Page Comparison

Generalalised Guifance Generalised Guidance for Creating new FAIR Clinical Data Sets

Dataset to be collected: Diagnosis data

from Scratch

This is based on the comment attached to “Actor perspective” written over here and should be further discussed/elaborated. Really drafty.

General Approach

1. Define FAIR Objectives & deliverables (Guidance:

...

Objectives

• make data available for future research

• ensure others can understand the data

• ensure data can be with other datasets

Deliverables

• Queryable FAIR diagnosis data

• publish dataset in Castor FDP, which is harvested by HRI catalogue

Collecting the data (Guidance: see Metroline XYZ)

• Build the Cator EDC study with the fields of interest (codebooks?)

• Communities already over here?

• Collect the data

Map (Guidance: see Metroline XYZ)

Publishing the dataset

...

1. Define your objectives for making data FAIR)

• Here we assume the goal is to

  • Create a new FAIR Clinical dataset

  • Publish this dataset to make it available for others

  • ?

2. Build study for data collection (Guidance: see Metroline XYZ)

• Communities already over here?

  • Assuming you want to already collect the data in a standardised way, build the study using knowledge from the communities? Reusing existing eCRFs and Codebooks?

  • (Find out whether work has already been done, which you can reuse. Perhaps a link to a page which talks about the communities? Or perhaps a page which talks about reusing what’s already out there?)

• Build the eCRFs with the fields of interest in the EDC

3. Collect the (meta)data (Guidance: see XYZ)

• Collect the necessary / desired metadata requirements;

  • Required: Core metadata mode

  • Desired: Health metadata model (is this extended core?)

...

• • Desired: Domain specific model(s) (Petals)

• Collect the data

  • Collect the data in the EDC

4. Map (Guidance: see Metroline XYZ)

5. Publish the dataset (Guidance: XYZ)

Applied Example

Use case: Researcher will collect Diagnosis data in Castor EDC and wants to make this dataset FAIR.

Step 1: Define FAIR Objectives & deliverables

Objectives

• make data available for future research

• ensure others can understand the data

• ensure data can be with other datasets

Deliverables

• Queryable FAIR diagnosis data

• Publish dataset in Castor FDP, which is harvested by HRI catalogue (is it?)

Step 2: Build study for data collection

• Researcher searches HRI codebook library for a preferred codebook with Diagnosis items

• Researchers builds a Diagnosis eCRF in Castor using suitable tools

• (In case items are missing, does she join something like the clinical community? Maybe she should, but I’m not sure whether most researchers would)