GNL FLAGSHIP GUIDE
How to Read a Research Study
A new diabetes headline, a manufacturer’s slide deck, a paper that says one thing while the clinic experience says another. This guide teaches you how to tell what the evidence forces you to say, what the evidence cannot tell you, and how to live comfortably in the gap.
Ask Grace
Want to ask Grace whether a study you have read holds up? Ask Grace.
How to use this guide
The seven parts walk an honest reader from “why this matters in clinic” through “the A-to-D grades Grace uses on every clinical claim” to “the auditor’s checklist you can run on any paper you encounter”.
Built on Ben Goldacre’s Bad Science, Bad Pharma, and I Think You’ll Find; on Nassim Nicholas Taleb’s Antifragile, Skin in the Game, and The Black Swan; on the Cochrane Handbook for Systematic Reviews of Interventions v6.5 (2024); on Trisha Greenhalgh’s How to Read a Paper, 7th edition; and on twenty years of paediatric T1D clinic experience. A reader who finishes the guide should be able to tell what the evidence forces them to say, what the evidence cannot tell them, and how to live comfortably in the gap.
The seven parts
Part 1, Why this guide exists
A clinic scene that repeats most weeks: the manufacturer’s 78% slide, the parent’s 51% lived experience, the gap between them. Why critical appraisal is a survival skill, not an academic exercise.
Part 2, The A-to-D evidence grades
How Grace classifies every clinical claim. Why each tier earns the qualifier it carries. Why Grade C is the tier most often confused with Grade A, and where industry pilot trials actually sit.
Part 3, Reading the structure of a paper
The Goldacre lens: selection bias, publication bias, the cover-the-options test, relative-vs-absolute risk, the healthy-user bias. The Reboxetine paragraph as the canonical anchor.
Part 4, Trade-offs and limits
The Taleb lens: ergodicity, the population-vs-individual gap, the Black Swan, the Lindy heuristic, Via Negativa, skin in the game. The Greenhalgh parallel critique in clinical language.
Part 5, From the clinic floor
What Grace actually does with a paper: the 5-paper synthesis, the comparator-equivalence test. The clinic patterns that repeat across twenty years that critical-appraisal textbooks do not name.
Part 6, The auditor’s checklist
Three worked examples (CGM in T2D, closed-loop AID, IOB modelling), the replication problem in plain language, and the ten-point TG Audit checklist any reader can run on any paper.
Part 7, Recommended reading
The ten primary sources this guide is built on, with one line each on why each earns its place. Free where free, accessible where accessible.