Letter to the Editor

Sir,

The study by Robertson et al is a fine example of nonsensical, misleading statistics.[ ¹ ] The authors demonstrate that, at certain time points after tonsillectomy, the Arnica-treated group experienced significantly less pain than the placebo group. Therefore, they draw an overall positive conclusion. If one uses their VAS data to draw a graph, one hardly needs a statistician to see that the pain experience in both groups did not relevantly differ ([Figure 1]). A statistician would have been useful, however, to adjust the results for major confounders: the marginal differences in favour of Arnica on three days may well have been caused by the high drop-out rate and/or the slightly higher intake of analgesics in the experimental group. Roberts et al conclude that “Arnica…provides a…significant decrease in pain…” I am not convinced!

Fig. 1 Arnica: . Placebo: .

• Reference

• 1 Robertson A., Suryanarayanan R., Banerjee A. Homeopathic Amica montana for post-tonsillectomy analgesia: a randomised placebo control trial. Homp 2007; 96: 17-21.