Article Text
Statistics from Altmetric.com
We read with great interest the article ‘The prognostic value of routinely performed minor salivary gland assessments in primary Sjogren’s syndrome’ by Risselada et al.1
One of the main findings of the paper is that lymphocytic focus score (LFS) ≥3 contributes significantly and independently to non-Hodgkin lymphoma (NHL) development in a standard multiple regression model (β=0.244; p=0.017).
Owing to the important practical consequences of this finding, we would like to comment on the methods used in the paper. In the current analysis, the risk of NHL is assessed as a dichotomous variable (NHL during follow- up, yes/no), which does not take into account the follow-up time (range: 10–408 months). Because the risk of NHL is time-dependent,2 differences in follow-up durations may account for apparent differences in the risk of developing NHL between patients with shorter and longer follow-up. In this context, we believe that a survival analysis taking into account follow-up time and censored individuals should be preferred.
Additionally, we would be pleased to know more about the statistical model used in the paper. Based on data presentation, we believe the authors have used a multiple linear regression model, which is not appropriate for predicting a categorical variable (occurrence of NHL, yes/no) based on another categorical variable (LFS ≥3, yes/no) and may therefore lead to inadequate estimates.
Footnotes
-
Competing interests None.
-
Provenance and peer review Not commissioned; externally peer reviewed.