This proposal addresses implementation of a new Pharmacy Benefits Management (PBM) guideline recommendation for the safety of aldosterone antagonist use in patients with heart failure. Aldosterone antagonists have been shown to prolong survival in patients with heart failure but they can also lead to dangerously high potassium levels (hyperkalemia). The PBM recommendation is early potassium level monitoring for patients with a new prescription for aldosterone antagonists to detect high levels before they become life threatening. Laboratory testing is recommended within 7 days of the prescription yet less than half of veterans with a new filled prescription for aldosterone antagonists have blood testing within 14 days. To improve care, PBM is initiating a program to notify VA pharmacists of patients at their facility recently started on aldosterone antagonists who may be in need of laboratory testing.
The indication for aldosterone antagonists in heart failure has expanded significantly in the 18 months. There are two aldosterone antagonists available in the VA system, spironolactone and eplerenone (1,2). Spironolactone is the most commonly used medication (>95% of prescriptions) and is inexpensive, while the newer and more expensive eplerenone has the advantage of less gynecomastia, an uncommon but sometimes irreversible side effect. Previously, aldosterone antagonists had been shown in randomized trials to improve survival and reduce hospitalization rates for patients with depressed left ventricular ejection fraction (LVEF) and severe symptoms of heart failure (New York Heart Association Class III or IV). Recently, a large trial was completed (Eplerenone in Mild Patients Hospitalized and Survival Study in Heart Failure (EMPHASIS-HF) that enrolled patients with only mild symptoms of heart failure and LVEF <=30%). This study was stopped due to a survival benefit with aldosterone antagonist treatment with the results published in mid-November 2010. The trial included a much broader population of heart failure patients: those with a reduced ejection fraction (less than 30% or less than 35% if the QRS duration was prolonged) and evidence of symptomatic heart failure. The main findings from EMPHASIS-HF were a significant improvement in survival, and a significant reduction in all cause and heart failure hospitalizations.
This Rapid Response Project evaluated the effectiveness of this PBM pilot program to improve monitoring for patients treated with aldosterone antagonists and performed a formative evaluation to guide further roll-out to other facilities. This issue is directly aligned with the CHF QUERI goal of improving the use of treatments known to prolong survival.
Design. Our evaluation for this aim was a facility level comparison of 10 PBM intervention sites versus 10 control sites, with additional comparisons to all remaining (n=134) VAMCs.
Intervention Group. As noted above, PBM selected 10 sites from 20 VA sites that expressed interest in the PBM intervention. While PBM did not randomly select sites, they selected facilities of varying sizes and region of the country. The expectation was 20-50 patients per site will be included.
Primary Control Group: The 10 facilities that have agreed to be part of the PBM program but were not selected served as the primary control group. This group was likely to be similar to the intervention group as they had also expressed interest in participating.
Primary Outcome. The primary outcome for Aim 1 is monitoring of potassium levels within 3 weeks of a new prescription for aldosterone antagonists. Monitoring within 1 week of starting the medication is recommended by the VA PBM guideline. Since we could not know when the patient started the medication, we allowed three weeks from the fill date to indicate acceptable monitoring. The patient lists created by PBM for the intervention sites served as the denominator. PBM determines patients with a new prescription for aldosterone antagonists (none in the prior 12 months). Patients are included in the facility's list if the prescription was filled between 1 week and 2 weeks ago and there has been no lab testing for potassium. The lists are generated on a rolling basis (every 2 weeks). The numerator was patients who subsequently had lab testing within 3 weeks of filling the aldosterone antagonist prescription. The same methodology and time periods were used to create the patients for the denominator and numerator for control sites
Aim 2 (Formative Evaluation): Methods: We examined four of the key elements of the Ottawa Model of Research Use (OMRU)-the practice environment, potential adopters, transfer strategies, and adoption (Figure 2) with a focus on specific variables identified by CFIR through quantitative email surveys and qualitative phone interviews. Of the other two key elements in the OMRU, innovation is defined (measuring of hyperkalemia) and outcomes are evaluated in Aim 1. Qualitative researcher and co-investigator Dr. Zickmund, and implementation researcher Dr. Mittman contributed to survey development and evaluation.
Eight sites participated for the duration of the project.
A total of 286 patients were identified by Pharmacy Benefits Management with a new prescription for an aldosterone antagonist and a history of heart failure.
166 of the 286 (58%) had no evidence of lab testing at least 2 weeks following a new prescription for an aldosterone antagonist.
120 (42%) had evidence of early lab testing.
129 of the 166 (78%) were still within 3 months of the initial aldosterone antagonist prescription. (Intervention cohort).
When compared to those that had appropriate early testing the type of medication (spironolactone, vs. eplerenone) was similar for the intervention cohort. Window pickup of of prescriptions was much more common in those with appropriate early testing (48% vs. 5%) than those in the intervention cohort.
Each site was asked to increase use monitoring for the intervention cohort patients.
Impact on Ordering:
Of those without initial lab testing, intervention (lab order; provider contact; patient contact) was deemed not necessary by the local pharmacists as the lab was either obtained or the lab order available by the time of site review (~ 6 weeks lag time);
An intervention did occur in a minority of cases as indicated by the site reviewer (in 7 patients the lab was ordered by the reviewer, in 7 patients the provider was contacted to order the lab).
Patient understanding of Rx or lab instructions did not appear to be a major reason for patients not having lab follow-up
1 site stated they implemented message field in CPRS when Rx spironolactone reminding provider to check K+ and sCr in 1 week if new Rx or dose increase, then monthly for 3 months
All 8 sites experienced delay with input of patient data at one or more of the 4 data entry deadlines (i.e., follow-up email reminder sent and/or deadline extended). Only one site did not meet any of the 4 data entry deadlines. Reasons for delay of data entry included: difficulty with initial access to Sharepoint site for data entry; awaiting presentation and approval of project concept to local P&T Committee; awaiting approval of participation in project from education department; project assigned to student or resident with questions regarding access, competing priorities, or where time assigned to site did not directly correspond with timeline for project; personnel relocation; competing priorities for time; illness; leave
Potassium Monitoring by Facility.
Initial review of monitoring data for patients with recent prescriptions shows no difference at baseline between the 8 sites that participated and the remaining 12 control sites. The data post intervention will be come available in 6 months.
Based on the results of this project Pharmacy Benefits Management is adding this metric (lack of potassium monitoring for new prescriptions of Adlosterone Antaognists) to their National Medication Use Evaluation Tracker (MUET) system where sites routinely sent patient lists of potential candidate patients..
None at this time.
Treatment - Efficacy/Effectiveness Clinical Trial