Essentials of a Successful Biostatistical Collaboration

Essentials of a Successful Biostatistical Collaboration by Arul Earnest is a unique approach to a biostatistics text in that the focus is not purely on study design and statistical analyses. While these topics are certainly discussed, equal attention is given to topics such as planning, project management, and effective communication that are important for any biostatistician who is collaborating with a research team.


The book begins with an overview of observational study designs as well as randomized control trials. Then the discussion turns to data management, power and sample-size calculations, and a variety of statistical analyses. Earnest provides an overview of statistical methods ranging from basics such as t tests and correlation to more advanced topics such as Cox regression and ARIMA models, and each topic is accompanied by the corresponding Stata commands and output. The following chapters give advice on both verbal and written communication, project management, and how to manage collaborations.


This book is a practical resource that will appeal to biostatisticians, epidemiologists, and other members of clinical research teams as well as students who plan to work in this field.



1. Observational Study Designs
1.1 Introduction
1.2 Comparative Features between Cohort, Case–Control and Cross-Sectional Studies

1.2.1 Cross-Sectional Study Design
1.2.2 Case–Control Study Design
1.2.3 Cohort Study Design
1.2.4 Ecological Study Design

1.3 Selecting the Appropriate Study Design
1.4 Key Questions to Ask a Clinician
1.5 Common Mistakes
1.6 Online Tools and Resources
1.7 A Collaborative Case Study
Study Summary
Alternative Study Designs
Key Learning Points


2. Randomised Controlled Trials
2.1 Introduction
2.2 Phases and Types of Trials
2.3 Types and Features of Randomisation
2.4 Controls and Blinding
2.5 Other Types of RCTs

2.5.1 Cluster RCT
2.5.2 Crossover Trials
2.5.3 Stepped Wedge Cluster RCT
2.5.4 Equivalence Trials

2.6 Data and Safety Monitoring Board
2.7 Choosing Outcome Measures for a Trial
2.8 Key Questions to Ask a Clinician
2.9 Tools and Resources
2.10 A Collaborative Case Study
Study Summary
Alternate Method of Randomisation
Significance of Trial
Rigorous and Appropriate Analysis
Key Learning Points


3. Form Design and Database Management
3.1 Introduction
3.2 Principles of Questionnaire Design

3.2.1 Open-Ended versus Closed-Ended Questions
3.2.2 Multiple-Response Questions
3.2.3 Double-Barrelled Questions
3.2.4 Wording of Questions
3.2.5 Ordinal Scales
3.2.6 Validation
3.2.7 Translation

3.3 Types of Variables and Scales of Measurement
3.4 Finding a Suitable Database Software for Your Study
3.5 Efficient Ways to Create and Manage an Excel Database

3.5.1 Multiple-Response Questions
3.5.2 Repeated Measurements
3.5.3 Missing Data
3.5.4 Data Validation in Excel
3.5.5 Limitations to Excel

3.6 Key Questions to Ask a Clinician
3.7 Common Mistakes to Avoid

3.7.1 Questionnaire Design
3.7.2 Database Design

3.8 Tools and Resources
3.9 A Collaborative Case Study
Key Learning Points


4. Sample Size and Power Calculations
4.1 Introduction
4.2 Linking Hypothesis Testing and Sample Size
4.3 Ingredients in a Sample Size Calculation

4.3.1 Objective of the Study
4.3.2 Type 1 Error (Level of Significance)
4.3.3 Type 2 Error (1 – Power)
4.3.4 Variability
4.3.5 The Effect Size
4.3.6 Two-Sided versus One-Sided Tests
4.3.7 Other Considerations

4.4 Commonly Performed Sample Size Calculations

4.4.1 Comparing Proportions between Two Independent Groups
4.4.2 Comparing Means between Two Independent Groups
4.4.3 Estimating Hazard Ratios in a Survival (Time to Event) Analysis
4.4.4 Estimating Coefficients in a Linear Regression Model
4.4.5 Estimating Odds Ratios in a Logistic Regression Model
4.4.6 Estimating a Kappa Coefficient in an Agreement Study
4.4.7 Repeated Measures Analysis
4.4.8 Cluster Randomised Trials

4.5 Key Questions to Ask a Clinician
4.6 Common Mistakes
4.7 Tools and Resources
4.8 A Collaborative Case Study
Key Learning Points


5. Statistical Analysis Plan
5.1 Introduction
5.2 Choosing the Appropriate Statistical Method
5.3 Common Statistical Hypotheses and Tests

5.3.1 Univariate Analysis Estimate Mean Estimate Proportion

5.3.2 Bivariate Analysis Compare Means in Two Groups Assumptions of Independent Student’s t-Test Non-Parametric Equivalent: Mann–Whitney U Test Compare Means in More Than Two Groups Assumptions of the ANOVA Test Non-Parametric Equivalent: Kruskal–Wallis Test Compare Means within the Same Group Calculate Correlation between Two Continuous Variables Compare Proportions Compare Survival Measure Agreement (Continuous Variables) Measure Agreement (Categorical Variables) Relative Risk Measures

5.3.3 Multi-Variate Analysis Continuous Outcome Measure–Linear Regression Model Assumptions of the Linear Regression Model Binary Outcome Measure–Binary Logistic Regression Model Predicted Probability Ordinal Outcome Measure–Ordinal Logistic Regression Categorical Outcome Measure–Multinomial Logistic Regression Count Data–Poisson Regression Model Survival Data–Cox Regression Model Generalised Estimating Equations Clinical Examples ARIMA Models Clinical Example

5.4 Issues to Note

5.4.1 Crossover Trials
5.4.2 Cluster RCT
5.4.3 Intention-to-Treat versus Per-Protocol Analysis
5.4.4 Missing Data
5.4.5 Multiplicity

5.5 Key Questions to Ask a Clinician
5.6 A Collaborative Case Study
Key Learning Points


6. Effective Communication Skills
6.1 Introduction
6.2 The Initial Meeting
6.3 Difficulty in Understanding Medical Jargon
6.4 Challenges in Explaining Statistical Concepts to Clinicians

6.4.1 Using the Venn Diagram
6.4.2 Using the Gaussian Curve
6.4.3 Simplifying Language
6.4.4 Employing Drawings

6.5 Effective Presentation Skills

6.5.1 Tips and Tricks within Microsoft PowerPoint for a Statistical Presentation
6.5.2 Actual Presentation
6.5.3 Preparing a Poster
6.5.4 Checklist for Preparing and Presenting an Effective Research Poster Content Layout and Format Narrative Description

6.5.5 Some Tips on Creating a Poster in Microsoft PowerPoint

6.6 Tools and Resources
6.7 A Collaborative Case Study
Key Learning Points


7. Effective Writing Skills
7.1 Introduction
7.2 Writing a Statistical Analysis Plan for a Grant Application

7.2.1 Specific Aims and Hypotheses
7.2.2 Background and Clinical Significance
7.2.3 Preliminary Studies/Progress Report
7.2.4 Methods/Approach CONSORT Statement STROBE Statement PRISMA Statement STARD Statement

7.3 Writing for a Publication – First-Author Publication

7.3.1 Selecting the Journal
7.3.2 Instructions for Authors
7.3.3 Tips for Formatting

7.4 Preparing a Manuscript

7.4.1 Abstract Tips for the Abstract Section

7.4.2 Introduction Tips for the Introduction Section

7.4.3 Methods
7.4.4 Results Tips for the Results Section

7.4.5 Discussion and Conclusion Tips for the ‘Discussion and Conclusion’ Section

7.5 Creating a Draft
7.6 Writing for a Publication – Collaborative Publication

7.6.1 Additional Tips When Writing for a Collaborative Publication

7.7 Resources
Key Learning Points


8. Project Management: Best Practices
8.1 Introduction
8.2 Creation of a Project File
8.3 Database Security and Confidentiality

8.3.1 Database Confidentiality
8.3.2 Database Security

8.4 Standard Operating Procedures
8.5 Ensuring Consistency and Reproducibility in the Results
8.6 Managing Multiple Projects
8.7 Obtaining Mentorship
8.8 Poor Project Management Skills (The ‘Not’s to Avoid)
8.9 A Collaborative Case Study
Key Learning Points


9. Managing Collaborations
9.1 Introduction
9.2 Getting the Most Out of a Collaboration
9.3 Providing Collaboration in a Large Complex Institution – Hub-and-Spoke Model
9.4 Providing Consultations
9.5 Seven Faces of Collaborators

9.5.1 The Auto-Pilot
9.5.2 The Pseudo-Statistician
9.5.3 The ‘Harry Houdini’
9.5.4 The p-Value Hunter
9.5.5 The Sceptic
9.5.6 The Passive Collaborator
9.5.7 The Faceless Collaborator

9.6 Useful Strategies to Adopt in Successfully Managing a Collaboration
9.7 Responding to Unreasonable Work Requests
9.8 Reasoning with a Collaborator Who Engages in Data Dredging
9.9 Coping with an Unreasonable Request on Turnaround Time
9.10 Negotiating Authorship
9.11 International Collaborators

9.11.1 General Statistical Conferences
9.11.2 Biostatistics Conferences

Key Learning Points


10. How Not to Design, Analyse and Present Your Study
10.1 Introduction
10.2 Choosing the Inappropriate Study Design
10.3 Selecting Too Few Subjects in the Study
10.4 Incorrect Use of Randomisation
10.5 Undertaking Incorrect Statistical Tests
10.6 Not Checking for the Assumptions Behind the Test
10.7 Data Dredging
10.8 Presenting Tables and Figures Inappropriately
10.9 Reporting and Interpreting Data Inappropriately
10.10 Conclusion
Key Learning Points


11. Views from the Ground: A Survey among Biostatisticians and a Chat with Clinicians
11.1 Introduction
11.2 Survey Methodology and Profile of Respondents
11.3 Problems Ever Faced in Collaborating with Clinicians
11.4 Training on Collaboration/Consultation Skills
11.5 Frequency of Performing Selected Tasks as a Biostatistician
11.6 Issues to Address in Order to Enhance Greater Collaborations
11.7 Skills Most Important to Gain to Improve on Collaborations with Clinicians

11.7.1 Formal Postgraduate Degrees
11.7.2 Short Courses
11.7.3 Online Courses
11.7.4 Journals (Development and Application of Statistics in Medicine)

11.8 Views from Clinicians Who Have Collaborated with Biostatisticians

11.8.1 Interview with Dr. Leong Khai Pang, MBBS, FRCPE, FAMS
11.8.2 Interview with Professor Nick Paton, MB BChir (Cambridge), MRCP (Internal Medicine) (London), MD (Cambridge), FRCP (Edinburgh), DTM & H (London)
11.8.3 Interview with Professor John Augustus Rush, MD, AB
11.8.4 Interview with Professor John McNeil, AM, MBBS (Adelaide), MSc (London), PhD (Melbourne), FRACP, FAFPHM

11.9 Conclusion
Appendix: Sample Survey Questionnaire
Key Learning Points


Author: Arul Earnest
ISBN-13: 978-1-4822-2698-0
©Copyright: 2017 CRC Press

Essentials of a Successful Biostatistical Collaboration by Arul Earnest is a unique approach to a biostatistics text in that the focus is not purely on study design and statistical analyses. While these topics are certainly discussed, equal attention is given to topics such as planning, project management, and effective communication that are important for any biostatistician who is collaborating with a research team.