May 6, 2013

Meaningful Use Stage 2 Exception vs. Exclusion Logic Explained

Introduced in Meaningful Use Stage 2 was the notion of both exception logic and exclusion logic with the Proportion-Based CQMs.  Based on numerous conversations over the past several months, there appears to be ongoing confusion in logical implementation of exception and exclusion logic for these CQMs as part of the Meaningful Use program.

The authoritative resource for this explanation is the Clinical Quality eMeasure Logic and Implementation Guidance Document that is publicly available on the CMS site.  The details for proportion-based CQM that I have paraphrased follows below:
  • Initial Patient Population: The set of patients (or episodes of care) to be evaluated by the measure.
  • Denominator Population: A subset of the Initial Patient Population.
  • Exclusion Population: The members of the Denominator that should not be considered for inclusion in the Numerator.
  • Numerator Population: A subset of the Denominator.  The numerator criteria are the processes or outcomes expected for each patient, procedure, or other unit of measurement defined in the denominator.
  • Exception Population: These are the members of the Denominator that were considered for membership in the Numerator, but were rejected, and meet the logic required for the exception criteria.
I have included an illustration that attempts to capture this updated logic for Meaningful Use Stage 2.

Meaningful Use Stage 2 Proportion-Based CQM Logic
Meaningful Use Stage 2 Proportion-Based CQM Logic

These final population counts are then used to calculate the performance rate of any CQM against a population of patients via the following formula:

Performance Rate = Numerator Count / (Denominator Count – Exclusion Count – Exception Count)


This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License. © Rob McCready, 2013.
Creative Commons License

13 comments:

  1. Hi Rob

    I am reading some documents and was quite confused but from the diagram above can we say following ?

    1. Denominator will always be subset of initial population
    2. Numerator will always be a subset of denominaotr
    3. Denominator exception and exclusion will be subset of denominator

    ReplyDelete
  2. Yes... for proportion-based measures. And there are subtle timing differences with exception vs. exclusion. Read about the other types here http://blog.agilehealthservices.com/2013/06/four-major-types-of-clinical-quality.html

    ReplyDelete
  3. Thanks for your very helpful post, as well as the CQM reference.

    ReplyDelete
  4. Thank u its too much helpful
    i need some clarification from you

    Denominator Exclusion are removed from denominator ? and same from exception ?

    ReplyDelete
  5. In other words
    Denominator=Denominator Exclusion-Denominator Exception ?

    ReplyDelete
  6. Thank you very much! As a software developer working on MU2, especially CQM, your blog really helps a lotttttttt!

    ReplyDelete
  7. I'm working with Cypress test data on NQF 0028. The Cypress site claims that measure contains an exception. I think I know why but do you know where I might be able to locate documentation that explains what exclusions and exceptions belong to the CQM's, or at least NQF 0028?

    ReplyDelete
    Replies
    1. The specific details of the exception logic and exclusion logic are separated out in the CQM definition for the MU Stage 2 measures when you download them. See http://projectcypress.org/faq.html#mu_stage2_hqmf There is human readable prose in the HTML view of the CQMs, and the HQMF defines executable logic in XML for the associated exclusion and exception logic.

      Delete
  8. Hello,

    Meaningful Use stage 2 requires that we should provide codified entries for reasons of denominator exclusion. Reasons can be Patient Reason, Medical Reason or Enviromental Reason. My question is how can we identify the reason of exclusion from NQF i.e. how can we identify that reason of exclusion is patient, medical or environmental?

    Please help!

    ReplyDelete
  9. Can the exclusion count be higher than the denominator count?

    ReplyDelete
  10. Is there a case where the denominator is really a subset of the initial pop? In the ones I've looked at the denominator is always the same as the initial pop.

    ReplyDelete
  11. Based upon your description, the graphic should show the Denominator Population as including the Exception population. While Exceptions are NOT part of the Numerator, this definition indicates they should remain in the Denominator. Definition: "Exception Population - These are the members of the Denominator that were considered for membership in the numerator but were rejected and meet the logic required for the exception criteria." This is also the case for the Exclusion Population. Basically everything within the scope of the Denominator in the graphic should be identified in your legend as part of the Denominator.

    ReplyDelete