Clinical Data Science in R · Course
Clinical AI and translational trauma research in four parts
EIGHT WORKSHOPS · CLINICAL DATA · R
Clinical Data Science in R
Clinical AI and translational trauma research in four parts. Read the evidence, test a model and defend a clinical decision.
A fracture alert can perform well on a test data set and still help little in the emergency department. A model can reliably recognise further images of the same laboratory animal and fail on a new one. This course starts from such decisions. Participants learn to read a finding, carry out the appropriate check in R themselves, and defend a limited recommendation to patients, researchers or hospital management.
| Course profile | |
|---|---|
| For | Medical students, clinical researchers and doctoral researchers who want to judge data from clinical care and run their own analyses in R. |
| Entry | Basic clinical concepts. A first R session is part of the first workshop; no prior programming, Python, Julia or machine learning needed. |
| Scope | Four parts, eight workshops (W1–W8), four teaching units of 45 minutes each: 32 units = 24 hours of contact time, plus about 36 hours of self-study. |
| Setting | Laptop with R, or printed outputs on paper. All required cases use synthetic data. No external AI account is needed. |
| Completion | An individual examination and a group project with an executable R analysis, a model passport and a reasoned decision. |
| Companion book | Clinical Data Science in R. Chapters B01–B28 are fixed identifiers; each workshop requires the core readings of two chapters. |
Explore the workshop and exercise network to browse by part, method or search term.
Four parts, eight workshops
I · Clinical questions and data (entry)
What exactly is being claimed here?
W1 · From clinical case to testable question · W2 · Language models and scientific evidence
II · Testing and comparing models (application)
Does the evidence hold in my setting?
W3 · Diagnostics, prediction and external evidence · W4 · What imaging, gait and tissue models measure
III · Justifying complex models (advanced)
Which biology or process lies behind the data?
W5 · Causal questions, trajectories and phenotypes · W6 · Pharmacometrics, forecasting and simulation
IV · Translating models into practice (independent application)
Which decision is defensible today?
W7 · Regulation, data protection and a security incident · W8 · Model colloquium, decision and review
In Part I participants formulate a clinical question and check where their data come from. In Part II they test a model, in Part III they compare explanations for their findings, and in Part IV they defend a decision about its use. One group project runs through all four parts. Start with the syllabus for workload, assessment and organisation.
How a workshop works
Each workshop page states the guiding question, observable learning objectives, the required reading, a table of the four teaching units with the evidence each one produces, the materials and the transition to the next workshop. The labs contain the full R code with real output; every lab also runs as a stand-alone script in a fresh R session and checks its target values with stopifnot(). Legal and role-play parts come as case sheets without code.
Every lab follows the same five steps: state the decision, look at the data, name the assumption, compute in R and report in one sentence for the person who has to act.
Connection to existing courses and books
This course does not repeat general statistics. Where a method is introduced only briefly, workshop and lab pages link to the place where it is explained in full.
Progression. The four parts follow the stages of the CTTIR statistics curriculum, which leads from introductory foundations (course 1) through intermediate regression and model diagnostics (course 2) and advanced study design, longitudinal data and causal inference (course 3) to specialist statistical learning (course 4). Here the same progression is applied to clinical decisions: entry (Part I), application (Part II), advanced (Part III) and independent application (Part IV). The five steps of every lab mirror the curriculum’s lab template, in which inference labs move from hypothesis through visualisation, assumptions and conduct to conclusion, and workflow labs from goal through approach, execution and check to report. Participants who have worked through the CTTIR courses will recognise the rhythm; those who have not can use them as a reference for every method used here.
| Resource | What it offers this course |
|---|---|
| CTTIR Courses | A four-course biostatistics curriculum in R with runnable labs. The workshop pages link to the matching labs, for example on diagnostic testing, mixed models or cross-validation. |
| CTTIR Tutorials | Stand-alone tutorials by topic, useful for looking up a single method during a lab. |
| Methods in R | A method reference for researchers and reviewers, with diagnostics, red flags and reporting templates such as TRIPOD-AI and ARRIVE. |
| Scientometrics in R | Case studies, labs and reference material on research literature, full text and reproducible pipelines; relevant to checking sources in W2. |
| Strategy in R | Simulation with explicit assumptions and decisions under uncertainty; relevant to W6 and to the ethics of automated decisions. |
| Clinical Data Science in R (companion book) | The cases, core readings and complete R labs behind every workshop. Course and book share cases, identifiers and figure style. |
Data status
All numbers are labelled as synthetic (generated in the course with a fixed seed), published (taken from a cited source), assumption (set for teaching; would need local checking) or open (a question the course raises but does not answer). No number describes a commercial product, a real hospital or unpublished project results. Fictitious animals and virtual pharmacokinetic parameters are used throughout; nothing on this site is a dosing recommendation.
Licence and citation
Texts are licensed under CC BY 4.0, code under the MIT licence.
Heller, R. (2026). Clinical Data Science in R · Course. https://cttir.github.io/clinical-data-science-in-r-course/
This course does not replace individual medical, legal or statistical advice.