[Equivalence and agreement in validation studies: A practical methodological review]

Accurate statistical analysis is essential in validation studies of instruments that quantify continuous variables against a reference standard. This article describes statistical approaches to evaluate equivalence between measurement instruments combining graphical methods and statistical tests. Its application is exemplified through a study that assessed the accuracy of a physical activity tracking wristband (Xiaomi Mi Band 4) for counting steps walked during different activities in patients with chronic respiratory diseases, and it was compared with a video-based reference method. Confidence intervals were used alongside predefined equivalence zones, TOST (two one-sided tests) procedures were applied, and both group-level and individual-level indicators of agreement were calculated, such as the mean error (ME), mean percentage error (MPE), mean absolute percentage error (MAPE), and root mean squared error (RMSE). In addition, some common errors were also discussed, such as the inappropriate use of scatter plots or correlations to assess accuracy. The article concludes that selecting appropriate statistical methods is a key aspect to ensure clinical and methodological validity in equivalence studies between measurement instruments that quantify continuous variables and a reference method.