In section 3.C. of the SPARCS application form (DOH-5132), applicants must list all of their planned linkages. Common linkages are listed and then a table is provided for the applicant to list other data sources. If more space is needed, applicants can use the Project Summary to list all planned linkages. Most data sources can be linked to SPARCS data so long as the data owner has provided approval for the linking and there is no risk of reidentification of individuals.
Publicly available data sources can also be linked to SPARCS data. Even though these data sources are available for anyone to use, they must still be listed as a planned linkage in your application. An example of a publicly available data source is the U.S. Census data.
When evaluating a planned linkage, SPARCS staff assess the feasibility of the match (are there enough common elements) and the risk of allowing the match (can patients be reidentified).
Please review the list below of circumstances before requesting a linkage,
- SPARCS will not approve any request for SPARCS linkage unless the applicant provides explicit documentation proving that the outside dataset can be matched to other data sources. This is to ensure SPARCS is following good governance practices and applicants are good data stewards.
- To protect the privacy of patients, SPARCS will not approve any requests where there is a risk that individuals can be reidentified based on the data available to the applicant already.
- New York City Vital Statistics will not allow the identifiable element ‘Address’ to be released if NYC birth or death data is requested. This is to protect the privacy of patients and reduce identification risk.
- SPARCS does not oversee other NYC or New York State data systems. As a result, SPARCS is not authorized to provide data such as vital statistics directly to an applicant.
- Identifiable elements will not be released to an applicant for them to do their own personal cohort matching or to create a control group to accompany their cohort. In order for an applicant to make a valid match, SPARCS would need to release multiple identifying elements and this represents an unnecessary risk to the privacy of patients.
- Identifiable elements (Date of Birth, etc.) will not be released so an applicant can validate a linkage performed by NYS SPARCS team. SPARCS does not allow the release of identifiable elements for this purpose.
- The Address identifiable data element will not be released for an applicant to receive the ZIP code extension (ZIP+4) to match SPARCS data to US Census or American Community Survey public datasets. This element is not necessary because these datasets do not contain ZIP+4.