Content uploaded by Joseph Tkacz
Author content
All content in this area was uploaded by Joseph Tkacz
Content may be subject to copyright.
Cost
Cost-
-effectiveness of Buprenorphine Medication Assisted Therapy among Opioid
effectiveness of Buprenorphine Medication Assisted Therapy among Opioid-
-Dependent Patients:
Dependent Patients:
Results from a Retrospective Analysis of Health Plan Membership Claims Data
Results from a Retrospective Analysis of Health Plan Membership Claims Data
Charles Ruetsch, Ph.D., Joseph Tkacz, M.S.
Health Analytics, Columbia, MD
Conclusions
Conclusions
Results
Results
Background
Background
N = 7,065 United Health Group HMO members with
integrated pharmacy and medical claims data were
accessed for this study
Inclusion criteria included
-16+ years of age
-continuously eligible in 2007
-one claim with a diagnosis of opioid dependence
using relevant ICD-9-CM codes (International Clas-
sification of Diseases, 9th Revision, Clinical Modifi-
cation)
Patients were categorized into 2 groups
-Buprenorphine group (n=2,026): must have filled at
least one prescription for buprenorphine
(Suboxone® or Subutex®) during the study period
-Non-buprenorphine group (n=5,039): the remain-
ing, non-buprenorphine cohort, which included in-
dividuals diagnosed with opioid dependence, but
were either untreated or treated with another mo-
dality
Methods
Methods
There has been a steady increase in both the number of
people who abuse and become dependent upon opioids
- though not causally related, the amount of prescrip-
tion opioids sold in the U.S., per capita, has increased
many fold over the past 10 years (US DEA)
Opioid dependence imposes a significant economic
burden on society—substantial costs arise from:
-increased health services utilization
-neglect of personal responsibilities
-greater criminal activity, justice and corrections
-other welfare expenditures (Doran, 2008)
Treatment options include psychotherapy, drug coun-
seling, abstinence treatment, and replacement therapy
The most commonly used medications in opioid re-
placement therapy are methadone and buprenorphine
-methadone is a long-acting, full opioid agonist, that
has abuse and diversion potential (Drake et al., 2002)
-buprenorphine is a newer form of replacement therapy
designed to minimize abuse and diversion while retai-
ning all positive aspects of methadone treatment
Buprenorphine treatment is likely cost-beneficial to pay-
ers (Kaur, 2008), but whether it is more cost-effective
than alternative treatments, remains in question
The objective of this study is to compare the health ser-
vices utilization and costs of opioid dependent patients'
treated with and without buprenorphine
It is hypothesized that treating opioid dependence with
buprenorphine will be lead to higher pharmacy and phy-
sician utilization, yet lower use of expensive health ser-
vices, resulting in a net economic benefit to the payer
Sample
Measures
Four health services utilization measures:
1. number of prescription fills
2. number of inpatient hospital days
3. number of emergency department visits
4. number of physician and outpatient visits
Five health services costs variables:
1. prescription costs
2. inpatient hospital costs
3. emergency department costs
4. physician and outpatient costs
5. total healthcare costs
Though buprenorphine patients incur higher pharmacy
and office visit costs, they use fewer expensive health
care services, resulting in an overall positive cost-
benefit conclusion for buprenorphine-replacement ther-
apy
These results support the primary hypothesis, and pro-
vide evidence that treatment of opioid dependence with
buprenorphine is cost effective for health plans
Assuming that the present results remain in direction
and magnitude across other health plans, lines of busi-
ness, and from year to year, health plans can experi-
ence considerable savings using buprenorphine as a
first line of treatment for opioid dependence supported
by outpatient services such as office based providers
and counselors
There are a few limitations to this study:
-Buprenorphine treatment is potentially endogenous-
identification of other factors that drive buprenorphine
and service utilization is essential to understanding
the relationship between buprenorphine and services
utilization
-Unmeasured characteristics correlated with both bu-
prenorphine treatment and the dependent variables
may lead to biased estimates
-Cohorts displayed different demographic and comor-
bidity profiles; therefore selection bias may also be
present
Next step will be to identify “other uncategorized medi-
cal costs” as their contribution is greater than are those
included in “typical services” indicator
The proposed groups would be: those treated with phar-
macotherapy, patients treated without pharmacotherapy,
and those untreated
Finally, alternative multivariate modeling approaches
will be considered: negative binomial regression (with
or without zero inflation) for count measures, and two-
part, log-transformed, or Gamma/Log link models for
costs measures
This study was funded by Reckitt Benckiser
2.3 1.73
6.36
4.24
11.15 10.7
0
2
4
6
8
10
12
Emergency
Department
Visits***
Inpatient
Hospital
Days***
Physician &
Outpatient
Visits
No Buprenorphine Buprenorphine
$48,402
$30,097
$0
$5,000
$10,000
$15,000
$20,000
$25,000
$30,000
$35,000
$40,000
$45,000
$50,000
Total Healthcare Costs***
No Buprenorphine Buprenorphine
$1,497 $1,469
$4,047 $3,241
$5,850 $6,105
$8,790
$5,971
$0
$1,000
$2,000
$3,000
$4,000
$5,000
$6,000
$7,000
$8,000
$9,000
Physician &
Outpatient
Costs
Emergency
Department
Costs***
Pharmacy
Costs*** Inpatient
Hospital
Costs***
No Buprenorphine Buprenorphine
Descriptive and bivariate analyses, and ordinary
least squares multivariate models of the nine de-
pendent variables were estimated
To remove the influence of cost-sharing, allowed
amounts were used for medical cost measures, and
pharmacy costs were gross
No outliers were trimmed from the dataset, and claim
reversals were retained
Finally, in addition to the main between groups vari-
able, buprenorphine treatment group, all regression
models included the same vector of covariates:
-age
-gender
-geographic region indicators
-medicaid enrollment flag
-health status as measured by the weighted Charlson
Comorbidity Index (Charlson, et al., 1987)
Data Analysis Figure 1. Healthcare Utilization Means
Figure 2. Healthcare Cost Means
Figure 3. Total Cost of Care
Multivariate Regression
Bivariate Analyses
Demographics
Buprenorphine vs. non-buprenorphine on demo-
graphics:
63% vs. 50% male
37 years vs. 40 years of age
5% vs. 10% Medicaid Beneficiaries (p’s < .01)
Patients receiving buprenorphine were in better
physical health as measured by the Charlson Co-
morbidity Index (0.48 vs. 0.74, p < .001)
An average patient treated with buprenorphine filled
7.63 buprenorphine prescriptions in 2007, covering
162 days of therapy at a gross cost of $2,065 (i.e.,
$12.75 per day of supply)
The buprenorphine group had significantly lower
emergency department visits and inpatient hospital
days (p’s < .01; Figure 1)
The buprenorphine group had higher pharmacy
costs, but lower emergency department costs and
lower inpatient hospital costs (p’s < .05, Figure 2)
Overall, buprenorphine patients cost significantly
less money per patient during 2007 for the health
plan (p < .01, Figure 3)
With just 5.21 more prescriptions filled ($1,208
costs), the use of other pharmaceuticals was lower
(5.21 vs. 7.63) among buprenorphine patients, per-
haps due to decreased use of prescription opioids
Individuals in the buprenorphine group spent 1.46
less days as a hospital inpatient (-$1,986 costs,
p<.01), presented in the emergency department 0.22
fewer times (-$301, not statistically significant), and
visited their physicians 0.88 more times in 2007
($160 costs, p<.01)
Overall, after controlling for demographics, Medi-
caid, and health status, patients receiving buprenor-
phine for opioid dependence had total healthcare
costs in 2007 that were $11,200 less (p<.01) than
non-buprenorphine
Buprenorphine—Suboxone
Intoxication from opioids results when the opioid
(heroin, prescription analgesics, etc.) binds to the
mu receptors in the brain, producing euphoria
When Suboxone is taken, it binds to the same mu re-
ceptors, but produces a mild euphoria, while simulta-
neously preventing other opioids from binding
Mu receptor
Opioid
Euphoria
Suboxone
< Euphoria