Coding and Data Standardization
Coding converts verbatim text (adverse events, medications, medical history) into standardized terms for analysis and regulatory submission. Know the dictionaries: MedDRA for adverse events and medical history (with System Organ Class and Preferred Term hierarchy, and version control), and WHO Drug (or equivalent) for concomitant medications (with ATC classification). Coding must be consistent, version-documented, and performed by trained coders.
Key concepts: verbatim term vs. coded term; the coding plan (which dictionary, which version, who codes, how often); autocoding vs. manual coding; and the requirement that coding is reproducible and auditable. Version upgrades mid-study require a documented migration plan because terms and hierarchies change between versions.
Traps: coding an AE to a term that changes its meaning (e.g., coding "head pain" to a more severe term); using different dictionary versions for different subjects without documentation; coding medications to MedDRA (wrong dictionary - MedDRA is for events, WHO Drug for medications); and failing to re-code when a new dictionary version is adopted. Another trap is assuming autocoding is always correct - it must be reviewed.
Rule of thumb: code to the most specific term that faithfully represents the verbatim, and never let coding alter clinical meaning. Worked example: verbatim "bad headache" is coded to the MedDRA Preferred Term "Headache," not to "Migraine" unless the source supports it. If the study upgrades MedDRA versions, the coding plan documents the migration and the database records which version coded each term.