July 01, 2019

The Carriere Distalizer does not work? The first research paper.


The Carriere Distalizer is a heavily promoted orthodontic appliance used to correct Class II malocclusion.  But does it work?  This is the first research paper on this appliance.

I have posted about this appliance before.I discussed the claims made by the manufacturer and their Key Opinion Leaders. I also pointed out that there is no refereed published research on this form of treatment.  It is suggested that this appliance shortens treatment time and is more comfortable than conventional appliances. One claim made is

“We correct the sagittal problem in 3 months and complete the whole treatment in 6-9 months”.

Essentially the appliance is a sectional appliance that is supposed to :

  • Distalise maxillary segments
  • Derorate and distalises the upper first molars.
  • Create a uniform biomimetric force to establish univectorial dental displacement. (I still do not know what this means).

Following the sagittal correction, a course of fixed appliance treatment is done to complete the treatment.

These claims are not based on any research. As a result, I was very interested in this new paper that Progress in Orthodontics published.  The study was done by a team from Southern California.

Evaluating the treatment effectiveness and efficiency of Carriere Distalizer: a cephalometric and study model comparison of Class II appliances.

K Yin et al.

Progress in Orthodontics. https://doi.org/10.1186/s40510-019-0280-2

This is an open access paper, so everyone can read it at no cost

What did they ask?

“How effective is the Carriere Distaliser in comparison to Class II elastics and the Forsus appliance.”

What did they do?

This was a rather tricky paper to read and interpret.  I decided that this was a retrospective study because of the way that the records were collected.  It certainly was not a trial.  This is important when we consider the results and conclusions. The PICO was

Participants: Patients with Class II Division 1 malocclusion aged 10-14 years old in the permanent dentition.

Intervention: Carriere Distaliser followed by fixed appliances

Comparison1:  Class II elastics and fixed appliances

Comparison 2: Forsus appliance

Outcome: Duration of phases of treatment and total treatment time, molar correction and multiple cephalometric measurements.

They collected the records of 78 patients from three private orthodontic practices. 18 subjects were assigned to each intervention.   It is essential to realise that one of the inclusion criteria was that complete pre and post-treatment records were available. All cases were completed.

They did elementary univariate tests across the groups.

What did they find?

When they analysed the records, they found no differences between the groups before treatment for any of the variables that they measured.

I felt that there were real issues with the statistics because they used inappropriate tests for a comparison of three groups.

I intend to make this post more focussed than the paper. As a result, I am only going to report on the comparison between the Carriere and the Class II elastics groups.

There were no differences between these two interventions for any cephalometric measurement of the amount of molar correction.  They did report differences for the duration of treatment.  I have put this in this table and calculated 95% confidence intervals.

OutcomeClass IICarriereDifference (95% CI)p
Treatment duration (months)23.9 (5.8)32.3 (8.4)8.4 (3.4-13.7)0.001
Class II correction (months)10.3 (3.9)6.3 (2.2)4.0 (1.8-6.1)0.001
Canine correction (mm)-3.5(1.5)-3.7 (1.7)0 (-1.0-1.0)1.0
Molar correction (mm)-3.8 (1.4)-3.7 (1.7)0.1 (-1.5-0.99)0.8

In summary, the sagittal correction with the Carriere appliance was faster than with Class II elastics. However, the total duration of treatment was much shorter with the Class II elastics.  When we look at the 95% confidence intervals, we can see that they are wide. This indicates a degree of uncertainty in the data. This is possibly a reflection of the small sample sizes.

The author’s conclusions

They wrote an extensive discussion. Their main point was that the difference between the total treatment time was possibly due to relapse of the molar derotation and the amount of finishing that the operators needed to do.

As there were no other significant differences between the interventions, they concluded

“It is not effective and efficient to treat Class II malocclusion using the Carriere Distaliser”.

What did I think?

Firstly, I thought that the results were interesting, and they certainly did not agree with the advertising hype and claims that are made about this appliance. However, I do have some concerns with the methodology that they used.

Firstly, this was a retrospective study of completed cases with full records. This means that there must be some selection bias. Usually, this is directed towards increasing the effect size.

My other concern was that three different operators treated the patients.  Importantly, we do not know the distribution of the treatments between the operators. As a result, we cannot discount that there may be differences in the skill and the experience of these operators. This means that the study suffers from proficiency bias.   This could have been explored by using regression analysis, but their statistical analysis was too simple.  This is a significant and fatal flaw in the study.

Final thoughts

I cannot help feeling disappointed with this paper.  Recently, I have been accused of being too negative in my posts. Nevertheless, I think that if I review an article and there are significant flaws, I should draw attention to these. I may be being harsh, but I still do not know whether the Carriere appliance works?

Related Posts

Have your say!

  1. Thank you for posting this useful critique. This appliance has been heavily marketed, with minimal to no data supporting several of it’s purported benefits. I’m not aware of any data to even support the claim of distalization. This would ideally (at a minimum) need a Tx and Control group of adequate sample size, and superimposed cephs (before and after the “distalization”).

    If one were to indulge in a Gedankenexperiment, biomechanically, with a line of action below the CRes of the molar (and no intra-slot couple), at best one might expect some distal tipping. How is distalization achieved when the mechanics preclude that? Proclination of the lower incisors (without third order control) is the more likely method of correction.

  2. Another exemple of a retrospective study trying to validate or invalidate a treatment option based solely on case reports. Orthodontics is in dire need of better designed research protocols if we ever want to be taken seriously as a profession.

  3. Thank you for pointing the flaws of this paper. This should help reviewers to be more critical when they a ask to review a paper before publication.
    However,I was please to see that , at least, they were unable to proove that Carriere Distalizer is better than cl II elastics.
    I am quite sure that will will see better studies in a near future that will likely support their conclusion.
    It also help me to not feel left alone because I never use this appliance yet in my practice, although I have an intro kit in my drawers. Chances are that it will stay in my drawers…

  4. Thank you for another wonderful post Kevin.

    Although the study was flawed, we have something to hang our hat on. No real surprise that Class II elastics are similar to Class II elastics with an expensive piece of hardware. The force is supplied by Class II elastics in both cases and this is delivered as pressure in the pdl. Perhaps I missed something in Ortho 101?

    Perhaps a biomimetric force is a biological force that mimics the measurement of a force in a biological system which does not have a pressure gauge embedded to monitor the induced pressure, and only when the propriety product has been acquired at considerable expense by the gullible practitioner who is fond of glossy brochures and gobbledegook? Or not!

    I believe that this revolutionary device works best with their proprietary force delivery system. Elastics, as we know them, are old hat. I have no doubts that the results would have been positive had Acceledent be used, preferably with a little wool over the eyes.

  5. Very honest consideration, as usual. Interesting to read the reply from dott. Carriere, dealing with statistic bias. Anyway no evidence still available.

  6. Dear Dr. O’Brien,

    In this ocasion I agree with your disapointment regarding this article.

    I this I share with you and your readers my analysis regarding the article that you link in your blog at Progress in Orthodontics.

    Although I am normally reluctant to participate in social media or blog discussions specially relative to the Carriere 3D Motion Appliance or the Carriere System, in this particular occasion I needed to exceptionally do so as unfortunately Progress in Orthodontics has confirmed in written that they do not accept letters to editor in reference to their published articles.

    I have also read and studied with great interest the article Yin K, Han E, Guo J, Yasumura T, Grauer D and Sameshima G: Evaluating the treatment effectiveness and efficiency of Carriere Distalizer: A cephalometric and study model comparison of Class II appliances. Prog Orthod 2019, 20:24.

    In the mentioned article, the authors search for scientific evidence of effectiveness and efficiency of the Carriere Class II 3D Motion Appliance (CMA) in comparison to Forsus Appliance (FA) and Traditional Elastic Traction (E).

    I thank to authors for their efforts in investigating the topic.

    However, the way their results are presented make very difficult for the readers of the journal to evaluate the results.

    In a nutshell, their study involves 3 time points and 3 conditions. The three time points are Pre-treatment (T1), End of Class II correction (T2), and Post-treatment (T3). According to the title of the article, what readers of this journal would expect to learn, are results at T2 depending on the characteristics of the patient in T1 for the three treatments, and in which results are T3 being modeled as a function of patient characteristics at T1, at T2, and the three treatments. The problem is obviously multivariate, as there are a number of patient characteristics involved. Armed with such results, clinicians could take the initial multivariate configuration of their patients and determine the expected results at T2 and T3 should each of the three treatments be applied. In statistical terms, a multivariate regression analysis is called for. Instead, we are simply provided with average results T3 vs. T1 for each pair of treatments and for each measure separately. At the very least, we would have liked to see results T3 vs. T1, T2 vs. T1, and T3 vs. T2 for each measure. In statistical terms, at the very least we would have liked to see a repeated measures ANOVA with 3 groups and 3 time points for each measure. Needless to point out, the different treatments considered operate between T1 and T2, with conditions between T2 and T3 assumed to be common between all treatments.

    However, we are given very little information on T2 results (which should be the main focus of the article) and very little information on the treatment applied from T2 to T3. Total treatment length (i.e., T3 vs. T1) is extraordinarily large and we are given very little information to account for this surprising result. Also, we are not given information on how patients were assigned to treatments in the first place, which is needed to evaluate treatment vs. clinician confounder. In other words, optimally, patients undergoing different treatments should be randomized across clinicians and patients initial characteristics to avoid confounders. We acknowledge the difficulty of performing such an experimental design and as a result we suggest that these information (patient initial characteristics, clinician, etc.) is taken into account in the analysis via a regression model.

    Additionally, the authors based their clinical approach on a different technical application of the CMA as the one prescribed on the System, the first difference was selecting a lingual arch (according to text and figures, Figure 8-E2) as a source of anchorage. The current anchorage indication (for the last decade) when using CMA in permanent dentition patients is to use a vacuum rigid 0.4” invisible retainer.

    Secondly, authors, according to “Figure 9”, seemed not to accomplish a genuine Class I Platform, when occlusal Class I posterior platform should be always exhibited as an intimate and active contact of the distal ridge of the upper canine towards the mesial ridge of the fist lower bicuspid, this is always prescribed before moving to the comprehensive fix appliance treatment or comprehensive aligner treatment.

    Fundamental data are missing in the methodology of this study, necessary data to clearly understand properly the methodology of this study like which type of elastic protocol is used, the force, type and timing of elastic use, protocols of the comprehensive bracket treatment and which are the reasons to over-extend so inefficiently the all treatment times in all the 3 Appliance categories.

    Focusing particularly on the CMA small sample of this study, when it is accomplished by the appliance the Class I Platform, the spacing produced in between the upper incisors and the absence of the sagittal discrepancy establishes an scenario in which those cases should be finished within a much shorter treatment time than in this article. As a point of reference might be useful to consult the results of the study on CMA by the University of Michigan and the University of Florence, accepted and scheduled to be published at the next November issue of Angle Orthodontist Journal, article accessible already online at: https://www.angle.org/doi/pdf/10.2319/121418-872.1

    The purpose and the title of the referred article in Progress in Orthodontics does not match with the conclusions of the study. Unfortunately authors confuse the conclusion based on opinion and manifest that the inefficiency or inability of the clinicians to correctly finish in T3 the bracket comprehensive treatment is related to the CMA, while this appliance resulted to be the most efficient appliance in their
    study when accomplishing its purpose, the Class I occlusal Platform.

    The conclusions stated in the introduction abstract do not even match with the final conclusions in the article, all of them written by the authors inside the same article.
    Possibly a more dedicated or profound editorial control and a deeper peer review would have helped very positively in this publication as a Major Revision would have given the scientific and clinical value of the publication for the benefit of all of us.

    The unusual express peer review of the current article possibly limited this possibility. (13 days between the receipt and acceptance of the manuscript and 25 days total between the 1st receipt and publication of the article). Annual journal metrics of Progress in Orthodontics https://progressinorthodontics.springeropen.com/submission-guidelines/prepare-supporting-information.

    Hope this small drop helps a little bit more to the clinicians that care for knowledge on this device.

    Luis Carriere

    • The article in the Angle Orthodontist is excellent, in illustrating how easily a study could be manipulated. Selected patients compared with untreated individuals from half century old growth studies. I thought that studies of this design had been abandoned, perhaps not at Michigan.

      Overjet was reduced (Class II correction), lower incisors proclined and there was a restriction of forward movement at Point A, with an increase in the vertical dimension. Sounds like Class II elastics were worn and anchorage was lost in the lower arch.

      The authors suggested that the treatment was efficient. Perhaps the selected cases were compared with no treatment.

  7. Ceph overlays I have seen show a full 8mm Cl II occlusion is “Sunday Bite” corrected to a “Cl I platform” in 10 weeks using Cl II elastics. This is mandibular re-positioning, (this time) with a Carriere Motion appliance.

    Any wonder why there is significant relapse with this expensive technique?

  8. When you consider the mechanics of the Carriere device, it just treats the class 2 molar element separately, the fixed phase is then entered subsequently. It stands to reason that this method will take longer. I suppose the question then becomes does it avoid the pitfalls of prolonged class 2 elastic use ( upper/lower incisor detorquing)??? I have not used it so I am awaiting some real evidence.

  9. I am surprised that a device used for more than a decade, only now has some published scientific data about its effects. I agree that published studies have methodological issues, but what I am asking myself at this time is why the author, who in his letter-response seems to demonstrate scientific knowledge, has never published auditable results about his own creation.

  10. I tried only a few cases and after a treatment-time of 6-9 months I compared a cast with the initial cast and made a ceph.
    the typical change was an expansion of the upper canines about 2mm with extrusion of the canines. This is a reaction I could also see on
    the pictures of a distalizer- presentation . I saw only minimum distalisationof upper molars but a strong reaction in the lower arch with protrusion of incisisors up to 4mm or mandibular repositioning. we need an evaluation of the distalizer without the following treatment with fixed apppliance to examine the changes wich are related to the distalizer.
    Dr.Michael Weber

Leave a Reply

Your email address will not be published. Required fields are marked *