Data source plan¶
Decide what to capture, from where, and at what granularity, before you go looking. This page goes with Chapter 4. Fill in a row for every field or concept your users need, then check the plan against real records before you design a single form.
| Entry | |
|---|---|
| Registry | |
| Condition | |
| Completed by (role) | |
| Date | |
| Version |
Validated instruments and common data elements
A validated instrument is a tested measure with a known scoring rule, administered as its developers intended; changing its items breaks comparison with studies that used it as intended. A common data element is a standard question with defined response options, asked the same way across studies; adding a response category generally does no harm, while deleting or collapsing categories makes comparison impossible.
Field by field¶
| Field or concept | Who needs it | Best source (health record, survey, diary or app, clinician, lab report, device) | How often | Granularity and unit | Standard (validated instrument, common data element, vocabulary) | Checked against real records? |
|---|---|---|---|---|---|---|
Representative record review¶
Look at real records, your own or a representative participant's with consent, and record what is there and what is not.
| Record reviewed (registry identifier, or my own) | What was there | What was missing | In structured fields, or in notes only |
|---|---|---|---|