Health Services Research & Development

Veterans Crisis Line Badge
Go to the ORD website
Go to the QUERI website

IIR 15-115 – HSR&D Study

New | Current | Completed | DRA | DRE | Portfolios/Projects | Centers | Career Development Projects

IIR 15-115
Factors and Outcomes Associated with Inappropriate Prescribing of Phosphodiesterase-5-Inhibitors for Pulmonary Hypertension
Renda Soylemez Wiener MD MPH
Edith Nourse Rogers Memorial Veterans Hospital, Bedford, MA
Bedford, MA
Funding Period: June 2016 - November 2020

BACKGROUND/RATIONALE:
Inappropriate care, such as prescribing ineffective medicines, costs VA billions and exposes Veterans to harm. In the Blueprint for Excellence, VA commits to de-implementing ineffective, low-value care. The American Board of Internal Medicine's Choosing Wisely campaign identified use of phosphodiesterase-5-inhibitors (PDE5I) for pulmonary hypertension (PH) due to left heart disease or hypoxemic lung diseases as wasteful, ineffective, and potentially harmful.

OBJECTIVE(S):
We propose to examine influences on and outcomes of use of PDE5I for PH as a test case to broadly inform de-implementation of inappropriate prescribing. Our specific aims are to: 1) Identify patient, clinician, and site level determinants of inappropriate use of PDE5I for PH; 2) Evaluate PDE5I prescribing in local contexts and identify best practices; 3) Establish the magnitude of harm associated with inappropriate use of PDE5I for PH.

METHODS:
Using VA-Medicare data we will build a SAS dataset of the cohort of Veterans treated with PDE5I for PH from 2005 until the most recently available year of data. Using pharmacy data, we will identify the cohort of patients who received at least one prescription for treatment of PH. We will use ICD9 codes to determine the likely underlying etiology of PH, categorizing prescribing as appropriate if the patient has a diagnosis associated with pulmonary arterial hypertension and inappropriate if he or she has PH associated with left heart disease or hypoxemic lung disease. We will validate performance of our algorithm through gold standard chart review of randomly selected patients. Using multivariable logistic regression, we will identify patient, provider, and site characteristics associated with inappropriate prescribing for PH. We will assess variation in rates of inappropriate prescribing by site, identifying high (n=3) and low performing (n=3) sites for evaluation. At the selected sites we will perform qualitative interviews with patients and VA staff (clinicians, pharmacists, leadership) to identify attitudes towards and practices associated with high or low rates of inappropriate prescribing. Finally, using propensity score matching, we will assess risk adjusted outcomes of patients treated inappropriately for PH, as compared to those who received appropriate prescribing, and those who did not receive advanced vasoactive agents for PH.

FINDINGS/RESULTS:
Compared to gold standard physician chart review on a randomly selected subset of PH patients treated with PDE5i (n=241), positive predictive value of our ICD-based algorithm for identifying inappropriate prescriptions was high (86%). Among 2790 Veterans treated with daily PDE5i therapy for PH between 2005-2012, 541 (19%) received appropriate treatment, 1,711 (61%) inappropriate treatment, and 358 (13%) treatment of uncertain value. The number of inappropriately treated patients increased more than 10-fold over the study period, from 53 in 2005 to 748 in 2012. Based on chart abstraction, only 46% of PDE5i-treated Veterans (111/241) underwent guideline-recommended right heart catheterization to confirm presence or type of PH. 40% of prescriptions were initiated outside of VA by a community clinician. To address these issues, we next assembled a cohort of Medicare eligible Veterans who received care for incident PH in VA between 2006-2015. Among 37,846 Veterans with Groups 2 or 3 PH, 1701 (4.5%) received daily PDE5i therapy from VA (inappropriate prescriptions). We have performed preliminary analyses of patient-, clinician-, and site-level characteristics associated with inappropriate prescribing. Factors significantly associated with inappropriate prescriptions include younger age, fewer comorbidities, Group 3 rather than Group 2 PH, care in the Southeastern US. Female patients and those on nitrates were less likely to receive inappropriate PDE5i prescriptions. The strongest predictors of receipt of inappropriate PDE5i therapy included right heart catheterization within the prior year (adj OR 4.0, 95% CI 3.6-4.5) and treatment with other PH medications (adj OR 6.2, 4.4-8.9).

IMPACT:
This study will establish the determinants of inappropriate use of PDE5I for PH, the magnitude of harms that result, and the means to prevent them. We intend to work with our partners in the Pharmacy Benefits Management office to develop an intervention to selectively de-implement inappropriate prescribing of PDE5I for PH, informed by the findings of this project. In particular, this test case will inform how to address common issues of: 1) tensions between guidelines, VA policies, and individual patient care, 2) non-VA prescriptions for dual use patients (a growing concern under the Veteran's Choice Act); 3) how to empower pharmacists as gatekeepers and opinion leaders, and 4) local implementation of national policies. At costs of $10,000-$13,000 per patient-year and no clear benefit, de-implementation of PDE5i treatment for Groups 2 and 3 PH is a potential high-yield target to improve value, safety, and quality of care for Veterans with an increasingly common clinical condition.

PUBLICATIONS:

Journal Articles

  1. Kim D, Lee KM, Freiman MR, Powell WR, Klings ES, Rinne ST, Miller DR, Rose AJ, Wiener RS. Phosphodiesterase-5 Inhibitor Therapy for Pulmonary Hypertension in the United States. Actual versus Recommended Use. Annals of the American Thoracic Society. 2018 Jun 1; 15(6):693-701.


DRA: Health Systems, Cardiovascular Disease, Lung Disorders
DRE: Treatment - Observational, TRL - Applied/Translational
Keywords: Practice Patterns/Trends, Quality Improvement, Patient Safety, Pharmacology
MeSH Terms: none