Table 1

Differentiating requirements for a representative ancillary study workflow
Step # Step Name Key Differentiating Requirements
1 Hypothesis generation a. Identify interesting categories of participants from primary study
b. Access existing data and specimen information for those participants
c. Provide sufficient information to external investigators for them to propose ancillary studies
2 Proposal review a. Review availability of existing specimens
b. Review priorities for use of remaining specimens, possibly reserving specimens
c. Evaluate overlap and comparability of existing data and proposed measurements/analyses
3 Creation of protocol or plan a. Decide which existing participant data and specimens to use.
b. Plan expectations for collecting new ancillary data complementary to existing primary data
4 Consent verification or acquisition a. Determine whether consent exists and is sufficient for desired analyses
b. Obtain additional consent if necessary
5 Retrieval of existing data a. Compile relevant subset of primary study data required for the ancillary study
b. If external investigators are leading the ancillary study, share this subset of data with them.
6 Delivery and analysis of specimens a. Request, locate, ship and track existing specimens, as well as manage material transfer agreements
b. Maintain identifiers relevant to primary study during further specimen analysis
7 Data integration a. Retain context and provenance from both primary and ancillary studies, including processing and quality control information
b. Retain origination information (primary vs. ancillary)
c. Resolve differences (representation, quality control, etc.) and join ancillary and primary data
8 Data/specimen repatriation a. Contribute ancillary study data (raw and/or processed) back to the primary study
b. Retain data context, provenance, processing and other metadata from ancillary study
c. Return unused specimens
9 Publication a. Coordinate preparation and review of publications across primary and ancillary investigators

This table calls out the key differentiating requirements for each step in the ancillary study workflow described in this paper. It aims to clarify how ancillary study requirements extend beyond those of primary studies and secondary data analysis. Note that this workflow addresses the “freezer study” scenario, where the ancillary study’s additional measurements come from analysis of existing specimens, not collection of new specimens or clinical data.

Nelson et al.

Nelson et al. BMC Medical Informatics and Decision Making 2013 13:5   doi:10.1186/1472-6947-13-5

Open Data