Missing data plan

Every blank has a cause, and the cause is usually somewhere a person can go and look. This page goes with Chapter 12. Agree the reason codes before collection, follow the plan in the order the work is done, and keep a register of every gap until it is recovered or explained.

	Entry
Registry
Condition
Completed by (role)
Date
Version

Registry identifiers only

Use registry identifiers on this page. Do not enter names, dates of birth, contact details, or anything else that could identify a person.

Reason codes

Code	Meaning	Use it when
Declined	The participant chose not to answer
Unreachable	Contact was attempted and did not succeed
Not applicable	The question does not apply to this participant
Not yet due	The scheduled time for this item has not arrived
Not asked	The item was not on the form version the participant completed, or skip logic hid it
Transfer failed	The data were authorized or sent and did not arrive
Unknown	The participant was asked and does not know
Other	Describe the reason in the gap register

The plan, in the order the work is done

Step	What we will do	Who	When	Done
Design for it
Reconcile expected against received
Profile what is missing
Investigate the cause
Recover what can be recovered
Check for bias
Decide how missing data will be handled, in the analysis plan
Report it

Handling strategy for each use

Finding the cause of each gap comes first. How the gaps are handled in an analysis can differ case by case, with the use, who is using the data, what they are trying to show, and the research question, so the strategy is set for each use and written into its analysis plan.

Use or research question	Who will use the results	What they are trying to show	How missing data will be handled	Where it is written in the analysis plan	Decided by (roles)

Gap register

Field	Records missing	Pattern (site, country, language, wave, form version)	Hypothesis about the cause	Kind (technical, process, design, clinical documentation, participant choice)	Who to ask	Recovered?	Reason code

Reporting statement

The STROBE guidelines ask authors to "explain how missing data were addressed" and to "indicate the number of participants with missing data for each variable of interest" (STROBE checklist).

Variable	Number missing	Why	How the analysis handled it