The EuroRec/Q-REC Repository of Certification Criteria

Figure 1 illustrates the way in which EHR quality criteria will be extracted from source materials and added to the repository, refined and indexed within it, and then subsequently extracted by end users when compiling test plans and procurement specifications. A similar approach will be used for adding and indexing complementary reference artefacts: standards, terminology systems, archetypes, XML schemata and open source components; this latter set will usually not be physically stored in the repository, but will be referenced to their originating locations and described in the repository.

This section of the specification elaborates the workflow process that applies to EHR quality and conformance criteria.

Original source documents are identified by the EuroRec Institute, national ProRec Centres or by other international experts. These should be formalised statements of requirement, functional specifications, conformance criteria or test plans that apply in whole or in part to EHR systems and preferably have been used within at least one jurisdiction, or at minimum have been subject to a formal peer review process (such as a ballot). The goal is to prioritise those specifications that have practical credibility, but it is also recognised that the implementation and adoption of EHR systems (as opposed to organisation-based health information systems or setting specific clinical applications) is relatively new: it is therefore expected that some areas of functional requirement will not yet exist within validated instruments and that other source documents such as de facto and de jure standards may be important to include. It is likely that some new specifications will be added to complete and maintain the EuroRec repository.

1. Candidate original specifications will be approved for inclusion by a Q-REC Editorial Board (to be appointed).
2. The original statements within approved documents will first be translated into English, together with relevant section headings or category names, aiming at faithful translation with no refinement or disambiguation (except as needed for the translation itself). The Q-REC central repository content will, for the foreseeable future, be exclusively in English. There might later be national repositories, maintained by ProRec Centres, with content translated from the English repository.
3. These translated original statements “Source Statements” (and their headings) will be imported into the repository, and indexed by each of the defined axes against high-level terms. Because Source Statements might at times have a complex or compound content, or be loosely expressed, this index will be deliberately over-inclusive rather than precise. The aim of this indexation is to assist with internal repository management and quality assurance. Being useful to external end-users when searching the repository is not as such a goal of this indexation. Each Source Statement will reference the original document from which it has come (although the source document might not always be physically stored in the repository, and might not always have been translated in full into English).
4. Each of these Source Statements will then be decomposed into one or more specific and singularly focussed Fine Grained Statements. The language might be re-expressed to avoid ambiguity. On occasions multiple very similar statements or sub-clauses from the same original source might be combined in making these new Fine Grained Statements. Each Fine Grained Statement will – at first - reference any primary Source Statement(s) from which it has been derived.
5. Each Fine Grained Statement will be indexed by each of the defined axes, using a detailed term in each case. For some axes, multiple classifications is permitted, but the goal should be to balance sensitivity and specificity in this process, so that subsequent searches do not overlook relevant statements but are also not overloaded with barely-relevant content.
6. Using the indexation as a means of reviewing and aggregating similar Fine Grained Statements across multiple sources, a further set of new broader and potentially richer Best Practice Requirements will be defined. Each of these will also be indexed, but such indexing may be straightforward as it might inherit the indices of the Fine Grained Statement(s) from which it has been derived.
7. Over time, some of these Best Practice Requirements will become validated through adoption within test plans. In such cases a corresponding Test Criterion will be added to the repository, and indexed.
8. With experience, certain groupings of Best Practice Requirements and subsequent Test Criteria will be found suitable for given EHR system modules and care settings. These might correspond to predefined queries on the repository based on certain index term values. These groupings will be defined as EuroRec Profiles, and will be made available as standardised sets to help foster consistent quality of EHR systems across Europe.
9. End users accessing the repository will be able to choose one of the EuroRec Profiles or define a customised search to identify relevant statements. These might then be exported so that the end user can localise them if necessary, and incorporate them into procurement specifications, test plans etc.
10. National ProRec Centres will be involved in supporting local end users, re-translating these exported statements, and working with vendors and purchasers to help validate and refine the Q-REC materials. The Centres will feed this experience back to Q-REC.