Confidence intervals should be used in reporting trials

doi:10.1136/bmj.321.7272.1351

BMJ 2000;321:1351 ( 25 November )

Letters

Confidence intervals should be used in reporting trials

The first 150 words of the full text of this article appear below.

EDITOR $---$ The dispute between Barber and Thompson,¹ who are advocating a t test, and Williams, Cohen, and Russell, who areadvocating a Mann-Whitney U test,² has its roots in the useof P values rather than confidence intervals. If Williams et alhad reported their results as a confidence interval for the differencein mean cost, as recommended for the results of clinical trialspublished in the BMJ, the question would not have arisen. Thesample size is surely large enough for the large sample Normalcomparison, which does not require data to follow a normal approximationand to which the t method approximates, to be valid, even withsuch highly skewed data. This gives a confidence interval forthe difference in cost of secondary treatment (routine minus openaccess) equal to -£180 to +£238, the point estimate being £29,and the same P value as the t. . . [Full text of this article]

CiteULike

Complore

Connotea

Del.icio.us Add to Digg

Digg

StumbleUpon Add to Technorati

Technorati What's this?

Relevant Article

Open access follow up for inflammatory bowel disease: Julie A Barber, Simon G Thompson, A Coomarasamy, D Van Der Berg, J G Williams, D R Cohen, and I T Russell
BMJ 2000 320: 1730. [Extract] [Full Text]

This article has been cited by other articles:

Williams, J G, Cohen, D R, Russell, I T (2002). Clarifications: open access follow up for inflammatory bowel disease. BMJ 324: 1586-1586 [Full text]
Morrison, J., Carroll, L., Twaddle, S., Cameron, I., Grimshaw, J., Leyland, A., Baillie, H., Watt, G. (2001). Pragmatic randomised controlled trial to evaluate guidelines for the management of infertility across the primary care-secondary care interface. BMJ 322: 1282-1282 [Abstract] [Full text]