Sheldon Friel Memorial Lecturer

EOS 2018 - Speaker - Fraser McDonald

Fraser McDonald

United Kingdom

Professor Fraser McDonald qualified from Birmingham University in 1980; he worked in general practice and junior hospital posts until he commenced postgraduate orthodontic training at the Royal Dental School in London under the direction of Professor W J B Houston. He completed specialist training in 1986 and was appointed as a lecturer in oral biology (orthodontics integrated with physiology). The work from his MSc project was awarded the European Orthodontic Society Research Prize. His PhD was completed in 1992, again under the supervision of Professor Houston. In 1999 he was appointed to a Chair in Oral Biology at King’s College London. He was appointed as Assistant Editor of the European Journal of Orthodontics in 1991 and full editor in 1993 until 2010. He was appointed Honorary Treasurer of the Society in 2012.

He has published over 120 papers and 160 abstracts together with 3 textbooks. He also holds the current role of Dental Examinations Convenor of the Royal College of Surgeons of Edinburgh.


Sheldon Friel Memorial Lecture

Growth and statistics

Many aspects of our research work require the use of statistics to plan and then to analyse the outcomes. This, as is common practice, involves analysis of the appropriate test numbers and the meaning/weight to be placed on the subsequent findings. Many clinicians believe that science and, in particular statistical examination, is a sharply demarcated and defined process. However statistics can be very varied and many statistical tests are used to develop the hidden meaning within the collected data. This presentation will focus on the value of statistics in reporting the data and the actual relevance of this to clinical practice, together with the necessary caution. For example, there is growing concern that P values do not report the ‘real’ value of outcomes. Additional statistics should include confidence intervals, Bayes factors, or effect sizes, together with other uncertainty measures in order to allow the strength of the study outcome to be evaluated (Chavalarias et al., 2016).

Statistical methodology will then be applied to one of the most important aspects of clinical orthodontics, 'growth'. As practising clinicians we are aware of the need to understand and to predict growth patterns and timing, to target certain treatments at periods of rapid growth and delay various types of treatment until growth is complete. We are also aware of the effect that growth may have on the stability of treatment outcomes. The main focus in areas of high patient expectation is the management of these within the biological framework and the basic principle that no individual patient is the absolute ‘average’. An analysis of a cohort of Australian patients with known dates of birth and extant lateral skull, dental panoramic and hand wrist radiographs, together with details of their standing height, has been carried out which shows the current uncertainty in growth prediction methodology.
D Chavalarias, J D Wallach, A Ho Ting Li, J P A Ioannidis (2016) Evolution of reporting P values in the biomedical literature, 1990-2015. Journal of the American Medical Association 315(11):1141-1148.