Article Text


Computer programs for epidemiologists. PEPI v. 4.0
  1. J Llorca
  1. Division of Preventive Medicine and Public Health, University of Cantabria School of Medicine, Santander, Spain

    Statistics from

    J H Abramson, P M Gahlinger. (Pp 305; $59.95). Salt Lake City: Sagebrush Press, 2001. ISBN 0-9703130-2-0

    The book is the manual of PEPI version 4.0, a collection of programs that includes a variety of programs for use in statistical analysis and planning of epidemiological studies, covering sample size estimation, contingency tables, standardisation, logistic regression, survival analysis—although no Cox regression—, smoothing of curves, and much more. Each program offers a number of options and outputs (the authors claim that “The programs may offer more options than you need, and most will display more results that you need”); this enlarges the range of possible users. The manual is clearly written and provides the main uses of each program as well as some mathematical details.

    Logistic regression programs read data files. All the other programs work on elaborated data (for example: rates or number of observations in each cell of a table); therefore, primary data must be tabulated or counted using other statistical software before using PEPI, and then elaborated data must be entered at the keyboard.

    Users of statistical packages (such as Stata, SPSS, or SAS) can find PEPI rather tedious because of this two phase procedure (tabulation in another program, analysis in PEPI). Furthermore, many programs in PEPI require reinitialisation each time you want to introduce new data. Nevertheless, my initial scepticism was modified after using it: when I needed to estimate the sample size for a matched case-control study, I could compare several packages and found that PEPI provides an output richer than others do. This feature is common to other programs in PEPI: they cover a variety of epidemiological tests wider than general purpose statistical packages.

    Epidemiologists can use PEPI with two main purposes when analysing data: as an alternative to statistical programs that are more expensive, or as a complementary toolbox when other programs are available. Teaching and learning purposes are also possible.

    View Abstract

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.