jkiddo/ccdascorecard

Automatic C-CDA scorecard to promote best practices

C-CDA Scorecard

Demo online

http://ccda-scorecard.smartplatforms.org

Promoting best practices in C-CDA generation

Achieving interoperability with Consolidated CDA documents means agreeing on "best practices" above and beyond the base specification. This tool captures best practices as rubrics in code, so you can automatically assess whether your documents conform. For example...

Example Best Practice: "Problem List entries should have one consistent status."

The C-CDA spec provides three distinct places to record a problem status. The best practice recommendation is to make sure they all agree! See below for an example of a rubric to help automate testing for this best practice.

Using and Extending

Setup

• nodejs
• mongodb
• libxml2

$ git clone https://github.com/chb/ccdaScorecard
$ cd ccdaScorecard
$ npm update

Run

node launch.js

Write new and better rubrics

Check out our initial examples in rubrics

Build a better client

This repo includes a simple example Web UI built agains the CCDA Scorecard REST API. It's implemented as a static HTML5 / JavaScript Web app at: public/ccdaScorecard/index.html

This can serve as a launch point for bigger and better.

The pieces and the REST API

Rubrics

GET /v1/ccda-scorecard/rubrics(/:rubricId)

Best practices are encoded as scoring rubrics that include a description, scoring table, and some procedural JavaScript code that checks a C-CDA for conformance to the rubric. Some rubrics feature binary grading (1 point = pass, 0 points = fail), while others provide point-based grading (say 0-3 points). So conformance to a rubric may not be all-or-nothing. For example there's a rubric to check whether problems have SNOMED codes. A problem list containing mostly SNOMED codes is much better than a problem list with none, and rubrics can take this into account with point-based grading.

Here's what the JSON description of a rubric might look like:

{
  "id": "labcodes",
  "scorecards": ["c-cda", "smart"],
  "category": ["Lab Results", "Codes"],
  "description": "Lab Results coded with LOINC's top 2K codes",
  "detail": "Lab results should be coded using LOINC. In pratice LOINC is huge, but 2000 codes cover 98% of real-world usage.  This means that most results should be covered by the 2000+ most common LOINC codes published by Regenstrief.",
  "maxPoints": 3,
  "points": {
    "3": "> 80% of lab results have a top-2K LOINC code",
    "2": "> 50% of lab results have a top-2K LOINC code",
    "1": "At least one lab result has a top-2K LOINC code",
    "0": "No lab results have a top-2K LOINC code"
  },
  "doesNotApply": "No lab results in document"
}

The URL above fetches a JSON description of all rubrics (or a single one).

Scorecard

POST /v1/ccda-scorecard/request

Submit a C-CDA document in the body of a scorecard request to obtain a scorecard in JSON. The scorecard will include a grade for each rubric on the CCDA Scorecard app.

Stats

As documents are validated, they're logged and used to calculate statistics about how often best practices are followed. Each rubric is associated with summary data, so you can easily get a count of how often each grade is achieved -- and get a better sense of how conformant your document is relative to the broader community.

For example, a rubric for a well-followed best practice might look like:

{
  "id": "vitals-using-loinc",
  "counts": {
    "3": 25,
    "2": 8,
    "1": 2,
    "0": 1,
    "N/A": 4
  }
}

The URL above fetches a JSON description of all stats (or stats for a single rubric).