The Adoption of Drug Eluting Coronary Stents by U.S. Hospitals

Document Sample
The Adoption of Drug Eluting Coronary Stents by U.S. Hospitals Powered By Docstoc
					The Adoption of Drug Eluting
  Coronary Stents by U.S.
   Hospitals, 2003-2004
  Peter W. Groeneveld, MD, MS
  Assistant Professor of Medicine
  Philadelphia Veterans Affairs Medical Center
  University of Pennsylvania School of Medicine

   Coauthors:
     Feifei Yang, MS

     Mary Anne Matta, MS

   Sponsor:
     Institute for Health Technology
      Studies (InHealth)
Background: A partially segregated
   national health care system
   22% of the nation’s doctors provide
    care for 80% of black Medicare
   25% of U.S. hospitals provide
    hospital care for 75% of the nation’s
    African Americans
   All providers, hospitals, and health
    systems are not created equal
     Differences in Innovation

   Hospitals with large numbers of black
    patients were slower to adopt
    cardiovascular innovations than
    hospitals with predominantly white
    patient populations
   This difference in hospitals’
    technology adoption practices has
    contributed to racial disparity in
    cardiac care nationwide
         Clinical background

   In April of 2003, FDA approved the
    drug-eluting coronary stents (DES)
    for treatment of coronary stenoses
   Drug eluting stents required minimal
    additional training on the part of
    health care providers, and limited
    capital investment by health care
    systems—in incremental innovation
         Research Questions

   For an incremental cardiovascular
    innovation, do minority-serving
    health care systems still lag “white”
   How is this effect modulated by
    academic hospitals, which tend to
    both be early adopters and are more
    likely to have large minority

   5% MEDPAR data from April, 2003-
    December, 2004
   Patients age ≥ 65 who underwent
    either DES or BMS
   No percutaneous coronary
    intervention in the prior 12 months
   n = 27,300 procedures at 1263
          Outcome variable

   Receipt of DES versus BMS indicated
    by the ICD-9 code and corresponding
   Little incentive to overcode, since
    DES triggered a registry requirement
   Little incentive to undercode, since
    BMS reimbursement was $2200 less
        Independent variables

   Total number of BMS+DES (offset term)
   Percentage of black patients hospitalized
    among all Medicare hospitalizations at
    each institution
   “Black” hospitals designated as hospitals
    in top quartile (≥ 16% black patients)
   Academic hospitals (Council of Teaching
          Model specification

   Negative binomial, log link, GEE
   DEScount = β0 +
                 β1   total_stent_count +
                 β2   acadhosp +
                 β3   blackhosp +
                 β4   blackhosp*acadhosp +
                 β5   avg_age +
                 β6   DM_prevalence + ε
 Hospital types and stent volume

“White,” non-        “White,” academic
academic hospitals   hospitals
n = 849 hospitals    n = 99 hospitals
Stent volume =       Stent volume =
332,000 (60%)        82,000 (15%)
“Minority-serving”   “Minority-serving”
non-academic         academic
hospitals            hospitals
n = 219 hospitals    n = 96 hospitals
Stent volume =       Stent volume =
80,000 (15%)         52,000 (10%)
         Results: Rate Ratios

        Calendar      Acad        Black        Black
Model   Quarter       Hosp        Hosp        x Acad
        1.50 (1.47-
 1      1.53)**

        1.50 (1.47- 1.15 (1.00-
 2      1.53)**     1.33)*

        1.50 (1.47- 1.17 (1.01- 0.85 (0.62-
 3      1.53)**     1.36)*      1.15)

        1.50 (1.47- 1.26 (1.01- 0.98 (0.70- 0.80 (0.49-
 4      1.53)**     1.55)*      1.37)       1.31)

   White academic hospitals implanted
    15% of all stents nationwide in
   White academic hospitals used drug
    eluting stents 26% more frequently
    than black academic hospitals or
    non-academic hospitals during the
    21-month technology diffusion period

   Access to new technology “early”
    may not necessarily be a good thing
   If early DES stent adoption was
    correlated with increased off-label
    use or other activities likely to
    diminish the effectiveness of the
    device, early access to technology
    may not have resulted in improved
    health care outcomes

   Despite the disproportionate number
    of academic centers that have large
    black populations, these centers
    were not leaders in cardiovascular
    device innovation
   Differential diffusion likely widened
    racial disparities in cardiac care
    during 2003-04

Shared By: