Download Cardiac resynchronization therapy utilization for heart

Congestive Heart Failure
Cardiac resynchronization therapy utilization for heart
failure: Findings from IMPROVE HF
Anne B. Curtis, MD, a Clyde W. Yancy, MD, b Nancy M. Albert, PhD, RN, c Wendy Gattis Stough, PharmD, d,e
Mihai Gheorghiade, MD, f J. Thomas Heywood, MD, g Mark L. McBride, PhD, h Mandeep R. Mehra, MD, i
Christopher M. O’Connor, MD, j Dwight Reynolds, MD, k Mary Norine Walsh, MD, l and Gregg C. Fonarow, MD m
Tampa, FL; Dallas, TX; Cleveland, OH; Research Triangle Park and Durham, NC; Chicago, IL; La Jolla and
Los Angeles, CA; Cambridge, MA; Baltimore, MD; Oklahoma City, OK; and Indianapolis, IN
Background Cardiac resynchronization therapy (CRT) has established efficacy for patients with systolic heart failure
(HF). Treatment rates and factors associated with CRT utilization among eligible patients in outpatient cardiology practices
have not been well studied.
Methods IMPROVE HF is a prospective cohort study designed to characterize current management of patients with
chronic HF and left ventricular ejection fraction ≤35% in a registry of outpatient cardiology practices located throughout the
United States. Baseline data were abstracted by trained chart review specialists from May 31, 2005, through June 22, 2007,
for 15,381 patients attending 167 outpatient cardiology practices. Multivariable analyses of patient and practice
characteristics identified predictors of CRT for eligible patients.
Results A total of 1,373 patients were eligible for CRT based on current guideline criteria, and 533 (38.8%) received a
CRT device, with 84.1% of these treated with a CRT-defibrillator. Cardiac resynchronization therapy use varied widely among
practices, with 11.1% at the 25th percentile and 53.4% at the 75th percentile. Patient age, insurance, longer QRS duration,
and practice location were independently associated with higher CRT utilization rates among eligible patients, whereas sex,
HF etiology, and other clinical and laboratory parameters were not.
Conclusions Despite being evidence based and guideline recommended, CRT is underutilized in eligible patients with
significant variations associated with age, insurance, QRS duration, and geographic location of practices. Practice-specific
performance improvement initiatives may be needed to reduce variations in use of CRT for eligible patients. (Am Heart J
2009;158:956-64.)
Heart failure (HF) is a chronic progressive disease
associated with substantial morbidity and mortality.1,2
Evidence from randomized controlled trials demonstrates
that cardiac resynchronization therapy (CRT) combined
with medical therapies significantly reduces morbidity
From the aUniversity of South Florida College of Medicine, Tampa, FL, bBaylor University
Medical Center, Dallas, TX, cCleveland Clinic Foundation, Cleveland, OH, dCampbell
University School of Pharmacy, Research Triangle Park, NC, eDuke University Medical
Center, Durham, NC, fNorthwestern University, Feinberg School of Medicine, Chicago, IL,
g
Scripps Clinic, La Jolla, CA, hOutcome Sciences, Inc, Cambridge, MA, iUniversity of
Maryland, Baltimore, MD, jDuke University Medical Center, Durham, NC, kUniversity of
Oklahoma Health Sciences Center, Oklahoma City, OK, lThe Care Group, LLC,
Indianapolis, IN, and mUCLA Medical Center, Los Angeles, CA.
Clinical trial registration information: www.clinicaltrials.gov; Unique identifier:
NCT00303979.
Submitted May 19, 2009; accepted October 9, 2009.
Reprint requests: Anne B. Curtis, MD, University of South Florida, 12901 Bruce B. Downs
Blvd., MDC 87, Tampa, FL 33612.
E-mail: [email protected]
0002-8703/$ - see front matter
© 2009, Mosby, Inc. All rights reserved.
doi:10.1016/j.ahj.2009.10.011
and mortality in patients with HF, reduced left ventricular
ejection fraction (LVEF), and prolonged QRS duration.3-13
A recent meta-analysis of 6 randomized controlled trials of
CRT for treatment of chronic, symptomatic left ventricular systolic dysfunction indicated that CRT was associated with a 28% reduction in all-cause mortality and a 37%
reduction in new hospitalizations for exacerbation of
HF.10 These results supported development of national
guidelines for patients with chronic HF including CRT for
those with LVEF ≤35%, QRS duration ≥120 milliseconds,
and New York Heart Association (NYHA) functional class
III or ambulatory class IV.2 Currently, few data are
available that characterize CRT utilization in clinical
practice other than estimates from randomized clinical
trials. The Registry to Improve the Use of Evidence-Based
Heart Failure Therapies in the Outpatient Setting
(IMPROVE HF) is a prospective cohort study of patients
with HF who are managed in outpatient cardiology
practices.14 The baseline data collected for IMPROVE HF
provide an opportunity to assess rates and predictors for
use of CRT in a contemporary patient cohort.
American Heart Journal
Volume 158, Number 6
Methods
The design and methods of IMPROVE HF and overall study
objectives have previously been described.14 Diverse community and academic single-specialty or multispecialty cardiology
outpatient practices from all regions of the United States were
invited to participate in the registry.14 Patients with systolic HF
or prior myocardial infarction and reduced LVEF were enrolled.
Patients were excluded from enrollment if they were b18 years,
did not have moderate to severe left ventricular systolic
dysfunction confirmed by qualitative or quantitative assessment
of LVEF ≤35%, were not expected to survive 1 year due to
conditions other than HF, were receiving palliative care, or after
heart transplantation. Baseline data were collected from medical
chart reviews by trained abstractors and included demographic
and clinical characteristics, medical history, previous treatments, NYHA functional status, laboratory results, diagnostic
tests, treatments, and QRS duration. Documented contraindications and other reasons for not prescribing evidence-based
therapy including CRT (such as medical, economic, social, and
religious reasons; noncompliance; and patient refusal) were
collected. Documentation of contraindications for CRT implantation was recorded, including medical reason(s) and patient
choice for not using CRT. Type of implanted device (CRTpacemaker [CRT-P] or CRT-defibrillator [CRT-D]) was also
recorded. Participating practices provided descriptive data
about geographic location, practice type, number of cardiologists and electrophysiologists, affiliations with hospitals or
transplant centers, presence of device or HF clinics, and average
annual number of patients. All analyses were prespecified in the
study protocol.
All participating practices were approved by local institutional
review boards (IRB), central IRBs, or received IRB waivers.
Highly trained, centralized chart review specialists completed
ongoing training, monitoring, and testing with oversight
provided by IMPROVE HF Steering Committee members to
maintain accuracy in data abstraction and ensure optimal data
quality. The average interrater reliability was substantial (κ =
0.82). In addition, an audit of all patient data compared with
source documentation was conducted for 20% of the entire
patient sample for a 10% random sample of participating sites.
Mean data concordance rate was 94.5% (range 92.3%-96.3%).
The average number of automated quality checks completed per
data field was 1.7 to ensure values entered met prespecified
ranges, formats, and units. Data quality reports were generated
monthly to monitor completeness and accuracy of data. The
registry coordinating center was Outcome Sciences, Inc (Cambridge, MA).
Baseline characteristics
This analysis was based on all baseline data collected between
May 31, 2005, and June 22, 2007, for a representative sample of
patients with systolic HF randomly selected from each practice.
Practices identified all patients with HF and documented LVEF
≤35%, and a random sample of records was selected for these
patients for each practice. This resulted in selection of
approximately 90 patients per practice (25th and 75th
percentiles, 58 and 107, respectively) with a median of 90
patients providing adequate statistical power to detect changes
in treatment rates within practices after the performance
improvement intervention component of IMPROVE HF.14
Curtis et al 957
Figure 1
Flow diagram for CRT eligibility and utilization. Determination of
eligibility for CRT in IMPROVE HF.
Medical records for all patients selected for the sample were
reviewed to confirm eligibility for study participation, and full
abstraction was completed for all patients determined to be
eligible. There were 1,257 patients screened for inclusion in
IMPROVE HF who were deemed ineligible for enrollment based
on exclusion criteria of a documented medical condition that in
and of itself would limit 1 year survival (eg, metastatic breast
cancer), direct physician documentation that life expectancy is
b1 year, patients were receiving palliative care, status post heart
transplantation, or other reasons for ineligibility.
Cardiac resynchronization therapy eligibility
Patients with LVEF ≤35%, evidence of electrical dyssynchrony
(QRS ≥120 milliseconds), and NYHA functional class III or
ambulatory NYHA class IV despite HF medical therapy were
considered eligible for CRT.2 All analyses of rates and predictors
of CRT included only patients with quantitative or qualitative
documentation of NYHA functional class and QRS duration at
levels defined by guidelines.2 Alternative CRT eligibility
constructs were also analyzed.
American Heart Journal
December 2009
958 Curtis et al
Table I. Baseline characteristics of patients eligible for CRT
Characteristic
Age, mean (SD)/median, y
Male, %
Race, %
White
Black
Race not documented/missing
Insurance, %
Medicare
Medicaid
Private
Other
Not documented/missing
None
HF etiology, ischemic, %
Prior myocardial infarction, %
History of coronary artery bypass graft, %
History of atrial fibrillation, %
History of peripheral vascular disease, %
History of hypertension, %
History of diabetes, %
History of chronic obstructive pulmonary disease, %
History of depression, %
NYHA class III/IV, %
LVEF, mean (SD)/median, %
Systolic blood pressure, mean (SD)/median, mm Hg
Diastolic blood pressure, mean (SD)/median, mm Hg
Resting heart rate, mean (SD)/median, beats/min
Rales on most recent examination, %
Edema on most recent examination, %
Sodium, mean (SD)/median, mEq/L
Blood urea nitrogen, mean (SD)/median, mg/dL
Creatinine, mean (SD)/median, mg/dL
QRS duration, mean (SD)/median, ms
QRS duration, %, ms
b130
130-150
N150
ACEI/ARB use, %
β-Blocker use, %
Aldosterone antagonist use, %
Digoxin use, %
Diuretic use, %
ICD use, %
Eligible, CRT
(n = 533)
Eligible, no CRT
(n = 840)
70.0 (12.0)/72.0
69.8
71.5 (12.0)/73.0
69.0
52.0
9.6
37.5
49.9
9.8
39.6
69.4
4.5
20.6
2.1
2.9
0.6
66.4
34.9
35.8
41.1
12.4
60.2
38.3
20.1
10.7
100
22.8 (7.4)/24.0
115.7 (18.3)/113.0
68.3 (10.8)/70.0
73.9 (10.4)/73.0
6.8
28.0
139.0 (3.5)/139.0
30.2 (17.8)/25.0
1.5 (1.0)/1.3
163.7 (28.1)/160.0
66.5
3.9
21.1
1.7
6.0
0.8
68.5
36.7
33.2
39.5
13.7
64.9
40.0
18.6
11.7
100
23.5 (7.3)/25.0
116.7 (19.4)/118.0
67.9 (11.3)/68.0
73.0 (11.2)/72.0
6.3
31.1
138.8(3.5)/139.0
30.4 (18.0)/26
1.5 (0.7)/1.3
158.7 (29.5)/154.0
9.8
25.1
65.1
78.2
85.2
37.9
50.1
84.1
84.1
15.1
30.7
54.2
78.08
82.7
32.5
46.0
82.6
50.0
P⁎
.024
.770
.868
Ineligible for CRT
(n = 14 008)
68.5
71.2
40.4
9.0
48.7
.082
.147
.505
.319
.564
.485
.081
.523
.490
.579
.492
.076
.374
.465
.073
.918
.462
.260
.716
.640
b.001
b.001
.947
.261
.042
.135
.506
b.001
59.3
3.5
25.2
3.7
6.9
1.3
64.9
39.7
30.5
29.8
11.1
61.5
33.5
16.2
8.5
19.6
25.7 (6.9)/25.0
120.9 (18.8)/120.0
70.6 (11.3)/70.0
72.0 (11.6)/71.0
3.4
18.7
139.3 (4.2)/140.0
25.3 (14.5)/21.0
1.4 (0.8)/1.2
124.3 (39.4)/116.0
60.3
14.3
25.3
77.6
83.9
26.0
35.7
70.2
40.7
⁎ Significance of comparison between patients eligible for CRT who received and did not receive a device.
Statistical methods
All statistical analyses were performed by independent
biostatisticians contracted by Outcome Sciences, Inc. Descriptive statistics were calculated for patient and practice characteristics for the full IMPROVE HF cohort and the subset of patients
eligible for CRT for all practices. Univariate general estimating
equation (GEE) hierarchical models identified patient and
practice characteristics that might be associated with variations
in CRT rates. All variables included in these analyses are shown
in Table III. The GEE analyses controlled for intrapractice data
correlations. Multivariable GEE models, based on generalized
linear model variables statistically significant at the .10 level in
the univariate GEE models, were calculated to identify factors
independently associated with CRT utilization. All analyses were
completed using SAS statistical software, version 9.1 (SAS
Institute, Cary, NC). Statistical inference testing was 2-sided,
and results were considered statistically significant if P b .05.
Results
Baseline clinical characteristics
A total of 15,381 patients attending 167 outpatient
cardiology practices met study inclusion criteria and were
included in this analysis. The NYHA class was quantitatively documented for 31.5% of patients and qualitatively
documented for 27.0%, comprising 58.5% of the total
cohort. QRS duration was documented for 67.7% of
American Heart Journal
Volume 158, Number 6
Curtis et al 959
Table II. Baseline IMPROVE HF practice characteristics in
patients eligible for CRT
Characteristic
Census region, %
South
Northeast
Central
West
Practice setting, %
University
practice setting
Nonuniversity
teaching setting
Nonuniversity,
nonteaching
setting
Multispecialty, %
HF clinic, %
Device clinic, %
Average number of
cardiologists in
practice, mean
(SD)/median
Average number of
electrophysiologists
in practice, mean
(SD)/median
Average annual number
of patients managed
by practice, mean
(SD)/median
Table III. Rates of CRT with alternative CRT measure
specifications
CRT quality measure definition
Eligible, CRT
(n = 533)
Eligible,
no CRT
(n = 840)
42.4
30.8
19.5
5.4
42.7
27.6
18.5
9.3
12.9
13.1
27.4
28.1
54.8
53.7
26.5
64.9
85.0
14.8 (9.3)/12.0
26.1
62.0
83.2
15.4 (13.1)/12.0
.877
.279
.367
.936
2.3 (1.8)/2.0
2.2 (2.0)/2.0
.144
P
.057
.935
2975.3
3163.6
.252
(3714.9)/1500.0 (4004.2)/1958.0
patients with QRS duration ≥120 milliseconds for 59% of
these. A total of 1,393 patients (9.1% of the total cohort)
were considered potentially eligible for CRT (Figure 1).
Twenty patients had documented contraindications or
other reasons for not implanting a CRT device, resulting
in 1,373 patients who met clinical criteria for CRT
implantation. Of these, 533 (38.8%) received a CRT
device, with 448 receiving a CRT-D (84.1%) and 85
receiving CRT-P therapy (15.9%). Among eligible patients
who did not receive CRT, 420 (50.0%) were treated with
implantable cardioverter-defibrillators (ICDs).
Patient characteristics are presented in Table I for those
eligible for CRT who did and did not receive a device and
patients not CRT eligible. Patients treated with CRT were
significantly younger than those who did not receive CRT.
Cardiac resynchronization therapy implantation rates were
not significantly different for patient's sex or race, although
racial designation was not assigned for nearly 40% of
patients. There were no significant differences between
patients receiving CRT and those not receiving CRT that
were attributable to HF etiology, comorbid health conditions, and clinical findings, although QRS duration was
significantly longer for patients receiving CRT. There were
no significant differences in treatment rates with angiotensin-converting enzyme inhibitors/angiotensin receptor
Treatment rates
Prespecified definition (LVEF ≤35%, QRS duration ≥120 ms, NYHA class III
or IV, no contraindications), n (%)
CRT
533 (39.4)
No CRT
820 (60.6)
Prespecified definition plus ACEI/ARB and β-blocker treatment, n (%)
CRT
360 (38.7)
No CRT
571 (61.3)
Prespecified definition plus ACEI/ARB, β-blocker, and aldosterone
antagonist treatment, n (%)
CRT
147 (43.1)
No CRT
194 (56.9)
Prespecified definition plus ACEI/ARB and β-blocker treatment or
contraindications/intolerance to ACEI/ARB and β-blockers, n (%)
CRT
395 (39.0)
No CRT
618 (61.0)
Prespecified definition plus ACEI/ARB, β-blocker, and aldosterone
antagonist treatment or contraindications/intolerance to ACEIs/ARBs,
β-blockers and aldosterone antagonists, n (%)
CRT
208 (41.0)
No CRT
299 (59.0)
blockers (ACEIs/ARBs), β-blockers, and diuretics between
patients implanted and not implanted, whereas administration of aldosterone antagonists was significantly higher for
patients treated with CRT. Geographic location of practices
was associated with a trend for differences in CRT rates with
the lowest rates noted for practices in the western United
States (Table II). Other practice characteristics were not
associated with significant differences for CRT.
Alternative CRT use measure definitions
Sensitivity analyses were performed to determine if
alternative measure specifications significantly influenced
rates of CRT. Table III presents rates of CRT for patients
with LVEF ≤35%, QRS duration ≥120 milliseconds, and
NYHA functional class III or ambulatory class IV who were
treated with (1) both ACEIs/ARBs and β-blockers; (2)
combination therapy with ACEIs/ARBs, β-blockers, and
aldosterone antagonists; (3) ACEIs/ARBs plus β-blockers
unless contraindicated; and (4) ACEIs/ARBs, β-blockers,
and aldosterone antagonists unless contraindicated. These
alternative specifications produced treatment rates similar
to those for the prespecified metric for CRT implantation.
If indications for CRT were expanded to include patients
with NYHA class I to IV, LVEF ≤35%, QRS duration ≥120
milliseconds, and no contraindications, then 1,207 (31.5%)
of 3,826 patients received CRT under these criteria.
Variations in CRT use
Cardiac resynchronization therapy use ranged from
0.0% to 100% between practices with 11.1% and 53.4% at
the 25th and 75th percentiles, respectively, for CRT in
eligible patients. Implantation rates ranged from 0.0% to
85.7% when the analysis was confined to practices with
960 Curtis et al
American Heart Journal
December 2009
Figure 2
Percent of eligible patients with CRT-D/CRT-P therapy by practice. Baseline CRT-D/CRT-P use for practices with ≥5 eligible patients.
Figure 3
The CRT-D and CRT-P use relative to total CRT use by practice. Baseline CRT-P compared with CRT-D use for practices with ≥5 eligible patients and
at least 1 patient receiving CRT.
American Heart Journal
Volume 158, Number 6
Curtis et al 961
Table IV. Estimated unadjusted and adjusted OR for CRT in eligible patients
Univariate
Patient and practice
characteristics
Age (10 y)
Sex (male vs female)
Race
Black vs white
Not documented vs white
Insurance
Medicaid vs Medicare
Private vs Medicare
Other vs Medicare
Not documented vs Medicare
None vs Medicare
Ischemic heart disease etiology
Myocardial infarction
Coronary artery bypass graft
Atrial fibrillation/flutter
Peripheral vascular disease
Diabetes
Chronic obstructive pulmonary disease
Depression
LVEF
NYHA class (III vs II)
QRS per 10 ms
Systolic blood pressure per 10 mm Hg
Creatinine
Sodium per 5 mEq/L
ACEI/ARB
Aldosterone antagonist
β-Blocker
Statin
HF clinic
Device clinic
Multispecialty
No. of cardiology physicians
No. of patients annually per 500
Practice setting
University vs nonuniversity, nonteaching
Nonuniversity, teaching vs nonteaching
Region
Central vs Northeast
South vs Northeast
West vs Northeast
Multivariable
OR (95% CI)
Wald m 2
P
OR (95% CI)
Wald m 2
P
0.90 (0.82-0.99)
1.04 (0.82-1.31)
5.04
0.09
.025
.770
0.89 (0.80-0.99)
–
4.38
–
.036
–
0.94 (0.64-1.38)
0.91 (0.72-1.15)
0.10
0.64
.754
.424
–
–
–
–
–
–
1.10
0.94
1.19
0.39
0.65
1.21
0.93
1.12
1.07
0.89
0.93
1.10
0.91
0.99
1.13
1.06
0.97
1.04
1.10
1.02
1.26
1.20
1.17
1.13
1.18
1.02
1.00
0.99
(0.64-1.89)
(0.72-1.23)
(0.53-2.64)
(0.20-0.74)
(0.17-2.52)
(0.97-1.53)
(0.74-1.16)
(0.89-1.41)
(0.86-1.33)
(0.64-1.23)
(0.74-1.16)
(0.84-1.45)
(0.64-1.28)
(0.97-1.00)
(0.80-1.59)
(1.02-1.10)
(0.92-1.03)
(0.91-1.19)
(0.93-1.29)
(0.78-1.32)
(1.00-1.58)
(0.89-1.62)
(0.94-1.45)
(0.90-1.42)
(0.83-1.67)
(0.80-1.31)
(0.99-1.01)
(0.98-1.01)
0.12
0.21
0.53
8.32
0.39
2.73
0.44
0.99
0.33
0.49
0.41
0.48
0.31
3.57
0.47
9.51
0.79
0.36
1.19
0.01
3.91
1.42
1.89
1.17
0.81
0.02
1.01
0.62
.733
.650
2.64
.004
.531
.098
.506
.319
.564
.485
.523
.490
.579
.059
.492
.002
.374
.547
.275
.910
.048
.234
.169
.2791
.367
.877
.314
.432
(0.51-1.62)
(0.60-1.12)
(0.55-2.86)
(0.15-0.66)
(0.14-2.15)
(0.93-1.53)
–
–
–
–
–
–
–
0.99 (0.97-1.01)
–
1.06 (1.02-1.10)
–
–
–
–
1.19 (0.94-1.51)
–
–
–
–
–
–
–
0.10
1.55
0.29
9.41
0.76
1.96
–
–
–
–
–
–
–
1.77
–
8.05
–
–
–
–
2.01
–
–
–
–
–
–
–
.747
.213
.592
.002
.383
.161
–
–
–
–
–
–
–
.184
–
.005
–
–
–
–
.156
–
–
–
–
–
–
–
0.97 (0.69-1.35)
0.96 (0.74-1.23)
0.03
0.12
.853
.725
–
–
–
–
–
–
0.95 (0.69-1.31)
0.89 (0.69-1.16)
0.53 (0.33-0.84)
0.10
0.76
7.15
.746
.383
.008
0.89 (0.64-1.25)
0.90 (0.68-1.18)
0.54 (0.33-0.87)
0.44
0.59
6.28
.509
.441
.012
≥5 patients eligible for CRT (Figure 2). The proportion of
CRT-P devices ranged from 0.0% to 100% between all
practices (0.0% and 25.6%, 25th and 75th percentiles,
respectively). When the analysis included only practices
with ≥5 patients eligible for CRT, the CRT-P relative to
total CRT devices ranged from 0.0% to 100.0% (Figure 3).
The unadjusted univariate odds ratios (ORs) and 95%
CIs for characteristics associated with CRT are presented
in Table IV. Older patient age was associated with lower
CRT use. Insurance status, ischemic heart disease, LVEF,
longer QRS duration, and aldosterone antagonist use
were significantly associated with CRT use. Practices
located in the West had significantly lower rates of CRT
compared with those in the Northeast.
0.91
0.82
1.25
0.32
0.54
1.19
Multivariable analyses of CRT use
The multivariable model demonstrated that increasing
age and lack of insurance compared with Medicare were
independently associated with lower CRT use, whereas
longer QRS duration was a significant predictor of CRT use
(Table IV). Treatment with aldosterone antagonists, HF
etiology, and LVEF were no longer significant predictors
of variations in CRT. Geographic location of practices
remained a significant predictor of CRT utilization.
Predictors of CRT-P versus CRT-D use
The adjusted multivariable model revealed several
patient and practice factors that were associated with
use of CRT-P versus CRT-D (Table V). Older patient age
American Heart Journal
December 2009
962 Curtis et al
Table V. Adjusted OR for CRT-P compared with CRT-D use in
eligible patients
Patient and practice
characteristics
Age (10 y)
Sex (male vs female)
Myocardial infarction
Coronary artery bypass graft
HF clinic
Device clinic
Multispecialty
No. of patients annually per 500
β-Blocker use
OR (95% CI)
1.64
0.54
0.69
0.57
0.82
4.11
0.31
1.02
0.66
(1.26-2.13)
(0.30-0.95)
(0.37-1.27)
(0.31-1.06)
(0.45-1.46)
(1.21-13.94)
(0.12-0.77)
(0.99-1.06)
(0.33-1.33)
Wald m 2
P
13.59
4.59
1.45
3.20
0.47
5.14
6.33
1.53
1.33
.0002
.0321
.2287
.0736
.4925
.0233
.0119
.2156
.2490
was significantly associated with CRT-P implantation,
whereas male sex was associated with significantly lower
CRT-P use. Implantation of a CRT-P was lower at
multispecialty compared with single-specialty clinics,
whereas practices affiliated with device clinics were
significantly more likely to use CRT-P compared with
those not affiliated with such clinics.
Discussion
This study evaluated CRT utilization in a large cohort of
patients with HF receiving care in outpatient cardiology
practices geographically distributed throughout the United States. Although CRT is indicated for only selected
patients with HF, these results suggest that a large
proportion of patients eligible for CRT based on medical
record information did not receive this intervention and
that use of CRT for eligible patients varied widely by
practice setting despite national guidelines and evidence
from clinical trials to support CRT. Notably, ICD
implantation rates were higher (50.0%) for patients
eligible for CRT but not treated with CRT, which is not
well aligned to evidence showing greater magnitude of
benefit from CRT.15
A QRS delay affects 30% to 60% of outpatients with
NYHA class III or IV.16,17 Cardiac resynchronization
therapy for appropriately selected patients has the
potential to enhance quality of life and reduce morbidity
and mortality associated with HF. The baseline characteristics of patients eligible for CRT in IMPROVE HF were
generally similar to patients enrolled in clinical trials of
CRT with respect to comorbid health conditions and HF
etiology. For example, 65% of the IMPROVE HF cohort
was documented to have an ischemic etiology, compared
with 58% in a systematic review of CRT trials.18 The use
of evidenced-based medical therapies in IMPROVE HF
was also similar to that of patients enrolled in randomized
controlled trials of CRT. Among patients treated with a
CRT device in this study, 78.2% received ACEIs/ARBs,
85.2% β-blockers, and 37.9% aldosterone antagonists,
compared with 95%, 70%, and 54%, respectively, in
CARE-HF.8 The mean age of patients eligible for CRT in
IMPROVE HF was 71 years, which was modestly older
than patients in CARE-HF.8
Predictors of CRT use
Many patient demographic and clinical characteristics
were not significant predictors of variations in CRT rates,
including sex, HF etiology, comorbid health conditions,
current symptoms, laboratory results, and treatment with
medical therapies. For example, after adjustment for
other characteristics, CRT implantation was equivalent
for women and men, which differs from other studies
that demonstrated lower use of CRT for female
patients.19,20 However, significantly higher rates of CRT
were associated with younger age, insurance status, and
longer QRS duration. The lower rates of CRT use in
practices located in the western United States may reflect
differences in training, familiarity with guidelines, or
variations in systems to facilitate implementation of
recommended care. This finding establishes the need
for interventions to increase adherence to guidelines for
the care of patients with HF throughout the United States.
It should be noted that some practices participating in
IMPROVE HF provided CRT to a large proportion of
eligible patients, suggesting that it is feasible to translate
the clinical evidence and guideline recommendations for
CRT into routine clinical practice.
The lack of quantitative or qualitative documentation
of NYHA functional class for 41.5% of patients and QRS
duration for 32.3% of the cohort is notable. This information is required to determine CRT eligibility, and the
absence of this information prevents determination of
the exact proportion of patients in the IMPROVE HF
registry who might have benefited from CRT. This
finding also suggests that physicians in outpatient
practices may not routinely consider CRT as a treatment
option for patients.
The variations in CRT use in IMPROVE HF may reflect
differences in documentation of current HF symptoms,
comorbidities, contraindications, intolerance, and patient
or physician reasons for not providing CRT. The decision
to proceed with CRT often requires multiple, detailed
discussions with patients that may not be documented in
medical records, especially when decisions are made to
forego use of an otherwise indicated device. Consequently, information in medical records may have suggested
that patients were eligible for treatment but that
appropriate decisions were made to withhold CRT.
This study has several strengths compared with other
registries of patients with HF. Detailed information on
patient eligibility for CRT was collected, which allowed
data analysis to include only patients eligible for CRT.
Unlike other device-based registries, the study population
was not limited only to patients who actually received a
CRT device, which facilitated identification of patient and
practice characteristics that were independent predictors
of CRT.
American Heart Journal
Volume 158, Number 6
Limitations
Certain limitations are inherent in the design of
IMPROVE HF and should be considered when these
findings are interpreted. Data were collected by medical
chart review, which depends on the accuracy and
completeness of documentation. Consequently, a proportion of patients considered eligible for CRT but not
treated may have had medical contraindications, comorbid health conditions, physician rationale(s), or patient
rationales (including treatment refusal) for the decision
not to utilize CRT that were not documented in the
medical record. An additional, undetermined number of
eligible patients may have received CRT after completion of baseline data collection and longer exposure to
cardiology care. Thus, it is possible that this study
underestimated the magnitude of utilization and variation in rates of CRT. The failure to document NYHA
functional class for 41.5% of patients and QRS duration
for 32.3% of patients is a significant limitation that
precluded determination of eligibility for CRT for the
entire cohort. It is likely that some of these patients
were eligible for CRT, which may have resulted in
overestimation of CRT rates. Furthermore, independent
verification of NYHA class was not done to validate the
NYHA class documented in the medical record. Residual
confounding variables may account for some of the
observations. Results demonstrated associations rather
than causal relationships between patient and practice
characteristics and CRT implantation. Furthermore,
patients treated at practices enrolled in IMPROVE HF
may not be entirely representative of the general
outpatient population of patients with HF, as IMPROVE
HF patients were followed up in self-selected cardiology
practices and ascertainment bias may relate to this point
of care.
Conclusions
Cardiac resynchronization therapy is an evidencebased, class I guideline-recommended treatment for
eligible patients with systolic HF. This study suggests
that CRT is underutilized in eligible patients receiving
care in outpatient cardiology practices with significant
variations associated with patient and practice characteristics, such as patient age, insurance, QRS
duration, and practice geographic location, whereas
comorbid conditions were not significantly associated
with CRT use. Additional research is needed to
identify the full complement of reasons to explain
these apparent differences in adherence to guidelinedriven, evidence-based CRT use at the patient,
physician, and practice levels. Performance improvement initiatives should be evaluated to determine if
they can reduce variations in CRT use in eligible
patients and improve the overall quality of outpatient
care provided to patients with HF.
Curtis et al 963
Disclosures
The IMPROVE HF registry and this study are sponsored
by Medtronic, Inc, Minneapolis, MN. The authors served
as consultants to Medtronic, Inc. As of June 2008 this
relationship expired for Dr Yancy. Dr McBride is an
employee of Outcome Sciences, Inc.
References
1. Rosamond W, Flegal K, Furie K, et al. Heart disease and stroke
statistics—2008 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Committee. Circulation
2008;117:e25-146.
2. Hunt SA, Abraham WT, Chin MH, et al. ACC/AHA 2005 guideline
update for the diagnosis and management of chronic heart failure in
the adult. Circulation 2005;112:e154-235.
3. Richardson M, Freemantle N, Calvert MJ, et al. CARE-HF Study
Steering Committee and Investigators. Predictors and treatment
response with cardiac resynchronization therapy in patients with
heart failure characterized by dyssynchrony: a pre-defined analysis
from the CARE-HF trial. Eur Heart J 2007;28:1827-34.
4. Cleland J, Freemantle N, Ghio S, et al. Predicting the long-term effects
of cardiac resynchronization therapy on mortality from baseline
variables and the early response: a report from the CARE-HF
(Cardiac Resynchronization in Heart Failure) Trial. J Am Coll Cardiol
2008;524:38-45.
5. Gasparini M, Regoli F, Ceriotti C, et al. Remission of left ventricular
systolic dysfunction and of heart failure symptoms after cardiac
resynchronization therapy: temporal pattern and clinical predictors.
Am Heart J 2008;155:507-14.
6. Young JB, Abraham WT, Smith AL, et al. Combined cardiac
resynchronization and implantable cardioversion defibrillation in
advanced chronic heart failure: the MIRACLE ICD Trial. JAMA 2003;
289:2685-94.
7. Cleland JG, Daubert JC, Erdmann E, et al. Longer-term effects of
cardiac resynchronization therapy on mortality in heart failure [the
CArdiac REsynchronization-Heart Failure (CARE-HF) trial extension
phase]. Eur Heart J 2006;27:1928-32.
8. Cleland JG, Daubert JC, Erdmann E, et al. The effect of cardiac
resynchronization on morbidity and mortality in heart failure. N Engl
J Med 2005;352:1539-49.
9. Fung JW, Chan JY, Kum LC, et al. Suboptimal medical therapy in
patients with systolic heart failure is associated with less improvement
by cardiac resynchronization therapy. Int J Cardiol 2007;115:
214-9.
10. Rossi A, Rossi G, Piacenti M, et al. The current role of cardiac
resynchronization therapy in reducing mortality and hospitalization
in heart failure patients: a meta-analysis from clinical trials. Heart
Vessels 2008;23:217-23.
11. McAlister FA, Ezekowitz J, Hooton N, et al. Cardiac resynchronization therapy for patients with left ventricular systolic dysfunction: a
systematic review. JAMA 2007;297:2502-14.
12. Turley AJ, Raja SG, Salhiyyah K, et al. Does cardiac resynchronization therapy improve survival and quality of life in patients with
end-stage heart failure? Interact Cardiovasc Thorac Surg 2008;7:
1141-6.
13. Abraham WT, Fisher WG, Smith AL, et al. Cardiac resynchronization
in chronic heart failure. N Engl J Med 2002;346:1845-53.
14. Fonarow GC, Yancy CW, Albert NM, et al. Improving the use of
evidence-based heart failure therapies in the outpatient setting: the
American Heart Journal
December 2009
964 Curtis et al
IMPROVE HF performance improvement registry. Am Heart J 2007;
154:12-38.
15. Cleland JGF, Tageldien A, Maarouf N, et al. Patients with heart
failure who require an implantable defibrillator should have cardiac
resynchronisation routinely. Heart 2008;94:963-6.
16. Galizio NO, Pesce R, Valero E, et al. Which patients with congestive
heart failure may benefit from biventricular pacing? Pacing Clin
Electrophysiol 2003;26:158-61.
17. Kashani A, Barold SS. Significance of QRS complex duration in
patients with heart failure. J Am Coll Cardiol 2005;46:2183-92.
O
18. McAlister FA, Ezekowitz JA, Wiebe N, et al. Systematic review:
cardiac resynchronization in patients with symptomatic heart failure.
Ann Intern Med 2004;141:381-90.
19. Hernandez AF, Fonarow GC, Liang L, et al. Sex and racial
differences in the use of implantable cardioverter-defibrillators
among patients hospitalized with heart failure. JAMA 2007;298:
1525-32.
20. Alaeddini J, Wood MA, Amin MS, et al. Gender disparity in the use
of cardiac resynchronization therapy in the United States. Pacing Clin
Electrophysiol 2008;31:468-72.
N THE MOVE?
Send us your new address at least 6 weeks ahead
Don’t miss a single issue of the journal! To ensure prompt service when you change your address, please photocopy
and complete the form below.
Please send your change of address notification at least 6 weeks before your move to ensure continued service. We
regret we cannot guarantee replacement of issues missed because of late notification.
JOURNAL TITLE:
Fill in the title of the journal here.
__________________________________________________________
OLD ADDRESS:
NEW ADDRESS:
Affix the address label from a recent issue of the journal here.
Clearly print your new address here.
Name ____________________________________________________
Address __________________________________________________
City/State/ZIP _____________________________________________
COPY AND MAIL THIS FORM TO:
Elsevier Health Sciences Division
Subscription Customer Service
3251 Riverport Lane, Maryland Heights, MO 63043
OR FAX TO:
OR PHONE:
OR E-MAIL:
341-447-8029
1-800-654-2452
Outside the U.S., call
341-447-8871
[email protected]