Validation of Heart Failure Events in ALLHAT Participants Assigned to Doxazosin and Chlorthalidone

1
Validation of Heart Failure Events in ALLHAT
Participants Assigned to Doxazosin and
Chlorthalidone

• L.B.Piller, B.R.Davis, J.A.Cutler, W.C.Cushman,
  J.T. Wright,
• J.D.Williamson, F.H.Leenen, O.Randall, J.S.Golden
• The University of Texas School of Public Health
  Houston, Texas

2
At Issue
The doxazosin group showed a doubling of risk of
CHF compared to the chlorthalidone group (RR,
2.04 95 CI, 1.79-2.32).
3
Objectives

• To describe the process by which the clinical
  reports of heart failure events were validated
• To describe the statistical analyses which led to
  the cessation of the doxazosin arm

4
Congestive Heart Failure
0.10
0.08
0.06
Cumulative Event Rate
doxazosin
chlorthalidone
0.04
0.02
9,541 5,457
5,531 3,089
2,427 1,351
9,541 5,457
9,541 5,457
13,644 7,845
0
0
1
2
3
4
C 15,256 D 9,061
Years of Follow-up
5
Relative Risks and 95 CICongestive Heart
Failure Doxazosin/Chlorthalidone
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
Age
Age
BNH
WNH
Hisp
BNH
WNH
Hisp
Nondi
Total
Diab
65
Men
Men
Men
W
W
W
a
2.32
2.53
2.41
2.81
2.40
4.14
3.21
2.62
3.97
2.59
2.37
High
1.79
1.51
1.78
1.47
1.58
1.11
1.67
1.49
0.85
1.76
1.67
Low
2.04
1.96
2.07
2.04
1.94
2.15
2.32
1.98
1.84
2.14
1.98
RR
6
Methods

• Comparison of baseline characteristics of CHF and
  non-CHF participants
• Comparison of pre and post-event medical
  management
• Endpoints Subcommittee evaluation of a sample of
  hospitalized and fatal heart failure cases using
  study criteria
• Determination and comparison of frequency and
  severity of systolic dysfunction through
  ascertainment of ejection fractions (EF)
• Comparison of case-fatality rates and causes of
  death
• Evaluate more stringent CHF-defined outcome

7
Baseline Characteristics

• Do baseline characteristics differ in expected
  ways between participants with and without CHF,
  and are these differences generally similar in
  the doxazosin and chlorthalidone arms?

8
Baseline CharacteristicsCHF vs. Non-CHF
Participants
9
Baseline CharacteristicsCHF vs. Non-CHF
Participants
10
Eligibility Criteria CHF vs. Non-CHF
Participants
11
Pre and Post-Event Treatment of Heart
Failure Participants

• Was pre-heart failure event step 1 treatment
  compliance equivalent in the chlorthalidone and
  doxazosin groups?
• Were the heart failure participants in the
  chlorthalidone and doxazosin groups managed
  similarly following the heart failure events?

12
Pre-Event Treatment ofHospitalized CHF
Participants
13
Pre-Event Treatment of Non-Hospitalized
but Treated CHF Participants
14
Post-Event Treatment ofHospitalized CHF
Participants
15
Post-Event Treatment of Non-Hospitalized
but Treated CHF Participants
16
Heart Failure Event Reporting

• Did the reports of hospitalized or fatal heart
  failure adhere to the study criteria?

17
ALLHAT Study Events
Management
Documentation

• Documentation required
• Discharge summaries for all hospitalizations
• Death certificates for all deaths
• Additional QC documentation for random 10 sample
  of MIs, strokes and fatal CHD

• Routinely reviewed for accuracy and
  appropriateness
• Queries sent to the sites for clarification of
  discrepancies
• A random 10 of MIs, strokes and fatal CHD
  reports reevaluated by the Endpoints Subcommittee
  for quality control

18
Endpoint SubcommitteeEvaluation of CHF Events

• 50 cases of reported hospitalized and/or fatal
  CHF
• Each case reviewed by two Subcommittee members
• Criteria for confirmation of CHF as described in
  the Manual of Operations

19
Criteria for CHFEvaluation
ALLHAT Manual of Operations, 5.3.4
20
New York Heart Association Functional
Classification of Congestive Heart Failure
I No limitations of activity ordinary
activity does not cause undue fatigue,
palpitations, dyspnea or anginal pain II Slight
limitations of activity asymptomatic at rest
ordinary activity results in fatigue,
palpitations, dyspnea or anginal pain III
Marked limitations of activity usually
asymptomatic at rest less than ordinary activity
causes fatigue, palpitations, dyspnea or anginal
pain IV Inability to carry on any physical
activity without discomfort symptoms at rest
increased discomfort with any physical activity
Criteria Committee, New York Heart Association.
Diseases of the heart and blood vessels.
Nomenclature and criteria for diagnosis. 6th ed.
1964114.
21
Framingham Criteria for the Diagnosis of
Congestive Heart Failure

• Major Criteria
• Paroxysmal nocturnal dyspnea (A)
• Neck vein distention (B)
• Rales (B)
• Cardiomegaly (B)
• Acute pulmonary edema (B)
• S3 gallop (B)
• venous pressure (B)
• hepatojugular reflux

Minor Criteria Extremity edema (B) Night cough
(A) Dyspnea on exertion (A) Hepatomegaly Pleural
effusion (B) Vital capacity ? by 1/3 of
nml Tachycardia 120 bpm (B)
Major or Minor Weight loss 4.5 kg over 5 days
treatment
At least two major or one major and two minor
criteria are required for the diagnosis of
CHF. KKL Ho et al, Circulation 88107, 1993.
22
Endpoints Subcommittee Evaluation of CHF
Sample (n50)
Cases with insufficient data 11/50 Remaining
cases with CHF 33/39 (85) Remaining
chlorthalidone cases with CHF 9/10
(90) Remaining doxazosin cases with CHF 9/10
(90)
23
Severity of Heart Failure

• What is the frequency of systolic dysfunction in
  doxazosin and chlorthalidone participants with
  heart failure?
• Is the systolic dysfunction equally severe in the
  chlorthalidone and doxazosin participants with
  heart failure?

24
Assessment of the Severityof Hospitalized and
Fatal Heart Failure

• Half of hospitalized or fatal heart failure
  events reviewed for ejection fractions (EF)
• If EF documented, further information sought
• How was it obtained?
• Was the EF equally severe across drug groups?
• Given possible differences between VA and non-VA
  sites in levels of diagnostic testing, was there
  a difference between the two groups in EF data?

25
CHF Ejection Fraction Data
26
CHF Ejection Fraction DataHospitalized or
Fatal CHF
27
CHF Ejection Fraction Data Hospitalized or
Fatal CHF (VA Clinics)
28
CHF Ejection Fraction Data Hospitalized or
Fatal CHF (Non-VA Clinics)
29
Fatality Rates and Cause of Death

• Is all-cause mortality similar for the
  chlorthalidone and doxazosin participants?
• Are case-fatality rates similar and as high as
  expected for chlorthalidone and doxazosin
  participants hospitalized for heart failure?
• Are causes of deaths for chlorthalidone and
  doxazosin participants previously hospitalized
  for heart failure similarly distributed?

30
All-Cause Mortality
Cumulative Event Rate
doxazosin
chlorthalidone
10,513 6,118
5,702 3,287
2,530 1,481
13,739 8,054
Years of Follow-up
C 15,268 D 9,067
31
Cumulative Mortality for Hospitalized CHF Cases
.5
.4
.3
Cumulative Mortality Rate
chlorthalidone
.2
doxazosin
.1
0
0
1
2
3
4
Years Hospitalized CHF to Death
32
Causes of Death of Participants
Hospitalized for CHF

33
Fatal and Hospitalized CHF Increased Risk?
Does the more stringent CHF outcome, fatal and
hospitalized CHF, show a similarly increased CHF
risk in the doxazosin group compared to the
chlorthalidone group?
34
Hospitalized or Fatal CHF
Rel Risk 95 CI 1.83
1.58-2.13 p .06
doxazosin
Cumulative Event Rate
.04
chlorthalidone
.02
0
0 0
0
1
2
3
4
Years of Follow-up
35
Relative Risks and 95 CIHospitalized or Fatal
Congestive Heart Failure Doxazosin/Chlorthalidone
36
Conclusions (1)

• The risk for the more stringent outcome of
  hospitalized or fatal heart failure was
  significantly higher in the doxazosin treatment
  group compared to the chlorthalidone treatment
  group (RR, 1.83 95 CI, 1.58-2.13).
• Case-fatality rates for participants with heart
  failure were high (20 in 2 years) and similar in
  both groups.
• Hospital discharge data for heart failure was
  largely consistent with the ALLHAT definition of
  heart failure.
• In both groups, two thirds of the hospitalized or
  fatal heart failure events with ejection fraction
  measurements had ejection fractions of 40 or
  less.

37
Conclusions (2)

• The treatment following the heart failure events
  in both groups was consistent with recommended
  treatment of heart failure in the community.
• Similar percentages of participants in both
  treatment groups were on step 1 medication prior
  to the heart failure event.
• Participants who developed heart failure had a
  greater history of coronary heart disease than
  participants who did not develop heart failure.