Although randomized controlled trials have demonstrated benefits of aldosterone antagonists for patients with heart failure and reduced ejection fraction (HFrEF), they excluded patients with serum creatinine >2.5 mg/dl, and their use is contraindicated in those with advanced chronic kidney disease (CKD). In the present analysis, we examined the association of spironolactone use with readmission in hospitalized Medicare beneficiaries with HFrEF and advanced CKD. Of the 1,140 patients with HFrEF (EF <45%) and advanced CKD (estimated glomerular filtration rate [eGFR] <45 ml/min/1.73 m 2 ), 207 received discharge prescriptions for spironolactone. Using propensity scores (PSs) for the receipt of discharge prescriptions for spironolactone, we estimated PS-adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for spironolactone-associated outcomes. Patients (mean age 76 years, 49% women, 25% African-American) had mean EF 28%, mean eGFR 31 ml/min/1.73 m 2 , and mean potassium 4.5 mEq/L. Spironolactone use had significant PS-adjusted association with higher risk of 30-day (HR 1.41, 95% CI 1.04 to 1.90) and 1-year (HR 1.36, 95% CI 1.13 to 1.63) all-cause readmissions. The risk of 1-year all-cause readmission was higher among 106 patients with eGFR <15 ml/min/1.73 m 2 (HR 4.75, 95% CI 1.84 to 12.28) than among those with eGFR 15 to 45 ml/min/1.73 m 2 (HR 1.34, 95% CI 1.11 to 1.61, p for interaction 0.003). Spironolactone use had no association with HF readmission and all-cause mortality. In conclusion, among hospitalized patients with HFrEF and advanced CKD, spironolactone use was associated with higher all-cause readmission but had no association with all-cause mortality or HF readmission.
The efficacy of aldosterone antagonists in patients with heart failure and reduced ejection fraction (HFrEF) has been established in multiple randomized controlled trials. These randomized controlled trials generally excluded patients with serum creatinine >2.5 mg/dl. Although post hoc analyses of randomized trials have suggested that spironolactone and eplerenone may improve outcomes in those with impaired renal function, they did not include patients with advanced chronic kidney disease (CKD). Because these drugs also increase the risk of hyperkalemia and worsening kidney function, the role of these drugs in patients with advanced CKD remains unclear, and their use is contraindicated in those with advanced CKD. In 1 study based on the American Heart Association Get With the Guidelines-Heart Failure data, among real-world patients with HFrEF that also excluded advanced CKD, the use of aldosterone antagonists had no association with mortality or cardiovascular readmission. These findings highlight the need for appropriate patient selection and monitoring so that the efficacy observed in randomized trials may be translated into clinical effectiveness in the real world. Because HF is the leading cause for hospital readmission and under the new healthcare reform law hospitals are facing billions of dollars of loss in Medicare payments for higher than average 30-day all-cause readmission, in the current analysis, we examined if a discharge prescription of an aldosterone antagonist was associated with lower all-cause readmission in older patients with HFrEF with advanced CKD.
Methods
Alabama Heart Failure Project was used for data analysis in the present study, the details of which have been described previously. Briefly, 9,649 charts of fee-for-service Medicare beneficiaries discharged from 106 Alabama hospitals from July 1, 1998, to October 31, 2001, with principal diagnosis of HF were identified and abstracted in 6 different 6-month periods. Of these, a unique cohort of 8,555 patients was identified, of which 8,049 were discharged alive. Of the 5,479 with data on EF, 3,067 had EF <45%, of which 1,142 had estimated glomerular filtration rate (eGFR) <45 ml/min/1.73 m 2 . After excluding 2 patients receiving potassium-sparing diuretics other than spironolactone, the final sample consisted of 1,140 patients, of which 207 (18%) received a discharge prescription for spironolactone. Extensive data on baseline demographics, medical history including use of medications, hospital course, discharge disposition including medications, and physician specialty were collected. The primary outcome of the current analysis was 30-day all-cause readmission. Secondary outcomes included 30-day all-cause mortality, HF readmissions, and combined end point of all-cause mortality or all-cause readmission. In addition, we also examined the association of spironolactone with these outcomes during longer follow-up. Data on outcomes and time to events were obtained from the Centers for Medicare and Medicaid Services Denominator File, Medicare Provider Analysis and Review File, and Inpatient Standard Analytical File.
Pearson chi-square and one-way analysis of variance were used for descriptive analysis as appropriate. We estimated propensity scores (PSs) for the receipt of spironolactone for each of the 1,140 patients based on 45 variables that were used to estimate PS-adjusted hazard ratios (HRs) for the association of spironolactone with outcomes. All statistical tests were 2 tailed with a p value <0.05 considered significant. Statistical analyses were performed using SPSS-21 for Windows (SPSS, Inc., 2012, Chicago, Illinois).
Results
Mean age of the patients (n = 1,140) was 76 years (±10), 49% were women, 25% were African-American. The mean EF was 28% (±9), mean eGFR 31 ml/min/1.73 m 2 (±10), mean serum creatinine 2.47 mg/dl (±1.59), mean serum potassium 4.5 mEq/L (±0.75). Patients receiving versus not receiving spironolactone at discharge were similar in respect to most baseline characteristics ( Table 1 ). However, patients in the spironolactone group were more likely to have prevalent HF and lower serum creatinine levels, and be prescribed digoxin, diuretics, and angiotensin-converting enzyme inhibitors or angiotensin receptor blockers than those not receiving spironolactone ( Table 1 ).
| n (%) or Mean (±SD) | Spironolactone on Discharge | ||
|---|---|---|---|
| No (n = 933) | Yes (n = 207) | p Value | |
| Age (years) | 76 ± 10 | 76 ± 8 | 0.860 |
| Female | 451 (48%) | 106 (51%) | 0.455 |
| African American | 228 (24%) | 55 (27%) | 0.520 |
| Smoker | 88 (9%) | 17 (8%) | 0.583 |
| Admission from nursing home | 50 (5%) | 9 (4%) | 0.552 |
| Past medical history | |||
| Prior heart failure | 771 (83%) | 186 (90%) | 0.010 |
| Coronary artery disease | 654 (70%) | 158 (76%) | 0.073 |
| Myocardial infarction | 322 (35%) | 72 (35%) | 0.941 |
| Percutaneous coronary intervention | 165 (18%) | 37 (18%) | 0.948 |
| Coronary artery bypass graft | 323 (35%) | 82 (40%) | 0.174 |
| Hypertension | 682 (73%) | 147 (71%) | 0.543 |
| Atrial fibrillation | 252 (27%) | 54 (26%) | 0.786 |
| Diabetes mellitus | 465 (50%) | 97 (47%) | 0.438 |
| Stroke | 214 (23%) | 46 (22%) | 0.825 |
| Chronic obstructive pulmonary disease | 306 (33%) | 61 (30%) | 0.354 |
| Dementia | 84 (9%) | 15 (7%) | 0.417 |
| Cancer | 21 (2%) | 7 (3%) | 0.342 |
| Clinical and laboratory data | |||
| Pulse (beats per minute) | 91 ± 23 | 90 ± 22 | 0.686 |
| Systolic blood pressure (mm Hg) | 144 ± 33 | 134 ± 30 | <0.001 |
| Diastolic blood pressure (mm Hg) | 79 ± 20 | 76 ± 18 | 0.122 |
| Respiratory rate (breaths/minute) | 24 ± 6 | 23 ± 5 | 0.070 |
| Peripheral edema | 665 (71%) | 164 (79%) | 0.020 |
| Pulmonary edema by chest x-ray | 679 (73%) | 149 (72%) | 0.816 |
| Serum sodium (mEq/L) | 138 ± 4.5 | 138 ± 5 | 0.040 |
| Serum potassium (mEq/L) | 4.5 ± 0.76 | 4.42 ± 0.70 | 0.269 |
| Serum creatinine (mg/dL) | 2.6 ± 1.7 | 2.1 ± 0.7 | <0.001 |
| Glomerular filtration rate (mL/min/1.73 m 2 ) | 30 ± 11 | 33 ± 8 | <0.001 |
| Blood urea nitrogen (mg/dL) | 42 ± 21 | 42 ± 22 | 0.720 |
| Hematocrit (%) | 36 ± 6 | 37 ± 6 | 0.033 |
| White blood cell (cell/μL) | 10 ± 9 | 9 ± 5 | 0.184 |
| In-hospital events | |||
| Pneumonia | 271 (29%) | 52 (25%) | 0.257 |
| Acute myocardial infarction | 58 (6%) | 13 (6%) | 0.973 |
| Pressure ulcer | 97 (10%) | 23 (11%) | 0.762 |
| Hospital and care characteristics | |||
| Rural hospital | 209 (22%) | 44 (21%) | 0.720 |
| Cardiology care | 610 (65%) | 156 (75%) | 0.006 |
| Intensive care | 47 (5%) | 11 (5%) | 0.870 |
| Length of stay (days) | 7.76 ± 6.5 | 7.77 ± 5 | 0.977 |
| Discharge medications | |||
| Renin angiotensin system antagonists | 492 (53%) | 133 (64%) | 0.003 |
| Beta-adrenergic blockers | 235 (25%) | 56 (27%) | 0.578 |
| Loop diuretics | 742 (80%) | 192 (93%) | <0.001 |
| Digoxin | 429 (46%) | 130 (63%) | <0.001 |
| Calcium channel blockers | 196 (21%) | 28 (14%) | 0.014 |
| Potassium supplementation | 348 (37%) | 78 (38%) | 0.918 |
| Anti-arrhythmic drugs | 174 (19%) | 48 (23%) | 0.136 |
| Antidepressants | 196 (21%) | 38 (18%) | 0.393 |
| Non-steroidal anti-inflammatory drugs | 62 (7%) | 11 (5%) | 0.479 |
Within 30 days postdischarge, unadjusted all-cause readmissions rates were 30% and 25% for patients receiving and not receiving spironolactone, respectively. PS-adjusted HR (95% confidence interval [CI]) associated with spironolactone use was 1.41 (1.04 to 1.90; Table 2 ). There was no association with all-cause mortality or HF readmission during 30 days postdischarge, although there was a near-significant association with 30-day combined end point of all-cause readmission or all-cause mortality ( Table 2 ). The risk of all-cause readmission (PS-adjusted HR 1.36, 95% CI 1.13 to 1.63) and the combined end point of all-cause readmission or all-cause mortality (PS-adjusted HR 1.30, 95% CI 1.09 to 1.54; Table 3 and Figure 1 ) during 1 year postdischarge were higher among patients in the spironolactone group. The adverse association of spironolactone use with 1-year all-cause readmission was significantly higher in the 106 patients with eGFR <15 ml/min/1.73 m 2 (HR 4.75; 95% CI 1.84 to 12.28) than in the 1,034 with eGFR 15 to 45 ml/min/1.73 m 2 (HR 1.34, 1.11 to 1.61, p for interaction is 0.003). Similar differences were observed for 1-year combined end point of all-cause readmission or all-cause mortality (p for interaction 0.007).
| 30-day Outcomes | % (Total Events/Total Patients) Spironolactone on Discharge | Hazard Ratio ∗ (95% CI); p-Value | ||
|---|---|---|---|---|
| No | Yes | Unadjusted | Propensity Score Adjusted | |
| All-cause readmission | 25% (237/933) | 30% (61/207) | 1.19 (0.90–1.57); p = 0.233 | 1.41 (1.04–1.90); p = 0.027 |
| Heart failure readmission | 12% (116/933) | 12% (25/207) | 0.97 (0.63–1.49); p = 0.872 | 0.90 (0.57–1.41); p = 0.635 |
| All-cause mortality | 8% (75/933) | 8% (17/207) | 1.01 (0.60–1.72); p = 0.961 | 1.05 (0.60–1.82); p = 0.866 |
| All-cause mortality or readmission | 31% (285/933) | 34% (70/207) | 1.13 (0.87–1.47); p = 0.352 | 1.31 (0.99 –1.73); p = 0.058 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
