Home > Articles > Books about clinical trials: a reading list for following research
Prepared by: Research Desk
Prepared by: Research Desk. Sources are linked for verification.
Following an investigational drug such as retatrutide means reading registry records, journal abstracts, press releases and regulatory statements, often within hours of each other. The skills that make that reading reliable are general. They include knowing how a trial is designed, what its statistics can and cannot show, and how published evidence gets distorted on its way to a headline.
This reading list collects seven books that teach those skills. None of them is about retatrutide, obesity or diet, and that is deliberate. The focus is how trials are designed, reported and registered, and how they sometimes go wrong.
Status: Investigational. As of 18 October 2026, the FDA states that retatrutide is not a component of any FDA-approved drug and has "not been found safe and effective for any condition". It also states that retatrutide "cannot be used in compounding under federal law" [1]. Nothing on this page describes or encourages its use. See our approval status page for the full regulatory record.
How we chose these books
We selected books that meet four neutral criteria:
Focus on method. Each book explains how evidence is generated or judged, not what to think about a specific product.
Named, verifiable authors. Every title is listed with its author or authors and publication year so readers can find the right edition.
A spread of levels. The list runs from a technical textbook to general-audience books that need no statistics background.
Honest limits. For each book we note what it does not cover, because no single title explains everything a trial report contains.
Our descriptions summarise each book's general subject and are not quotations from it. We earn nothing from any drug developer, and the selection follows the sourcing standards described on our about page.
Comparison table
Title | Author(s) | Year | Focus | Level |
|---|---|---|---|---|
Fundamentals of Clinical Trials | Lawrence M. Friedman, Curt D. Furberg, David L. DeMets | Multiple editions | Trial design and conduct | Advanced / textbook |
How to Read a Paper | Trisha Greenhalgh | 1997 (later editions) | Appraising published studies | Intermediate |
The Art of Statistics | David Spiegelhalter | 2019 | Statistical reasoning and uncertainty | Beginner to intermediate |
Statistics Done Wrong | Alex Reinhart | 2015 | Common statistical errors | Intermediate |
Bad Pharma | Ben Goldacre | 2012 | Missing data and reporting bias | General audience |
Rigor Mortis | Richard Harris | 2017 | Reproducibility in biomedical research | General audience |
Science Fictions | Stuart Ritchie | 2020 | Problems in how science is done and reported | General audience |
Group 1: Trial design fundamentals
Fundamentals of Clinical Trials (Friedman, Furberg and DeMets)
What it teaches. This is a standard textbook on how randomized trials are planned and run. It covers the path from defining the research question through randomization, blinding, sample size, monitoring and reporting.
Who it suits. It suits readers who want to understand every field on a trial registry record and why it matters. That includes students of research methods and anyone writing about trials professionally.
Limits. It is long and technical, and it assumes some comfort with basic statistics. Editions differ, so confirm which one a library or bookstore holds.
How it connects to trial tracking. A ClinicalTrials.gov entry such as the TRIUMPH-1 record (NCT05929066) lists the allocation method, masking, outcome measures and enrollment [2]. A design textbook explains why each of those choices exists and what goes wrong when one is weak.
How to Read a Paper (Trisha Greenhalgh)
What it teaches. First published in 1997 and updated in later editions, it is a guide to appraising published research. It shows how to recognise the type of study and judge how far its conclusions can be trusted.
Who it suits. It suits readers who already skim abstracts and want a more systematic way to read the full paper. It bridges the general-audience books and the textbook above.
Limits. It covers many study types, so trial design gets less depth than in a dedicated textbook. Older editions contain dated examples.
How it connects to trial tracking. The phase 2 obesity study of retatrutide randomized 338 adults for 48 weeks [3]. In general terms, a phase 2 trial of that size is built to explore doses and gather early safety signals. Larger phase 3 trials are designed to confirm findings. Knowing the phase behind a headline changes how much weight it can bear.
Group 2: Reading results and statistics
The Art of Statistics (David Spiegelhalter)
What it teaches. Published in 2019, it is a general introduction to learning from data. Topics include the difference between correlation and causation, how uncertainty is expressed, and how risk is communicated.
Who it suits. It suits readers with little or no statistics training. It is the most approachable title on this list and works well as a first read.
Limits. It covers statistical thinking in general rather than trials specifically, so trial-specific mechanics get little space.
How it connects to trial tracking. The TRIUMPH-2 paper describes itself as a "double-blind, parallel-group, randomised, placebo-controlled, phase 3 trial" [4]. Each term maps onto a basic idea:
A placebo group provides a comparison.
Blinding limits expectation effects.
Random allocation lets differences be attributed to the intervention rather than to systematic differences between groups.
The role of chance is then handled by the statistical analysis.
Statistics Done Wrong (Alex Reinhart)
What it teaches. This short 2015 book catalogues common statistical mistakes in published research. Examples include misreading p-values, relying on small studies and running many comparisons.
Who it suits. It suits readers who know the basics and now want to spot errors. It can be read in a few sittings.
Limits. It assumes familiarity with p-values and confidence intervals. It also focuses on errors rather than on designing a good study from scratch.
How it connects to trial tracking. Trials often report many outcomes and several comparison groups. The phase 2 type 2 diabetes study, for example, was both placebo-controlled and active-controlled [5]. Every additional comparison raises the chance of a spurious finding. That is why the pre-specified primary outcome deserves more attention than a striking secondary one.
Group 3: When evidence goes wrong
Bad Pharma (Ben Goldacre)
What it teaches. Published in 2012, it is a critique of how trial evidence can be distorted through unpublished studies and selective reporting. It also argues for registration and full results reporting.
Who it suits. It suits general readers who want to understand why trial registries exist. It needs no statistics background.
Limits. Transparency rules have changed since it was written, so some specifics are dated. It is an argument, not a neutral survey.
How it connects to trial tracking. TRIUMPH-2 has a public registry entry (NCT05929079) that lists its planned outcomes [6]. There is also a published phase 3 report that names the trial [4]. A reader can therefore compare what was planned with what was reported.
Rigor Mortis (Richard Harris)
What it teaches. This 2017 book by a science journalist examines why much biomedical research fails to replicate.
Who it suits. It suits readers interested in the laboratory stage that comes before human trials. It is narrative and accessible.
Limits. It focuses largely on research that precedes human studies, so it says less about registry records or trial reporting.
How it connects to trial tracking. Explanations of how a molecule works often rest on laboratory and animal findings, which carry more uncertainty than headlines suggest. For this reason, our trial-by-trial guide separates proposed mechanisms from what human trials have actually measured.
Science Fictions (Stuart Ritchie)
What it teaches. Published in 2020, it is a broad overview of problems in how science is conducted and reported, along with proposed reforms.
Who it suits. It suits general readers who want one well-organized overview of the replication problem across science, not only medicine.
Limits. It ranges across many fields, so only part of it bears directly on drug trials.
How it connects to trial tracking. Exaggerated headlines are a recurring problem in research reporting. TRIUMPH-5 has its own registry record, which lists the trial's comparator and primary outcome [7]. Any headline about that trial can be checked against those registered fields, and against a peer-reviewed report where one exists.
Putting the books to work on trial news
The seven books share a common thread: a short set of questions that apply to any new trial report.
Is there a registry record? Look up the NCT number on ClinicalTrials.gov and note the pre-specified primary outcome [2][6].
What phase is it? In general terms, phase 2 trials mainly explore doses and early safety, while phase 3 trials confirm findings in larger groups. The retatrutide phase 2 obesity study [3] and the phase 3 TRIUMPH-2 trial [4] are examples of each.
What was the comparison? It may be placebo, an active comparator, or both [4][5].
Is the source primary? A peer-reviewed paper or registry entry carries more weight than a press summary or an industry forecast.
What does the regulator say? Regulatory status comes from the regulator itself. The FDA's own page on unapproved GLP-1 drugs addresses retatrutide by name [1], and our approval status page links that record.
Tools and further reading
Readers new to research methods often get the most from pairing one general-audience book with one more technical title. A good sequence is a statistics primer alongside a book on study appraisal, followed by a book on reporting bias once the basics feel familiar.
When comparing editions and formats, check the publication year and whether the edition has been revised. Also check whether an index and reference list are included, and whether an ebook or audiobook keeps tables and figures intact.
Public and university libraries often hold research-methods textbooks. Interlibrary loan is a low-cost way to sample a title before buying it.
Price comparisons should include shipping and edition differences. Older editions of methods textbooks can be cheaper but may lack updated chapters on reporting standards.
Frequently asked questions
Which book is the best place to begin?
For readers without a statistics background, The Art of Statistics by David Spiegelhalter (2019) is the most approachable. How to Read a Paper by Trisha Greenhalgh is a natural second step.
Do these books cover retatrutide or other specific drugs?
No. They teach general methods for judging trial evidence. Our retatrutide guide applies those methods to the published trials, and the approval status page tracks the regulatory record.
Are the older books still relevant?
Mostly, yes. The design principles in Fundamentals of Clinical Trials and the statistical pitfalls in Statistics Done Wrong remain current. Bad Pharma (2012) describes a reporting landscape that has since shifted, so its specific examples read best as history.
Why does a trial registry record matter when a paper exists?
The registry entry records what investigators planned to measure before results were known. Comparing it with the published paper shows whether the reported outcomes match the planned ones. That comparison is possible for TRIUMPH-2 [4][6].
Last updated: 18 October 2026.
Related on this site
Is retatrutide FDA approved? Legal status and approval timeline
Retatrutide: mechanism, trial findings and side effects explained
How to read retatrutide trial results: endpoints, placebo, dropouts
Not medical advice. This page reports published research and regulatory records for information only.
Research-only notice: Content on this site is provided for research and educational purposes only and is not intended for human use. It is not medical advice.
Translation notice: Some content on this site may have been machine-translated using Google Translate and may contain errors.