An occasionally irregular blog about orthodontics

We have a new trial on orthodontic retention!

We have a new trial on orthodontic retention!

Orthodontic retention, we have a new trial!

In this post I am going to return to the subject of orthodontic retention and review a new study on retention methods.  This is an update on a previous post.  With the addition of this paper I think that we are beginning to build on our knowledge of retention.  Maybe it is, also, time for Simon Littlewood to stop jetting around the world and get his team to update their Cochrane review on retention?

This paper is published in the EJO and the full text is denied to most people because it is not an open access journal.

Unknown-3Five year post-retention outcomes of three retention methods-a randomised controlled trial.

Gudrun Tynelius and Harvigland Stubbarp

DOI: 10.1093/ejo/cju063

What did they do?

In this paper they presented the results of a long-term study of retention and I think that this new publication goes some way to address the criticism that most other studies have been concerned with short term outcomes.

This was a single centre RCT carried out by one team.  The sample of patients were essentially Class I crowding cases requiring four pre molar extractions and straight wire fixed appliance treatment.  This could be described as straightforward traditional orthodontic treatment based on sound mechanics and philosophy!

When the treatment was completed the patients were randomly allocated to one of three retention regimes, these were

  • Upper vacuum formed retainer (VFR) and a lower canine to canine bonded retainer
  • Upper VFR and stripping of lower incisors with no retainer
  • A positioner.

The removable retainers were worn mostly at nights in the first year and then every alternate night for the remaining 12 months of a two year retention regime.

Dental casts were measured at the start of treatment, at debond, at the end of retention at two years and then five or more years out of retention.

They measured several dental variables but I am only going to mention the findings for Little’s Irregularity Index.

Randomisation was done by a ballot system; there was good blinding of data recording and analysis. They carried out a sample size calculation.

75 patients were randomised and it was important to find that 26 failed to complete the study.

What did they find?

They found that there were no significant differences in Little’s Index at the start, debond or at the end of retention.  After five years out of retention there was some relapse but this was not marked (1.8-2.6mm).  There were no differences between the groups at any point of data collection.

What did I think?

I felt that this was a good, well carried out and reported study that adds to knowledge. However, as with most studies we do need to interpret the results carefully.  I think that there are two important factors that we need to consider.

Firstly, there was a significant “drop out” rate. The authors acknowledge this, but state that this did not result in a lack of statistical power. While this is a good point, the high number of patients who did not complete the study does introduce some bias.  We could argue that this is a fact of life for all long-term orthodontic studies, as many of our patients move away from our clinics. Nevertheless, it does introduce a degree of uncertainty because we know nothing about the patients who did not return for final records.

The other important factor that we need to consider is that the patients were Class I crowding cases treated to a high standard with extractions.  Consequently, the findings may only be relevant to this type of case and treatment. I am not sure that we can extrapolate this to current trend in non-extraction treatment based around arch expansion, proclination of lower incisors, short-term orthodontics, “super fast”  and other compromise treatment …

Overall, I thought that this was a good well-conducted study that will reinforce the practice of many operators. It provides further evidence that VFRs are effective.  I am less convinced about carrying out stripping, it may be a form of preventing relapse, but to me it is bit invasive. I am not sure why, I just don’t like it!

I will continue with my current VFR night only retention for my patients and this study supports my practice.  But what do I do about the expansions and proclinations that are resulting from the current non-extraction phase of my treatment provision?
Edman Tynelius, G., Petren, S., Bondemark, L., & Lilja-Karlander, E. (2014). Five-year postretention outcomes of three retention methods–a randomized controlled trial The European Journal of Orthodontics DOI: 10.1093/ejo/cju063

(Visited 1,313 times, 6 visits today)

Tags: , , , ,

There Are 11 Comments

Trackback URL | Comments RSS Feed

  1. Simon Littlewood says:

    Kevin, you will be delighted to know that the Cochrane Retention review update is underway! You are quite right that it would be very timely to update this now, particularly with the publication of high quality studies like the one led by Gudrun.

  2. Jonathan says:

    Surely the authors didn’t factor in for a >33% dropout. Is it even possible that if more than 1/3 of the patients drop out of a study, this cannot have an effect on the ‘statistical power’?

    Can’t we at least know the group split of the patients who dropped out for us to draw our own conclusions?

    Was the lower bonded retainer removed at ‘end of retention’ or did the bonded retainer continue, past the ‘end of retention’ period . . . for this group only?

    • Kevin O'Brien says:

      Thanks, in their power calculation they built in a projected drop out rate and they enrolled more patients than they needed to give sufficient power. While the high drop out rate does not influence the power it does introduce bias, as we know nothing about the patients who dropped out.

      They did include data on the patient flow through the study in the form of a CONSORT diagram and this was clear.

      The lower bonded retainer was removed at the end of two years according to the paper.

      I hope that I have answered your questions

      • Fang Hua says:

        Professor, yes the exact number of dropped out patients in each group could be found in Figure 2.

        But the CONSORT for Abstract guideline (Hopewell, 2008) did recommend reporting both the number of paticipants randomised to each group and number of participants analysed in each group. I think this information would be easier to find if reported in the abstract, especially for those outside of the paywall.

        I found that the authors said in their abstract that “49 patients were randomly assigned to …”. I think this might be a little misleading as actually 75 patients were initially randomised while 49 was the overall number of people analysed.

        • Kevin O'Brien says:

          Hi, yes, I agree that the abstract should contain as much information as possible, particularly if there is a paywall that stops people reading the paper.

  3. Chris Kettler says:

    Forty years of specialist practice have taught me:
    1. Lower retainers are necessary. At one time the NHS did not fund lower retainers in Practice.
    2. Removable retainers can be fitted a week after debond in all cases.
    3. Removable retainers do not need to be worn full time ever. They can be worn just at home for one week and then at night only. After twelve months they only need to be worn enough to keep them fitting.
    4. Hawley retainers are superior. Patients who have tried both, prefer Hawleys. Hawleys can be repaired, Hawleys that do not fit because they have not been worn enough, can often be adjusted.
    5. I believe the current vogue for Essix retainers is because they are easy to make, easy to fit and cheap. The occlusal coverage limits occlusal settling.
    6. Bonded retainers inevitably break and are not a substitute for removable retainers. They must be backed up with a removable retainer which is worn enough to keep it fitting against the day the fixed retainer inevitably breaks.
    7. Bonded retainers should be fitted on the visit prior to debond. A poor bonding technique will be revealed at the next visit before you debond. The retainer impression is taken over the bonded retainer and the removable retainer conforms to it.
    8. Relapse in unpredictable. Many patients do not mind a minor relapse. The more minor the original malocclusion, the more the patient complains about minor relapse. Lesson: do not treat minor malocclusions, let your other local orthodontic practice have the grief.

  4. Philip Benson says:


    Some wise words from an experienced clinician – I agree with nearly all of this, except the assertion that patients prefer Hawley retainers to vacuum-formed retainers. Is there any evidence for this? I use Hawley retainers in my patients with hypodontia and vacuum-formed retainers, part time from the start, for nearly everything else. It just shows the need to have proper patient-reported outcomes in all trials.

    There is also the issue of cost-effectiveness. Our lab can produce two vacuum-formed retainers in about an hour (mostly waiting for the plaster to set) for about half the cost of one Hawley retainer without this impacting on their other work. We usually fit the retainer the same day the appliance is removed.

  5. I read with care the data in the publication and have several comments: 1. In my country, minor changes in the position of the front teeth are not acceptable by the patients therefore most of the orthodontists use lower as well as upper fixed retainers.
    2. Most of the complaints are those of space opening between the incisors, tiny rotations or minute buccal or palatal movements of the laterals, (which neither the Hawley not the VFR can hold. Teeth move in those 2 appliances slowly and surely. If the laterals were initially positioned palataly, they tends to go back, until they meet the lower incisors. It almost does not matter what you do, those tiny changes brings the patients back and forth to the clinic. Rotated upper cuspids will always rotate backwards etc, Therefore, fixed retainers as well as (I mean with) upper Hawley or VFR are almost mandatory. In my 30 years of being professional I saw so many changes even with fixed retainers that nothing is a surprise to me (open spaces, rotations, intrusions and extrusions of upper centrals and much more). I do involve in many court cases (defend the doctors and the profession) related to changes following treatment and the things I saw are incredible. I say: we can almost always bring teeth to the place we want them to be, but we cannot hold them there (forever)!

  6. Chas Lister says:

    An interesting paper. As Dr Kettle suggests, my protocols have evolved over 24 years of patient funded practice. As suggested by Dr Brezniak, relapse is unwanted by patients and parents.

    I generally place a fixed retainer to lower 3×3, but this has to be a robust wire, because I want to retain until late lower third facial growth has completed. So until early to mid twenties for males, likely earlier for females.

    The upper will be a Hawley, for reason of maintaining arch form through the late teen – early twenties period of facial maturity. If rotations were present, a fixed upper 3×3 is placed inside the Hawley. Worn full time for 6 months, night only 6 months, then 2-4 nights per week. Compliance is the critical element!

    VFR have their place but do interfere with the occlusion. They seem to promote night parafunction, and the posterior-mainly contact places load on the maturing TMJ structures.

    The Hawley allows the lower posterior occlusion to Self-establish against the uppers.

    This paper while interesting is also limited. It seems to confirm that all retention works reasonably well, and is not overly dependent upon methodology. It seems to confirm minor relapse will happen if compliance slips. It also depends how far apart one sets the goal posts. A 1.8 – 2.6mm Little index could be construed as overall retention failure after 5 years. We would not accept such an outcome in restorative dentistry for example. A chipping composite at 5 years is a failure; clinical significance is a different matter to the cold evaluation of the technical procedure.

    The fact remains: if the teeth move within 5 years, the retention has failed.

    As always therefore, it is down to the clinician and his or her control of the post treatment environment. If you take proper account of the pace of late lower third facial growth, and the establishment of adult parafunction habits, if any, and set a high expectation of no relapse by say age 21 female / 23 male, a clinician can engineer effective retention with ease.

    This paper seems to suggest how you do that is less critical than the period over which it is undertaken. Managing compliance, as always, may be more important than choice of retainer appliance.

  7. I agree with you all. After 30 years of practice, I choose upper and lower fixed retainers (the lower is rigid) and two rigid splints thermoformed but without occlusal resin (to settle the occlusion) and linked together by a metal W system to avoid sagittal relapse (PUL W retainer). It has high compliance and is worn every night for 12 months and one night less every 3 months, so minimum 2 years.
    You can see photos on these two links:

  8. Wise words by clinicians, most of whom have greater experience than my decade of working in both specialist practice and hospital.
    A couple of points:
    1) I agree with Phil Benson, that in hypodontia cases, there is great value in fitting a retainer on the same day. I normally leave a sectional fixed on the two teeth either side of the site of the absent tooth/teeth until the Hawley retainer is ready. This scenario and generalised spacing cases aside, I agree with Chris Kettler, that most other cases can wait a week for removable retainers to be fitted.
    With adult cases, I routinely use fixed retention, as I find these patients will look for an excuse to not wear a removable retainer, but this does come with its perils of breakages etc.
    2) I have become an advocate of VFR’s in the last 3-4 years, and in general I do think compliance is better. I also think they provide a better guide for detecting early minimal relapse at a 3 month check. Of interest, I see that upper laterals often intrude by 1mm or so during this period- whilst this is of little clinical relevance, this vertical change can’t be detected when using a Hawley retainer.

Post a Comment

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


Pin It on Pinterest