Prospective Evaluation and Identification of Cardiac Decompensation in Patients with Heart Failure b

1
Prospective Evaluation and Identification of
Cardiac Decompensation in Patients with Heart
Failure by Impedance Cardiography Test (PREDICT)

• Packer M, Abraham WT, Mehra MR, Yancy CW, et al.
• Utility of Impedance Cardiography for the
  Identification of Short-Term Risk of Clinical
  Decompensation In Stable Patients With Chronic
  Heart Failure. J Am Coll Cardiol. 2006472245-52.

2
Noninvasive Hemodynamic MonitoringImpedance
Cardiography (ICG)

• 4 dual sensors with 8 lead wires placed on neck
  and chest
• Current transmitted by outer electrodes and seeks
  path of least resistance blood filled aorta
• Baseline impedance (resistance) is measured using
  inner electrodes
• With each heartbeat, blood volume and velocity in
  the aorta change
• Corresponding change in impedance is measured
• Baseline and changes in impedance are used to
  measure and calculate hemodynamic parameters

3
ICG Hemodynamic Parameters
Stroke Volume / Index (SV / SI) Cardiac Output /
Index (CO / CI) Systemic Vascular Resistance /
Index (SVR / SVRI) Systolic Time Ratio
(STR) Pre-ejection Period (PEP) LV Ejection Time
(LVET) Velocity Index (VI) Acceleration Index
(ACI) Thoracic Fluid Content (TFC)
4
ICG Device and Report
5
Heart Failure
Definition Inability to provide cardiac output
necessary to meet bodys demands under normal
ventricular filling pressures Prevalence ? 5
million Americans with chronic heart failure at
age 40, lifetime risk of developing HF is
20 Incidence 550,000 new cases/year Morbidity 1,0
93,000 hospital discharges (2003) Most frequent
cause of hospitalization in elderly Mortality Caus
es or contributes to 286,000 deaths/year Cost
30 billion (2006) 5,912 per Medicare
discharge (2001)
AHA Heart and Stroke Disease Statistics 2006
Update.
6
Clinical Profiles of Heart Failure Wet/Dry and
Cold/Warm
Congestion (Wet)?
NO
YES
Signs/Symptoms of Congestion Orthopnea / PND JV
Distension Hepatomegaly Edema Rales Abd-Jugular
Reflex
B
A
YES
Warm Wet
Warm Dry
Adequate Perfusion (Cold) ?
(Complex)
(Low Profile)
C
L
NO
Cold Wet
Cold Dry
Evidence of Low Perfusion Narrow pulse
pressure Cool extremities Sleepy /
obtunded Hypotension with ACE inhibitor Low
serum sodium Renal/hepatic dysfunction
Nohria A, et al. J Am Coll Cardiol. 2003411797.
7
Reliability of Clinical Assessment for
Estimation of Hemodynamics

• 50 correct clinical judgement to determine PCWP
  as wet or dry or CI as warm or cold 1,2
• 58 sensitivity for clinical signs for elevated
  PCWP 3
• 51 accuracy for CO, 47 for SVR 4

1 Connors et al. 2 Eisenberg et al. 3 Stevenson
et al. 4 Steingrub et al.
8
PREDICT Background and Objective

• Prior studies have shown the prognostic value of
  hemodynamics over periods too long to allow for
  immediate intervention to prevent an imminent
  clinical event
• Purpose of study was to assess the utility of ICG
  in predicting short-term clinical deterioration
  in ambulatory patients with HF
• The prespecified primary hypothesis was that
  changes in ICG variables combined into a
  composite score would predict the occurrence of a
  major clinical event

Packer M, et al. J Am Coll Cardiol.
2006472245-52.
9
PREDICT Investigators
Principal Investigator Milton Packer,
MD University of Texas SW Medical Center, Dallas,
TX

• William T. Abraham, MD
• The Ohio State University Heart Center, Columbus,
  OH
• Mandeep R. Mehra, MD
• University of Maryland, Baltimore, MD
• Clyde W. Yancy, MD
• University of Texas SW Medical Center, Dallas, TX
• Christine E. Lawless, MD
• DuPage Medical Group, Chicago, IL
• Judith E. Mitchell, MD
• SUNY Downstate Medical Center, New York, NY
• Thierry H. Le Jemtel, MD
• Albert Einstein Hospital, New York, NY
• Frank W. Smart, MD

• Ileana L. Pina, MD
• Case Western Reserve University, Cleveland, OH
• Barry H. Greenberg, MD
• University of California, San Diego, San Diego,
  CA
• James B. Young, MD
• Cleveland Clinic Foundation, Cleveland, OH
• Daniel P. Fishbein, MD
• University of Washington, Seattle, WA
• Paul J. Hauptman, MD
• St. Louis University, St. Louis, MO
• Robert C. Bourge, MD
• University of Alabama Birmingham, Birmingham, AL
• John E. Strobeck, MD, PhD

10
PREDICT Inclusion and Exclusion Criteria

• Men or women, aged 18 and over, NYHA Class II,
  III, or IV with chronic HF 2 months duration
• Emergency department visit, unscheduled clinic
  visit, or hospitalization for HF lt 3 months
• Clinically stable, receiving medications for HF
  at doses considered appropriate
• Main exclusion criteria
• Height lt 47 or gt 91 weight lt 66 lbs or gt 342
  lbs
• Hemodynamically significant aortic regurgitation
• Myocarditis, cor pulmonale, hypertrophic/restricti
  ve myopathies
• Severe renal (Cr gt 5 mg/dl) or liver disease
  (ALT, AST gt 3X nl)
• Planned CRT, LVAD therapy
• Recent ACS, MI, sustained VT without ICD

Packer M, et al. J Am Coll Cardiol.
2006472245-52.
11
PREDICT Data Collection

• Visits to outpatient HF clinic every 2 weeks for
  26 weeks
• Recorded variables
• Baseline patient characteristics
• Age, gender, race, etiology, EF, medications
• Clinical
• Vital signs (HR, SBP, DBP) and weight
• Patient assessment (symptom score)
• On a scale of 1 to 100, how do you feel, with
  100 being the best?
• Functional assessment with NYHA classification
• Impedance cardiography
• BioZ ICG Monitor (CardioDynamics, San Diego, CA)
• ICG data was blinded
• Patients cared for in usual manner

Packer M, et al. J Am Coll Cardiol.
2006472245-52.
12
PREDICT Event Definition and Adjudication

• Reportable events
• Death, hospitalization, ED visit due to any cause
• HF events
• Death due to any cause, hospitalization/ED visit
  due to worsening HF
• Three cardiologists independently evaluated each
  hospitalization and ED visit event description
• Each cardiologist independently rated each event
• Related to worsening HF
• Not related to worsening HF
• Events in which all three cardiologists did not
  agree were jointly discussed and a final rating
  was determined by group consensus

Packer M, et al. J Am Coll Cardiol.
2006472245-52.
13
PREDICT Enrollment and Visits

• Sites 21
• Patients enrolled 212
• Visits completed 2316
• Average visits per patient 10.9 3.8

Packer M, et al. J Am Coll Cardiol.
2006472245-52.
14
PREDICT Baseline Characteristics

• Characteristic Value
• N 212
• Age 58.5 14.7
• Sex - male 145 (68.4)
• Race white 132 (62.3)
• Race black 74 (34.9)
• Ischemic etiology 98 (46.2)
• EF 27.1 13.5
• NYHA Classification 2.7 0.5
• Frequency Class II 67 (31.6)
• Class III 139 (65.9)
• Class IV 6 (2.8)
• Meds
• ACEI or ARB 177 (83.5)
• BB 152 (71.7)
• Aldosterone antagonist 86 (40.6)
• Diuretic 203 (96.2)

Packer M, et al. J Am Coll Cardiol.
2006472245-52.
15
PREDICT Results HF Events

• HF Event Type N Patients ()
• Death 16 16 (7.5)
• Hospitalization 78 50 (23.6)
• ED Visit 10 10 (4.7)
• Total 104 59 (27.8)
• At initial visit, no study variable could predict
  patients with or without a HF event during the
  study
• 77 HF events within 14 days of a preceding study
  visit
• 14-day HF event rate 77 events / 2316 visits
  (3.3)

Packer M, et al. J Am Coll Cardiol.
2006472245-52.
16
PREDICT Results Differences in Clinical
Variables Before HF Events
Packer M, et al. J Am Coll Cardiol.
2006472245-52.
17
PREDICT Results Differences in ICG Variables
Before HF Events
Packer M, et al. J Am Coll Cardiol.
2006472245-52.
18
PREDICT Results Multivariate AnalysisBaseline
Characteristics, Clinical Variables, ICG
variables, Change in Clinical ICG Variables
Association to HF Event lt 14 Days
Packer M, et al. J Am Coll Cardiol.
2006472245-52.
19
PREDICT Derivation of Independent ICG Predictor
Variables

• Thoracic fluid content index
• Inverse of baseline impedance, from which
  pulsatile change (delta Z) originates
• Velocity index
• Maximum deflection of the first derivative of the
  impedance waveform (C)
• Left ventricular ejection time
• Aortic valve opening (B) to closing (X)

Packer M, et al. J Am Coll Cardiol.
2006472245-52.
20
PREDICT ICG Score Calculation

• The three independent ICG variables from
  multivariate analysis were combined into a single
  composite score
• Higher thoracic fluid content index higher risk
• Lower velocity index higher risk
• Shorter left ventricular ejection time higher
  risk
• Regression score created and translated to log of
  odds, probability for HF event, and score value
• Score scale, 0 to 10
• Low risk 0 to 3
• Intermediate risk 4 to 6
• High risk 7 to 10

Packer M, et al. J Am Coll Cardiol.
2006472245-52.
21
PREDICT Results Multivariate AnalysisBaseline
Characteristics, Clinical Variables, ICG Score,
Change in Clinical ICG Variables
Association to HF Event lt 14 Days
Packer M, et al. J Am Coll Cardiol.
2006472245-52.
22
PREDICT Results Risk Stratification with ICG
Score
HF Event lt 14 days
P lt 0.0001 high vs. low p lt 0.001 intermed. vs.
low p lt 0.01 high vs. intermed.
Packer M, et al. J Am Coll Cardiol.
2006472245-52.
23
PREDICT Results Hemodynamic Quadrants
High to Low Risk Quadrant RR 7.0 plt0.0001
Thoracic Fluid Content (/kOhm)
gt 35
lt 35
gt 35
Stroke Index (ml/m2)
lt 35
HF Event lt 14 Days
Packer M, et al. J Am Coll Cardiol.
2006472245-52.
24
Conclusions

• The ICG composite score was the most powerful
  predictor of a short-term HF event in the study
• True even when baseline characteristics, clinical
  variables, and changes in clinical variables were
  considered first
• Patients with a low stroke index and high
  thoracic fluid content identified those at
  seven-fold risk of a HF event within 14 days

Packer M, et al. J Am Coll Cardiol.
2006472245-52.