Bookmarkers beware: Bookmarks to pages other than the home page may not work after we change servers in April Editor's Choice | This Week in BMJ | Press releases BMJ No 7125 Volume 316 Medicine and multimedia Saturday 10 January 1998 Arcus Quickstat Iain Buchan Research Solutions, £249 ISBN 0582 945 65 8 Once the preserve of statisticians, who typed incomprehensible command lines to a blank computer screen, statistics in medicine is increasingly open to mere mortals. This is partly because of the availability of personal computers packing more punch than the NASA mainframes that put the first men on the moon but mostly because of the development of accessible and intuitive statistical packages like Arcus Quickstat. Arcus Quickstat was developed with the user in mind. The standard worksheet is compatible with Microsoft Excel and will accept data pasted from other Windows applications (my first analysis was conducted on data pasted into the worksheet from a Word 7 table). There is a consistent approach across different analyses, a sample data sheet to help you get started, and a fully referenced help menu that describes the different analyses, how to do them, and how to cite them. It also helps you to decide what approach to use and, importantly, warns you when to seek help. The package includes facilities for standard parametric and non-parametric analyses, regression and correlation, survival analysis, and meta-analysis. There is also a useful graphing function, which produces graphs as Windows metafiles that may be pasted directly into Windows word processing or presentation packages, where they can be further refined if required. The host of quick to use facilities rank among its best aspect - for example, "number needed to treat," ttests from summary data, derivation of series of random numbers, Fisher's or Gart exact test for 2 x 2 tables, power calculations, and many more. Although they are straightforward to use, there seem to have been no short cuts in the analyses. All the functions tested gave compatible results to industry standard software. The main limitation is that, in order to achieve ease of use and accessibility, some degree of flexibility is lost. Specifically, there is no facility for interactive programming, although such activities remain largely the preserve of specialist statisticians and most of us are delighted not to see a command line and blinking cursor. The axis scaling function for log scales is quirky but will doubtless improve in later versions. Using Arcus Quickstat was described by a colleague to be as "easy as falling off a log," and most people who are only moderately numerate (based on a small convenience sample in an academic department) seem able to get started without going through hours of futile mouse clicking and head scratching. With this in mind, Arcus comes with a "health warning" that conducting complex analyses without statistical advice could seriously damage your credibility. For those with access to the internet, a time limited but otherwise fully functional version of Arcus Quickstat is available at http://www.camcode.com. Nick Freemantle, senior research fellow, Centre for Health Economics, University of York, York Rating: **** Conflict of interest: I have used Arcus Quickstat as a beta tester for the past year and suggested to Iain Buchan that it would be useful to incorporate meta-analysis methods in the package. We worked closely to develop this part of the package, but neither my department nor myself received any financial or other direct reward from the company or any related benefit other than the use of the package. Home | Current issue | Past issues | Classified ads | Career Focus | Feedback Collections | About this site | About the BMJ | BMA | Medline