DECOMPENSATED HEART FAILURE: LESSONS LEARNED FROM THE ADHERE REGISRY

1
DECOMPENSATED HEART FAILURELESSONS LEARNED FROM
THE ADHERE REGISRY

• Maria Rosa Costanzo, M.D.
• Midwest Heart Specialists
• Medical Director, Edward Center for Advanced
  Heart Failure
• Naperville, Illinois, U.S.A.

2
Why Focus on Acute Heart Failure?

• Despite advances in diagnosis and treatment of HF
  over 1 million patients will be hospitalized this
  year
• HF hospitalizations continue to be one of largest
  expenses for CMS1,2
• There are currently no national guidelines for
  acute heart failure management
• Hospital readmissions
• 20 at 30 days
• 50 at 6 months
• Mortality
• 11.6 at 30 days3
• 33.1 at 12 months3
• Clinical trials in heart failure
• Focus on Omit
• Stable outpatients
  Criteria for admission to hospital
• Systolic dysfunction
  Treatments for acute heart failure
  Enroll relatively younger
  Diastolic dysfunction
• pts and exclude many
• pts with co-morbidities

1American Heart Association. 2003 Heart and
Stroke Statistical Update. Dallas, Tex American
Heart Association 2002. 2Hunt SA et al. ACC/AHA
guidelines for the evaluation and management of
chronic heart failure in the adult. 2001. 3Jong
P et al. Arch Intern Med. 200216216891694.
3

• Observational Studies
• Advantages
• All inclusive. Patients with co-morbidities,
  women of child bearing potential, elderly
  included. Real-world
• Can provide detailed information of patient
  characteristics, treatment strategies, and
  outcomes of interest
• With large numbers of patients can allow
  assessment of infrequent events or unusual
  patient populations
• Multiple analyses can be performed on same
  cohort. Assess interventions with and without
  commercial value
• Disadvantages
• Potential selection, observational, and
  investigator bias and can be confounded by
  variety of factors

• The ADHERE
• (Acute Decompensated Heart Failure National
  Registry) Registry
• Phase IV
• Multicenter
• Observational
• Open label
• Electronic web-based
• Registry of the management of patients
  treated in hospitals for acutely decompensated
  heart failure in the US

4
Goals of the ADHERE Registry

• Describe demographics and clinical
  characteristics of patients hospitalized with
  acutely decompensated heart failure (AHF)
• Characterize current management of hospitalized
  patients with AHF
• Define treatment strategies associated with best
  clinical outcomes and most efficient use of
  resources
• Assist in evaluating and improving the quality
  of care

5
The ADHERE Registry
gt150,000 pts from 263 US Hospitals
Electronic Data Capture (EDC) System
6
Characteristics of Heart Failure Patients
Enrolled in the ADHERE Registry

• Average age 72.5 years
• Women 52
• Ischemic etiology (CAD) 60
• Renal insufficiency 30
• Diabetes 44
• Preserved LV systolic function ?50
• Atrial fibrillation 31
• Diabetes 44

7
Crucial Link Between LV Assessment and ACEI Use
ADHERE Past Medical HistoryAll Enrolled
Discharges in the Last 12 Months
(07.01.2002-06.30.2003)
The Nation n58919 75 45 (n26719) 58
Prior Heart Failure () Prehospital LVEF Assessed
() lt40 or Mod/Sev Impairment ()
66
59
55
53
48
28
Newman 97
Philbin 98
Senni 99
8
Utilization of Evidence-based Therapies in Heart
Failure
History of HF and LVEF Documented and ? 0.40
?
Excludes patients with documented
contraindications.
2300/7883 Patients hospitalized with HF prior
known dx of systolic dysfunction HF outpatient
medical regimen. ADHERE Registry Report Q1 2002
(4/01-3/02) of 180 US Hospitals Presented by GC
Fonarow at the Heart Failure Society of America
Satellite Symposium, September 23, 2002.
9
ADHERE Variation in ACEI Use
ORYX Core Measure HF 3 - LVEF lt 40 prescribed
ACEI at discharge
100
80
60
Rate ()
40
20
0
ADHERE Hospitals
ADHERE Dec 2002, 206 Hospitals 23,193 patients
(subset with LVEF lt .40, no CI) Fonarow J Card
Fail 20039S79
10
Demographic Characteristics, Clinical
Characteristics and Outcomes
11
First (Geographic) Point of Care at Registry
Hospital
All Enrolled Discharges (n105,388) October 2001
to January 2004
Observation Unit lt1
12
IV Vasoactive Use
Door to first vasoactive treatment
27 of patients are given IV vasoactives
The ADHERE Registry 2nd Quarter 2003 National
Benchmark Report Scios Inc.
13
IV Vasoactive UseImportant Where Begun
Length of Stay

• Common vasoactives used include
• Nesiritide 10
• Nitroglycerin 10
• Dopamine 6
• Dobutamine 6
• Milrinone 3

7.0
4.5
The ADHERE Registry 2nd Quarter 2003 National
Benchmark Report Scios Inc.
14
IV Vasoactive UseImportant Where Begun
In-Hospital Mortality
10.9

4.3
plt0.0001 vs inpatient unit
The ADHERE Registry 2nd Quarter 2003 National
Benchmark Report Scios Inc.
15
Most Common IV Medications All Enrolled
Discharges (n105,388) October 2001 to January
2004
16
IV Diuretic Use

• 64 of patients are given an IV diuretic only
• Common diuretics used include
• Furosemide 84
• Bumetanide 7
• Torsemide 2

88
64
The ADHERE Registry 2nd Quarter 2003 National
Benchmark Report Scios Inc.
17
Complications of Diuretic Therapy for Heart
Failure
Diuretic Therapy
Saluresis and Diuresis
? Plasma Volume
? Cardiac Output
? Renal Blood Flow
? PRA
? GFR
? Aldosterone
Postural Hypotension
? Distal CaReabsorption
? ProximalReabsorption
Kaliuresis
Pre-renalAzotemia
Hypokalemia
? Uric AcidClearance
? CalciumClearance
Glucose Intolerance
Hypocalcemia
Hyperuricemia
GFR glomerular filtration rate PRA plasma
renin activity. Kaplan NM. Treatment of
hypertension drug therapy in clinical
hypertension. In Kaplan NM, Lieberman E, Neal
WW, eds. Clinical Hypertension. 1994203.
18
Marked Activation of the RAAS by Loop Diuretics
50
10
Mean Confidence Interval
Plasma Renin Activity (ng/mL/h)
2.5
0.5
After Diuretic (n 11)
Before Diuretic (n 12)
Bayliss J et al. Br Heart J. 1987571722.
19
Diuretic Therapy Significantly Decreases
Glomerular Filtration Rate
N 16 NYHA II (19) and III (81) mean
baseline creatinine clearance, 108 51
µg/mL. GFR was estimated using a 7-hour
creatinine clearance.Gottlieb SS et al.
Circulation. 200210513481353.
20
Vasodilation Is Required to Normalize
Ventricular Filling Pressures
IV Diuretic Monotherapy Causes Reflex
Vasoconstriction, Increased Afterload, and
Decreased Cardiac Index
100
IV furosemide nitroprusside
80
IV furosemide alone
60
Maximal Stroke Volume ()
40
20
0
0
10
20
30
40
Pulmonary Capillary Wedge Pressure (mm Hg)
25 class IV patients furosemide alone or with IV
nitroprusside. Adapted from Stevenson LW,
Tillisch JH. Circulation. 19867413031308.
21
Diuretic Use and the Risk of Mortality in
Patients with Left Ventricular Dysfunction
Mortality Risk by Diuretic Use at Baseline
SOLVD database Cooper HA et al. Circulation.
1999 100(12) 1311
22
Renal Insufficiency Chronic Diuretic Therapy
Mortality
9
7.8
8
7
of pts
6
No ChD
ChD
5
4
P lt 0.0001
3
2
Cr gt 2.0
Cr lt 2.0
Cr lt 2.0
Cr gt 2.0
Creatinine level
Costanzo MR et al. JACC 2004 43 (5) Supl. A 180A
23
Renal Insufficiency Chronic Diuretic Therapy
Odds Ratio of Mortality
Costanzo MR et al. JACC 2004 43 (5) Supl. A 180A
24
Most Common IV Medications All Enrolled
Discharges (n105,388) October 2001 to January
2004
25
Intravenous Inotropic Agents During
Hospitalization for Decompensated Heart Failure
OPTIME-CHF
Event Rate ()
20
HR 6.0 P lt 0.001
Milrinone
HR 3.3 P lt 0.001
Placebo
15
HR 3.1 P 0.004
10
HR 1.7 P 0.19
HR 3.8 P 0.18
5
0
Adverse Event
Sustained Hypotension
Acute MI
Mortality
Afib
Cuffe MS et al. JAMA. 200228715411547.
26
VMAC PCWP Through 48 Hours
0
-1
NTG
Nesiritide
-2
Plt0.05 pooled nesiritide compared to
nitroglycerin
p lt 0.05 versus NTG
-3

-4



-5

-6


-7

-8
-9
-10
-11
Time
Publication Committee for the VMAC Investigators
JAMA 2002 2871531-1540
27
VMAC Primary Endpoint
Dyspnea at 3 hours
100
p-values are based on Van Elteren Test with
7-point ordinal scale
P0.034
90
P0.191
80
70
60
Improved ()
50
40
30
20
10
No change
0
Worsened ()
Placebo
Nesiritide
NTG
-10
Publication Committee for the VMAC Investigators
JAMA 2002 2871531-1540
28
Adverse Events in VMAC During Placebo-Controlled
Period
Adverse Event
NTG
Nesiritide
Placebo
P value
(n 143)
(n204)
(n 142)
39 (27)
36 (18)
20 (14)
0.015
Any adverse event
Headache
17 (12)
11 (5)
3 (2)
0.003
Hypotension
6 (4)
5 (2)
0 (0)
0.031
2 (1)
1 (0.5)
0 (0)
0.481
Symptomatic hypotension
Abdominal pain
4 (3)
0 (0)
0 (0)
0.014
Publication Committee for the VMAC Investigators
JAMA 2002 2871531-1540
29
Mortality Data Risk Adjustment Process
30
ADHERE In-Hospital Mortality and Use of
Parenteral Vasoactive Medications
Results of Comparative Mortality Analysis of
Nesiritide Treatment vs Nitroglycerin, Milrinone
or Dobutamine
Favors Other Agent
Favors Nesiritide
Nitrogycerin n 5,902(Nesiritide n 4,573)
P0.300
Milrinone n 1,631(Nesiritide n 4,830)
P0.0001
Dobutamine n 3,437(Nesiritide n 4,431)
P0.0001
1
2
0
ADJUSTED Odds Ratio(and confidence intervals)
Journal of Cardiac Failure, October, 2003 9(5)
(Suppl) S81 (Abstract 298) presented at HFSA,
2003
31
Predictors of Mortality in ADHERE

• ADHERE is one of the largest and the most
  comprehensive datasets on patients hospitalized
  with acutely decompensated heart failure
• CART (Classification and Regression Tree)
  analysis to identify clinical variables
  predictive of lower, intermediate, and higher
  mortality risk
• Covariate and Propensity Adjusted Analysis of
  Mortality Analysis by IV Therapy
• Analysis of Variation in Processes of Care and
  Relationship to Clinical Outcomes

32
CART Analysis Variables Analyzed
Fonarow Circulation 2003108IV-693
33
ADHERE CART Predictors of Mortality
BUN 43 N33,324
Greater than
Less than
2.68 n25,122
8.98 n7,202
SYS BP 115 n24,933
SYS BP 115 n7,150
6.41 n5,102
15.28 N2,048
5.49 n4,099
2.14 n20,834
Cr 2.75 2,045
Highest to Lowest Risk Cohort OR 12.9 (95 CI
10.4-15.9)
21.94 n620
12.42 n1,425
Fonarow Circulation 2003108IV-693
34
Mortality Rates Comparison
Fonarow Circulation 2003108IV-693
35
ADHERE CART Analysis

• ADHF patients at low, medium, and
  highin-hospital mortality risk are easily
  identifiable from vital sign and laboratory data
  obtained on presentation
• This ADHERE Risk Tree provides a practical beside
  tool for mortality risk stratification
• Three variables are the strongest predictors
• BUN gt 43 mg/dL
• SBP lt 115 mmHg
• Serum creatinine gt 2.75 mg/dL

36
ADHERE CART Analysis

• The ADHERE CART analysis provides insights into
  individual risk variables for in-hospital
  mortality
• Renal insufficiency, volume status, and systemic
  perfusion have major prognostic importance
• The cardiorenal syndrome is a key determinate of
  ADHF prognosis
• This research sets the stage to define optimal
  treatment strategies to improve outcomes for ADHF
  patients at intermediate and high risk

37
Goals of the ADHERE Registry

• Describe demographics and clinical
  characteristics of patients hospitalized with
  acutely decompensated heart failure (AHF)
• Characterize current management of hospitalized
  patients with AHF
• Define treatment strategies associated with best
  clinical outcomes and most efficient use of
  resources
• Assist in evaluating and improving the quality
  of care

38
Impediments to the Uptake of Evidence Based
Medicine

• Inconsistent definition of heart failure
• No AHA/ACC/HFSA guidelines for acute heart
  failure
• Once patients symptoms have improved often
  viewed as no longer having heart failure
• Under-recognition that patients are at high-risk
  for disease progression
• Poor communication between EMC physician,
  cardiologist and primary care physician, expect
  therapies to be started as outpatient, but does
  not happen
• Lack of systems

39
Performance Indicators for Heart Failure Patient
Care (JCAHO)
Patients Treated ()
Performance Indicator
HF-1 n28,776 HF-2 n34,397 HF-3 n12,725
HF-4 n5,475 Fonarow J Card Failure 20039S79
40
ADHERE Critical PathwaysPerformance Improvement
Process
Find and Support a Champion
Assess HF Treatment Rates Enter Data into ADHERE
Registry
Implement Refined Protocol Hospital Team
Coordinates Implementation of Refined Protocol
and Tools
Evaluate and Assessment Hospital Team Reviews
ADHERE Reports
Refine Protocol Hospital Team Identifies Areas
for Improvement and uses Tool Kit
41
Trends in Treatment for AHF in ADHERE Q1 2002 to
Q4 2003
Use of Nesiritide
Use of Inotropes
20
16
18
14
16
12
14
10
12


10
8
8
6
6
Inotrope
Nesiritide
4
4
2
2
0
0
2002.1
2002.2
2002.3
2002.4
2003.1
2003.2
2003.3
2003.4
2002.1
2002.2
2002.3
2002.4
2003.1
2003.2
2003.3
2003.4
42
Trends in Clinical Outcomes for AHF in ADHERE Q1
2002 to Q4 2003

• Need for mechanical ventilation decreased from
  5.3 to 3.6 (RR 0.73, Plt0.0001)
• ICU LOS decreased from mean 4.4 to 3.4 days
  (Plt0.0001)
• Hospital LOS decreased from mean 6.3 to 5.8 days
  (Plt0.0001)
• In-hospital mortality decreased from 4.5 to 3.9
  (RR 0.86, P0.03)

43
Trends in Quality of Care at Discharge in
ADHERE Q1 2002 to Q4 2003
Q1 2002 n 8,198 Q2 2002 n11,289 Q3 2002
n14,430 Q4 2002 n16,925
Q1 2003 n17,735 Q2 2003 n16,719 Q3 2003
n13,984 Q4 2003 n 10,265
Baseline Characteristics Similar All 8 Quarters
44
Key Elements to Quality Improvement

• Why Do Some Hospitals Succeed?
• Access to current and accurate data on treatment
  and outcomes
• Have stated goals
• Administrative support
• Physician champion, support among clinicians
• Use of pre-printed orders, care maps
• Use of data to provide feedback

Bradley JAMA 20012852604-2611
45
Goals of the ADHERE Registry

• Describe demographics and clinical
  characteristics of patients hospitalized with
  acutely decompensated heart failure (AHF)
• Characterize current management of hospitalized
  patients with AHF
• Define treatment strategies associated with best
  clinical outcomes and most efficient use of
  resources
• Assist in evaluating and improving the quality of
  care