SIGNIFICANT VARIATION IN MEDICAL THERAPY TREATMENT RATES FOR BPH RELATED TO LOCAL MARKET FACTORS

1
SIGNIFICANT VARIATION IN MEDICAL THERAPY
TREATMENT RATES FOR BPH RELATED TO LOCAL MARKET
FACTORS
J.C. Sung1,2, L.H. Curtis2, R.L. Sur1, C.G.
Marguet1, J.O. LEsperance1,W.P. Springhart1,
G.M. Preminger1, K.A. Schulman2 and D.M.
Albala1 1Division of Urology, Department of
Surgery and 2Center for Clinical and Genetic
Economics, Duke Clinical Research Institute,
Duke University and Veterans Medical Centers,
Durham, NC, USA
Abstract
Methods
Results
Results
Methods
Explanatory Factors for Finasteride Rates

Explanatory Factors for Doxazosin Rates

• We performed a county level cross-sectional study
  of New York, New Jersey, North Carolina, Florida
  and Pennsylvania in 2000, which represent 26 of
  the nations seniors
• Drug dispensings were obtained from the National
  Drug Code Directory. Invasive and minimally
  invasive surgeries were identified from the
  Healthcare Cost and Utilization Project State
  Inpatient and Ambulatory Surgery Databases.
  Health resources and socio-demographic
  characteristics were determined from the Area
  Resource File.
• Our sample size included 50,182 annual
  finasteride dispensings, 81,270 annual tamsulosin
  dispensings, 70,124 annual adjusted doxazosin
  dispensings and 38,769 BPH-related surgeries for
  an eligible population of 6,556,866 (defined as
  men over age 54) in 317 counties.
• We used Weighted Coefficient of Variation (WCV)
  and Systematic Component of Variation (SCV) as
  our measures of variation.
• WCV is the ratio of the standard deviation of the
  county treatment rate to the mean treatment rate,
  weighted by the study population in each county.
  SCV is the variance across counties that cannot
  be explained by the variation within the county.
• 42 percent of total doxazosin dispensings was
  attributed to the treatment of benign prostatic
  hyperplasia in 2000 (JAMA, 2004)

We used multivariable linear regression to create
models describing the relationships between each
drug and local market factors. Best models were
chosen by adjusted R-squared selection and
removal of statistically insignificant factors.
County-level weights were used to correct for
differences in population sizes among counties
Results
Introduction
Conclusions
Medical therapy is important a-blockers are the
first-line of treatment 5-a-reductase
inhibitors can reduce the risk of surgery in
certain patients Combination therapy is very
effective However, no studies have
analyzed Geographic variations in BPH medical
therapy rates Sources of variation Our
objectives were to Measure geographic variation
in finasteride, tamsulosin and doxazosin
rates Examine relationships between each drug
and local market factors
Explanatory Factors for Tamsulosin Rates
Significant variation exists in medical therapy
rates which can be explained by market factors
including treatment patterns, health resources
and socio-demographic characteristics Finasteride
was negatively associated with surgical rates,
while tamsulosin was positively related to
surgical rates. These relationships may be a
result of their clinical effectiveness in
reducing the risk of surgery for
BPH. Finasteride, tamsulosin and doxazosin
dispensing rates were positively related to one
another as well as to higher education levels and
urologists per capita.
Significant Variation The calculated variation
for finasteride, tamsulosin and doxazosin using
both SCV and WCV measures was significant for
each state, exceeding the 95 percent confidence
level of variation that would be expected by
chance.