Define the unit before entering data
If one person can have multiple episodes, decide how those will be represented. Otherwise, repeated reports may be mistaken for separate cases or genuine episodes may be merged.
Use a stable investigation identifier and keep identifying information appropriately protected.
A focused starting structure
| Field group | Example content |
|---|---|
| Identification | Unique case or episode code |
| Timing | Onset, specimen, and report dates |
| Classification | Current case status and version |
| Clinical information | Relevant symptoms and outcomes |
| Exposure | Defined exposures in the relevant period |
The actual fields depend on the investigation. Collecting every imaginable variable can slow work and increase privacy risk.
Unknown is not no
Use explicit codes for unknown, not asked, not applicable, and absent when those distinctions matter. A blank cell should not silently become a negative exposure.
Preserve updates
Case status and dates may change as information improves. Record the source and update history rather than overwriting important uncertainty without a trace.
A line list is a working investigation tool, not a static spreadsheet created once. Clear definitions, validation checks, controlled access, and regular review make later curves and comparisons more trustworthy because the underlying records remain interpretable.
