Scientific practice in medical research.

Abstract

The brief report by Kuitunen1 published in this issue of Acta Paediatrica raises the problem of misinterpretation of sampling uncertainty and misuse of p-values and confidence intervals in medical research. Scientific journals and societies worldwide have long emphasised the urgency of the matter and suggested possible remedies,2 but their message seems to have remained largely unheeded so far. The brief report by Kuitunen1 duly reinforces this message. In addition to outlining the problem, the report summarises the results of an extensive review of articles published in Acta Paediatrica in the year 2022. The review gauged the current practice in reporting and interpreting research findings, focusing on the use and interpretation of p-values and confidence intervals. The results of the investigation were congruent with what might be expected of high-quality medical journals. The brief report by Kuitunen1 also suggests possible changes to the current practice and encourages the use of confidence intervals instead of, or along with, p-values, arguing that the former summary measure better facilitates understanding of sampling uncertainty. The report tacitly implies that reading the information confidence intervals provide should not equate to merely checking whether they contain the value of no association, like, for example, whether a confidence interval for a relative risk contains the value one. The issues listed above and the many more that are not may form an initial basis for starting a comprehensive assessment of the limitations of the current practice that might lead to finding possible remedies, to which the brief report by Kuitunen1 would represent a relevant contribution. I end this editorial by excerpting the enlightening conclusions offered in the statement released by the American Statistical Society.2 ‘Good statistical practice, as an essential component of good scientific practice, emphasizes principles of good study design and conduct, a variety of numerical and graphical summaries of data, understanding of the phenomenon under study, interpretation of results in context, complete reporting and proper logical and quantitative understanding of what data summaries mean. No single index should substitute for scientific reasoning’. None.