome of this study is pretty meaningless.
Good point!
Excellent article, thank you for sharing the paper and your thoughts. I would be curious to know what the impact might be of compliance monitoring systems, both as a method of compliance, and the contribution of (lack of) compliance on treatment time. I would also be very interested in your thoughts about the role of the operator and specifically in the nature of the treatment setup, as one of the variables could be the nature of staging and anchoring in the treatment setup. Clear aligner systems have wide variability in several important factors – compliance, fit of aligners, and analysis of the treatment setup. The wear schedule mentioned by others (2 weeks vs 1 week) is not an independent factor, but rather closely related to the setup (are all teeth being moved simultaneously in smaller increments that do not compress the periodontal ligament). The discrete movements of clear aligners vs the continuous force of braces is very important to study, and I hope to see more studies that explore the differences of each system.
Yes a compliance monitor may be helpful., However, other studies have shown that they do not appear to influence our patients. But at least we would have information on whether the patients wore their aligners for the prescribed times.
Thank you for sharing Dr. O’Brian. Interesting RCT which raises some questioning:
1. In an earlier article published in 2014 in the Angle Orthodontist (Comparative time efficiency of aligner therapy and
conventional edgewise braces), author Bushang et al. came with the following results:
“Compared to ALT, CEB required significantly (P , .01) more visits (approximately 4.0), a
longer treatment duration (5.5 months), more emergency visits (1.0), greater emergency chair
time (7.0 minutes), and greater total chair time (93.4 minutes). However, ALT showed significantly
(P , .01) greater total material costs and required significantly more total doctor time than CEB
(P , .01).”. 150 patients were treated with ALT and 150 with CEB. All by the same orthodontist. With similar cases used in both articles (class I patients with less than 4-5mm crowding), I am wondering how in 8 years, they could come to a result where it took more time to treat the same types of patients with clear aligners?! I surely know that I personally have a better understanding and better results with clear aligners than I had 8 years ago.
2. In an RCT published in the Angle Orthodontist in 2021 “Effect of clear aligner wear protocol on the efficacy of tooth movement”, Al-Nadawi et al. concluded: “Achieving a clinically similar accuracy between the 7-day protocol and 14-day protocol in half the treatment time suggests a 7-day protocol as an acceptable treatment protocol.”. I wonder why in this article from Lin et al. they did not use a 7 days protocol? What were the reason to stick with the 14 days protocol for all clear aligner patients?
I have been using clear aligners since 2004. I started with the 14 days protocol. That was suggested by the aligner company I was using. By the way, it would be interesting to explain one day how they came up with that protocol.
The other thing we could question: what was the velocity of movement in each aligner used? Because we can play with the wear time (5, 7, 10, 14 days) but we can also adjust the velocity of movement programmed in each aligner.
It’s always easy to play with numbers and make them say what we want. I am surprised, Dr O’Brian, that you were not more critical of this RCT. But yet, I am not an orthodontist. Maybe there are things that slips my understanding.
Anyway, I have to thank you again Dr. O’Brian because it is by following you and reading your blog that I can now take the time trying to understand the articles and read more than the abstract. And question them.
We all have access to the same scientific literature. It is sometimes the interpretation that we make of them that can be different. And that will distinguish experts in the same field.
I will finish with some words of a very respected microbiologist from Montreal, Dr. Jean Barbeau (freely translated in english by myself)
“Everyone wants to be right. The pitfall to avoid in science is not wanting to be wrong. And for that we need humility. For all.”
Thanks for your comments, I thought that they were very interesting and relevant. I hope that I was sufficiently critical of this study and I did draw attention to some relatively severe problems.When I write a blog post, I intend to highlight the “good and the bad” aspects of a study with the aim of enabling any readers to come to their own conclusions about the paper.
I have looked at the two studies that you mentioned. The first was a retrospective study. As a result, it would be subject to a degree of selection bias. This is why we can be less certain about the results. There is also a tendency for retrospective studies to over estimate the effect of a treatment under investigation. I think that this explains the discrepancy.
As regards the second paper, I am not sure why they did not use the 7 day protocol. I am not sure if the authors would be aware of the second paper when they started their study. Unfortunately, they did not include information on the start and finish dates of their study and this information would have been useful. This is one of the CONSORT recommendations on writing up an RCT. While the Angle does publish a number of trials, they are not very strict on applying CONSORT. I hope that I have answered your questions and thanks for your nice comments.
Dear Dr. O’Brien (got it this time!),
First, let me apologize for misspelling your name in my comment. I appreciate the time you took to answer my questions. I really learn a lot following your blog (recommended to me by a good friend, Dr. Gerry Samson). It made me become more critical or careful when reading articles or when presented with “facts” from other dentists/orthodontists. It also gave me a better understanding of the difference between clinical opinions and scientific facts. Not that clinical opinions are not good, but they are what they are and they should be presented as well. I always enjoy reading your posts (and the ones from your collaborators as well): I constantly learn something and they always make me think.
Hello Stephane – disclosure, I speak for Align Technology for the last 23 years, never held shares unfortunately!
I can explain how Align Technology “the company” – the only computer programmed aligner manufacturer in 2004 – came up with the 14 day protocol as it was Prof Robert Boyd, myself and internal orthodontist Ross Miller working on the feasibility of such an appliance, from the outset in 1998. We started with conventional orthodontic knowledge concerning rate of orthodontic tooth movement. At that time, and still today as far as I can glean form literature, the all too convenient paradigm of 1mm per month for bodily tooth movement was the base line. So as the pioneers in computer programmed aligner technology, whatever was programmed needed to be even more conservative than the “model.” In 1998, the max velocity per aligner was 0.33mm (used for distalization, 0.25,, per aligner for other linear movements). So at 14 days aligner change, we were well under the paradigm that was considered biologically and physiologically compatible for “bodily” tooth movement and it was an easy interval for patients to remember. There were some infant studies form the University of Washington, (one of the authors was a graduate that assisted us in the initial UOP (Pacific) studies), that favored a 14 day aligner change; however there were many other variables tested and the strength of that particular finding was not high. More recently, Al-Nawadi et. al. published in the Angle in 2021 a comparison of 7, 10 and 14 day aligner change rate with no clinically significant difference, a small statistical difference favoring 14 days in some posterior movements. They concluded by recommending 7 day change rate as acceptable, and up to clinicians discretion for slowing rate of change for some posterior movements.
Keep in mind that Align Technology has evolved in many ways since the 14 day protocol was initially adopted 24 years ago (almost a quarter of a century!!) , in terms of programming; maximum velocity of activation (that cannot be exceeded – even if you switch aligners sooner, the activation remains the same); staging or order of movement, aligner material and optimized (active, computer engineered) attachments that did not exist when the original 14 day aligner change rate was adopted and as a result Align Technology currently suggest aligner change rate of 7 days.
Personally, I pay attention to the manufacturers recommendations, believing that they would desire clinicians using their system to achieve the optimal clinical outcome possible. However, I am also cognizant of the fact that this is 1 paradigm that on the surface fits all, in reality it likely fits no-one optimally. What I mean by that is that every patient is different in terms of bone structure, root anatomy, movement programmed, occlusion and finally compliance – huge variable. I balance this with my knowledge of tooth movement biology (currently in awe of LR Iwasaki) and what I understand is that movement of a canine root “bodily” still fits neatly within a 1mm a month model, awareness that not all movements are comparable to a canine root activated to move bodily and that keeping an aligner in a patients mouth longer than 1 week rarely solves a movement issue- compliance usually drops in my experience. Invisalign Aligners (can only speak to these) are not engineered as a shape driven system , but a force driven system (so if you are trying to match Clinical outcome to ClinCheck stages, you will always lose sleep) and plastic is plastic, even the smart ones. I wonder at what point does the material stop influencing tooth position and tooth position begins to influence the plastic? Bottom line is that we are clinicians and we should be diagnosing, planning , monitoring and evaluating every patient, with manufacturer recommendations as a start point, not the only point. Do you change arch wires at the same interval for each patient and each movement?
Thanks for sharing with the update of this research article.
This is my humble personal opinion. I understand there are different protocols on the number of days in changing the aligners, but even if the patient changes the aligner every 7 days, my experiences with Conventional / Lingual fixed appliances treatments vs Aligners still tells me that fixed appliance will always treat cases faster than Invisalign. This applies to both non-extraction and extraction cases.
The only exception would be using Aligners for distalization of upper posterior teeth, which also prevent the side effects of proclination of the front teeth. This protocol is not possible for Fixed appliances without the use of mini-screws anchorage, so if the patient doesn’t want mini-screws and don’t mind about the length of treatment, aligners would be the chosen option.
For my patients who want invisible treatment options, I always offer both lingual fixed appliance and aligners.
It is refreshing to see any study comparing braces to clear aligner treatment (CAT), especially a reasonably well done one.
I expect that the KOLs will start their ad hoc arguments shortly. The critiques will likely be that the practitioner was not experienced enough, they did not use the latest technology, and the aligner changing interval was not the latest. The problem is that, all of these really ad-hoc arguments. They must be presumed false until proven true.
So far, the KOLs just don’t seem to have time to do any substantive comparative research. That the practitioner or practitioners doing these studies typically are very experienced with CAT seems to hold no weight with them. Lastly, there is no evidence that a more frequent aligner changing interval is any more efficient at achieving movements considered more difficult with CAT. Said otherwise, sometimes a more frequent aligner changing interval just causes them to “stop tracking” sooner and necessitates that a “refinement scan” is needed sooner.
In any event, the burden of proof is still with those going against this imperfect, though still well done research. As yet, they have not met this.
Figure 1, the flow diagram does not state how many patients were assessed for eligibility to take part in the study. This information should be included according to the CONSORT guidelines. A previous RCT comparing clear aligners with fixed appliances found that 77% of individuals (184 out of 240) screened for recruitment into the trial were deemed not eligible to take part (White et al 2017 https://doi.org/10.2319/091416-687.1). The reason stated was that they did not fulfil the inclusion criteria, presumably because their malocclusion was too severe, so it is difficult to apply the results of these studies to the treatment of more significant malocclusions.
One of the things that is argued that Invisalign is faster than fixed appliances is that in the latter they say 50% of the time is lost correcting problems that the same specialist created, due to lack of correct Dx, lack of planning or both.
And since Invisalign through the Clínckhec an excellent approach and staging is done, the treatment is completed in a timely manner without wasting time, but this is assuming that a good Dx and a correct staging have been done.
But in fixed appliances it should not be different, there are Orthodontic philosophies that follow a detailed Dx, treatment plan and treatment sequence, as everyone knows we have the VTO, in this each aspect of the treatment can be planned in detail, although ordinarily it is in a sagittal view, planning can also be carried out in the frontal and axial planes, and lastly there is also the Set-up to corroborate or perfect what is seen in the VTO, of course, with the manipulation of images this is much easier
And another aspect, according to multiple evidences, Invisalign treatments should be refined in a 2nd stage in 70% of cases, this is truer the more complex the case.
In addition, the estimated movement of tooth movement with aligners is 0.25mm per appliance, if the appliance is changed as fast as one per week the total effective movement is 1mm per month, no different from fixed appliances.