GWTG HFSA Poster 2006

1
Hospital Variation and Characteristics of ICD
Implantation in Eligible Patients with Heart
Failure Data from the GWTG-HF Registry Bimal
Shah, MD, MBA1, Adrian Hernandez, MD1, M.H.S, Li
Liang, PhD1, Sana M. Al-Khatib, MD1, M.H.S.,
Clyde W. Yancy, MD2, Gregg C. Fonarow, MD3, Eric
D. Peterson, MD, MPH1 on behalf of the Get With
the Guidelines Steering Committee 1 Duke
Clinical Research Institute, Durham, NC, 2 Baylor
Heart and Vascular Institute, Dallas, TX, 3
Division of Cardiology, University of California
Los Angeles, Los Angeles, CA
Hospital Characteristics Stratified By ICD Rates
Background
Results
Comparison of Patient Baseline Characteristics
Stratified By Hospital ICD Rates

• From January 2005 to June 2007, 54,750 patients
  were discharged from 234 GWTG-HF hospitals.
• Six interventional hospitals with no procedures
  recorded and 81 hospitals with patients were excluded.
• The final analysis cohort consisted of 147
  hospitals with 15,492 eligible patients.
• The mean ICD rate (prior, new, or planned) was
  29, but only 13 for new or discharge
  prescription.

• The ACC/AHA Guidelines recommend implantable
  cardioverter-defibrillators (ICD) in patients
  with symptomatic heart failure (HF) and left
  ventricular dysfunction on optimal medical
  therapy.
• However, studies have shown incomplete
  application of ICD use in eligible patients.
• Hospital variation and characteristics
  associated with adherence to recommendations for
  ICD therapy are largely unknown.
• We examined hospital-level variation in ICD use
  and hospital characteristics associated with low
  rates of ICD use in eligible patients.

Data presented as percentages. expressed as
median (25th, 75th percentile)
Performance Measures Stratified By Hospital ICD
Rates
Characteristics of ICD and non-ICD Capable
Hospitals
Data presented as percentages. ACE
angiotensin-converting enzyme inhibitor, ARB
angiotensin II receptor blocker, LV left
ventricular activity level, diet, discharge
medications, follow-up appointment, weight
monitoring, and instructions for worsening
symptoms.
Limitations
Methods
Data presented as percentages. PCI percutaneous
coronary intervention, CABG coronary artery
bypass grafting expressed as median (25th, 75th
percentile) serum creatinine 2.0 mg/dL

• The voluntary GWTG-HF registry may include
  hospitals with a higher likelihood of following
  evidence-based recommendations, thus limiting
  extrapolation of our results to broader community
  practice.
• Data were recorded via a standardized case
  report, but ICD contraindications, medical
  reasons for not implanting, and patients offered
  but declining ICD therapy may have been
  underreported.

• Using the Get With the Guidelines-Heart Failure
  (GWTG-HF) registry, we identified patients with
  Class I or IIA indications for ICD therapy (LVEF
  35).
• Patients were excluded if they had documented
  contraindications for ICD therapy. Hospitals with
  interventional hospitals with no procedures
  reported were excluded.
• The outcome variable of interest was the
  implantation of an ICD during hospitalization or
  a discharge prescription for ICD therapy.
• Characteristics of ICD and non-ICD capable
  hospitals were compared as well when stratified
  into tertiles based on ICD rate.
• Multivariate hierarchical model was used to
  account for each hospitals adjusted case-mix ICD
  rate.

Observed New or Planned ICD Rates in GWTG-HF
Hospitals Among ICD Eligible Patients
Data presented as percentages. expressed as
median (25th, 75th percentile)
Case-mix Adjusted Predictors of Hospital ICD Rates
Conclusions

• ICD therapy for eligible HF patients varies
  widely potentially suggesting there may be
  differential incentives or barriers for use.
• Although overall rates of ICD therapy are low,
  the extent to which hospitals offer surgical or
  invasive procedures is associated with higher
  rates of ICD therapy.
• Further studies are needed to determine the
  barriers for ICD therapy in eligible patients
  meeting ACC/AHA guideline recommendations.

DISCLOSURE INFORMATION The following
relationships exist related to this presentation
BRS (none), AH (research funding from Medtronic,
GSK, Scios/JJ), SMA (research funding and
speaking fees from Medtronic), CWY (research
funding, consultant fees, and/or honorarium from
GSK, Medtronic, CardioDynamics, Scios/JJ,
AstraZeneca, and NitroMed), GCF (research
funding, consultant fees, and honorarium from GSK
and Medtronic), EDP (Research funding from
BMS/Sanofi-Aventis Merck/Schering-Plough).
Error bars represent standard deviation. p