January 04, 2016

Do orthodontic extractions change soft tissue profiles?

Do premolar extractions influence soft tissue profile in Class II cases?

As we all know the extraction/non-extraction debate  continues in orthodontic and dental circles. My feeling is that most of the opinions on either side tend to lack an evidence base, because there is a dearth of research on orthodontic extractions. I was, therefore, pleased to see this systematic review published in the European Journal of orthodontics. I thought that this would be a good start to my blog posts for this New Year. Sadly, I was wrong, but read on…

Soft-tissue changes in Class II malocclusion patients treated with extractions: a systematic review.

Janson, G., Mendes, L. M., Helena, C., Junqueira, Z., & Garib, D. G. EJO doi:10.1093/ejo/cjv083


The authors pointed out in the literature review that the extraction debate, particularly for class II malocclusion, is important. This is because there are concerns that premolar extractions might cause greater lip retrusion than other methods, for example, functional appliances and headgear.

What did they do?

The aim of the review was to evaluate the soft tissue changes after orthodontic treatment with premolar extractions in class to Division I malocclusion is. I felt that  this review was important because any evidence that is derived will fuel the debate.  They adopted systematic review methodology in an attempt to provide high-level evidence. Their inclusion criteria in the form of a PICO  was:

I would like to point out that they included retrospective studies and I will return to this later.

They classified the risk of bias in the studies using the Cochrane Collaboration tool for assessing risk of bias.

What did they find?

After the usual filters based on the PICO they obtained the final sample of 13 papers. Of these, eight were retrospective, three were prospective and two were unclear.

They provided a large amount of data in supplementary tables and I looked at these studies closely. I found that only one study was a randomised clinical trial. As a result, we have a systematic review that includes only one RCT out of 13 papers.  This is important.

They did not combine any of the data and simply listed various cephalometric measures derived from individual papers. They discussed these in a long discussion that I  struggled to understand. They finally concluded “that according to studies with high scientific evidence that when class II malocclusion is treated with extractions there are various levels of upper lip retrusion”.

What did I think?

I was disappointed in this paper and I am sorry to be critical. However, this review attempted to answer a controversial clinical problem and was published in a high-quality journal. There is also a danger of misinterpretation of the quality of this paper and hence the results.

I have several points that I think are important:

  • The whole purpose of a systematic review is to systematically review the high-quality evidence. This means that the systematic review should be confined to prospective  controlled clinical trials or randomised clinical trials, in which bias is minimised. If the reviewers include retrospective studies, that are poorly controlled,  biases are introduced and the findings of the review are not high quality.   In effect, by including studies of high risk of bias and low quality all that is happening is a recycling of low levels of evidence.  When this is then repackaged in a systematic review readers, may be less critical and the review may have undeserved impact.
  • The authors used the Cochrane Collaboration risk of bias tool to evaluate bias in the final sample of papers.  This is designed for use with randomised trials and I cannot see how they used this for retrospective studies.  This is a major problem and I have checked and rechecked the paper to be absolutely clear on their methodology.  I’m not sure how the EJO  referees did not raise this as an issue.
  • They suggested that the studies provided a high level of evidence, yet they included eight studies that they classified as high risk of bias ( although I do not know how this was calculated).

In summary, I cannot help thinking that this paper exhibited significant problems and to be honest it does not add to our knowledge. I have covered “how to read a systematic review” in a previous blog post and my previous comments are relevant to this review.

We also need to be cautious in interpreting studies that evaluate the non-extraction/extraction decision. This is because it may have been a clinical decision to use mechanics that open the naso labial angle.  We should always remember that the effects of treatment are a combination of the treatment decision and our mechanics.  Simply combining data from a mixture of studies completely ignores the “real world’ clinical situation.

Finally, I would like to call on journal editors to be more critical when assessing some of the systematic reviews that are submitted. The use of this methodology is widely accepted and we should welcome the publication of good review. However, we should make sure that the standards are high. I do worry that there is a danger of the systematic review becoming the cephalometric analysis review of the 1980s;  a lot of data meaning very little.

Related Posts

Have your say!

  1. outstanding blog entry! kudos!

  2. “At last” I thought, “perhaps we are going to get an authoritative view on maybe just one aspect of the oldest of all orthodontic conunudra – to extract premolars or not”

    But no, all we got was a critique on how the review was designed and how the methodology was employed with comments about the possibility of poor interpretation by the review’s authors. No comment whatsoever on the crucial and ancient issue emblazoned in the heading of this blog – “Do premolar extractions influence soft tissue profile in Class II cases?”

    I suppose I should not be surprised; it seems to have become the norm in many blogs for some time – methodological criticism without any really useful information.

    What does surprise me, though, is why scientists feel they have to keep re-examining subjects that have already been examined and tested “ad nausiam”.

    We all know that if you remove hard structure from beneath or behind soft tissue, that will affect the shape of that tissue. To try to prove otherwise with ‘science’ does nothing more that highlight the defensive agenda behind the study.

    Sadly Kevin’s Blog has tended to become a series of lessons on how to read and assess a systematic review, with little information regarding clinical or practical value.

    • Thanks for your comments and I am sorry that you feel my blog has simply become a “series of lessons on how to read systematic reviews with little information regarding clinical or practical value”. I am not sure that this is the case. I know that I cast a critical eye over the research evidence and I do this to help readers understand the way to interpret the literature. If I review my last few postings, I have highlighted clinically useful information on fixed vs removable functional appliances, the effectiveness of myofunctional and standard functional appliances, methods of treating orthodontic ulceration, the effects of Acceledent on pain and whether we can speed up orthodontic treatment. I also consider that it is my role to teach people how to read systematic reviews and I had hoped to cover this as part of my blogs.

      In answer to your point about what the answer is to the question on the effect of extractions on soft tissue profile, my feeling is that extractions do not influence the profile and the greatest effect is likely to be the mechanics that we choose to use.

      Has anyone any comments on the post by Noel Stimson?

    • Totally support your opinion

  3. On this side I am a fan of Prof. O’Brien’s thought provoking reviews and admire the substantial time it takes to craft this blog. This is a valuable service for orthodontists and the dental profession. As well, I especially enjoy this type of snappy exchange. Noel Stimson has seen fit to weigh in on the O’Brien blog with “blunt trauma”. Given Stimson’s authoritative position as Editor of Cranio UK I wonder if he will meet the challenge of these sharp criticisms by offering his own clear views on the subject.

  4. Dear Kevin,
    I feel compelled to comment on your we-just-don’t-know assertion. I pray that we all live to see an ethical prospective comparison of extraction and non extraction. By that I mean a trial in which the subjects are fully informed, have no preference (we flipped a coin; it came up extraction. Sign here), equally susceptible, and participating in a study without data-peeking and “Big Brother is Watching” surveillance. As I contemplate eternal life, perhaps the following might warrant consideration:

    Paquette, D.E., Beattie, J.R., and Johnston, L.E., Jr.: A long-term comparison of non-extraction and bicuspid-extraction edgewise therapy in “borderline” Class II patients. Am. J. Orthod. Dentofac. Orthop., 102:1-14, 1992.

    Cassidy, D.W., Jr., Herbosa, E.G., Rotskoff, K.S., and Johnston, L.E., Jr.: A Comparison of Surgery and Orthodontics in “Borderline” Class II, Division 1 Adults. Am. J. Orthod. Dentofac. Orthop., 104:455-70, 1993.

    Beattie, J.R., Paquette, D.E., and Johnston, L.E., Jr.: The functional impact of extraction and nonextraction treatments: A long-term comparison in “borderline,” equally-susceptible Class II patients. Am. J. Orthod. Dentofac. Orthop. 105:444-9, 1994.

    Hagler, B.L., Lupini, J., and Johnston, L.E., Jr.: Long-term comparison of extraction and nonextraction alternatives in matched samples of African American patients. Am. J. Orthod. Dentofac. Orthop. 114:393-403, 1998.

    Johnston, L.E., Jr.: A comparative analysis of Class II Treatments: A retrospective/prospective alternative. Clin. Orthod. Res. 1:142-6, 1998.

    Scott, S, and Johnston, L.E., Jr.: The perceived impact of extraction and nonextraction treatments on matched samples of African American patients. Am. J. Orthod. Dentofac. Orthop. 116:352-8, 1999.

    Johnston, L.E., Jr., Paquette, D.E., Beattie, J.R., Cassidy, D.W., Jr., McCray, J.F., and Killiany, D.M.: The reduction of susceptibility bias in retrospective comparisons of alternative treatment srategies. In: Clinical Research as the Basis of Clinical Practice. Katherine Dryland Vig and P.S. Vig (eds.), Ann Arbor, Center for Human Growth and Development, 1991. (pp. 155-77)

    • Gerry Samson is correct: providing commentary on the state-of-our-art can be useful and commendable. However, occasionally those opinions are substituted for the evidence; thereby, subverting the purpose of the evidence-based process in clinical practice. In addition, if we are all expected to depend upon strictly overwhelming statistically significant results produced ONLY from “gold standard” RCT’s, I suspect that we will soon distil our publications down to a single journal that is published about once a year (truly made for light reading).

      It seems such a casual exercise to push decades of diligent work, appearing in the form of thousands of non-RCT studies, into the dustbin simply because someone might feel there could possibly be some “biases.” Just maybe, there isn’t . . . (besides, RCT’s aren’t immune to poor prosecution and biases, either).

      In addition, Lysle Johnston has reminded us, “Everything in orthodontics works . . . at least well enough to keep the lights on. Since everything works, nothing matters (least of all the work product of academics).” Taken to the extreme, if only some RCT’s are deemed worthy and all other research isn’t to be considered, then just about any old crazy idea in orthodontics could find a comfortable home. This is especially true since it is unlikely that many desired RCT’s can be ethically performed (or they may require an inordinate amount of time to answer questions that were previously answered (e.g., RCT’s for early treatment).

      Consequently, requiring an RCT to confirm or deny many concepts performed in daily clinical practice (e.g. extraction/nonextraction) just ain’t going to happen. So, what’s a mother to do? I’d refer the reader back to the “pyramid of evidence” and take a step down to look at what else we have to reference.

      The question posed in the paper questioned in this blog was, “Does extraction change facial profiles?” This is actually a rather bizarre query because it is not the act of extracting teeth, but rather (herein lies the rub) the subsequent treatment that can potentially alter the profile and an RCT isn’t required to answer that.

      The real clinical question is actually, “Does extraction (and the subsequent treatment) routinely harm the profile?” The evidence-based answer appears to be overwhelmingly, “No.”

      Like Lysle Johnston’s reply above, I offer a few non-RCTs as my ante into the discussion. I direct those in your blogosphere that rather than read my stuff, take a look at the references in these articles and seek out the sage advice directly for themselves.

      Bowman and Johnston. “The Esthetic Impact of Extraction and Nonextraction Treatments for Caucasian Patients” with Dr. Lysle E. Johnston, Jr., Angle Orthodontist, 70:3-10; 2000.

      Bowman. “More than Lip Service: Facial Esthetics in Orthodontics,” Journal of the American Dental Association (Cover Story), 130:1173-81,1999.

      Bowman “Pulsus a Mortuus Equus.” In: Age Appropriate Orthodontics – Part I. Seminars in Orthodontics. 2014 Mar;20(1):36-45.

      Bowman. “Altering the Extraction Decision with Miniscrews Part I,” OrthoTown, April, 42-49, 2011.

      Bowman, Ramos AL, Sakima ML, Pinto AS. “Upper Lip Changes Correlated to Maxillary Incisor Retraction-A Metallic Implant Study” with Angle Orthodontist, 75:499-505, 2005.

      Bowman and Johnston. “Orthodontics and Esthetics,” Progress in Orthodontics, 8(1):112-129, 2007.

    • I enjoyed reading Dr. O’Brien blog about soft tissue changes following orthodontic extraction treatment. First, I would like to point out that the authors, in the abovementioned systematic review, did not use the domains of Cochrane collaboration’s tool for assessing risk of bias; however, they adapted the general classification of the Cochrane collaboration’s tool (“unclear, low, high”). Modification of a risk of bias tool is likely to affect the validity and applicability of the results. Without validating the newly developed tool or group of quality items, interpretation of the review findings could be questioned. However, this is a common pitfall in dental systematic reviews. Evidence from methodological studies showed that non-validated risk of bias assessment tools (consisting of items extracted from a variety of tools) were used in 38% of dental reviews (in 256 out of 684). This is concerning because of the limited ability to assess how the estimates of effect may have been biased due to study conduct; clinical decision making may therefore not be based on valid findings provided by the best evidence reviewed.
      Second, I agree with Dr. O’Brien regarding his comment about including retrospective studies. One of the limitations of retrospective studies is selection bias. In retrospective studies on soft tissue changes following extraction treatment, patients probably were different at the outset and thus subject to one treatment or the other (susceptibility bias). Retrospective cohort studies usually fail to control susceptibility bias, which is usually controlled in randomized controlled clinical trials.
      Although Dr O’Brien’s definition of a systematic review “to systematically review the high-quality evidence” is appreciated, given that there is no high-quality evidence published on the topic, I think that summarizing the best AVAILABLE evidence, by the authors, is still considered acceptable evidence and qualified to be published.

      • Thanks for the comment. Yes, I agree it is acceptable to publish a review that discusses the best available evidence. However, this was not the case in this review. They stated that the evidence in the review was from studies that represented high levels of evidence, this was not the case. My feeling is that the review would have been improved if they reviewed the evidence and then provided a critique of the studies, and then point the way forwards.

  5. Hi Jay, you make very good points and I think that I will address these in a future blog post and we can have a debate about this. But for now, I agree with your comment on the fact that we should not look at extractions in isolation, as the mechanics adopted during the treatment will also influence facial profile.

    I also take your point about not rejecting retrospective studies but I also feel that we should not classify retrospective studies as “high levels of evidence” (which is what was done in this paper). My feeling is that they do provide some evidence but the important fact is that research findings from all studies should be interpreted with caution. For example, if there are only retrospective studies available we should be able to inform a child and parents that a form of treatment might have an effect and when the evidence is available from trials we can be more certain in the information that we give to people. I will expand on this in my next blog post when I shall also discuss the decision process and strength of evidence for my cervical spine surgery, which is a far more risky procedure than orthodontics…

  6. I have followed with interest the commentaries that this post has generated. I would like to add a grain of salt.
    Maybe we are missing an additional factor in the discussion: Communication. How much are we worry that the reader is incapable of critically appraise a publication? How much can we or should we protect the individual clinicians of making their own decisions based on available evidence? One of the problems I do perceive is the fact that some clinicians may only read a publication title, like the topic and jump into the abstract conclusions without reading/critiquing anything else and then apply the conclusions into practice decisions. That is where Kevin’s concerns apply, I believe. Not carefully crafted conclusions are dangerous. Myself I have not been too careful in this regard over the years. I am staring to realize that. So for the clinician that just reads the conclusions then not making it clear how dangerously unsupported are some conclusions is a big problem. Maybe we should add a color background to those conclusions from red “use at your own peril”, amber “be aware of the risks” to green “solid, well founded”, or something like that. That is when it comes back to communication. Somewhere in an obscure location in the discussion or results there may be a warning line, but who really notices it? Maybe we should have stronger “how to read the literature” training among us clinicians to take some of the heavy weighting from the editor(s), reviewers, etc. Even orthodontic systematic review consistently warns us how “poor” the evidence is. The burden at the end of the day is in each one of us that make an informed decision with our patient’s believes and expectations.
    On another note I do not support the idea that we should only rely on RTCs. It is clear that for some clinical scenarios it will never ethically apply. Then there is the issue of long-term results and how unlikely is that the original well-calculated RCT sample will participate again. That is where well-conducted retrospective studies may have an edge. At the end of the day in this specific topic the point of discussion would be if the extraction was made to diminish dentoalveolar protrusion or just to relieve crowding. In the latter the impact on the soft tissues may be minimal. Selection criteria seems to be a common weakness in this type of study. I am already aware that the flavour of the day is impact of extractions in future OSA. But I guess that will be a topic for a different day.
    Thanks Kevin and others for this lively useful discussion. Great references to read carefully.
    Happy New Year! Carlos

    • I am quite sure what you say about the clinically useful information in your blogs is correct since you will know better than I what you have written. However, I have usually found your blogs to be predominantly about the quality of the science rather than the significance of the clinical content. Obviously the former affects the latter, but it would be nice to know what you actually believe to be the case, the science notwithstanding.

      I find it sad that, as you say, we still do not know the answer to so many questions. Sad because orthodontic science has been around for almost 150 years and even with today’s knowledge and facilities, we are still arguing about the fundamental questions.

      In response to Dr Samson’s challenge, I am more than happy to “offer my own clear views” on the subject under discussion, but not by using the science as my sole raison d’etre.

      Why? Because I believe that there is very little solid science to back up what is being said on either side of most of orthodontics’ most vexatious arguments.

      Some time ago I trawled through ten of the most significant recent Cochrane Reviews on orthodontics. Subjects ranged from “the use of splints in bruxing children” through “orthodontic intervention in TMD” and “Class III treatment” to “treatment of deep bite Class II division 2”. All reviews except one concluded there was “insufficient evidence” or “insufficient research data”, etc. in the 650-odd studies that were included. The only study to have a firm outcome was one that looked at early versus teenage treatment and which concluded that early treatment was better only in the avoidance of trauma to prominent upper incisors, especially if Twin Blocks were used. Many conclusions stated “no treatment recommendations can be made on the basis of these trials”. Every single review demanded that more research was urgently required. Not a satisfactory state of affairs in my view.

      For these and other reasons I have always tried to temper my use of science with common sense. The modern Hippocratic Oath makes it clear that there is more to medicine than science alone; the art of dentistry is just as important and that it should include keeping our minds open. The original Oath goes on to say “I will take care that they (the sick) suffer no hurt or damage”.

      My view of orthodontics is very simple. While we do not yet fully understand the fundamental causes of malocclusion, we should understand that dragging teeth back into the skull just at the age when the skull and facial bones are trying their utmost to grow forwards (the so-called growth spurt) simply makes no sense. It is so much more sensible to make use of natural growth forces; to fight them is to ensure defeat eventually (did I hear someone say ‘relapse’?).

      It makes no sense either to pointlessly remove healthy teeth simply to make the orthodontist’s job easier in straightening the remaining teeth into a nice smile. If we are given 32 teeth it is because we are supposed to have 32 teeth. That dreadful old lie “your child has too many teeth” no longer has any meaning, thank God. Hopefully we all understand that the crowding problem generally lies in jaws that are too small with inadequate arch length. Ipso facto the solution is to help the jaws to grow. Simple.

      Of course there are many parts to the complex story of malocclusion, but the above is where I started once I realized the mistakes I had made in my earlier years.

      • Noel, thanks for the comments. You are correct in my blog posts I concentrate on the evidence and I draw attention to the quality of the research. I then draw a conclusion based on my interpretation, but I also finish most of them by suggesting that readers make their own conclusions. This is because I would not be so presumptuous to suggest my opinion that may not necessarily agree with other people’s interpretation of the literature. This is what reading the literature is all about and we should move on from clinicians basing their practice on what a certain guru, authority or Professor says. This is nicely illustrated by the debate that I had with Lysle, Jay and Carlos in the comments of this post.

        Your comments on Cochrane reviews are interesting but it does not mean that you can simply ignore them, as they do point a way forwards. This is where we are in orthodontic research and we need to find ways of improving so that we have a stronger evidence base.

        Your final comments are your own opinion, but as I said I do really give my opinion I just point people in the direction of the evidence and encourage them to interpret it. Maybe you would like to let me have some references to underpin your beliefs?

        • Kevin,
          This is all most interesting; I now realize that this is primarily a blog about the science and its validity or acceptability. Any conclusions or comments on the clinical aspects of a study or review are based purely on your interpretation of the evidence. This in turn forms your opinion, I would have thought, so is your ‘conclusion’ (which you willingly offer) not the same as your ‘opinion’ (which you are reluctant to suggest)?

          Forgive me if I sound confused. Though like you, my opinion is also based on my interpretation of the science, coloured very much by 40 years of experience and, yes, listening to a number of extremely talented and experienced people from all over the world who were not hide-bound by the limitations and strictures of NHS orthodontic thinking.

          Of course we should not ignore the Cochrane reviews; they have all been filtered through the gold standard of scientific excellence. But I do not agree that they show the way forward. On the contrary, they show at the very highest level that orthodontic science has hit the buffers with a massive vote of inconclusiveness and an appeal for more and better science. If as you say this is really ‘where we are in orthodontic research’ after 150 years of argument, it is not surprising so many of us rely on our own ‘experience’.

          Why does such research need ‘interpretation’? If the science is good then the evidence is clear and unequivocal; no interpretation is needed. Opinions are then either invalid or unnecessary and it is absolutely clear what we should be doing or not doing. Sadly, nothing is further from the truth at this time.

          Kevin, it may take me a while to compile a list of suitable references for you that support what many of us regard as ‘progressive’ orthodontics. I am also aware that much of it will fail to impress you, but hey, ‘this is where we are in orthodontic research’!

          • Thanks for the comment. As I have stated before, in my blog posts, I outline a paper and highlight aspects of the research. At the end of the posts I encourage readers to interpret the findings and decide if they help with their own clinical practice.

            I totally disagree with your query as to whether “research requires interpretation”. This is exactly why research papers are published. We should question and interpret all sources of evidence that range from clinical experience to randomised trials.

            I look forward to receiving the list of the papers that outline “progressive orthodontics”. Although I have no idea what this is!

          • Kevin

            You asked me for some references to support my views on premolar extractions and their possible effects on facial profiles; I have listed six below. Interestingly the most recently published study I could find is 16 years old.

            Melson B, Hansen K and Hagg U. 1999. Overjet reduction and molar correction in fixed appliance treatment of class II division 1 malocclusions: Sagittal and vertical components. American Journal of Orthodontics and Dentofacial Orthopedics 115:13-23.

            Stellzig AS, Basdra EK, Kube C. and Komposch G. 1999 “Extraction Therapy in Patients with Class II/2 Malocclusion”. Journal of Orofacial Orthopedics. 60:39-52.

            Kahl-nieke.B, Fischback.H, and Schwarze.C.W. “Post retention crowding and incisor irregularity: A long-term follow-up evaluation of stability and relapse”. BJO.22: 249-257. 1995.

            Harris EH, Gardner RZ, and Vaden JL. 1999. American Journal of Orthodontics and Dentofacial Orthopedics 115:77-82.

            Forsburg, C. and Odorick, L. 1979, Changes in the relationship between the lips and the aesthetic line from eight years of age to adulthood. European Journal of Orthodontics 1: 265-370.

            Little, R.M. Riedel, R.A. & Artun, J. “An evaluation of changes in mandibular anterior alignment from 10 to 20 years post-retention.” American Journal of Orthodontics and Dento-Facial Orthopedics. 93:423-428. 1988.

  7. I think that Johnny Nash had it all worked out in 1972 when he sang:

    “There are more questions than answers, and the more I find out the less I know…”

    Wouldn’t life be boring if we had all the answers?

  8. This blog has been a truly enriching experience for me…to learn and understand the thinking process of some of the big wiggies of orthodontics…thanks dr kevin for such an experience.what i could understand was that
    1.Cochrane reviews are the gold standard. You cannot just take whatever evidence you have and say this is the best possible i could do. When you strive for the best then only you reach some where near it. I enjoy reading cochrane reviews and ur blog has taught more about it.
    2. Dr Noel s comment that the reason for crowding is as simple as 32 teeth in a small sized jaw seems oversimplified and unfounded. And the management of crowding as “growing the jaws” is even more bewildering.
    3. Reasons for arch length discrepancy could vary from jaw size discrepancy or even tooth size discrepancy. The jaws could be normal sized andtooth size might be increased. This has been corroborated by dr begg and kesling…that as humans are evolving, we move from raw to softerdiet , reduced interproximal attrition leads to increased crowding. This has been one of the main reasons for extracting premolars.
    4. In patient with normal jaw size how much space can we expect to achieve just by expansion and how do we resist the strong muscular forces leading to relapse?
    5. Its my opinion that there cannot be one simple answer to a question like this. As many eminent orthodontists above have stated selection criteria and the mechanics used are important considerations to be made while treating a case.
    6. Last but not the least, if “forces of nature” could take care of all the malocclusions, then we orthodontists would have been jobless i guess!!;)

  9. I think that what Kevin defines high quality research (RCTs) is good for clinical practice if the clinical reality to which it is applied is not be misrepresented to satisfy the demands of the research methodology.
    The choice of which treatment is to be evaluated greatly influences the “scientific process” which follows; this choice can be not very scientific. Are RCTs always the right tool to find our clinical answers?

  10. I have been reading these posts with some interest, especially because my work has been used as a reference in some portion of them.

    First off I would like to point out in addition to the comments that Lylse made on ethics and consent, that although an RCT is the gold standard for some endeavors such as handing out different colored pills in a medical study, in the case of most orthodontic treatments no one involved (patients, operators nor examiners) can be blinded to the treatment variable, for instance whether teeth were extracted or not. No matter how the study is designed (prospective or retrospective) there will be inherent bias. That is why we used discriminant analysis to retrospectively match samples in an effort to control for bias.

    The other comment that I would like to make is the questionable basis of the argument about whether extractions affect profiles. Of course they do. Even in our matched sample of borderline patients there was about a 2 mm difference in the position of the lips between groups post treatment. That may not seem like much and may be insignificant for one patient yet the 3D volumetric loss associated with that change could also be detrimental to another patient with similar cephalometric findings yet entirely different soft tissue resilience, elasticity and muscle tone. My point is that we as orthodontists should use the information gleaned from the numerous sources available to make the best decisions we can for each individual patient based on their characteristics rather than treating every patient the same because we have established some dogmatic approach derived from findings in the literature, or to describe that approach as Lysle has been known to say ,”the treatments don’t vary, only the patients do.”

    In summary I believe that evidence has its place along with experience but overlaid on both of those is the practitioner’s desired outcome and let’s face it, at the end of the day it is what each of us considers a beautiful result that drives our treatment decisions. It is that inability to agree on what constitutes beauty and therefore a single best outcome that leads to our century plus old arguments over which approach is “better”.

    • Amen Dave. Not only do we have to have to take into account the great variability found in our patient populations and our own aesthetic preferences, we have to give credence to our patients preferences and desires. I frequently make the recommendation to extract 4 bi’s with full confidence that at a minimum I will not harm the profile, and will likely improve it, only to have the patient refuse to have teeth removed.

  11. I would like to congratulate Kevin for his blog. I read all postings with interest and I would like to add the following thoughts for whatever they worth.
    -It seems to me that we often turn the discussion on evidence into a yes-no.
    -I believe that all evidence is important and useful and should be considered.
    – A critical appraisal of the evidence allows us to correctly use and interpret the evidence and I believe that we probably all agree that not all evidence is created equal.
    -Different types of evidence is required or is acceptable for different questions and circumstances; however, for any design there are rules which we need to follow as the chosen “design label” does not guarantee the quality.
    – There is a lot of literature in medicine and dentistry on incomplete design, reporting and synthesis of the evidence which results in waste. See Lancet series 2014 on the topic.
    -I think it is important that we include in our educational system appropriate coursework that would allow us to critically review the literature and to abstain from company presentations at the exhibit halls of national meetings! I do not think that I received the skills to critically appraise the literature when I was in ortho school.
    – Evidence based orthodontics/dentistry/medicine does not ignore the importance of clinical expertise and patient preferences and circumstances

  12. Dear Kevin,
    While we await the outcome of your Delphic meditations (“Ignore that man behind the curtain!”), I think I need clarify my position on RCTs. I am not a Luddite; I don’t question their position on the great pyramid of truth. But with respect to extraction/nonextraction, I would invoke Matthew 7:3-5 King James Version: “And why beholdest thou the mote that is in thy brother’s eye, but considerest not the beam that is in thine own eye?” In other words, one can easily see major defects in many of the most vaunted orthodontic RCTs, all of which are seen as providing high-value information, apparently by dint of their group membership. This classification seems to classify all other sources as low level and thus only capable of providing questionable data that cannot even be discussed. Let them eat cake! (In effect, Marie Antoinette is saying, “Treat any way you want; the question can never be answered to our satisfaction.”) Were there to be a discussion, however, I would suggest that the Illuminati consider two points: is the question worth the time and treasure of an RCT and can it be addressed by randomization? Considering all the devious nonsense that surrounds the “perils” of extraction, I would opt for the highest level of proof that can be obtained ethically. Therein lies the rub: can extraction be randomized to a fully-informed subject? Baumrind came up with a method of defining uncertainty and, in so doing, permit the secret randomization of the un-randomizable. It was never implemented, and I doubt that it would pass IRB muster today. Accordingly, I would suggest that the use of discriminant analysis as a method of confounder summarization is and should be considered to be “the next best thing.” It provides a management framework in which nobody involved in the treatments knew that someday their results were going to be examined. It has been applied to several different populations and questions and has produced consistent (2mm, 2mm, 2mm…) and orderly results. Further, it allows the timely generation of long-term post-treatment data, a major benefit given that most of those who prosecute RCTs will be dead (along with the major question) or retired when the time comes really to see how things came out. Indeed, my experience in contacting former patients is that only about 1 in 10 are willing to return for records. Given the samples used in many of the RCTs, a 10 percent recall boils down to a million-dollar case report. Here is the sticking point: discriminant analysis defines equally susceptible sub-samples in which treatment assignments already have been made, in some cases decades ago. Technically speaking, assignment is haphazard, rather than random. The actual significance of this deviation needs to be discussed in the context of the importance of the question and the absence of obvious alternatives. We ought to chat about this someday….

  13. Dear Prof, As a StR I love your blog as it’s unravelling the mystery of research to me! After reading this blog and the replies I have a question. Since RCTs are the best evidence and an SR of RCTs the highest level- how would we collectively evaluate studies that look at treatments which would be unethical to withhold (so RCT design not appropriate) e.g an alveolar bone graft. Lots of surgical procedures have come about through experience and error and retrospective assessment of surgeon success- would you say these studies are not worth systematically reviewing? Or is it that this kind of review can’t claim to be of high evidence, just best available. Thank you!

  14. Noel Stimpson says growing the jaws is simple. How is it done?

  15. I have responded to Kevin’s invitation to provide references to underpin my comments and I am currently compiling them.

    Dr Ray Reed (Jan 13th) asks me how I believe ‘growing the jaws’ is done. There are many types of functional appliance around and abundant opportunities to learn how to use them. The current favourites with many of today’s orthodontists seem to be the Advanced Lightwire Functional appliance (ALF) and the well-established Clark Twin Block appliance, sometimes used in combination.

    Cochrane Reviews will usually dismiss the idea of maxillary or mandibular ‘growth’, saying any orthopaedic changes are minimal and limited to dento-alveolar change only. That is OK as the alveolus is part of the maxilla or mandible anyway!

    My experience has been that the mandible is the difficult one and that the majority of the change has been the result of re-modelling of the temporomandibular joint area. Maxillary development is much easier – I recall successfully developing a ten year-old’s maxilla from an inter-premolar distance of 19mm to 29mm over 2 years with an ALF.

    In response to Professor Agarwal’s comments (Jan 8th) on my statement that “dental crowding is due to small jaws”, I would say this:
    1. I agree. Of course Cochrane should be the gold standard, but even so science is never the whole answer. Nash’s song about “the more I find out the less I know” is actually a paraphrase of a famous quotation from scientist and father of radar, Professor Henry Tizard:

    “I hold in particular that no true scientist can be an atheist. The more his knowledge grows the more apparent is the depth of his ignorance”.

    The gold standard Cochrane Review is simply a high-quality snapshot of what we (think) we know today having had a close look at what others have worked on recently. A true scientist has to accept that things can change. The difficulty for most of us is staying within this context.

    2. I am astonished that Professor Agarwal views my statement about dental crowding and the human dental compliment of 32 teeth as “over-simplified and unfounded”. 32 teeth has been the human norm for several million years and is represented as such in the vast majority of the human race. The only reason the crowded teeth cannot fit into the arch is because the arch is not big enough (except maybe in those rare cases of macrodontia). The solution is to “grow the jaws” not reduce the number of teeth.
    The causes of under-developed jaws are many and complex, but the potential for growth is usually still present up to age 20 and will normally be released with appropriate functional appliances once the cause(s) have first been addressed. This is crucial if relapse is to be minimized or avoided.

    3. I agree with what you say about arch length and tooth size, but I cannot agree that reduced interproximal attrition is a justification for extracting premolars. Such views come from a lack of understanding of the causes of crowding; crowding rarely dental in origin; crowding starts outside of and independently of the dental arches primarily due to environmental factors. The dental crowding is the result not the cause. Simply straightening the teeth (the symptom) will not solve the problem (the cause) and it will inevitably relapse.

    4. If the patient has a normal sized jaw, perhaps further expansion should not be required! But in answer to your question, I have seen many cases where lost premolar spaces have been reclaimed by arch length development (often inaccurately referred to as expansion) and the spaces treated with implants or bridges, providing the patient with a normal size jaw(s) and a full complement of teeth. I know several orthodontists who specialize in this kind of treatment; many of their patients are mature adults who are displeased with the appearance of their ‘treated’ teeth.
    The muscle forces that can lead to relapse also have to be addressed and should not be under-estimated; they are a crucial part of the cause of the problem. We have all been taught that if muscle and bone are in conflict, the muscle always wins – deviate swallow, hyper-active mentalis muscle, incompetent lips, mouth breathing, etc. These all play a part in malocclusion and all need to be addressed if they are present. Myofunctional therapy was first introduced seriously into orthodontics by Daniel Garliner in 1971 but it is only relatively recently that the impressive ‘Myobrace’ system, which incorporates intensive myofunctional therapy, has appeared.

    5. In my opinion, the principle behind treating malocclusion is indeed very simple – treat the underlying causes, of which there are many, as you so rightly say, not just the obvious result. Achieving this is not always simple or straight forward and requires an understanding of what those causes are and an ability to spot them in a patient.
    One thing we have found out. If the muscles are left behaving abnormally and the facial and jaw bones are not in the correct position and of the right size, the dental result will fail.
    I firmly believe that malocclusion is simply a muscle-bone conflict or imbalance; the displaced teeth do no more than tell us what is going on.

    6. Malocclusion in ancient and primitive peoples is comparatively rare, but in modern westernized humans disturbingly common. I think the reasons are obvious and that is why orthodontists will never be jobless!

  16. Dear Kevin,
    This last post is a classic example of the thought crime that is licensed in perpetuity when the most respected academics say “we just don’t know,” when, in fact, we do know as well as we ever will.

  17. Interesting point about “if we are given 32 teeth, we are meant to have 32 teeth”…seems very reminiscent of the (now discredited) position assumed by E.H. Angle. Now, let’s assume for a minute that we can indeed “fit” all the teeth into the jaw. At what cost? Severe detriment to facial balance, esthetics, stability, TMJ and periodontal health? How is that ever an acceptable treatment objective? In the absence of well-defined treatment goals and outcome measures, I suppose any treatment plan (however illogical) is plausible.
    Also, what about the 3rd molars? Is it acceptable to extract them? The diminished jaw size has occurred as an evolutionary adaptation owing to the change in diet, and the significant reduction in masticatory muscle size. These attempts to revive Zinjanthropus have failed numerous times (as noted by Tweed and others). Orthodontic extractions should not be erroneously conflated with inappropriate diagnosis or poor mechanics, both of which can lead to detriment in facial esthetics. The attached paper below provides useful information on changes in profile in cases of extraction vs. no treatment.
    The proponents of “non-extraction at any cost” unfortunately are fueled by emotion and not evidence. I would like to see some data supporting some of these more interesting theories of growth and development proposed by our erstwhile colleagues.

    Rathod AB, Araujo E, Vaden JL, Behrents RG, Oliver DR. Extraction vs no treatment: Long-term facial profile changes. Am J Orthod Dentofacial Orthop. 2015 May;147(5):596-603

  18. One additional point – a change in profile is not always undesirable and in several situations is a treatment goal. So, extractions CAN change the profile, but it would depend on several factors including the anchorage and mechanics set-up, degree of crowding/protrusion, lip thickness/tonicity etc.

  19. What is your opinion on biobloc orthotropics? He maitains teeth extractions changes face profile and he works on restoring the extracted teeth back.

Leave a Reply

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