| | 1 | = Data Validation in Onyx Questionnaire = |
| | 2 | |
| | 3 | Most data received in the questionnaire is patient-reported, and therefore cannot be easily validated without calling the patient and repeating the questions. However, notes-derived data can, and therefore should, be validated. |
| | 4 | |
| | 5 | Following discussion with Philippe Laflamme and Vincent Ferretti, the plan for validating the notes-derived content in Onyx is as follows: |
| | 6 | |
| | 7 | Create an additional stage within Onyx questionnaire containing exactly the same questions as the Notes Data stage. The new stage to be optional. A different member of staff reviews the notes and enters data into the new stage. At export time, the answers to the two identical stages are exported together to a location (excel?) for comparison and analysis. |
| | 8 | |
| | 9 | We can decide a certain proportion of cases to process for validation. Provisionally we have decided to validate the notes data for 10% of the participants. |
| | 10 | |
| | 11 | == Notes to support BRICCS onyx data validation process == |
| | 12 | |
| | 13 | In Onyx, select the 'all participants' view. Sort by appointment time, then every tenth case should be selected for validation. If the tenth case has been completed, use the next available 'in progress' interview, but continue to count ten from the 'correct' case, to find the next one, i.e. if case 30 is completed, validate 31 but next one is 40 not 41. |
| | 14 | |
| | 15 | Copy the enrollment ID and participant ID for the selected cases into the Excel sheet: 'BRICCS validation cases to do.xls' in the UHL BRICCS drive. |
| | 16 | |
| | 17 | When a case has been validated, a date should be entered in the Excel sheet and the case can then be marked as completed in Onyx. |
| | 18 | |
| | 19 | Cases should not be completed in Onyx unless either they are NOT selected for validation OR they have been validated and a date entered in the Excel sheet. |
