Abstract P156: Kidney Function and Serum Potassium as Predictors of Spironolactone Initiation in Congestive Heart Failure

Abstract

Introduction: There are a paucity of data on real-world spironolactone use among patients with congestive heart failure (CHF), and whether spironolactone initiation differs by kidney function and serum potassium. Hypothesis: Patients with reduced kidney function and elevated serum potassium are less likely to be prescribed spironolactone than patients with normal kidney function and serum potassium. Methods: We identified patient cohorts with incident CHF and available laboratory data in the Truven MarketScan Commercial Claims and Encounters database (MS) from 2010-2015 (N=22,881), and the Geisinger Health System Integrated Electronic Health Record (GHS) from 2004-2016 (N=17,470), and assessed medication use through prescription order and dispensing data, and comorbidities using ICD codes. We measured incidence, initial dose, and duration of incident spironolactone use, overall and by levels of kidney function and serum potassium. We used Cox proportional hazards regression models to quantify the association between measures of baseline and time-varying estimated glomerular filtration rate (eGFR) and serum potassium, and incident spironolactone use, adjusting for other medication use and comorbidities. Results: Among 22,881 incident CHF patients in the MS database, mean age at CHF was 54.3 (±8.7) years, 47.2% were female, and the mean eGFR was 78.9 ml/min/1.73m 2 (±21). The 17,470 incident CHF patients in GHS were older (74.0 years ±12.9), more likely to be female (51.5%), and have a lower mean eGFR (62.5 ±24.3). Both cohorts had a mean baseline serum potassium of 4.3 mEq/L (±0.5). There were 1,341 (1,774) incident prescriptions for spironolactone over 34,384 (52,187) person-years in MS (GHS). Median starting dose was 25 mg in both cohorts. Women were less likely to start spironolactone in both cohorts, while older patients were less likely to start spironolactone only in GHS. Compared to patients with eGFR >90 ml/min/1.73m 2 , those with eGFR <30 ml/min/1.73m 2 were less likely to initiate spironolactone in MS (hazard ratio [HR] = 0.57, 95% confidence interval [CI]: 0.38-0.84) and GHS (HR = 0.69, CI: 0.48-0.99). Patients with serum potassium <3.5 mEq/L were more likely to initiate spironolactone compared to those with 3.5-4.9 mEq/L (normal) in both cohorts (HR=1.99 [95% CI: 1.57-2.52] in MS; HR=2.70 [95% CI: 2.34-3.26] in GHS). Conclusion: Consistent with the labeled guidelines, patients with CHF and reduced kidney function were less likely to use spironolactone, while those with lower serum potassium levels were more likely to use spironolactone.