While the States have the technical capacity for providing data to a national registry, it is clear that a detailed design of the system—from both a technical and a policy point of view—is critical to implementing such a registry.
A national registry would be most useful if, at a minimum, it included perpetrator characteristics needed to produce high probability matches. While 100 percent accurate matches would not be attainable, the registry should be capable of producing high probability matches that can then be verified with individual States. In addition, it should be able to minimize the number of false negatives so that users can have a reasonable expectation that nonmatches indicate the absence of a substantiated maltreatment record in other States.
The prevalence analyses indicate that a registry that includes name, sex, and date of birth may be sufficient to the task. States collect and maintain this information electronically, although date of birth may not have been verified.
Three important aspects of the technical design of a registry to consider include:
- What would the process entail for identifying authorized users of the registry? How would these persons be identified, registered, and periodically confirmed as authorized users?
- What is the minimal data set that would be the most efficient and effective to maintain on a system?
- Would it increase usage if certain types of cases or certain types of perpetrators were excluded from the system?
Another critical feature of the system's design would be the ability to provide data in a timely and accurate manner. If data were found to be out-of-date or inaccurate, the utility of the system would be highly compromised. Given that States also will need to expunge data from a national registry, the design of the system would need to incorporate an ability of each State to update the system at any time. The best way for States to ensure that the data are accurate may be for them to refresh their entire list of perpetrators during each data submission rather than adding and expunging individual names from the registry. More work will be needed to determine the most efficient design of a system, which could be updated at frequent intervals, approximating real-time data.
While State data systems appear to have the capacity to provide acceptably accurate and timely data to a national registry, the effort is likely to require additional resources to accomplish. Although voluntary, if a significant number of States do not provide data relatively quickly to a national registry, the registry would not be useful. To participate within a relatively short time, States will need adequate resources and technical assistance.
"ResearchReport.pdf" (pdf, 4.61Mb)
"ReportMethodology.pdf" (pdf, 738.21Kb)
"CaseLawReview.pdf" (pdf, 598.85Kb)