Treatment of venous leg ulcersBMJ 2000; 320 doi: https://doi.org/10.1136/bmj.320.7232.446 (Published 12 February 2000) Cite this as: BMJ 2000;320:446
Nice study, pity about the sample size
- Frank Dobbs, senior lecturer in primary health care
- University of Plymouth, Plymouth PL4 8AA
- Vascular Surgery Office, Department of Surgery, University of Edinburgh, Royal Infirmary of Edinburgh, Edinburgh EH3 9YW
EDITOR—Dale et al based their calculation of the sample size (200 patients) on the minimum size required to find the large effect observed in a previous very small study.1 The confidence interval for this large effect would have been very large.
A better strategy would have been to decide on a detectable difference that would be clinically significant and design the study to be capable of detecting this difference. The difference that they found (64% healing with pentoxifylline v 53% healing with placebo) would probably be clinically significant in view of the high …