February 08, 2016

What do we know about orthodontic retention? A new Cochrane Review tells us something?

Can a new Cochrane review help us with deciding on the best orthodontic retention?

One of the most challenging areas of orthodontic care is deciding on how we are going to retain the great treatment that we have done. I also find that most of my patients ask many questions on retention. Can the  answer to our problems be found in this new Cochrane Review on orthodontic retention?

Retention procedures for stabilising tooth position after treatment with orthodontic braces.

Littlewood S, Millet D, Doubleday B, Bearn D, Worthington H

Cochrane Database for Systematic Reviews.  DOI: 10.1002/14651858.CD002283.pub4

I have discussed retention before on this blog and in my last posting I asked whether Simon Littlewood would stop going round the world lecturing and get round to updating his systematic review. I am pleased to see that he and the rest of the team, based in the North of England and Scotland,  have found some time and  updated their Cochrane review, which was originally published 8 years ago in 2006. This is a picture of the North of England, it is bleak up North!

What did they do?

They carried out this review to evaluate the effects of different retention strategies used to retain orthodontic treatment. As with all Cochrane reviews they only included randomised trials. The other inclusion criteria were

Participants: Adults and Children

Intervention: Any orthodontic retainer at the end of all orthodontic treatment

Outcome measures: Any measure of tooth irregularity

They carried out an extensive search of the literature using standard Cochrane methods.

Two authors selected the papers, data was extracted onto a customised data form and assessment of bias was carried out with the Cochrane risk of bias tool.

What did they find?

They initially found 441 articles. They then filtered them for relevance and quality to leave a final sample of 13 papers. They provided a lot of data on each trial. These studies compared different types of retainers including vacuum formed, bonded and Hawley retainers.

They decided that the most important outcome to report was Little’s Irregulariy Index.

When they looked at risk of bias they found that four studies were at low risk and ten had a high risk of bias.

This was a complex review with many comparisons and interventions being included and I have concentrated on those that I think are most important to me.

Lower fixed retainers vs lower thermoplastic.

They found an overall 4% failure rate and there was some evidence that bonded retainers were slightly better at reducing relapse than thermoplastic retainers. There was also some evidence that patients were more satisfied with bonded than thermoplastic retainers.

Thermoplastic v Hawley retainers

There was limited evidence to suggest that thermoplastic retainers were better than Hawley retainers and patients were happier with thermoplastic retainers.

Full vs Part-time wear

There was some evidence that part-time wear was as effective as full-time wear.

What did I think?

This is a Cochrane review and the findings are of high quality. One of their conclusions was that most of the studies they included were of high risk of bias. However, one of the concerns that I have with the Cochrane risk of bias assessment is that it is very unforgiving and when I looked at the reason for the classification of some of the studies, I felt that this was rather harsh. As a result, it is important to evaluate the risk of bias tables in these reviews and come to your own decisions on whether findings are going to influence your practice.

So, how does this review influence my practice. In previous postings I have emphasised that when we evaluate a study we need to look closely at the treatment effect. I have looked closely at the differences that they calculated for this review and all these are very small. As a result, I cannot help feeling that when we look at the effectiveness of the different regimes they all seem to “work”.  It is also important to point out that the most important factor in retention is patient co-operation and I presume that the trials included this as an outcome, but I could not find any detailed information.

My next step was to look at patient preferences and acceptability The analysis showed that patients tended to prefer bonded retainers. So at this point I am beginning to think that bonded retainers are the best? But I really do not like them because of hassle with failures. I was, therefore, surprised to see that they reported low failure rates.

I was in a dilemma and I wondered how I could practice evidence based orthodontic retention. But I got round this by considering that evidence based care is built around a combination of scientific evidence, clinical experience and patient opinion. If we factor this in, I can conclude the following:

  • The research evidence shows that it all works.
  • My clinical experience of bonded retainers is not good, I get too many debonds.
  • Patients like thermoplastic retainers and they can be worn part time.

As a result, I am going to stick with thermoplastic retainers which are worn at night only. Or…have I dodged the issue?



Related Posts

Have your say!

  1. Hi Kevin
    i use bonded retainers almost exclusively in all my patients; NHS and private, using essix / VFR when needing to retain posterior corrections eg expansion. We audited our failure rate last year and out of over 240 bonded retainers placed in a 6/12 period our failure (ie partial debond, our total loss of retainer occured 3 times in this sample) was 23% in first 3 months, dropping very rapidly to less than 2% by 12 months. My own bonded retainer has lasted 25years. To me the trick for long term success with bonded retainers is tooth preparation, flowable restorative composite using a lab formed wildcat wire retainer. see http://www.windmillorthodontics.com/for-dentists.htm for how we fit DBRs – less than 5 minutes chairside time
    best wishes and keep up the great blog

  2. It is always reassuring for me, a simple little g.p. who has a great passion for orthodontics, to see that I think like you do. From what I understand, the ” what kind of retainer is best” issue is a bit like the “which braces are better” issue. Everybody will say that what works best is what they use. Some other will say that what is best is what they sell. But personnal preferences (dentist and patient) and clinical experience is probably what we all should trust in taking our decisions. Everytime this kind of questionning arises, I always have the voice of my master Gerry Samson saying: “There is not only one good way to do things, otherwise we would all be doing it!”. Thanks again Dr O’Brien for keeping us instructed and entertained.

  3. J’apprécie énormement vos chroniques qui suscitent beaucoup d’interets dans ma pratique orthodontiques quotidiennes.Bravo et merci.

  4. Dear Kevin
    Thank you for your ever excellent observations. You state in your introduction “I am pleased to see that he and the rest of the team, based in the North of England and Scotland, have found some time and updated their Cochrane review, which was originally published 8 years ago in 2006”.
    As a matter of interest, however, is Cork (Professor Millett’s base) in the north of England or is it in Scotland??
    Keep up the great work.
    Kind Regards

    • Hi Jonathon, thanks for spotting the typo! However, Declan did used to work in Scotland and Cork is a shirt flight to Manchester…

  5. Thank you for another nice review. A small point that is regularily overlooked in “Littlewood” papers is the cost to patients. Within the hospital setting where most papers originate there is often no cost to the patient for a replacement or repair. In NHS specialist practice there is no cost to the patient for repair BUT there is £62.50 cost for replacement of one retainer. Various rates of replacement for Vacuum formed retainers are reported some them surprisingly high – This is because it is generally not possible to repair a broken VFR. On the other hand it is often possible to repair a broken Hawley retainer. You can immediately see that in NHS specialist practice patients are likely to be financially better off having Hawley retainers and this may be a factor in deciding which type of retainer to provide.
    (Unfortunately as the cost of repairs is down to the practice not the patient there is a pressure on providers to go more for replacements not repairs and thus choose VFR’s. This is not in the patients best interest)
    It would be nice to see the patients costs factor taken into account when papers compare different types of retainer for use in NHS.

    • Yes, you are correct and I hope that as orthodontic trials improve their methodology that cost effectiveness is factored in.

    • Very good point and also one I suspect which drives the type of treatment selected for patients. Eg extraction vs non extraction. Certainly I see a significant shift pattern away from seemingly a standard 4 x four extraction treatment to a non extraction bias amongst private practitioners in our area. Is this because the practitioners have suddenly seen the light, learned new techniques, feel the treatment is better, or patients demand this treatment when they have the choice because they are paying?
      Of course this is a phenomenon that has also long been observed in general practice, where treatment modalities are adapted to the fee scale. How many pregnant nursing mothers ended up with 3 anterior crowns in the 80’s for example or how many more bite plane appliances worth 12 UDA’s are made nowadays to fit in with the system. For the benefit of the patient or the practitioner?

  6. Regarding complications with fixed lingual retainers Dr. Tim Shaughnessy has published a remarkable set of case observations “Inadvertent tooth movement with fixed lingual
    retainers” in Am J Orthod Dentofacial Orthop 2016;149:277-86
    Considering this publication the need for patient/doctor informed consent is evident.

    • Thanks for this link, it is very useful and I may add this to another blog posting later on this month, if I get time. It is the unwanted movement of teeth in lingual retainers which causes me a load of hassle and I should have mentioned it in my post.

  7. Hello – I’m a first-time “commenter”, and a little nervous! I have followed with relish your blog, being a fan since I first met you as an ortho resident in about 1993 in Australia. Thank-you so much for your courage, wisdom and dedication.
    I look forward to reading the article, as the whole topic of retention, I feel is handled quite poorly by our profession and in particular our specialty. It staggers me that not only most of the public but even some clinicians still believe that we have the ability to “stabilize” the healthy dentition. Considering no tissue in our living bodies is stable, in fact the entire universe is continually in flux, this is a tall order! (yes, I know many orthos consider themselves God like, but…) If we accept that this is unreasonable , then I am supposing that we must consider some kind of life-time retention if we intend to maintain our treatment results. This requirement would likely favour a fixed type of retainer, however I’m with you Kevin, I don’t prefer them and only place them as a second option, based on individual preference and initial malocclusion (spacing mainly). I believe however, that I came to this clinical decision for slightly differing reasons compared to those of O’Brien’s. I too dislike the breakage issue, and even if we minimize it with great techniques as suggested by the first commenter, Ross , I find that when they do break it is often insidious. The patient may not realize as only 1 or 2 teeth may come away. By the time the patient or clinician recognizes movement, it is too late – however with a removable, the patient knows immediately if lost or broken so action may be taken prior to significant movement taking place – if patients are educated well.
    So this brings me to the reason that I was prompted to blog. I hypothesize that the pattern of relapse occurring after a broken fixed compared to a lost removable is different. My thought process for better or worse: Fixed retainers rarely totally fail in my experience, so that 1 or 2 teeth move, often labially slightly, rarely do they cross as the other teeth are still being held so they cannot. Also most patients with bonded do not have posterior arch width maintenance, so this dimension has been “used to” intra-oral forces and will not change as rapidly when compared to loss of a removable – often Hawley or Essix type. When such appliances are lost or broken, the entire arch form is suddenly vulnerable to intra-oral forces, genetic, inter-stellar etc. A change in arch width and lack of total control over anterior segment here , I feel, is much more likely to produce overlap of incisors within smaller circumference rather than a labial movement of 1 or few teeth being held in given circumference, or even the whole anterior segment that has not had posterior width held prior.
    So all was going well for me in Kevin’s review until he mentioned that the authors “decided” that the most important outcome was based on Little’s irregularity index. Way back in the day when I was studying occlusal indices, this index measured the deviation of contact point to contact point, so that if there was no “overlap”, and simply butterfly imbrication pattern, a score of 0 still applied. Possibly things have changed, not sure how they used the index. I am a fan of objective measuring, but we are only as good as our ruler, so I am wondering if the Index used was possibly biased towards the “crowding” pattern resulting from failed bonded, rather than lost or broken removable? I look forward to seeing if this is addressed within the article.
    My next query is whether the type of appliance that patients were treated with was taken into account – not because 1 is superior to the other (as Stephane points out well), but because I feel that patients treated with a removable type of appliance tolerate removable type retainers better than those treated with fixed appliances, simply because they have already been complying for years. When you offer a removable to a newly debonded fixed patient, they require new skill sets; whereas the removable appliance patient already has that skill set, and I feel tolerate a removable retainer better.
    As to the outcome revealing patients preferring bonded over removable, I personally feel that is difficult to assess (as is all) – have the sample used both, what was the information provided? The outcome to me rather points to the preference of the clinicians, rather than the patients. That most clinicians preferred bonded, within their sample. We can easily influence our patients preference, I do almost every day as I have patients enquiring about a bonded wire. They usually believe that this is a 1 stop answer to retention. When I expose my bias and explain retention as a long-term insurance policy, that bonded wires also require repair and maintenance, they do not hold posterior arch form and may be detrimental to periodontal health long-term; patients begin to consider a nightly removable, used not only to hold the entire artificial position we have created, (no matter if we maintain original arch form, canine width etc), but may also protect against clenching and bruxing effects, and may be used to carry dental medicaments if indicated, or adjusted if required in case of a Hawley type- they often change their mind about a wire. So I guess I am asking if the finding indicates more the original appliance type and the clinician preference rather than the patient preference and performance in retainer type?

    Concluding thoughts are, that when such a review is made, in my mind, one of the major variables influencing the performance of a retainer may actually be the performance of the patient, biased by type of treatment appliance used as well as the education provided to them; and the second is whether the measurement tool in itself may be biased towards 1 type of retainer?

    • Thanks for the great comment and no one should be nervous about posting. I monitor the comments very closely and keep out the aggressive ones etc! These are all really good points that you have made and clinically they make a lot of sense. I think that one of the other reasons that I stopped using bonded retainers was the unwanted tooth movement.
      You are also right about Little’s index and I think that I was a little unfair to the authors when I wrote that they chose to report this. It seems that this is the most popular outcome measure that was used in the trials and so they stuck with this. I think that one advantage of using such a coarse measure is that when a difference is detected it is likely to be real rather than minor.
      Finally, I agree with you totally about the greatest factor in relapse in retention being the patient. I would hope that this was measured in the trials, but I could not find anything in the papers that I looked at. If a patient does not co-operate with their retainer they should still be kept in the trial in an intention to treat design, and this would be reflected in the results.
      Again, thanks for the great and very helpful comments

  8. Bonded retainers are a problem for NHS practitioners, especially in the salaried sector. It is not possible to discharge a patient as they are liable to return, sometimes years later if there is a problem. GD Practitioners understandably, will not maintain them.

  9. Kevin,

    This is an intriguing report of the Cochrane review for a particular reason; while it covers all the significant features of retention appliances and relapse from a purely dental perspective, I believe some fundamental factors are omitted.

    You state that “the most important factor is patient co-operation” implying that failed retention and relapse are the fault of the patient. I am not sure I agree; vital though co-operation obviously is, relapse is much more complex than that and hangs on a number of factors, one the most significant being the timeline that is absent from both the review and the report.

    How long were these patients in retention and how soon after retention was discontinued was Little’s Irregularity Index finally applied? A year? 5 years? 20 years? To answer your question, I would say, yes, you may have dodged the issue!

    I would have thought that without this timeline information the results are of limited value.

    On the subject of bonded retainers constantly requiring repair, the osteopathic view is that this may be due to natural bony flexing, mainly of the mental symphysis and the palatal midline suture.

    But the bottom line is that orthodontists have not yet discovered or understood the causes of malocclusion, though it is clear to many that irregular teeth are simply a signpost pointing to the actual problem; they are a symptom and not the problem itself. Until this is accepted and understood, the unwelcome spectre of relapse will always be with us.

  10. Kevin,
    Good comments thus far, seems time for a US thought, something other than boring presidential politics. I do think that you side stepped the issue a bit in that the matter of time of retention was not adequately considered. If we agree that it is difficult to place teeth where they will remain stable without any retention then, thinking of what will work best for the patient, we have a couple of major obligations: first to structure our treatment such that it places the teeth in a position that best leads to long term stability (see Jim Boley and Jim Vaden articles and presentations following Tweed teachings) and second to provide our patients with the retainer(s) that, being aware of human fallibility, gives them the best chance of maintaining long term what we orthodontists have provided and as suggested by Bob LIttle that being a fixed lower retainer (can’t be left at grandmas). This is where the time issue raises its head. Bonded retainers, if placed properly, can remain in place for several years even if the patient moves cross country and in the case of the lower bonder canine to canine retainer, it greatly assists, if proper overjet has been achieved, in maintaining the upper alignment. A lost or broken upper retainer when night time only wear is prescribed can easily and relatively inexpensively be replaced with a thermoplastic substitute. But the key and what I teach my residents is to give the patient the best chance of long term stability and for that it is hard to beat the fixed lower retainer especially for our American teenagers.

  11. Kevin,

    Thank you again for another excellent and thought-provoking post. I am a first time contributor to your blog, so pardon me if I am less ‘scientific’ than is probably required.
    For my part as a humble GP with a specialist interest in orthodontics for over 20 years, I have changed and re-changed my orthodontic retention regime several times and have now settled on the following based on my own experience ( non-scientific ) and listening to lecturers for many years.
    I have just finished four years attending Dr. Derek Mahony’s Mini-Residency in London and his advice was that for stable long-term retention that bonded retainers were the way to go. Teeth just inevitably move over time and if patients wanted a copper-fastened result, bonded was indicated. What I now do is this – bonded lingual retainers for life, vacuum formed retainers for daytime wear over the top of the bonded as these are ‘socially’ more acceptable for the patient, and then Hawley retainers for long-term,night-time wear to allow some ‘occlusal settling’ which I believe vacuum formed do not allow. I suggest to my patients that retention is a lifetime deal, and that they should continue to wear their Hawleys perhaps once or twice a week in bed to maintain arch form. Fortunately to date my debond rate with the retainers is very low, and we give a two year guarantee on this component of the retention. Any breakages after that incur a fee and patients are advised to attend ASAP if a debunk does occur. As I say Kevin, not entirely scientific, but seems to be a regime which semi-agrees with you and the Cochrane review.

    • Just to add to the discussion, Simon Littlewood mentions a Scandinavian orthodontist who does not use any retainers and he had to admit that this orthodontists results appeared to remain surprisingly stable. The only explanation he had was that it may have been the type of cases treated, IE Class I mild to moderate crowding. He certainly felt further investigation was warrented.

  12. Dear Prof – I came across the post when thinking how best to manage retention long term. I notice the preference and evidence supporting bonded retention. The challenge in the real world even within a private practice like my own is the cost effectiveness for both practice and patient. Bonded retainers fail and follow your patients long enough and this reality can be depressing not least because with time compliance with annual Inspection drops. As pointed out few dentists will want to take on the responsibility either. In the short term success rates maybe high but 5-10 years down the line things fail. The paper referenced above about inadvertent movement is very valid as this occurs as the wire can become active through distortion or fracture. With the added risk of adverse incidents – following on from a BOS report of ingestion we probably do need to think about the effectiveness (clinically and cost wise) as well as risks long term. The advent of cloud based digital model storage and ease of replication from this with vfrs may make for a practical solution? For both patient and orthodontic practice ? Yours still looking for the solution

Leave a Reply

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