How to Use This Book¶
The chapters follow the life of a registry: planning before anyone enrolls, running the registry, working with the community and the field, and using AI on top of all of it. Each chapter opens with a case study from my own work, followed by Lessons for Registry Leaders, the rules the chapter teaches, what AI can and cannot do for that part of the work, and a take-home message. You can read from beginning to end or go directly to the chapter you need.
The Slow Registry Workbook holds every template the book describes. Fill in each page online, or print it and work by hand.
This is a living book. It is published here and revised as it grows, and each new version adds chapters, case studies, and corrections. If you lead or support a registry, I would like to hear from you: a case study from your own registry, a rule you would add, or a question the book should answer. Contributions are anonymized and used only with permission; the contribute page explains how.
A note on case studies in this book. For the case studies in this book, I make changes to protect the identities of individuals and institutions involved. These adjustments do not alter the substance of the lessons.
A note on words. I use "participant" for a person enrolled in a registry, "patient" only where a person is receiving medical care, and "person with lived experience" for people living with a condition and their caregivers. I avoid "subject."
What each chapter holds¶
| Part of the chapter | What it holds |
|---|---|
| Case study | A first-person account from registry work, followed by lessons for registry leaders |
| The rules | Numbered, concrete, and short enough to act on |
| What AI can do, and what only a person can do | Two short lists, side by side |
| Take-home | One paragraph you could repeat to a board member |
| Practice | An exercise on the Lantern Registry, the book's hypothetical registry with synthetic data; the workbook page for your own registry; and questions to check your understanding, with the answers folded until you open them |
Practice materials¶
The Lantern Registry is a hypothetical registry with synthetic data: an invented condition, Epilepsy Syndrome X, invented genes, invented participants, and dates set in the 2040s so that nothing in it can be mistaken for a real export. The data are built with the problems real registry exports have. Every chapter's exercise uses it, and the interactive lab lets you click through its export and flag what needs a query or a correction. The Slow Registry Workbook holds the templates for your own registry, and the AI prompt library has a prompt for each chapter's topic, paired with the step only a person can do. If you are reading with a board, a team, or a class, using this book with a group has discussion prompts and the answer key.
Reading by role¶
| If you are | Start with |
|---|---|
| A parent or advocacy group leader starting a registry | Chapters 1, 2, 3, 5, 7, and 14 |
| A coordinator | Chapters 5, 11, 12, and 13 |
| A board member | Chapters 1, 2, 3, 11, and 16 |
| A data scientist new to registries | Chapters 2, 6, 8, 9, 12, 17, and 18 |
| A clinician or researcher advising a family-led registry | Chapters 1, 2, 6, 8, and 15 |
What you need¶
A web browser and a spreadsheet program are enough for every exercise. If you work in R, Python, or REDCap, the practice files open there too, and no programming is required. Some exercises ask you to search PubMed or ClinicalTrials.gov, both free.
A note on privacy¶
The Lantern Registry is synthetic: its condition, its genes, its participants, and its dates are invented, and no real person's data appear anywhere in the book. Nothing you click in the lab or type in the workbook leaves your browser. When you bring your own registry into an exercise, use counts, registry identifiers, or invented examples, never names, dates of birth, contact details, or anything else that could identify a person.