Research methods guide
How to define exposures and outcomes
Document each concept and its operational definition. Include source fields, coding rules, time window, units, missingness, and validation.
To study initiation of remote BP management and 90-day unplanned admission, define reproducible algorithms and separate positive, negative and indeterminate cases.
Exposure, outcome and time dictionary
Field names are illustrative and must be mapped to actual data.
| Item | Worked example | Verification |
|---|---|---|
| Exposure | Active enrollment and first transmission timestamp; use an eligible encounter date for the comparator. | Inspect activation lag and address grace periods and immortal time if using future uploads. |
| Outcome | First unplanned admission in days 1–90, deduplicated by type and event time. | Exclude planned stays and duplicate transfers; state out-of-system coverage. |
| End of observation | Keep death, end of coverage and day 90 separately. | Death may compete with admission; unknown follow-up is not a confirmed non-event. |
Apply the method to your materials
Review boundary and discordant records
Test same-day admissions, day 90, transfers and coverage gaps with a subject-matter reviewer.
Preserve primary and sensitivity definitions
Version code, units, windows, missing codes and sample review. Assess the influence of plausible alternatives.
Questions that arise in practice
Does an ICD code list define the outcome?
Also specify field position, record type, timing, duplicates and coverage, then compare with chart review or validation evidence.
Verified sources
The official sources below were checked on 2026-11-12. Access conditions and update dates are rechecked before publication.
RigorMed workspace
Start with your research materials
Bring your research question, available materials, and the steps you want to check.
