What did they ask?
They did this study to find:
“The number of failures of direct and indirect bonded retainers at 5 years follow up and investigate the stability of tooth positions”.
What did they do?
They did a single centre 2 arm parallel randomised controlled trial with a 1:1 allocation.
The PICO was
Participants
Orthodontic patients with a fixed retainer bonded directly or indirectly between September 2012 and June 2013.
Interventions:
Fixed mandibular retainers made from 0.0215 multistrand wire bonded directly.
Control:
Identical retainer bonded indirectly.
Outcome:
The primary outcome was 5-year survival of the retainers. Secondary outcomes were inter-canine and inter-premolar distances.
This was a follow-up study to a previously reported trial. The authors had reported details of this study before. However, they outlined that they used pre-prepared randomisation and did allocation concealment by contacting a trial coordinator. They could not blind the operator or patient, but they collected the data blind.
They used Kaplan-Meier survival analysis to look at the retainer’s survival. Then they used a relevant multivariate analysis for the other data.
What did they find?
In any long-term study, it is essential to look closely at the dropout rate. This is because a high dropout rate can indicate potential bias, lead to loss of statistical power, and reduce the generality of the results. At the start of the study, 64 patients were randomised to have their retainers bonded directly or indirectly. At the end of the study, 52 participants remained (26 in each group). This meant that there was a dropout rate of 18.7%. I will come back to this later.
The most important finding was that 14 participants in each group had a bonding failure. This was a very high 54%. There were no differences between the groups. The five-year survival rate was 46%.
Interestingly, 83% of the failures occurred in the first year and 86% in the first two years. From then on, the failure rate dropped significantly.
They then looked at the dental relapse for the 24 failure-free patients. Again, there were no differences between the groups. Unfortunately, they did not report the deterioration of the patients whose retainers had failed.
The authors reported also reported on unexpected changes. Interestingly, at T2 they found that 5 participants showed an increased lingual crown torque of the mandibular left canine. However, in only one of these was the change felt to be severe.
Their overall conclusions were:
- The 5-year survival rate for the retainers was 46%
- Most failures occurred in the first year (83%).
- There was no marked change in intercanine and inter-premolar distances.
- There was no difference between the direct and indirectly bonded retainers.
What did I think?
This was a well carried out and nicely reported trial. The authors should be congratulated for completing this study with a long follow-up time. This was an outstanding achievement. While some may consider that the dropout rate was high, I feel this is good for a long-term orthodontic study, and it would be hard to do this any better. As a result, we should accept the results.
The results were both interesting and disappointing. While there were no differences in the failure rate between the techniques. The failure rate was very high. While we may all feel that our retainer success rate is higher than this. We need to remember that in this study, all the patients were accounted for, and we tend to be optimistic when we evaluate our own performance. There are no other comparable studies over this period. As a result, we should accept their findings.
However, while the number of failures may be concerning. We must consider that this trial was done in a post-graduate orthodontic clinic. As a result, we cannot ignore the effect of proficiency bias. It would be reasonable to suggest that a highly experienced operator should have a higher success rate.
Final thoughts
This trial provides helpful information on bonded retainers’ failure rates. Notably, most retainers fail in the first two years, which suggests that once a patient has worn a retainer for two years, we may not need to review them on a 12-month basis. Perhaps, we should also warn our patients that failure rates are high.
