Aldosterone receptor antagonists have been shown in randomized trials to reduce morbidity and mortality in adults with symptomatic systolic heart failure. We studied the effectiveness and safety of spironolactone in adults with newly diagnosed systolic heart failure in clinical practice. We identified all adults with newly diagnosed heart failure, left ventricular ejection fraction of <40%, and no previous spironolactone use from 2006 to 2008 in Kaiser Permanente Northern California. We excluded patients with baseline serum creatinine level of >2.5 mg/dl or a serum potassium level of >5.0 mEq/L. We used Cox regression with time-varying covariates to evaluate the independent association between spironolactone use and death, hospitalization, severe hyperkalemia, and acute kidney injury. Among 2,538 eligible patients with a median follow-up of 2.5 years, 521 patients (22%) initiated spironolactone, which was not associated with risk of hospitalization (adjusted hazard ratio 0.91, 95% confidence interval 0.77 to 1.08) or death (adjusted hazard ratio 0.93, confidence interval 0.60 to 1.44). Crude rates of severe hyperkalemia and acute kidney injury during spironolactone use were similar to that seen in clinical trials. Spironolactone was independently associated with a 3.5-fold increased risk of hyperkalemia but not with acute kidney injury. Within a diverse community-based cohort with incident systolic heart failure, use of spironolactone was not independently associated with risks of hospitalization or death. Our findings suggest that the benefits of spironolactone in clinical practice may be reduced compared with other guideline-recommended medications.
Clinical practice guidelines for heart failure management have designated β blockers, angiotensin-converting enzyme inhibitors (ACEIs), angiotensin II receptor blockers (ARBs), and aldosterone receptor antagonists as class I therapies for patients with heart failure and reduced left ventricular ejection fraction (LVEF). However, in clinical practice, aldosterone receptor antagonists (e.g., spironolactone) are used significantly less frequently compared with the other guideline-recommended medications. The lesser use of spironolactone likely stems from concerns regarding serious adverse events and questionable effectiveness outside of clinical trial settings. Three major randomized clinical trials of aldosterone receptor antagonists have shown improved morbidity and mortality in patients with an LVEF of <40% and at least class II heart failure symptoms. Because these randomized studies were conducted in highly selected populations with structured treatment and follow-up protocols, questions remain about whether the benefits and risks of aldosterone receptor antagonists can be fully translated to clinical practice. To help clarify the uncertainties between randomized trial findings and clinical practice observations, we studied the safety and effectiveness of spironolactone compared with other guideline-recommended heart failure medications in a diverse cohort of adults with newly diagnosed systolic heart failure.
Methods
The study was conducted within the Kaiser Permanente Northern California, a large integrated health-care delivery system that provides comprehensive care to >3.2 million members within the San Francisco and greater Bay Area. The study cohort was identified from the previously established Kaiser Permanente Northern California Chronic Heart Failure Registry. We included subjects if they met any of the following criteria from January 1, 2006 to December 31, 2008: ≥1 inpatient admission with a primary discharge diagnosis of heart failure ( International Classification of Diseases, Ninth Edition codes 398.91, 402.01, 402.11, 402.91, 428.0, 428.1, 428.22, or 428.9) or ≥3 outpatient encounters, excluding emergency department visits, with any diagnosis of heart failure. We excluded patients if they had any outpatient, emergency department, or hospital discharge diagnosis of heart failure before January 1, 2006, those with left ventricular systolic function of >40%, and those with any spironolactone or eplerenone use in the 12 months before study entry. Left ventricular systolic function was determined from review of relevant diagnostic test results found in the electronic medical record surrounding index date. To be consistent with randomized trials for ACEI, ARB, and aldosterone receptor antagonists, patients with a baseline serum creatinine level of >2.5 mg/dl or a baseline serum potassium level of >5.0 mEq/L were excluded from the study cohort. In summary, the present analysis was focused on the subset of patients with newly diagnosed heart failure with documented LVEF of <40% who had no aldosterone receptor antagonist use in the 12 months before study entry. The study was approved by the institutional review boards of the Kaiser Foundation Research Institute and Stanford University. Waiver of informed consent was obtained because of the nature of the study.
Receipt of spironolactone was identified based on filled prescriptions found in pharmacy databases using previously described methods. We also controlled for use of β blockers, ACEI, ARB, loop diuretics, potassium supplementation, digoxin, calcium channel blockers, and statins. As described previously, longitudinal medication use was estimated from drug refill patterns using the calculated day supply for each prescription. Briefly, for any 2 consecutive prescriptions, the patient was classified as continually taking the medication if the second prescription was filled within 30 days of the projected end date of the first. If the second prescription was filled >30 days after the projected end date of the first, the patient was classified as not taking the medication from day 31 until the start date of the next prescription. If a non-fatal hospitalization occurred during follow-up, the length of stay (in days) was added to the estimated days of supply for any prescription crossing that hospitalization because patients were unlikely to take their own medications while hospitalized. Baseline medication use was defined as a filled prescription found in the outpatient pharmacy database within 120 days before or 30 days after patient entry into the study cohort.
The primary outcomes of interest were all-cause mortality, all-cause hospitalization, severe hyperkalemia, and acute kidney injury occurring during follow-up until December 31, 2010. Severe hyperkalemia was defined as a serum potassium concentration of ≥6.0 mEq/L identified from ambulatory or inpatient laboratory databases. Acute kidney injury was defined using modified criteria from the Acute Kidney Injury Network as an increase above outpatient baseline renal function in serum creatinine level of >150% and/or ≥0.3 mg/dl measured during a hospital admission or emergency department visit. Baseline serum creatinine level used for defining the occurrence of acute kidney injury was designated as the outpatient nonemergency department value closest in time from 7 to 365 days before the associated inpatient admission or emergency department visit. All-cause hospitalizations were defined as any inpatient admission, excluding emergency department visits not resulting in admission, and were identified from hospital discharge and billing claims databases. All-cause hospitalization was chosen as the outcome rather than cardiovascular- or heart failure–specific hospitalization because it would capture all serious adverse events from spironolactone. Deaths were identified from the Kaiser Permanente hospitalization files, administrative records (e.g., proxy reporting), the California state death certificate files, or the Social Security Administrative vital status files.
Age, gender, and self-reported race or ethnicity were identified from health plan databases. We ascertained information on coexisting illnesses based on diagnoses or procedures using International Classification of Diseases, Ninth Edition codes, laboratory results, or specific therapies from health plan hospitalization discharge, ambulatory visit, laboratory, and pharmacy databases; diabetes mellitus registry ; and regional cancer registry. This included baseline and follow-up diagnoses of acute myocardial infarction, unstable angina, stroke, peripheral arterial disease, diabetes, hypertension, systemic cancer, chronic liver disease, chronic lung disease, mitral or aortic valvular disease, dementia, depression, ventricular tachycardia or fibrillation, and atrial fibrillation or flutter. Estimated glomerular filtration rate (eGFR, ml/min/1.73 m 2 ) was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation. The outpatient nonemergency department laboratory value closest in time before cohort entry was used to determine baseline eGFR, serum potassium level, and hemoglobin level. Time-varying values of laboratory results were ascertained throughout the follow-up period using health plan laboratory databases.
Analyses were performed using the SAS software version 9.2 (SAS Institute Inc., Cary, North Carolina). Baseline characteristics are presented as means with SDs, medians with interquartile ranges, and frequencies with percentages. We compared baseline characteristics among patient subgroups using the t test or Wilcoxon rank sum test for continuous variables and the chi-square test for categorical variables.
We used the “new user” design and multivariate extended Cox regression models with time-dependent covariates to examine the independent association between spironolactone use during follow-up and outcomes of interest. Specifically, we quantified initiation of and longitudinal exposure to spironolactone and other relevant heart failure medications, along with time-updated pertinent laboratory measurements. Covariates for Cox regression were selected a priori based on previously published studies, clinical relevance, and differences between spironolactone and nonspironolactone users in characteristics listed in Table 1 . Results were qualitatively similar using intention-to-treat and propensity score-matched Cox regression models, so only the main analytic results are presented.
| Variable | All Subjects (n = 2,358) | Spironolactone Use | p | |
|---|---|---|---|---|
| Yes (n = 521) | No (n = 1,837) | |||
| Age (yrs) | 69.2 (14.6) | 63.5 (14.7) | 70.9 (14.1) | <0.001 |
| Women | 800 (33.9) | 159 (30.5) | 641 (34.9) | 0.06 |
| White race | 1,620 (68.7) | 347 (66.6) | 1,273 (69.3) | 0.36 |
| Nonwhite race | 715 (30.3) | 167 (32.1) | 548 (29.8) | |
| Hypertension | 1,478 (62.7) | 290 (55.7) | 1,188 (64.7) | <0.001 |
| Diabetes mellitus | 721 (30.6) | 152 (29.2) | 569 (31.0) | 0.43 |
| Dyslipidemia | 1,539 (65.3) | 308 (59.1) | 1,231 (67.0) | <0.001 |
| Previous myocardial infarction | 294 (12.5) | 46 (8.8) | 248 (13.5) | <0.01 |
| Previous unstable angina | 54 (2.3) | 14 (2.7) | 40 (2.2) | 0.49 |
| Coronary artery bypass graft surgery | 27 (1.1) | 4 (0.8) | 23 (1.3) | 0.36 |
| Percutaneous coronary intervention | 110 (4.7) | 17 (3.3) | 93 (5.1) | 0.09 |
| Implantable cardioverter defibrillator | 23 (1.0) | 3 (0.6) | 20 (1.1) | 0.29 |
| Atrial fibrillation or flutter | 623 (26.4) | 97 (18.6) | 526 (28.6) | <0.001 |
| Ventricular fibrillation or tachycardia | 77 (3.3) | 12 (2.3) | 65 (3.5) | 0.16 |
| Peripheral arterial disease | 84 (3.6) | 10 (1.9) | 74 (4.0) | <0.05 |
| Mitral or aortic valvular disease | 223 (9.5) | 36 (6.9) | 187 (10.2) | <0.05 |
| Dementia or depression | 359 (15.2) | 73 (14.0) | 286 (15.6) | 0.38 |
| Chronic lung disease | 592 (25.1) | 144 (27.6) | 448 (24.4) | 0.13 |
| Chronic liver disease | 33 (1.4) | 6 (1.2) | 27 (1.5) | 0.59 |
| Cancer | 328 (13.9) | 55 (10.6) | 273 (14.9) | <0.05 |
| Previous gastrointestinal hemorrhage | 50 (2.1) | 6 (1.2) | 44 (2.4) | 0.08 |
| Number of outpatient visits in the previous yr | 7.0 (3.0–14.0) | 6.0 (3.0–11.0) | 8.0 (4.0–14.0) | <0.001 |
| LVEF | ||||
| 30%–39% | 1,302 (55.2) | 203 (39.0) | 1,099 (59.8) | <0.001 |
| <30% | 1,056 (44.8) | 318 (61.0) | 738 (40.2) | |
| eGFR (ml/min/1.73 m 2 ) | 60.2 (47.7–75.4) | 64.9 (52.5–79.0) | 59.1 (46.5–73.9) | <0.001 |
| ≥60 | 1,030 (43.7) | 252 (48.4) | 778 (42.4) | <0.001 |
| ≥45 and <60 | 595 (25.2) | 133 (25.5) | 462 (25.1) | |
| ≥30 and <45 | 343 (14.5) | 42 (8.1) | 301 (16.4) | |
| <30 | 71 (3.0) | 7 (1.3) | 64 (3.5) | |
| Potassium (mEq/L) | 4.3 (4.1–4.6) | 4.3 (4.0–4.6) | 4.4 (4.1–4.6) | <0.05 |
| <4.0 | 350 (14.8) | 93 (17.9) | 257 (14.0) | <0.05 |
| 4.0–4.4 | 813 (34.5) | 161 (30.9) | 652 (35.5) | |
| 4.5–5.0 | 778 (33.0) | 160 (30.7) | 618 (33.6) | |
| Hemoglobin level (g/dl) | ||||
| <10.0 | 91 (3.9%) | 13 (2.5%) | 78 (4.2%) | <0.01 |
| 10.0–11.9 | 337 (14.3) | 55 (10.6) | 282 (15.4) | |
| ≥12.0 | 1,475 (62.6) | 333 (63.9) | 1,142 (62.2) | |
| Systolic blood pressure, mm Hg | ||||
| <140 | 1,687 (71.5) | 379 (72.7) | 1,308 (71.2) | 0.25 |
| ≥140 | 152 (6.4) | 39 (7.5) | 113 (6.2) | |
| ACEI or ARB | 2,026 (85.9) | 489 (93.9) | 1,537 (83.7) | <0.001 |
| β Blocker | 1,942 (82.4) | 467 (89.6) | 1,475 (80.3) | <0.001 |
| Loop diuretic | 1,798 (76.3) | 433 (83.1) | 1,365 (74.3) | <0.001 |
| Potassium supplementation | 785 (33.3) | 203 (39.0) | 582 (31.7) | <0.01 |
| Calcium channel blocker | 507 (21.5) | 84 (16.1) | 423 (23.0) | <0.001 |
| Digoxin | 465 (19.7) | 142 (27.3) | 323 (17.6) | <0.001 |
| Statin dyslipidemia medication ∗ | 1,464 (62.1) | 329 (63.1) | 1,135 (61.8) | 0.57 |
| Nonaspirin antiplatelet medication | 302 (12.8) | 53 (10.2) | 249 (13.6) | <0.05 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
