Aldosterone antagonists such as spironolactone substantially reduce risk of death, HF hospitalization, and improve functional status. Yet, only 1/3 of eligible patients in the U.S. receive these drugs. Optimal implementation of aldosterone antagonists represents the greatest potential to save lives among evidence-based HF therapies. There is also known overuse of spironolactone in patients at risk of hyperkalemia and such use has led to alarming rates of hyperkalemia in the U.S. and Canada. Within the VA Health Care System, there are unfortunately no published data on degree of underuse and only preliminary data on inappropriate use of aldosterone antagonists in HF patients. Further, preliminary studies have found that there is inadequate potassium and creatinine monitoring in VA patients with heart failure, indicating a potential gap in patient safety. It is important to not only to document gaps in evidence-based use of spironolactone but also to understand the complex reasons for these gaps.
1a) Identify gaps in evidence-based use and laboratory monitoring of spironolactone therapy in HF patients at two VAMCs in Phoenix and Providence
1b) Because we had access to national level VA data via Providence VA, we broadened our initial objective (Aim#1a) to identify gaps in spironolactone therapy by measuring appropriate and inappropriate use after HF hospitalization in an national cohort of Veterans with HF.
2) Identify determinants of gaps in spironolactone use and monitoring among providers at Phoenix VAHCS.
1a) We linked Phoenix and Providence VA patient data from the VA Patient Treatment File, External Peer Review Program, Decision Support System databases to obtain clinical, laboratory and prescription data. From patients admitted with a principal diagnosis of HF at each VAMC, we determined the proportion of patients who appropriately and inappropriately received an aldosterone antagonist. Next, we calculated the rate of serum potassium and creatinine monitoring at 15 and 30 days, respectively, after prescription of an aldosterone antagonist.
1b) Using the linked data as described above for a nationwide cohort of Veterans from 2002-2009, we calculated the proportion of recently hospitalized HF patients who received an appropriate or inappropriate (e.g., had contraindications) prescription for spironolactone within 90 days of discharge. We also measured temporal trends and hospital variation using the medians odds ratio (MOR). A MOR > 1 indicates significant hospital-level variation.
2) To understand reasons for care gaps at a single VAMC , we conducted a mixed methods study at Phoenix VA. First, we conducted an internet survey with a convenience sample of providers from internal medicine, cardiology, pharmacy, and primary care regarding their knowledge, familiarity, and perception of barriers. We then conducted six focus groups to encourage discussion of shared experiences with aldosterone antagonists. Member check interviews were then conducted to corroborate study findings. Focus groups and interviews with audiotaped, transcribed, and coded (NVivo) by two qualitative-trained coders with interrater reliability checks.
1a) Assessment of Local Treatment Gap - From 2002-2009, 39% of aldosterone-antagonist-eligible Phoenix VA patients (34/87) and 16% of Providence VA patients (9/55) were appropriately prescribed an aldosterone antagonist. Further, during the study period, 18% of Phoenix VA patients and 12% of Providence VA patients received an inappropriate prescription, i.e., patients had an contraindication the drug. With regard to monitoring of serum potassium, at 15 days, 27% of Phoenix VA (n=263) and 25% of Providence VA patients (n=73) received monitoring after a prescription of an aldosterone antagonist. For serum creatinine, at 15 days, 27% of Phoenix VA and 26% of Providence VA patients received lab monitoring.
1b) Assessment of National Treatment Gap - From a sample of 37,126 patients (n=131 hospitals), 9,355 were Ideal MRA candidates, and 4,056 were Non-Ideal candidates. Among Ideal candidates, 36% received an MRA, but there was a temporal decline in MRA use between 2003-2009 (41% in 2003 to 31% in 2009, p<0.001). Of Non-Ideal candidates, 27% received an MRA with a temporal decline in use (34% in 2003 to 22% in 2009, p<0.001). Hospital-level rates of MRA prescription ranged from 0-71% for Ideal Candidates and 0-100% for Non-Ideal candidates. The MORs of MRA prescription for Ideal and Non-Ideal candidates were 1.44 and 1.36, respectively, suggesting significant hospital-level variation.
2) Assessment of Barriers - The survey of providers (n=50) revealed unfamiliarity with eplerenone, especially among primary care providers. 1/3 of respondents reported they would order a lab test more than 2 weeks after drug prescription, whereas guidelines recommend testing within 1 week. Less than half of respondents identified NYHA class 2 HF patients as eligible for an aldosterone antagonist.
42 subjects participated in one of six focus groups, and 6 participated in member check interviews. Eight barriers to aldosterone antagonist use were found across three general sources: Patient-based, Provider-based, and System-based. Patient-based barriers included Polypharmacy/Comorbidities, Adverse Effects of Drug Therapy, and Perceived Non-Adherence. Provider-based barriers included Unclear Provider Roles and Responsibilities, Coordination and Transitions of Care, and Lack of Familiarity or Experience with MRA Use. System-based barriers included System-Overload/Provider Time Constraints, and Lack of Systematic Follow-up Procedures.
With the findings of this report, the VA is able to know the quality of aldosterone antagonist adoption and monitoring in Veterans with heart failure. It is clear from our local and national level analyses that the aldosterone antagonist treatment gap does exist in the VA with 1/3 of patients appropriately receiving the drug. There is also a substantial portion of patients who inappropriately received an aldosterone antagonist and could be harmed. The presence of wide hospital variation has not been previously reported. This finding suggests a gap between high- and low-performing VA hospitals and need for addressing system-level barriers.
Our qualitative findings are the first in the cardiology literature to conceptualize a framework to understand these gaps and the existing barriers to aldosterone antagonist use. The VA will be able to use this information identify approaches to improve quality of HF care and patient outcomes among Veterans. Further, these findings may inform efforts to improve adoption of spironolactone and other evidence-based HF therapies in non-Veteran populations.
- Dev S, Hoffman TK, Kavalieratos D, Heidenreich P, Wu WC, Schwenke DC, Tracy SJ. Barriers to Adoption of Mineralocorticoid Receptor Antagonists in Patients With Heart Failure: A Mixed-Methods Study. Journal of the American Heart Association. 2016 Mar 31; 5(3):e002493.
- Dev S, Lacy ME, Masoudi FA, Wu WC. Temporal Trends and Hospital Variation in Mineralocorticoid Receptor Antagonist Use in Veterans Discharged With Heart Failure. Journal of the American Heart Association. 2015 Dec 23; 4(12).
- Dev S, Hoffman T, Kavalieratos D, Schwenke D, Heidenreich P, Wu W, Tracy S. Barriers to Adoption and Monitoring of Mineralocorticoid Receptor Antagonists in a VA Medical Center. [Abstract]. Journal of the American College of Cardiology. 2015 Mar 1; 65(10_S):doi:10.1016/S0735-1097(15)60847-6.
- Dev S, Lacy ME, Masoudi F, Wu WC. Characteristics Associated with Appropriate Use of Aldosterone Antagonists in a National Cohort of Veterans Hospitalized with Heart Failure. [Abstract]. Circulation. Cardiovascular quality and outcomes. 2014 Jul 1; 7:A158.