...
| Table of Contents |
|---|
What Data can be Migrated?
Here is an overview of data that can be migrated:
...
The types of customer data which can be migrated include the following
Volunteer Data
This is the information about the volunteer, such as contact, and medical history.
Appointments
Past or future appointments.
Study Participation
Summary info about a volunteer’s previous study participation. This is extremely limited, does not include clinical data, and is meant to support advanced volunteer search.
Configuration
Configurations required to support migrated data include initialization of lists of supported medications, substance uses, and medical conditions.
Often these appear in source systems as data drop-downs.
What are the mappings required to migrate any source data into ClinSpark?
Some data maps cleanly into ClinSpark, like DOB. Any transformations required must be noted in the specification, such as from source system race definitions to those categories used in ClinSpark.
An example is volunteer phone data. Migration is an opportunity to clean source data. We need very explicit instructions with example errors on the source side, and a definition of what transformations are requested.
Another common example is medical history. Substance use data for instance frequently needs to be “cleaned” so that it can be transformed into the ClinSpark representations. To document these transformations between data representations in the source DB and the desired representation in ClinSpark, we suggest using a spreedsheet or alternative method of presenting tabular data.
This approach is ideal since it can be ingested programmatically by the transformation program.
Migration Specification
This document identifies the places in ClinSpark which can receive migrated customer data. For each type of data, it is the customer’s responsibility to identify where in their source system this data can be found.
This document is intended to be used as a worksheet, specifying all migration work which must be done. Any work to be performed by FH IQVIA engineering must first be clearly specified in this document. Upon completion, this document will serve as a record of migration work which has been performed, and as a guide for user testing and validation.
About the Specification and Process
Collaborative Document
The specification document is collaboratively filled out over the course of the migration process. It is not unusual to see multiple editors working on different sections of the docdocument. And a critical part of the workflow is inline comments which can be assigned to readers and support threaded discussions about particular portions of the docdocument.
To support this, the spec specification itself needs to be on a collaboration platform. Foundry Health can provide the spec using the Foundry G Suite docs app, which meets this requirement. If for some reason this is not acceptable to your organization, if there’s another collaborative platform such as Office 365, then it is fine to use that. As long as all collaborators have access to a commenting-enabled collaborative platform things will work fine.
Iterations of specifications and implementations
Each iteration starts with a dedicated specification document. The first iteration tends to be by far the biggest, since this is where the full scope of the migration is spec’d out and a first comprehensive implementation is attempted.
The output of this implementation is then reviewed by the customer. Any gaps, errors or other changes are identified in the review of the iteration deliverable. And the all of these are then documented in a new specification document for iteration 2.
The spec documents for later iterations do not need to adhere to the format of the primary specification. They tend to be focused on particular issues, and they should clearly articulate the change and most importantly, show test cases with specific examples from the test volunteers.
Spec Design
Spec Templates
You can download the Migration Spec Template as a MS Word document here. This should be copied to a collaborative platform such as G Suite (Foundry can provide this) or one chosen by the customer.
About the Requirements
What Data can be Migrated
Here is an overview of data that can be migrated:
INSERT VIDEO HERE
The types of customer data which can be migrated include the following
Volunteer Data
This is the information about the volunteer, such as contact, and medical history.
Appointments
Past or future appointments.
Study Participation
Summary info about a volunteer’s previous study participation. This is extremely limited, does not include clinical data, and is meant to support advanced volunteer search.
Configuration
Configurations required to support migrated data include initialization of lists of supported medications, substance uses, and medical conditions.
Often these appear in source systems as data drop downs.
. We may be able to offer e.g. Teams / Sharepoint / Excel as a collaboration platform to certain customers, and if not, we may be able to use yours.
Example
The following example specification has been designed to be read by a script during the migration process…
...
| View file | ||
|---|---|---|
|
Specification Sections
The specification is broken up into sections, tabs in the workbook where each section tab is a different place in the ClinSpark UI that can potentially receive volunteer information.
For each section, the following areas must be documented:
ClinSpark UI
You’ll see that this is already populated in the specification. A portion of volunteer data will be taken from the ClinSpark UI, and annotated with numbered dots which identify individual fields.
...
This defines the scope of this section of the spec. The remainder of this section will cover other details about these fields.
Customer Source UI
For the numbered dots defined above in ClinSpark, show where in the source UI this data lives.
...
Trace Fields to the Customer Source DB
For each field we need to know how the data shown in the source UI is represented in the source DB.
...
To demonstrate that this is correct, provide the sql example:
Source System Table/Screen
Source System Column/Atttibute
CS Table/Screen
CS Column/Atttibute
Migrate (Y/N)?
Status (Dev Ready/Dev Complete/Validated)
Migration Instructions/Rules
Comments
Then for each data type (Medication Form, Medication Target, Medication Frequency etc.) all the possible options for Source must be listed against their intended target.
Screenshots
Screenshots of target and source should not normally be needed unless we run into difficulties…
...
SQL
It will not normally be needed to provide the SQL for each field in case we run into difficulties…
...
If required, provide the SQL in a textual form which can be copied and later executed by engineers as examples. Provide the execution output showing the test data highlighted in this section’s example to demonstrate it is correct.
...
...
Some data maps cleanly into ClinSpark, like DOB. Any transformations required must be noted in the specification, such as from source system race definitions to those categories used in ClinSpark.
An example is volunteer phone data. Migration is an opportunity to clean source data. We need very explicit instructions with example errors on the source side, and a definition of what transformations are requested.
Another common example is medical history. Substance use data for instance frequently needs to be “cleaned” so that it can be transformed into the ClinSpark representations. To document these transformations between data representations in the source DB and the desired representation in ClinSpark, we suggest using a google sheet. Here is an example Customer Medical History Mapping Definition Template. This approach is ideal since it can be ingested programmatically by the transformation program.
Questions, Issues, Decisions
Place topics for discussion here. Add comments and assign them as appropriate to members of either the Foundry or customer team.
Implementation Evidence
...
Iterations of specifications and implementations
Each iteration starts with the completion of the dedicated specification document. The first iteration tends to be by far the biggest, since this is where the full scope of the migration is spec’d out and a first comprehensive implementation is attempted.
The output of this implementation is then reviewed by the customer. Any gaps, errors or other changes are identified in the review of the iteration deliverable. And then all of these are then updated in the specification document for iteration 2.
The specification changes for later iterations should focus on particular issues, and they should clearly articulate the change and most importantly, reference test cases with specific examples from the test data.
Data Without an Existing Field in ClinSpark
Frequently a source system will have a field for volunteer data which doesn’t have an explicit field in ClinSpark.
For these one option is to create within ClinSpark custom volunteer questions, and migrate the source data into these questions. For example here are some custom questions from a demo system:
...
Project Tracking
In addition to the specification, each migration effort will be managed and tracked (questions, issues, decisions etc.) in a Basecamp sub-project…
...