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

Health Services Research & Development

Go to the ORD website
Go to the QUERI website

RRP 12-492 – HSR&D Study

RRP 12-492
Evaluation of glycemic over and under-treatment indicators for high-risk Veterans
Leonard M Pogach MD MBA
East Orange Campus of the VA New Jersey Health Care System, East Orange, NJ
East Orange, NJ
Funding Period: November 2013 - April 2015

One in four veterans have diabetes, of whom 40% are 60-69 years of age (the Vietnam era cohort); and about 25% are over 70 years of age. Serious co-morbid conditions are common (31%) even in younger (<65 years) veterans. About one in five veterans >65 years have cognitive impairment or dementia, and about 30% have an estimated Glomerular Filtration Rate of <60 ml/min/1.73m2. About 30% of veterans receive insulin, and use of basal insulin has increased by 41% since 2008. Recent American Diabetes Association Guidelines agree with VHA-DoD Diabetes Guidelines that target values should be individualized based upon factors such as life expectancy, comorbid conditions, patient preferences, and risk for serious hypoglycemia. However, the current National Committee for Quality Assurance (NCQA) <7% A1c measure for persons less than 65 years of age neither stratifies by insulin nor excludes serious non-diabetes related or mental co-morbid conditions. The NCQA <8% A1c measure applicable to persons 65 to 74 years of age has no exclusion criteria. There is also no recommendation to assess glycemic management in individuals with diabetes >75 years of age. Due to these concerns, VHA only tracks A1c >9% (poor control). This gap led the DM-QUERI R&M Committee to recommend development of clinical indicators for potential over- and under treatment of glycemic control as a key strategic goal. Based upon our published work (continuous measures, glycemic regimen complexity and co-morbid conditions), we will work in partnership with VACO partners to refine and evaluate such indicators.

Concurrent with the funding of the RRP, the draft (now final) Department of Health and Human Services National Action Plan for Prevention of Adverse Drug Events stated: "Specifically, the [NQF measures] do not exclude patients for whom HbA1c <8 percent would be inappropriate according to new guidelines, or stratify by medications (such as insulin). Neither do they address potential overtreatment in high-risk groups." The investigators conceptualized that rather than separate measures for undertreatment and overtreatment, that a composite measure of in range (IR) -which is guideline concordant; and out of range (OOR) -defined as the current 9% poor control and the <7% overtreatment level, would better convey appropriate treatment in the elderly.

The long term objective of this proposal is to improve the appropriateness of glycemic management. Our specific aims are: (1) To develop technical specifications for clinical indicators of under- and overtreatment. a) To evaluate the impact of different inclusion and exclusion criteria on measure performance characteristics (e.g., sensitivity and specificity). b) To work with our VACO Program Office partners to finalize measure specification. (2) To assess variation at various VA organizational levels (CBOCs, facilities, and VISNS). 3) To work with our VACO Program Office partners to develop clinical indicators for over- and under-treatment that can be utilized to inform quality improvement, surveillance and direct patient care.

Using the CDW outpatient, medication, laboratory files, we identified the denominator of patients at high risk for hypoglycemia as having - received insulin (I) or sulfonylurea (SU) in FY 2012; age 65 or greater; last serum creatinine 2.0 mg/dl; ICD-9CM diagnosis of cognitive impairment (CI) / dementia; at least one A1c in FY 2013. The patient level outcomes were the percentage of patients with A1c <7%, <8%, >9%, OOR (<7% or >9%), and IR (7.5%-8.5%) stratified by 65-75 years and 75+ years. The facility level outcomes were the facility rankings correlation of dichotomous measures with OOR and IR. (Spearman's Rank correlation coefficient). Changes in facility level percentile rankings of facilities in the top and bottom two deciles based on the <8% compared to rankings based upon OOR measure.

We identified 435,078 patients on I/SU, of whom 112,541 (25.8%) met the inclusion criteria at 130 facilities (average number of patients: 866; range 110 - 3,518). The average age was 78.5 years. 98.8% were male. 64.3% were married. 25.5% were Poor and 50.2% were disabled. The A1c was Out Of Range (OOR) in 44.1% of patients: OT (< 7%) in 31.6% of patients, UT (> 9%) in 12.5%. Only 30.4% were in the range recommended by ADA (A1c 7.5-8.5%). The OOR measure ranged from 33.2% in the best performing decile (of the 130 facilities) to 56.8% in the worst performing decile. The ranges for OT and UT alone in best and worst performing deciles were 20.9%-49.1% and 7.5%-19.4%, respectively. Facility rankings for OT or UT were poorly correlated. Regarding facility rankings, 16 of 26 worse performing facilities (bottom 2 deciles) improved to higher deciles; 20/26 (77%) of best performing (top 2 deciles) worsened to lower deciles.

Conclusions: We conclude that nearly half of our study population of elderly or ill patients, most of whom would not be included in current HEDIS measures, were OT or UT, and therefore at risk for short term harms. Among older/sicker patients on Sulfonylurea and/or Insulin, over-treatment (<7%) is 2.2 fold more common than under-treatment (>9%). There was significant facility level variation, and even the best performers had more than a third of high risk patients with OOR A1c. There was marked facility level variation.

We recommend replacing the current <8% measure with an OOR measure that more appropriately focuses physician action and better reflects "patient safety". Additionally, an In-Range (IR) interval measure can guide quality improvement efforts consistent with current guidelines and the limitations in the precision of the A1c test from clinical laboratories.

External Links for this Project

NIH Reporter

Grant Number: I21HX001191-01

Dimensions for VA

Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.

If you have VA-Intranet access, click here for more information

VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address.
    Search Dimensions for this project


Conference Presentations

  1. Pogach L, Maney M, Soroka O, Tseng C, Aron D. An Out of Range (OOR) Glycemic Population Health Safety Measure. Paper presented at: VA HSR&D / QUERI National Meeting; 2015 Jul 8; Philadelphia, PA. [view]
  2. Pogach LM, Maney M, Soroka O, Tseng C, Tseng C, Aron D. An Out of Range (OOR) Glycemic Population Health Safety Measure. Poster session presented at: American Diabetes Association Annual Scientific Session; 2015 Jun 1; Boston, MA. [view]

DRA: Diabetes and Other Endocrine Disorders
DRE: Treatment - Observational
Keywords: none
MeSH Terms: none

Questions about the HSR&D website? Email the Web Team.

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.