+1 Recommend
0 collections
      • Record: found
      • Abstract: found
      • Article: not found

      Statistics review 12: Survival analysis


      Read this article at

          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.


          This review introduces methods of analyzing data arising from studies where the response variable is the length of time taken to reach a certain end-point, often death. The Kaplan–Meier methods, log rank test and Cox's proportional hazards model are described.

          Related collections

          Most cited references5

          • Record: found
          • Abstract: found
          • Book: not found

          Essential Medical Statistics

          <b>Blackwell Publishing is delighted to announce that this book has been Highly Commended in the 2004 BMA Medical Book Competition. Here is the judges' summary of this book:</b><p>"This is a technical book on a technical subject but presented in a delightful way. There are many books on statistics for doctors but there are few that are excellent and this is certainly one of them. Statistics is not an easy subject to teach or write about. The authors have succeeded in producing a book that is as good as it can get. For the keen student who does not want a book for mathematicians, this is an excellent first book on medical statistics."<p><i>Essential Medical Statistics</i> is a classic amongst medical statisticians. An introductory textbook, it presents statistics with a clarity and logic that demystifies the subject, while providing a comprehensive coverage of advanced as well as basic methods.<p>The second edition of <i>Essential Medical Statistics</i> has been comprehensively revised and updated to include modern statistical methods and modern approaches to statistical analysis, while retaining the approachable and non-mathematical style of the first edition. The book now includes full coverage of the most commonly used regression models, multiple linear regression, logistic regression, Poisson regression and Cox regression, as well as a chapter on general issues in regression modelling. In addition, new chapters introduce more advanced topics such as meta-analysis, likelihood, bootstrapping and robust standard errors, and analysis of clustered data.<p>Aimed at students of medical statistics, medical researchers, public health practitioners and practising clinicians using statistics in their daily work, the book is designed as both a teaching and a reference text. The format of the book is clear with highlighted formulae and worked examples, so that all concepts are presented in a simple, practical and easy-to-understand way. The second edition enhances the emphasis on choice of appropriate methods with new chapters on strategies for analysis and measures of association and impact.<p><i>Essential Medical Statistics</i> is supported by a web site at <b>www.blackwellpublishing.com/essentialmedstats</b>. This useful online resource provides statistical datasets to download, as well as sample chapters and future updates.
            • Record: found
            • Abstract: not found
            • Book: not found

            Statistics at square one

              • Record: found
              • Abstract: found
              • Article: not found

              Survival of patients with bronchiectasis after the first ICU stay for respiratory failure.

              Respiratory failure (RF) is a frequent cause of death among patients with bilateral bronchiectasis. An ICU admission is commonly required, and neither short-term or long-term outcomes have been studied. We performed a retrospective study over a 10-year period (January 1990 to March 2000). All patients with bilateral bronchiectasis admitted for the first time in the medical ICU for RF were reviewed. Patients with cystic fibrosis were excluded. Forty-eight patients (mean age +/- SD, 63 +/- 11 years; mean simplified acute physiology score [SAPS] II, 32 +/- 12) of whom 25% received long-term oxygen therapy (LTOT) were identified. All the patients were treated with intensive medical care, associated with noninvasive ventilation in 13 patients (27%), and 26 patients (54%) required intubation. Nine patients (19%) died in the ICU. The 1-year mortality rate was 40%. Among the variables recorded at ICU admission, age > 65 years (p = 0.002), SAPS II score > 32 (p = 0.012), use of LTOT (p = 0.047), and intubation (p = 0.027) were associated with reduced survival in univariate analysis by Cox regression. Multivariate analysis by Cox proportional hazard model showed that age > 65 years (relative risk [RR], 2.70; 95% confidence interval [CI], 1.15 to 6.29) and use of LTOT (RR, 2.52; 95% CI, 1.15 to 5.54) were independently associated with reduced survival. We performed the first study providing information related to the impact of the first ICU stay for RF on long-term outcomes for patients with bilateral bronchiectasis. Age > 65 years and prior use of LTOT were associated with reduced survival.

                Author and article information

                Crit Care
                Critical Care
                BioMed Central (London )
                6 September 2004
                : 8
                : 5
                : 389-394
                [1 ]Senior Lecturer, School of Computing, Mathematical and Information Sciences, University of Brighton, Brighton, UK
                [2 ]Senior Registrar in ICU, Liverpool Hospital, Sydney, Australia
                Copyright © 2004 BioMed Central Ltd

                Emergency medicine & Trauma
                log rank test,kaplan–meier method,cox's proportional-hazards model,survival function s(t),hazard ratio,cumulative hazard function h(t)


                Comment on this article