Statistics for Oncology - A course for Scottish trainees
Authors: Aviva Petrie and Caroline Sabin
Publisher: Wiley-Blackwell
Why recommended: Part of the popular ‘At a Glance’ series, this concise text presents statistical concepts in a visual, easy-to-digest format. Ideal for quick revision and understanding key concepts at a glance. Particularly useful for FRCR exam preparation as it covers the core statistical methods commonly encountered in oncology literature.
Authors: Stephen J. Walters, Michael J. Campbell, David Machin
Publisher: Wiley
Why recommended: Written by UK-based professors from the University of Sheffield, this text explains medical statistics with minimal technical detail while maintaining rigour. It features real examples from clinical trials and includes multiple-choice exercises. Particularly relevant for UK/Scottish oncology trainees as it uses NHS and UK clinical examples.
Author: Martin Bland
Publisher: Oxford University Press
Why recommended: A well-established introductory text widely used in UK medical schools. Clear explanations of fundamental concepts with medical examples throughout.
Author: Harvey Motulsky
Publisher: Oxford University Press
Why recommended: Focuses on conceptual understanding rather than mathematical formulas. Excellent for developing intuition about statistical reasoning before tackling more technical material.
Authors: Girling, Parmar, Stenning, Stephens & Stewart
Year: 2003
Publisher: Oxford University Press
Why recommended: Definitive text on cancer clinical trials methodology. Essential for understanding trial design, randomization, endpoints, interim analyses, and interpretation of results. Directly relevant to FRCR examination questions on clinical trial interpretation and critical appraisal.
Authors: Coggon, Barker & Rose
Edition: 5th Edition, 2003
Publisher: BMJ Books
Why recommended: Clear, concise introduction to epidemiological principles essential for understanding cancer research. Covers study design, bias, confounding, and interpretation of epidemiological data. Compact and accessible for busy trainees.
Author: Mark Elwood
Edition: 4th Edition, 2017
Publisher: Oxford University Press
Coverage: Focuses on critical appraisal skills essential for evaluating published research. Develops the analytical approach needed for evidence-based oncology practice and FRCR examination questions on study interpretation.
Website: www.consort-statement.org
Essential guidelines for reporting randomized controlled trials. Understanding CONSORT is crucial for critically appraising trial publications.
Website: www.ct-toolkit.ac.uk
Comprehensive online resource for clinical trials methodology, including sample size calculators, randomization tools, and guidance documents.
Website: http://www-users.york.ac.uk/~mb55/pubs/pbstnote.htm
Martin Bland’s collection of statistical notes published in the British Medical Journal. Excellent short explanations of key statistical concepts and common pitfalls. Highly recommended for exam revision.
These texts bridge basic concepts and advanced applications, with emphasis on practical implementation in cancer research and clinical trials.
Author: Frank E. Harrell Jr.
Publisher: Springer
Software: R (with rms package)
Why recommended: Good reading for oncology researchers conducting multivariable analyses. Addresses real-world challenges including missing data, model validation using bootstrap methods, survival analysis, and logistic regression. Comprehensive R implementation makes it ideal for UK trainees using RStudio. Particularly valuable for cancer prognosis studies and clinical trial analysis.
Key topics: Survival regression, ordinal regression, model validation, handling missing data, avoiding categorization of continuous variables
Author: Ewout W. Steyerberg
Publisher: Springer
Why recommended: Ideal companion to Harrell’s text. Provides additional explanations, examples, and simulations of modern statistical methods. Highly relevant for developing and validating prognostic models in oncology.
Key topics: Prognostic model development, internal and external validation, risk prediction, model updating
Author: Bernard Rosner
Publisher: Cengage
Why recommended: Comprehensive intermediate text balancing theory with practical application. Includes software implementation guidance for R, SAS, and Stata. Over 300 medical research examples and problems with solutions.
These advanced texts cover sophisticated methodologies for complex cancer research questions and clinical trial design.
Author: Frank E. Harrell Jr.
Why recommended: Suitable for Masters’ and PhD-level training. Advanced chapters cover Bayesian modelling, bootstrap confidence intervals, generalized least squares for longitudinal data, and robust analysis using ordinal regression. Essential for methodologically sophisticated cancer research.
Advanced topics: Bayesian inference, model uncertainty quantification, advanced survival models, repeated measures analysis
Authors: Eric Vittinghoff et al.
Publisher: Springer
Why recommended: Advanced regression modelling with emphasis on practical suggestions for real-world data challenges. Note: Has Stata emphasis, but concepts transfer to R.
The following are a good intro to conducting statistical analysis using the R programming language.