[CQM-981] Harmonization is needed for (a) defining denominator populations for preventive care and screening measures, (b) definition (value sets) of tests and conditions across measures. Examples provided Created: 12/26/13  Updated: 03/22/19  Due: 01/24/14  Resolved: 03/22/19

Status: Closed
Project: eCQM Issue Tracker
Component/s: None

Type: Intent/Governance affecting more than 1 eCQM Priority: Minor
Reporter: Floyd Eisenberg Assignee: Mathematica EC eCQM Team
Resolution: Answered Votes: 1
Labels: LOGIC, VSharmonization, ValueSet, logicharmonization

Attachments: Microsoft Word Value Sets HIV Screening Measures - Proposal 26Dec2013.xls    
Solution: Thank you for the suggestions. We will take these into consideration during the next annual update phase.
Solution Posted On:
Tracker Notification:
Floyd Eisenberg
Impact: Lack of harmonization in definition (value sets and logic) will cause significant confusion and excess work effort in (a) defining new measures or HeD rules, and (b) implementing such measures and rules, and (c) comparing performance results among EPs and EHs. Specifically, in writing rules for the QIR HeD project, choosing among conflicting value sets is causing rework as clinical SME's are confused about which (if any) existing logic to reuse.
Comment Posted On:

 Description   

1) Inconsistency with definition of (a) encounters used to specify an encounter that qualifies a patient for a measure denominator is exclusively one preventive care visit in some measures and either one preventive care visit or two "non-preventive" care visits in other measures. The SMEs (CDC) question why there is no standard method across all eCQMs
2) Inconsistent definition of data elements with competing value sets (see attached Excel Spreadsheet for a listing of value sets with competing and at least non-harmonized content. The spreadsheet was developed as part of a new eCQM and HeD development project for ONC:
(a) Definition of an ambulatory encounter (value sets) differ among measures - NCQA and AMA-PCPI are often similar but have identical value sets with different OIDS. QIP and Behavioral Health measure steward use their own value sets which are different from the first two stewards.
(b) HIV tests - one value set is HIV load only, one include many SNOMED-CT and a few LOINC codes, LOINC has many codes that apply and is the preferred terminology for lab testing. SNOMED-CT is not preferred to describe the observable entity, but only the result.
(c) STI - sexually transmitted infection value sets are incomplete or conflicting

 Comments   
Comment by Floyd Eisenberg [ 01/09/14 ]

This response is encouraging. However, there is a directly related parallel project title "QIR" developing clinical decision support rules in HeD for the measures using these value sets. The HeD activities are in progress now. To accommodate an agile process to (1) create rules directly derived from the existing measures (CQM 138 and CQM 165), and (2) creating eCQMs for new measures not previously engineered as eCQMs waiting for annual update time frames is problematic. The value set changes impact the HeD rules and the new eCQMs should reuse existing value sets where possible. Waiting for annual update changes will either delay the new eCQM and all HeD work or new value set development will be needed. If harmonization is to be successful, coordination of parallel contracts is required. Changes coincident with the next annual update may significantly impact deliverable timeframes for the QIR work and may cause significant rework.

Comment by Jeff McCartney (Inactive) [ 01/09/14 ]

These issues related to several measures, including all 11 of the MU-2 measures being developed under the ONC HITECH contract. We have forwarded this to the Measure Developers. Regarding the inconsistencies in how an encounter is defined , a broad discussion, perhaps through eMIG or a Governance call,at is warranted. We will examine the value set issues and assess any changes necessary in time for the next annual update.

Generated at Tue Mar 18 19:29:14 EDT 2025 using Jira 9.12.16#9120016-sha1:6bee0863f3e6dbb91e4be2d992a3b6761c21c9e0.