Professional Documents
Culture Documents
3 2007
Clinical Investigations
ABSTRACT
Background: Spironolactone use for heart failure (HF) has increased dramatically after the publication of
the Randomized Aldactone Evaluation Study trial; yet, few studies have examined its real-world impact.
We aimed to determine the population effect of spironolactone use on mortality in HF patients discharged
from hospital.
Methods and Results: All patients discharged alive between October 1997 and December 2001 in Nova
Scotia, Canada, with a primary diagnosis of HF were enrolled in the Improving Cardiovascular Outcomes
Study. Two year, all-cause mortality was the primary end point. A total of 7816 patients were identified, of
whom 644 (8%) were discharged home on spironolactone. After adjusting for differences in clinical covariates, spironolactone use did not emerge as an independent predictor of long-term survival (OR 0.97,
P 5 .80). When only the subgroup of patients enrolled in a HF clinic were included (n 5 990), spironolactone use was associated with reduced rates of all-cause mortality at 2 years (OR 0.52, P 5 .003).
Conclusions: Although spironolactone use was not associated with improved long-term survival in the
general HF population, it was associated with improved long-term survival in patients enrolled in HF
clinics. These data highlight the challenges of knowledge translation from a clinical trial into practice.
(J Cardiac Fail 2007;13:165e169)
Key Words: Aldosterone antagonism, Mortality, Congestive heart failure.
165
Statistical Analysis
Methods
Study Cohort and Protocol
Expired at 2 Years
Discharged on Spironolactone
Variable
Mean age, yrs (SD)
Male gender (%)
EF performed
within 60 days (%)
EF ! 35 % (%)
Diabetes (%)
Hyperlipidemia (%)
Smoking (% ever)
Peripheral vascular
disease (%)
Creatinine O
2.5 mg/dL (%)
Admission K O
5 mmol/L (%)
Loop diuretics (%)
ACE inhibitors (%)
Angiotensin receptor
blockers (%)
Digoxin (%)
b-blockers (%)
Statins (%)
NSAIDs (%)
No
(n 5 7172)
Yes
(n 5 644)
P Value
Variable
75.4 (11.4)
3377 (47.1)
3123 (43.5)
73.0 (12.8)
344 (53.4)
289 (44.9)
!.0001
.002
.51
795
2515
1668
3148
737
128
263
162
297
66
!.0001
.003
0.28
.28
.98
(41.7)
(35.1)
(23.2)
(43.9)
(10.3)
(62.8)
(40.8)
(25.2)
(46.1)
(10.2)
639 (9.1)
40 (6.3)
.02
735 (10.4)
60 (9.4)
.32
5277 (73.6)
4577 (63.8)
468 (6.5)
562 (87.3)
485 (75.3)
59 (9.2)
!.0001
!.0001
.01
2144
3565
1239
182
278
360
144
13
!.0001
.003
.001
.42
(29.9)
(49.7)
(17.3)
(2.5)
(43.2)
(55.9)
(22.4)
(2.0)
No
(n 5 4975)
Yes
(n 5 2841)
P Value
2046 (41.2)
2411 (48.5)
1698 (60.0)
1001 (35.2)
!.0001
!.0001
606
1726
1368
2676
458
317
1052
462
1690
345
(50.8)
(37.0)
(16.3)
(59.6)
(12.1)
!.0001
.04
!.0001
!.0001
!.0001
281 (5.8)
398 (14.3)
!.0001
421 (8.6)
375 (13.4)
!.0001
3612
3310
336
1404
2681
1055
1295
(40.7)
(34.7)
(27.5)
(53.8)
(9.2)
(72.5)
(66.5)
(6.8)
(28.2)
(53.9)
(21.2)
(26.0)
126 (2.5)
405 (8.1)
2227
1752
191
1018
1244
328
785
(78.4)
(61.7)
(6.7)
(35.8)
(43.8)
(11.6)
(27.6)
69 (2.4)
239 (9.2)
!.0001
!.0001
.96
!.0001
!.0001
!.0001
.12
.78
.67
Results
A total of 7816 patients were discharged from hospitals
in Nova Scotia with a primary diagnosis of HF during the
observation period. Their average age was 76 years; 52%
were female; and their median EF was 40%. Of these patients, 644 (8%) had been discharged on spironolactone.
Patients discharged from hospital on spironolactone were
younger and were more likely to be male and have diabetes
(Table 1). Their mean EF was 31.5%, although less than
half of those discharged on spironolactone had had a measure of their EF within the 60 days preceding their discharge. Furthermore, 6% had a creatinine O 2.5 mg/dL,
9% had K O 5.0 mmol/L on admission, and 5% were
given concomitant nonsteroidal anti-inflammatory drug
therapy. Of the 644 patients who were prescribed spironolactone at hospital discharge, 326 (51%) either had a documented serum creatinine O2.5 mg/dL, a serum K O5.5
mmol/L, or no documentation of left ventricular EF
!35% within 60 days of hospital admission.
At 2 years, rates of all-cause mortality were similar between patients who had been discharged on spironolactone
and those who had not been (37.1% vs. 36.3%, P 5 .67).
After adjusting for differences between the 2 groups, we
found that spironolactone use did not emerge as an independent predictor of long-term survival (OR 0.97, 95% confidence interval [CI] 0.79e1.20) (Table 3). A time-to-event
analysis yielded similar results (Fig. 1).
A total of 990 patients with a discharge diagnosis of HF
were followed in 1 of 4 specialized HF clinics during the
study period (Table 4). As a group, they were slightly younger than the overall HF population. Furthermore, clinic
Effect
Spironolactone
Age (per year)
Female gender
Smoker (ever)
Diabetes
Hyperlipidemia
Peripheral
vascular disease
Serum creatinine
(per mg/dL)
Loop diuretics
Angiotensin-converting
enzyme inhibitor
b-blocker
Statin
Point
Estimate
95%
Confidence
Limits
1.24)
1.05)
0.92)
1.34)
1.46)
0.96)
1.92)
P
Value
0.97
1.04
0.79
1.16
1.16
0.77
1.57
(0.76,
(1.03,
(0.69,
(1.00,
(0.93,
(0.61,
(1.28,
.80
!.0001
.002
.05
.19
.02
!.0001
1.20
(1.11, 1.30)
.001
1.20
0.80
(1.02, 1.41)
(0.69, 0.94)
.03
.005
0.79
0.73
(0.68, 0.91)
(0.57, 0.95)
.001
.02
Ouzounian et al
167
patients were noted to have higher rates of concurrent treatment with other HF medications, including loop diuretics,
ACE inhibitors, and b-blockers. Of these patients, 11%
had been discharged from hospital on spironolactone. Of
the 107 patients who were prescribed spironolactone at hospital discharge, 18 (17%) either had a documented serum
creatinine O 2.5 mg/dL, a serum K O 5.5 mmol/L, or
no documentation of left ventricular EF !35% within 60
days of hospital admission. After fully adjusting for differences in baseline variables, spironolactone emerged as an
independent predictor of reduced long-term mortality (OR
0.52, 95% CI 0.34e0.79) (Table 5).
In an attempt to dissect out the mechanism for the incongruent effects of spironolactone use on mortality in our
population, we repeated our logistic regression model for
only those patients who met RALES criteria, with an EF
of less than 35%, a serum creatinine less than 2.5 mg/dL,
and a serum K less than 5 mM (n 5 628).
Spironolactone use in this subgroup was not associated
with improved long-term survival (OR 1.20, 95% CI
0.68e2.11).
Discussion
We found that treatment with spironolactone among all
patients admitted for HF was not associated with improved
survival. In contrast, spironolactone use in patients followed in HF clinics was associated with a significant
decrease in the relative risk of all-cause mortality.
In RALES, 1663 patients with NYHA III-IV HF and an
ejection fraction ! 35% were randomized to receive either
a low dose of spironolactone (25 mg/day) or placebo. Exclusion criteria included a plasma potassium concentration
O 5.0 mmol/L and a serum creatinine concentration O 2.5
mg/dL. Importantly, only 11% of the treatment group had
been receiving a b-blocker, and 25% of the patients were
diabetic.8 The RALES protocol called for monitoring of serum potassium levels 1 week after initiating therapy, and
then again at 1 month and every 3 months thereafter. An
increase in serum potassium O 5.5 mmol/L at any time
prompted a review of concomitant medications associated
with hyperkalemia, such as potassium supplements or nonsteroidal anti-inflammatory drugs, and a reduction in the
dose of study medication to 25 mg every other day. Serious
hyperkalemia occurred in 1% of the placebo group and 2%
of the spironolactone group (P 5 .42).
There are several possible reasons for our studys failure
to detect the positive effect of spironolactone use on survival observed in the RALES trial. To begin with, the differences in outcome likely reflect differences between the
2 studies in terms of patient selection and extent of
follow-up after discharge. Our patient population represents
a broader group of HF patients, with less severe HF than in
RALES (mean EF 31% vs. 25%). Also, the closeness of follow-up dictated in the RALES trial may have minimized
the risk of hyperkalemia and other sources of mortality associated with spironolactone use in an unselected HF
No
(n 5 883)
Yes
(n 5 107)
P Value
213 (24.1)
552 (62.5)
729 (82.6)
15 (14.0)
68 (63.6)
91 (85.0)
.02
.83
.52
338
257
417
487
107
65
43
47
65
11
(53.8)
(29.1)
(47.3)
(55.2)
(12.1)
41 (5.8)
522
41
465
323
256
11
(59.1)
(4.6)
(52.7)
(36.6)
(29.0)
(1.2)
(77.4)
(40.2)
(43.9)
(60.4)
(9.9)
5 (4.7)
90
8
73
59
37
1
(84.1)
(7.5)
(68.2)
(55.4)
(34.6)
(1.0)
!.0001
.02
.51
.34
.54
.65
!.0001
.20
.002
.0004
.26
.00
Several smaller studies have warned of a higher risk of hyperkalemia associated with spironolactone in the community
than what was observed in RALES.9e12 Several hypotheses
have been proposed: inappropriate monitoring of potassium
levels; neglect of predisposing patient attributes, such as
diabetes; neglect of conditions that develop during therapy,
including renal insufficiency; inappropriate doses of
spironolactone; deliberate increases in dietary potassium
intake by patients; and the inappropriate extrapolation of
RALES findings to a broader patient population.13 Although
we do not have information regarding rates of hyperkalemia
or the cause of death in our patients, inappropriate prescription of spironolactone may have resulted in higher rates of
Table 5. Logistic Regression Model for Mortality
in Clinic Patients
Variable
Spironolactone on
discharge
Age (per year)
Peripheral vascular disease
Creatinine (per mg/dL)
b-blocker
Statin
Loop diuretic
Point
Estimate
95%
Confidence
Limits
P Value
0.52
(0.34, 0.79)
.003
1.03
1.46
1.35
0.60
0.44
1.54
(0.99,
(0.78,
(0.94,
(0.39,
(0.27,
(0.83,
.08
.24
.39
.01
.0007
.17
1.05)
2.73)
1.95)
0.93)
0.70)
2.87)
Ouzounian et al
169
References
1. OConnell JB, Bristow M. Economic impact of heart failure in the
United States: time for a different approach. J Heart Lung Transplant
1993;13:S107e12.
2. Hunt SA, Baker DW, Chin MH, et al. ACC/AHA guidelines for the
evaluation and management of chronic heart failure in the adult: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll
Cardiol 2001;38:2101e13.
3. Liu P, Arnold JM, Belenkie I, et al. The 2002/3 Canadian Cardiovascular Society consensus guideline update for the diagnosis and
management of heart failure. Can J Cardiol 2003;19:347e56.
4. Pitt B, Zannad F, Remme WJ, et al. The effect of spironolactone on
morbidity and mortality in patients with severe heart failure. N Engl
J Med 1999;341:709e17.
5. Cox JL, Ramer SA, Lee DS, Humphries K, Pilote L, Svenson L,
Tu JV, on behalf of the Canadian Cardiovascular Outcomes Research
Team Investigators. Pharmacological treatment of congestive heart
failure in Canada: a description of care in five provinces. Can J Cardiol 2005;21:337e43.
6. Canadian Cardiovascular Outcomes Research Team Investigators.
Burden of cardiovascular disease in Canada. Can J Cardiol 2003;19:
997e1004.
7. Cox JL, ICONS Investigators. Optimizing disease management at
a health care system level: the rationale and methods of the Improving
Cardiovascular Outcomes in Nova Scotia (ICONS) study. Can J
Cardiol 1999;15:787e96.
8. Pitt B, Perez A. Spironolactone in patients with heart failure. N Engl J
Med 2000;342:133e4.
9. Berry C, McMurray JJ. Serious adverse events experienced by patients
with chronic heart failure taking spironolactone. Heart 2001;85:E8.
10. Schepkens H, Vanholder R, Billiouw JM, Lameire N. Life-threatening
hyperkalemia during combined therapy with angiotensin-converting
enzyme inhibitors and spironolactone: an analysis of 25 cases. Am J
Med 2001;110:438e41.
11. Svensson M, Gustafsson F, Galatius S, Hildebrandt PR, Atar D. Hyperkalaemia and impaired renal function in patients taking spironolactone for congestive heart failure: retrospective study. BMJ 2003;327:
1141e2.
12. Juurlink DN, Mamdani MM, Lee DS, et al. Rates of hyperkalemia
after publication of the Randomized Aldactone Evaluation Study.
N Engl J Med 2004;351:543e51.
13. McAlister FA, Stewart S, Ferrua S, McMurray JJ. Multidisciplinary
strategies for the management of heart failure patients at high risk
for admission: a systematic review of randomized trials. J Am Coll
Cardiol 2004;44:810e9.