2017 Project Plan

This is where I keep notes on what needs to get done this year, well, at least the big things.

Patient Record Maintenance

This focuses on three supplements: Metadata Update, Delete Metadata, Delete Documents

On Metadata Update - the supplement needs major work to ready it for final text later this year. Now that the REDOC effort has laid in some supporting work it might be possible to finally clean up MU. While it is not possible to list all the things it needs, here are some highlights. The work is structured as a collection of separate requests that can be encoded in a transaction. We have not yet found a decent way to convey this structure in documentation. Next is the structure of the separate requests. They are documented sections: preconditions, trigger, postconditions etc. They are easy enough to understand in isolation but very hard to comprehend when used together. They need to be linked and supported with lots of examples. The rules for the separate requests are supported by a collection of common rules and conditions. These need to be reformulated/reorganized. Publication: August

Tooling - we currently have no tooling to support this profile beyond our Document Registry simulator. The primary need is in visualization at two levels - intra-object and inter-object. The intra-object visualization would display/highlight the differences between two versions of an object. The inter-object visualization would display/highlight the differences in the collection of objects making up the patient record for a single patient. The would also support Metadata Delete in that the same view would cover both.

Tests - while we will be able to get some examples introduced into the profile, the test cases end up being an executable collection of example of how to use the profile to get things done. I anticipate around 40 test cases are needed.

On Delete Metadata - this shares a lot of structure with Metadata Update in that there is common metadata information model that must be maintained and validated against. Much of the description of this model is through non-normative diagrams in the TF. For Delete and Update Metadata to be feasible this needs formalizing. How much is not yet determined.

Tooling - we currently have no tooling to support this profile. First it needs to be added to the Document Registry simulator. Then the visualization support for Metadata Update would also cover Delete Metadata.

Tests - Two separate test suites are needed - one where Metadata Update is present and one where Metadata Update is not present. Each of these is likely around 30 tests.

On Delete Documents - this is a separate work that is rather simple and will need little effort at the profile level.

Tooling - add support to the Document Repository simulator.

Tests - three or four tests are necessary.

FHIR bindings

This covers the MHD and PIXm profiles.

I have no responsibility for profile work in this area.

MHD Tooling - translators between the MHD and XDS metadata and message formats. These will be offered separately and bound to the existing Document Registry simulator. This will support cross MHD/XDS testing. While there are existing libraries for FHIR validation, these will need extending to handle the FHIR extensions built into MHD.

MHD tests - first is a translation of all current applicable XDS Registry Simulator tests. Additional tests may be needed but this analysis has not yet been done.

PIXm tooling - patient identity feed tool to support testing of MHD based registries.

PIXm tests - do not intend to get into this area.

UI

The work on updating the toolkit UI went well in 2016. There are additional areas that need work this year.

There are UI implications from the above Patient Record Maintenance and FHIR work items above. Beyond that...

Simulator log browser - this tool need a total rewrite. There is a great deal of addition functionality needed and the presentation is horrible.

Conformance tool updates - having gotten a half season's feedback on this new tool there are a significant number of updates needed. Some of these are UI related and some are related to the need to formalize the orchestration model that underlies the tool.

Basic tools (queries/retrieves/submissions) - the UI presentation has not kept up with the needs of this group of tools. There are grouped here because they have a common support code base. Some of the UI code from the Conformance tool can be used here.

Depth of testing

I need to do a walk-through/analysis of the current tests and validators and compare them to the updates made to the TF over the last few seasons. Many enhancements are missing. This includes both positive and negative tests.

Documentation

The GitHub wiki has worked well last year as a base for documentation. There is still a lot to write to catch up the documentation to the code base.

Summary

This is more than a year's work and I haven't yet had a chance to set priorities.