2023 HSR&D/QUERI National Conference

4088 — Development and Validation of Quality Measures for Testosterone Prescribing in VA

Lead/Presenter: Ann Borzecki,  COIN - Bedford/Boston
All Authors: Borzecki AM (Center Healthcare Organization and Implementation Research), Conti J (CHOIR) Reisman J (CHOIR) Vimalananda V (CHOIR) Nagy M (Milwaukee VAMC) Paluri R (VA Bedford Healthcare System) Linksy A (CHOIR) McCullough M (CHOIR) Basin S (Harvard Medical School) Jasuja G (CHOIR)

Objectives:
Testosterone therapy is recommended for the treatment of male hypogonadism, defined as the presence of signs and symptoms of testosterone deficiency, along with low morning testosterone levels on at least two occasions. Due to the potential risks associated with testosterone therapy, including cardiovascular events and prostate cancer, it is important that testosterone is prescribed in accordance with best practices. Studies have found significant guideline-discordant prescribing of testosterone. In 2018, the Office of Inspector General identified that VA clinicians infrequently adhered to Endocrine Society guidelines-based VA Criteria for Use for testosterone therapy. Thus, accurate measures to assess appropriateness of testosterone prescribing are needed to improve prescribing practices. Our objective was to develop and validate electronic health record (EHR) based quality measures around initiation and monitoring of testosterone prescribing.

Methods:
This retrospective cohort study included male veterans who initiated testosterone during January or February 2020, were regular users of VA care and were alive at the study period end. We used national EHR data from the VA’s Corporate Data Warehouse (CDW) and chart review data from individual patient EHRs. We developed 9 initiation and 9 monitoring measures based on the 2018 Endocrine Society guidelines, expert opinion, prior work, and that could be operationalized using national VA EHR data. Initiation measures included lack of documentation of appropriate laboratory tests (e.g., lack of low morning total or free testosterone levels done on two occasions, and presence of contraindications around initiation, such as having a prostate cancer ICD-10 diagnosis). Monitoring measures similarly included lack of documentation of appropriate laboratory testing (e.g., lack of a follow-up testosterone, and presence of contraindications between the initial testosterone prescription and the second (or follow-up) testosterone prescription). We assessed criterion validity for these 18 measures by comparing findings from CDW data to chart review data from 142 charts. We examined positive and negative predictive values (PPVs, NPVs), overall accuracy (OA), and Matthews Correlation Coefficients (MCCs).

Results:
We found high PPVs (> = 78%), NPVs (>98%), OA (> = 94%) and MCCs (>0.84) for the 12 measures based on laboratory data (5 initiation and 7 monitoring). For the 6 measures relying on ICD-10 diagnostic codes, we similarly found high NPVs (100%) and OAs (> = 98%). However, PPVs for measures of acute conditions occurring prior to testosterone initiation (i.e., acute myocardial infarction or stroke) or new conditions occurring after initiation (i.e., prostate or breast cancer) PPVs were much lower (0 to 50%) due to few or no cases.

Implications:
We developed and validated several clinically relevant, EHR-based quality measures for assessing testosterone prescribing practices. Our measures were highly accurate when compared to the gold standard of chart review, especially those based on laboratory values. Since these measures are electronic, they may be readily deployed system-wide to evaluate the quality of testosterone prescribing at the level of an individual clinician, facility, or the entire healthcare system.

Impacts:
Widespread adoption of these quality measures in the VA and other healthcare systems can help identify gaps in the quality of testosterone prescribing , thereby helping improve care of men with hypogonadism.