Retrospective descriptive cohort
What happened to a defined cohort in existing records?
When to use this project
Use a retrospective descriptive cohort to describe care, events or clinical trajectories over a historical follow-up period. The question is about the experience of an eligible population, rather than the causal effect of choosing one treatment over another.
How it differs
Unlike a cross-sectional snapshot, this design requires a time origin and follow-up. Unlike a comparative cohort, it does not require a causal treatment contrast. Retrospective describes when the records are examined, not whether the outcome preceded cohort entry.
Decisions before you start
- Define entry independently of future outcomes and specify available look-back.
- Separate people, episodes, first events, recurrent events and person-time.
- Specify censoring, competing events and how incomplete outcome capture limits estimates.
How the flows work together
Each flow below links to its diagram, steps and tables. Flow order organizes the project; it does not imply that all work is strictly sequential. Data assessment, analysis planning and protocol refinement may inform one another.
1. Cohort overview
Overview · Question through longitudinal study reporting
Use this overview to state the purpose and link to detailed decisions. Keep a versioned protocol and decision register, and add evidence references for the rationale.
2. Sources and data preparation
Data · Establish suitable analysis-ready data
A large dataset can still omit the clinical variables or follow-up needed for the question. Review source coverage, linkage and changes in coding before finalizing eligibility and analysis. A recorded absence may mean no capture rather than no disease.
3. Population and observation periods
Selection · Define entry and follow-up rules
Define time zero and the population at risk before counting events. Distinguish unique people from episodes and person-time. Death can be a competing event rather than ordinary loss to follow-up, depending on the outcome.
4. Variables and outcomes
Design · Define measurements and event ascertainment
Operational definitions translate clinical concepts into recorded variables. Retain code versions, look-back windows and validation evidence. Distinguish a missing measurement from a measured absence and examine differential outcome capture.
5. Estimation and sensitivity plan
Statistics · Specify estimates and robustness checks
Choose the measure before the denominator: prevalence, cumulative risk and incidence rate answer different questions. Report descriptive estimates with uncertainty and avoid attributing observed outcomes to a treatment without a comparative causal design.
Equations and statistical interpretation
These explanations describe notation and assumptions. Flow does not calculate these values.
Incidence rate
D is the number of eligible incident events and Tᵢ is observed time at risk for person i. State the time unit and whether the outcome is first occurrence or recurrent events.
A rate is not a probability. An event count divided by the initial cohort size is not a valid fixed-horizon risk when follow-up is incomplete; specify a survival or competing-risk approach as appropriate.
Published example
Use the clinical-course description as the example for this starter. The paper also contains risk-factor analyses, which go beyond a purely descriptive aim. Examine its restriction to patients with completed hospital outcomes and the resulting selection limitations.
This publication illustrates a study design, not an endorsement of every method or a reproduction of its results. Follow its protocol, supplement and update history when studying the example.
Hypothetical scenario
A hospital wants to describe one-year readmissions among adults discharged after a first recorded heart-failure admission using its existing linked records.
- Cohort overview: Define discharge as time zero and one-year descriptive outcomes.
- Sources and data preparation: Check linkage to readmissions and deaths, data permissions and changes in coding.
- Population and observation periods: Define first recorded admission, required look-back, repeat episodes and observation end.
- Variables and outcomes: Version heart-failure and readmission algorithms and assess records outside the hospital network.
- Estimation and sensitivity plan: Estimate event frequencies or cumulative incidence with death handled according to the question; report capture limitations.
Deliver a populated protocol, decision log, data specification, and planned outputs. Enter numerical findings only after analysis and verification. Keep decisions made before seeing results distinguishable from later amendments.
Sources
- STROBE checklist for cohort studies
- RECORD checklist for studies using routinely collected health data
- HMA–EMA Catalogue of RWD studies: registration and protocol guidance
Source check: 2026-09-17. Record the version of each reporting or appraisal tool used in your actual protocol.
Compare all Observational and RWE studies project types · Open editor in a new tab