I read with great interest the study of S ugar et al . Their aim was to compare the outcomes following fixation of mandibular angle fractures with a combined transbuccal and intra-oral technique, with the standard intra-oral technique. Although the study question is interesting, there are several limitations in this study.
The study is a randomized controlled trial (RCT); however, important quality aspects are absent or not reported. There are no power calculations for the required sample size. In medical fields where outcomes are soft and susceptible to misinterpretation, small studies are often underpowered to detect differences between the compared arms . Unfortunately, this is not unusual in RCTs from our specialty . Furthermore, in the current study, allocation concealment is not ensured and there is no mention about blinding. But, the most striking finding is the fact that randomization was mixed up. As the authors admit: “The procedures were coded A and B for the randomization process, which caused some confusion… As a result, some patients were allocated to the wrong groups .” However, they try to explain that this hasn’t influenced the study, since “the intervention and control groups were reasonably matched across a range of characteristics” . This is a common misunderstanding about the usefulness of randomization. Randomization uses the play of chance to assign participants to the intervention groups. The unpredictability of this process, if not subverted, should prevent systematic differences between the groups, provided that sufficient number of people is randomized . Thus, it helps to minimize differences that may arise from unknown confounders . In a properly designed RCT, baseline characteristics may differ (because of pure luck), but this wouldn’t reduce the quality of the study or influence the final results. I am afraid this is not the case in the current study; the randomization sequel has been broken and the accuracy of the results can know be questioned. This “ little unbalance ” between the compared groups, can be responsible for the marginal difference observed.
Furthermore, there is another important aspect that needs clarification. According to Table 1, five patients have been excluded due to unstable fractures requiring 2 osteosynthesis plates. How this was guessed prior the operation? I assume that the patients were initially enrolled. Instability was then discovered intra-operatively, they had 2 plates’ fixation and then excluded. If I am correct, these patients shouldn’t have been excluded–they should have been included in an intention to treat analysis. Or, on the other hand, if this was a reason for exclusion, it should have been clearly described in the inclusion/exclusion criteria. Additionally, there is not a single mention about NNT (number needed to treat – i.e. how many patients do we have to treat to prevent one complication), and this is another weakness of the study.
The only statistically significant difference observed in the current study is the infection rate which seems to be higher in the intra-oral approach. However, the difference is only 3 patients, and this could easily be the result of the significant biases described above. I wouldn’t expect the two approaches to have different infection rates; both of them employ intraoral incision and contamination with the oral flora. I therefore believe that conclusions in the current study need to be drawn with great caution.
A very interesting finding in this study is the surgeons’ preference towards the transbuccal approach. The authors are to be congratulated for including this questionnaire into the study, because it highlights this particular trend. Therefore, despite the limitations, I believe that the current report adds to the current body of evidence regarding mandibular angle fractures. Debate is essential for improving the quality of this literature.