Reply to Drs Moran and Pedrera-Zamorano

Dear Editor,

I thank Drs Moran and Pedrera-Zamorano for their attention to our work. However, I would have expected more accuracy in their criticism, given that the letter evidently refers to our latest publication: Homeopathic therapy in paediatric atopic diseases: Short-and long-term results, while in the text they twice mention a previous study (2012): Homeopathy in paediatric atopic diseases: long-term results in children with atopic dermatitis.

Going to the heart of their criticisms, in table 5 it probably was not necessary to indicate the statistical significance of the data, since worsening of the prognosis for children with initial allergy was not in the original aims of the research, but only an incidental observation arising from our clinical practice. It does not refer at all to the effectiveness of homeopathic therapy. We chose the chi squared test (despite one cell with value under 5) because we decided to consider the variable ‘complete remission of symptoms’ not as dichotomous one, but as a continuous variable with two levels of measurement (complete remission and any other level of pathology).

The Wilcoxon test (used to determine the p-value in table 3) can be used for paired samples when a continuous variable (like the symptoms of patients or their state of health) is measured by rank (it is also indicated as first choice from the statistical package SPSS). In table 3 we preferred to place in the rows each initial level of the symptomatology because the use of only the two rows showing the simple distribution ‘pre-post’ of the dependent variable would make it impossible to identify the evolution from initial to final level (e.g. for allergic rhinitis the most resistant cases are the initially most severe ones, whereas for dermatitis the resistant cases originate also from mild or moderate initial levels).

Therefore, we reiterate the methodological correctness and accuracy of our choices.