Development of a spreadsheet for the calculation of new tools to improve the reporting of the results of medical research.

Abstract

Primary objective: Confidence levels, clinical significance curves and risk - benefit contours have recently been developed with the aim of improving the interpretation of clinical studies. It has been suggested that their use may complement traditional methods of result reporting (i.e. p values and 95% confidence intervals), and subsequently improve medical decision-making. These new methods have been used to report study results; however, widespread use may be hampered by the lack of available computer software. Our objective was to design user-friendly software to enable researchers to use these statistical methods.

Research design: A spreadsheet was designed to calculate these statistics for the assessment of two arm medical studies. The spreadsheet was tested using a wide range of input data and operating systems. User-friendliness was tested by researchers who had no background in statistics.

Main outcomes and results: The confidence calculator was successfully designed, and found to be user-friendly. The spreadsheet is MS Excel-based to allow wide usage. The spreadsheet is freely available from the author and from the web site http://www.primercollaboration.com/Tools/tools.html.

Conclusions: We have successfully designed a spreadsheet that is simple to use and freely accessible for researchers. This is the first such software that calculates confidence levels, clinical significance curves, and risk - benefit contours.