Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

VA Health Systems Research

Go to the VA ORD website
Go to the QUERI website
HSRD Conference Logo



2019 HSR&D/QUERI National Conference Abstract

Printable View

4109 — E-quality measure specifications and assessment of quality of care

Lead/Presenter: Jeremy Sussman,  COIN - Ann Arbor
All Authors: Sussman JB (Center for Clinical Management Research, University of Michigan, Ann Arbor)), Ratz DP (Center for Clinical Management Research, University of Michigan, Ann Arbor), Wiitala WL Center for Clinical Management Research, University of Michigan, Ann Arbor) Burns JA (Center for Clinical Management Research, University of Michigan, Ann Arbor) Youles B (Center for Clinical Management Research, University of Michigan, Ann Arbor) Keusch J (Center for Clinical Management Research, University of Michigan, Ann Arbor) Hayward RA (Center for Clinical Management Research, University of Michigan, Ann Arbor)

Objectives:
The introduction of eQuality Measures has allowed the automation of core outpatient quality measures. Currently, these measures are modeled on those from the Healthcare Effectiveness Data and Information Set (HEDIS). Since HEDIS uses data from many healthcare organizations, the metrics rely on one-year billing data. Here, we sought to evaluate the impact of small changes in measure construction on overall quality assessment. We specifically focused on the measure requiring moderate-strength statins or greater for people with a history of atherosclerotic cardiovascular disease (ASCVD).

Methods:
We used VA CDW data to construct the eQM Statin use measures (specifically, at least one statin fill) for active VA patients with a history of ASCVD. We looked at 5 measure specifications: 1) the current measure; 2) do not exclude myalgia, myositis, myopathy, or rhabdomyolysis patients, which are a poor-reliability diagnostic codes; 3) include any statin, not just moderate/high dose statins, which is more consistent with VA guidelines; 4) include patients who had a diagnostic code of ASCVD in the prior 5 years instead of just 1 year; and 5) listed statin allergy or intolerance in the electronic health record.

Results:
503,402 patients from 130 VA sites met the initial measure criteria. Excluding patients with documented statin allergies decreased the number eligible by 17% to 416,097. The 5-year look back increased the number to 595,672 (18%). Across the specifications, between 67.6% and 75.3% of patients met the measure. Allowing use of any strength of statin to meet the measure shifted site rankings by 13 positions or more for 35% of sites. Of 20,105 patients with at least 3 allergies listed in the electronic health record, 6,398 took a statin for at least 80% of the calendar year.

Implications:
Current eQMs are somewhat limited by an attempt to mirror HEDIS and are sensitive to clinically-minor choices in definition. These would vary how many patients are eligible substantially, but do not seem to alter the final statin use rate by a large amount.

Impacts:
Formal specification of eQM can alter care substantially and should align with organizational values.