Segments are a powerful tool that allows you to track participant milestones and drive actions. See below for segment best practices, as well as examples of how segments may be useful for your project.
Table of Contents
Many teams find a benefit from monitoring participants grouped by their current milestone. Consider the various milestones of your project, e.g. Baseline Survey Complete, Surgery and Post-Surgery Survey complete, etc. You’ll likely want to create segments for each of these milestones, especially if coordinator intervention is required.
Once you have defined your milestones, you may find that you have two kinds:
- Milestones based on days, survey completion, or other events/data in MyDataHelps
- Milestones based on events external to MyDataHelps
For milestones related to MyDataHelps activity, refer to the Filter Criteria Types article for information on how to define these segments. If you have milestones that depend upon events external to MyDataHelps, consider creating a coordinator survey to record when the milestone occurs. Then, you can use Survey Answer filter criteria to determine when the milestone occurred.
MyDataHelps Designer organizes segments in alphabetical order. To order segments in a manner that is functional and defines the segment, consider prefixing your segment names with numbers (e.g. “01. Segment A”, “02. Segment B”, etc.) to ensure they appear in an order that works for you and your team.
Create Segments: Examples
The following sections review practical applications for segments that may be useful for building your study.
You may wish to send reminder notifications to participants invited to your study, but have not yet enrolled. This example reviews creating an email notification, creating a segment, and creating a schedule to notify all participants who were invited to participate over 1 week ago and have an invitation status of pending.
First, create the email notification to send to participants.
Next, define the participant segment. In this scenario, participants were invited to the study by the coordinators, so you can use the Creation Date and Invitation Status criteria to determine if a participant was created in MyDataHelps over a week ago and has a status of pending.
Finally, create a Schedule to send the email reminder to participants on entering this segment. This results in a one-time notification for any participants that were invited over 1 week ago and have an invitation status of pending.
You may wish to divide participants into separate cohorts and send notifications based on those assigned cohorts. In this example, participants are assigned to a cohort based on a survey response, and the cohorts are used to determine the frequency of notifications.
- Cohort A receives 0 notifications
- Cohort B receives 1 notification per week
- Cohort C receives notifications every other day
First, create a custom field to store the cohort survey response in.
Once the custom field has been created, create a segment for each cohort (Cohort A example below).
Next, define the survey step in which you will assign the cohort. In this example, participants are placed into a cohort based on their activity level.
Assuming the notification has already been created, the last step is to schedule the notifications for each cohort. Since cohort A will not receive notifications, start by scheduling notifications for cohort B. Cohort B should start receiving notifications 1 week after Enrollment and continue to receive notifications every week for 10 weeks.
Cohort C should start receiving notifications 1 week after Enrollment and continue to receive notifications every other day for 10 weeks.
Monitoring Linked Accounts
For studies that collect EHR or Fitbit data, teams often desire to monitor participants' Linked Accounts, determining if any require re-connecting. A simple segment with the following criteria can allow your team to monitor when accounts need to be reconnected.
Device (e.g. Fitbit) reconnect
Survey Creation and Due Date Segment
Segments can filter by survey creation and due dates. This can be useful to investigate participants with recently created and delivered surveys, as well as determine which participants have overdue surveys.