Longitudinal vs Repeating Design
- Melissa Hall
- Terry Nguyen
Longitudinal data collection allows instruments to be reused in multiple places (arms/events)
Repeating instruments allows instruments to be reused in the same place
Repeating events allows the same set of instruments to be reused
Longitudinal Data Collection
Enabling longitudinal data collection allows the data collection instruments to be organized into events.
In the example above data collection has been broken into 5 events, most of which are based chronologically. The ‘Quality of Life Survey' and 'Check-in’ forms appear multiple times for each record, but can only be filled out once for each event.
Repeating Instruments
Enabling repeating instruments allows a form to be filled out again an unlimited number of times.
Repeating instruments can be used with or without longitudinal data collection enabled.
In the example above, the ‘Quality of Life Survey' and ‘Check-in’ forms can be filled out multiple times using the '+’ button. The record home page displays all the instances for easy viewing. In this case, labels are added using dates to make the instances easy to tell apart from one another.
Repeating Events
Enabling repeating events allows a longitudinal event to be filled out again an unlimited number of times.
Repeating events are only available if longitudinal data collection is enabled.
In the example above data collection has been broken into 3 events. The ‘Quality of Life Survey' and 'Check-in’ forms appear multiple times for each record.
Rather than define each ‘Monthly’ event ahead of time and number each, the event is allowed to repeat by clicking the ‘+ Add new’ button. This allows the set of instruments belonging to the 'Monthly Data' event to be reused.
Design Considerations
Is the same form going to be filled out multiple times for the same record?
Example: patient visits, check-ins, lab results that are conducted multiple times, adverse events
No: don’t enable longitudinal data collection or repeating instruments
Yes, and:
Scenario | Example | Possible REDCap design |
|---|---|---|
The form reused a set number of times by all records | 2, 4, and 6 week visits | Longitudinal events |
The form is reused as needed - or - The form will be reused an unknown number of times | Unscheduled visits, complications | Repeating instrument |
The form is reused in different kinds of 'events' | Comparing metrics before/after an intervention | Longitudinal events |
The form is mostly reused, but there are slight differences | Comparing metrics before/after an intervention, now including additional questions | No special design considerations needed. Copy and create a separate instrument - or - Longitudinal events, but certain questions are hidden using branching logic |
There is a series of forms that will be repeated |
|
|
Are some forms not applicable to certain records?
Example: form applicable to certain age groups,
No: all records can fill out the forms
Yes, because:
Scenario | Example | Possible REDCap design |
|---|---|---|
The study populations are different | Intervention group and control group | Longitudinal arms |
The form that’s used changes depending on the participant | Different forms for adult participants vs children | Form display logic |
The form may or may not be used | Study termination, complications | No special design considerations needed. Some records will contain no data for that form. |
Some forms change depending on the study site | Consent forms differ by region | Longitudinal arms - or - Form display logic |