I was interested to read the article by Grünheid et al in the February 2014 issue. The authors aimed to assess the accuracy and reproducibility of dental measurements on 3 types of digital models. As they pointed out, emodels (GeoDigm, Falcon Heights, Minn), SureSmile models (OraMetrix, Richardson, Tex), and AnatoModels (Anatomage, San Jose, Calif) were made for 30 patients. Mesiodistal tooth-width measurements taken on these digital models were timed and compared with those made on the corresponding plaster models, which were used as the gold standard. Accuracy and reproducibility were assessed with the Bland-Altman method.
It is important to mention that reliability (precision) and validity (accuracy) are 2 completely different methodologic issues that are assessed by different statistical tests and should not be confused with each other. Why did the authors not use the well-known intraclass correlation coefficient (agreement type and not consistency) to assess reliability? For validity analysis, we need at least 80 samples comparing our measurements with the gold standard using appropriate tests, depending on type of the variable; however, for reliability (interobserver or intraobserver), around 25 to 30 samples will be enough.
The authors reported that measurements on the SureSmile models were the most accurate, followed by those on emodels and AnatoModels. Measurements taken on the SureSmile models were also the most reproducible. Scientifically, it is not acceptable to confuse validity and reliability by using 1 statistical test; moreover, statistically significant results have nothing to do with the clinical importance of the findings. They also concluded that tooth-width measurements on the digital models can be as accurate as, and might be more reproducible than, those taken on plaster models. Such misinterpretations and misconceptions should be avoided; otherwise, we will face mismanagement of the patients in our routine clinical care.