Year : 2018 | Volume
: 148 | Issue : 6 | Page : 763--764
Presenting medical statistics from proposal to publication
P. S. S. Sundar Rao
Formerly Professor, Department of Biostatistics, Christian Medical College, Vellore 632 004, Tamil Nadu, India
P. S. S. Sundar Rao
Formerly Professor, Department of Biostatistics, Christian Medical College, Vellore 632 004, Tamil Nadu
|How to cite this article:|
Rao PS. Presenting medical statistics from proposal to publication.Indian J Med Res 2018;148:763-764
|How to cite this URL:|
Rao PS. Presenting medical statistics from proposal to publication. Indian J Med Res [serial online] 2018 [cited 2020 Jun 4 ];148:763-764
Available from: http://www.ijmr.org.in/text.asp?2018/148/6/763/252167
2nd edition, J.L. Peacock, S.M. Kerry, R.R. Balise (Oxford University Press, Oxford, UK) 2017. Price: Not mentioned.
This handy book is a second edition being published 10 years after the first edition, and will be useful to many students, especially young researchers, who have some basic understanding of statistics and use of standard statistical computer software. The authors are experienced teachers of biostatistics and also researchers who have a good idea of the requirements of students and those who need help in preparing research protocols and writing scientific papers.
The book may be broadly considered in three sections: the first four chapters explain the research process and how to write a research proposal; the next seven chapters deal systematically with the statistical analyses from simple study designs to complex multivariable studies and constitute the main focus of this book. The final two chapters describe the presentation of medical statistics in a randomized controlled trial and in a meta-analysis. Thus the 13 chapters lead the researcher on a composite tour of how to present statistics elegantly and correctly in a user-friendly language. The book has a list of pertinent references for those interested in further studies and ends with an Index.
As stated by the authors, this is not a textbook in statistics but a practical reference book. Biostatistics has grown rapidly over the recent few decades and it has become necessary to use the best statistical expertise available to help medical research. This expertise requires not merely routine computer software that are now widely available, but a deeper understanding of the concepts and theories involved in medical investigations. This book tries to provide the reader with the essential background to be able to then use the right study design and appropriate statistical techniques.
A major difference between this book and other books available on medical statistics is the dynamic presentation of the various steps using relevant boxes in each chapter with the appropriate computer outputs from the three major statistical software-the SPSS (Statistical Package for the Social Sciences), the R and the SAS (Statistical Analysis System). For a few topics, two other popular software, the Epi Info™ and the Stata are used so that the reader gets familiar with using computer outputs and how they could be meaningfully integrated into the report.
Chapter 1 introduces the book as well as the importance of presenting medical statistics in a simple, straightforward and meaningful manner to different audience. It also mentions what statistical programmes are used to present various results. For an uninitiated researcher, some of the statements may seem somewhat strange or difficult to understand at this stage but the authors have made an effort to explain the problems in presenting medical statistics. It is therefore, essential to read this chapter to know the scope and limitations of using this book as well as its strengths.
Chapter 2 is entitled as 'Introduction to the research process', but focusses mainly on the beginning stages of a research project, namely the data collection process, as well as its management using computers. This is a detailed and useful chapter for all the readers, and cautions on the pitfalls and dangers of not starting the research carefully, and collecting relevant data in a manner that facilitates proper data processing that would help in subsequent statistical analyses and reporting. The authors have given many examples and have tried to guide the researchers on this crucial step. Every researcher should benefit by reading and utilizing the suggestions made.
Determining minimum sample size is a vital question of every researcher, and this is covered in chapter 3 along with guidance on how to formulate the aims and objectives of a research study. The title is a bit of a misnomer since the chapter deals more with sample size rather than how to write a research protocol. The titles of the first 4 chapters can be misleading and the contents somewhat overlapping or restricted to a particular aspect, but these are important chapters which together guide the researcher in preparing a composite research protocol.
From chapter 5 onwards, the book takes the reader through the step-by-step process of presenting descriptive statistical analyses from the initial to final stage of reporting the findings appropriate to the aims of the research. Chapter 5 sets the stage and is a critical reading for the researcher to decide the manner and content of the numerical as well as qualitative data. The language and description are quite lucid and will be helpful for any type of researcher. Anticipating the tremendous varieties of study designs used by medical researchers to answer specific research questions or objectives, chapter 6 deals with single group studies, chapters 7 and 8 cover the comparisons between two groups, whether unpaired or paired.
Chapters 9 and 10 deal with analyzing relationships between variables from simple to multiple correlations and regressions, thus guiding the reader to choose the best statistical analyses. It is difficult to demonstrate the use of every statistical analysis or a computer programme, however, the authors have judiciously used a variety of procedures that will guide the researcher to expand further as needed.
Chapter 12 on survival analysis is well written and offers further reading. The interventional studies have become an integral part of most medical studies and the fundamental principles of a randomized controlled trial (RCT) are set out in this chapter, as enshrined in the international guidelines (CONSORT; Consolidated Standards of Reporting Trials). The final chapter on meta-analyses or systematic reviews will be useful for most readers at least in their review of literature as a prelude to their own research.
The authors have succeeded in condensing the entire gamut of medical statistics and their presentation in a conveniently sized book, which also includes an index and selected references for further reading. One wishes that the authors could have presented a few outstanding books on Biostatistics written by international experts, not focusing entirely on their own publications or websites.
Despite limitations in including all types of study designs or statistical analyses that are now used in investigating various research questions or testing a plethora of research hypotheses in single book, the authors have included the most popular ones and the reader can extend the ideas given to their own research as required. To that extent, this book is useful not only for the students but their supervisors as well. This book must be commended for the manner in which it has succeeded to bring a fresh look at presenting medical statistics in preparing research proposals that may facilitate faster publication of scientific papers.