RigorMed
← Research resources

Research methods guide

Retrospective cohort protocol template

A retrospective cohort protocol starts with the source population, index date, exposure, comparator, and follow-up. Then specify outcomes, confounding, censoring, and analysis.

Study questions and design

Compare two outpatient strategies in a synthetic cohort. Align eligibility, assignment and follow-up to avoid crediting pre-exposure event-free time to the exposed group.

Retrospective cohort protocol skeleton

Replace examples with auditable definitions in your own data.

Retrospective cohort protocol skeleton
ItemWorked exampleVerification
Entry and time zeroInclude eligible new users with no use in the preceding 180 days. Index at first qualifying initiation of either strategy.Verify comparable eligibility, start dates and data coverage.
Outcome and follow-upUnplanned admission within 90 days; retain death and exit separately.Prespecify the algorithm, administrative cutoff, competing events and censoring.
Bias and analysisConsider pre-initiation severity, age and prior utilization; report risk difference with interval.Prespecify overlap, missingness and alternative definitions; avoid treating mediators as baseline confounders.

Apply the method to your materials

  1. Run a feasibility table first

    Check populations, observable windows, key missingness, duplicate outcome codes and site coverage dates before estimating effects.

  2. Freeze protocol and log deviations

    Version definitions, primary and sensitivity analyses, and document data-driven changes to comparators and their implications.

Questions that arise in practice

Can the first completed follow-up be treatment start?

If eligibility begins earlier, account for exposure classification and observation in the intervening period before selecting an analysis.

Verified sources

The official sources below were checked on 2026-11-12. Access conditions and update dates are rechecked before publication.

  1. STROBE Checklists

RigorMed workspace

Start with your research materials

Bring your research question, available materials, and the steps you want to check.