A132
activity at least three times a week. Obesity (OB) was defined by a binary indicator if
body mass index was greater than 30. The structural parameter, average treatment
effect (ATE) was defined as the impact of PA on obesity, if individuals are randomly
assigned to PA. The second parameter average treatment in the treated (ATET) mea-
sured the impact of PA on obesity amongst those who became physically active, rather
than for the population as a whole. To control for the unobservable factors affecting
PA, we specified a recursive bivariate probit model. To avoid identification based on
functional form, instruments added were presence of any limitations, and injury.
Covariates included age, gender, race, ethnicity, geographical and metropolitan area
location, smoking status, comorbidities and perceived physical and mental health.
RESULTS: Based on naïve probit model, the probability of obesity, evaluated at the
means of the data, was 0.099 lower amongst those who were physically active (p  disorders are frequently treated with immune globulin, administered intravenously or
0.001). Effect of selection bias was positive and significant (rho  0.55 p  0.001).
Based on the recursive probit model, ATE was a 27.7% decrease (95% CI
0.279 to administered immune globulin products in patients in the home setting. METHODS:
0.275) while ATET was a 38.8% decrease (95% CI
0.391 to
0.385) in probability A retrospective, longitudinal cohort analysis of patients (18 years and older) provided
of obesity amongst those who were physically active. CONCLUSIONS: Unobservable
heterogeneity may be masking the true effect of physical activity on obesity. Account-
ing for this bias confirms a significant protective effect of physical activity against the
likelihood of obesity.
PSY12
TREATMENT OF TRANSFUSIONAL IRON OVERLOAD IN PATIENTS
WITH MYELODYSPLASTIC SYNDROME OR SEVERE ANEMIA: DATA
FROM MULTI-CENTER CLINICAL PRACTICES
p A1, Duh MS2, Wang ST2, Dial E2, Fanourgiakis I1, Fortner B3, Paley C4, Mody-Patel N4,
Raptis
Corral M4, Scott J5
1
University of Pittsburgh School of Medicine, Pittsburgh, PA, USA, 2Analysis Group, Inc.,
Boston, MA, USA, 3P4 Healthcare, Lakeland, TN, USA, 4Novartis Pharmaceuticals
Corporation, East Hanover, NJ, USA, 5P4 Healthcare, Ellicott City, MD, USA
OBJECTIVES: Patients with myelodysplastic syndrome (MDS) or severe anemia
requiring repeated transfusions of red blood cells (RBCs) risk developing transfusional
iron overload (TIO), which can cause organ damage and reduce survival. Iron chela-
tion therapy (ICT) has been shown to improve survival and quality of life in patients
with TIO; however, ICT utilization in clinical practices is not well understood.
METHODS: The medical records of patients diagnosed with MDS or severe anemia
q6 months before data extraction, aged q 21 years at their diagnosis, received q1 RBC
transfusion were reviewed. ICT-eligibility was defined as q20 units of RBCs transfused
or q2 serum ferritin (SF) tests q1000 mcg/L. Study endpoint was ICT-treatment rate
among ICT-eligible patients with lower-risk MDS [IPSS (low or intermediate-1); WHO
(RA, RARS, RCMD, RCMD-RS or 5q); FAB (RA or RARS)]. Characteristics and
survival of treated and untreated groups were described. RESULTS: Medical records
data for 283 patients were extracted. Among 78 ICT-eligible patients with lower-risk
MDS, only 32 (41%) received ICT. At ICT-initiation, treated patients received on
average 13.3 transfusions (27.6 units) and mean first SF near-ICT-initiation was twice
the recommended level at 1949 mcg/L. Median overall survival for all ICT-eligible
patients was significantly longer for those ICT-treated than untreated (8.7 versus 4.7
years, log-rank p  0.02; multivariate hazards ratio  0.372, p  0.03). CONCLU-
SIONS: This observational study finds only 41% of ICT-eligible patients with lower-
risk MDS received ICT in clinical practice, and their treatment was initiated later than
recommended. Among all ICT-eligible patients, those who received ICT had a signifi-
cantly better overall survival than untreated patients.
PSY13
FREQUENCY AND BOTHERSOMENESS OF SIDE EFFECTS IN PAIN
PATIENTS TAKING OXYCODONE IMMEDIATE RELEASE: IMPACT ON
PRESCRIPTION AND OVER-THE-COUNTER MEDICATION USE
Anastassopoulos
p K1, Chow W2, Ackerman SJ3, Tapia CI3, Benson C2, Kim M2
1
Covance Market Access Services, Inc, Gaithersburg, MD, USA, 2Ortho-McNeil Janssen
Scientific Affairs, LLC, Raritan, NJ, USA, 3Covance Market Access Services, Inc, San Diego,
CA, USA
OBJECTIVES: Oxycodone immediate release, alone or in combination (oxycodone),
is widely used to treat pain. However, oxycodone is often associated with bothersome
side effects, which may lead to increased medical resource use, including prescription
and over-the-counter (OTC) medications. The objective of this first analysis was to
assess the frequency and bothersomeness of side effects and other medication use
among patients taking oxycodone. METHODS: An online survey was completed by
a nationwide convenience sample of patients currently taking oxycodone for non-
malignant pain. Detailed data on patient experience with oxycodone were collected.
A minimum sample size of 600 was determined to ensure reasonably accurate esti-
mates around the frequency of side effects and medical resource utilization. RESULTS:
Among the 601 respondents [mean age of 45 years (range 18–86), 85.0% Caucasian,
69.1% female], almost half, 45.6%, were taking oxycodone for back/neck pain,
16.8% for osteoarthritis/rheumatoid arthritis, 14.3% for pain due to injury/trauma,
10.5% for recent surgery, 7.2% for fibromyalgia, and 5.7% for neuropathic pain. The
mean daily dose was 16.5 mg (range 2.5–200). Overall, 83.5% were bothered by side
effects with 29.9% being moderately/extremely bothered. Over half, 53.1%, were
bothered by constipation, almost one-third, 31.3%, by nausea, 27.6% by pruritus,
and 14.8% by vomiting, among others. A significantly higher proportion of respon-
dents bothered by side effects reported use of prescription (13.1% vs. 0%; p  0.001)
or OTC (20.5% vs. 9.1%; p  0.007) medications to manage those side effects, com-
pared to respondents not bothered by side effects. CONCLUSIONS: The majority of
survey respondents experienced side effects of oxycodone, some of which led to the
Abstracts
use of prescription and OTC medications. These results point to the potential benefits,
both humanistic and economic, of better-tolerated pain medications. Further analyses
of this data will assess the health status, pain intensity, and other resource utilization
among oxycodone users.
PSY14
TOLERABILITY OF INTRAVENOUSLY ADMINISTERED IMMUNE
GLOBULIN IN THE HOME SETTING
Tankersleyy MA
Accredo Health, Inc, Memphis, TN, USA
OBJECTIVES: Patients with primary immune deficiencies and many neurological
subcutaneously. The aim of this study was to evaluate tolerability of intravenously
intravenous immunoglobulin (IVIG) using data from the Accredo Therapeutics elec-
tronic medical record was conducted. Inclusion criterion was infusion of at least one
dose of IVIG during the study period. Patients were followed from July 1, 2007 to
June 28, 2008. Three components to estimate tolerability were evaluated. The first
was aggregate adverse drug event rate. The second was infusion completion rate (ICR),
calculated by dividing the number of successful infusions by the number of attempted
infusions. The third measure was a managed therapy completion rate (MTCR-A) for
patients that had an identified ADE and were still managed to completion of therapy.
RESULTS: The sample size for review was 33,065 doses of IVIG dispensed during the
study period. The results based upon the three measures of tolerability were: 1) an
ADE rate of 2.35% (95% CI: 1.79%–2.93%) based on 778 reported adverse drug
events; 2) an infusion completion rate (ICR) of 99.74% (95% CI: 99.65%–99.82%);
and 3) a managed through completion of therapy (MTCR-A) of 88.6% (95% CI:
80.8%–92.35%). Gender, age, diagnoses and BMI were also evaluated for their effect
upon tolerability of infusion. CONCLUSIONS: Intravenous administration of immune
globulin is an important alternative infusion option for patients. Patients can be well
managed in the home on intravenously administered immunoglobulin. These findings
contribute to previous research related to safety of administration of immunoglobulin
in the home.
SYSTEMIC DISORDERS/CONDITIONS – Cost Studies
PSY15
THE IMPACT OF ADHERENCE ON THE COSTS AND BENEFITS OF
INTENSIVE LIFESTYLE MANAGEMENT (ILM) IN OVERWEIGHT AND
OBESE PATIENTS AT HIGH RISK FOR TYPE-2 DIABETES MELLITUS
(T2DM)
Willis M1, Asseburg C1, He J2, Neslusan C2
1
IHE – The Swedish Institute for Health Economics, Lund, Sweden, 2Johnson & Johnson
Pharmaceutical Services LLC, Raritan, NJ, USA
OBJECTIVES: The Diabetes Prevention Program (DPP) demonstrated that ILM
improves lipid values, reduces blood pressure and reduces the risk of developing
T2DM. Controversy over magnitude of whether these benefits can be achieved cost-
effectively has diverted attention from questions about the generalizability of the
results. The high rate of treatment adherence in the DPP may not be reproducible in
actual practice. The objective of this study was to assess the impact of treatment
adherence to ILM on estimates of health benefits and costs for a cohort of overweight
and obese patients at high risk of developing T2DM. METHODS: The IHE/JNJ
weight management model, a Markov-based, micro-simulation model that includes
relevant co-morbidities and risk factors, was used to simulate the costs and health
outcomes of ILM over 25 years for 500 cohorts of 1,000 hypothetical overweight and
obese pre-T2DM patients. Efficacy and baseline population characteristics were taken
primarily from the DPP. Costs for ILM and care associated micro- and macro-vascular
complications and other co-morbidities as well as quality-of life data was obtained
from existing literature. Four scenarios were assessed: adherence as observed in the
DPP and reductions in the DPP adherence rate by 25%, 50%, and 75%. RESULTS:
In all, ILM resulted in 19.96 undiscounted life years (LYs), 18.03 undiscounted
quality-adjusted life years (QALYs), at a cost of $78,965, assuming DPP-like adher-
ence. Forty percent of the cohort ultimately developed T2DM. Reducing adherence
by 25%, 50%, and 75% reduced LYs by 0.17, 0.29, and 0.40, QALYs by 0.36, 0.68,
and 1.03, and increased costs by $2154, $4190, $6813, respectively. The rate of
T2DM transition increased by 5, 10, and 15 percentage points, respectively. CON-
CLUSIONS: Patient adherence is an important driver of the benefits and costs of ILM
and should be considered explicitly in cost-effectiveness estimates.
PSY16
A CANADIAN BASED PHARMACOECONOMIC ANALYSIS OF SELECTED
ANTICONVULSANTS, SNRIS AND TCAS IN TREATING NEUROPATHIC
PAIN
Iskedjian
j M, Walker J
PharmIdeas Research and Consulting Inc, Oakville, ON, Canada
OBJECTIVES: Neuropathic pain starts as or is caused by a primary lesion or dysfunc-
tion in the nervous system. It impacts use of health care resources and may incur
employment disruptions. The primary goal in managing neuropathic pain is to make
it more tolerable. Three classes of atypical medications, anticonvulsants (ACs), sero-
Abstracts
tonin-norepinephrine retake inhibitors (SNRIs) and Tricyclic Antidepressants (TCAs)
are used for the short-term treatment of neuropathic pain. The cost-effectiveness of
these three medications in Canada was examined. METHODS: A decision-tree ana-
lytic model compared the three drugs, over 18 weeks. Response rates were determined
from two meta-analyses, “From-Placebo” and “Through-Placebo”. Outcomes of
interest were rates of achieving 50% reduction in baseline pain (full-response) or 30%
reduction (partial-response). A secondary outcome was the number of pain-controlled
days (PCDs) applying full-response. The Ministry of Health (MOH) perspective
applied direct medical costs: medications, physician visits, diagnostic tests, and hos-
pitalizations; societal perspective (SOC) also applied indirect costs (productivity loss);
all in CN$2007. RESULTS: The “From-Placebo” approach yielded MOH (SOC) costs
of $380 ($1808); $448 ($2030); and $560 ($2063) for TCAs, SNRIs and ACs, respec-
tively, while primary outcomes were 79.3%; 76.4%; and 77.8%, and secondary
effective outcomes were 48 PCDs, 43 PCDs; and 46 PCDs for the same sequence. The
“Through-Placebo” approach yielded MOH (SOC) costs of $331 ($1513); $459
($1868); and $519 ($2125) for TCAs, SNRIs, and ACs, respectively, with primary
(secondary) effectiveness outcomes of 88.0% (60 PCDs), 80.6% (41 PCDs); and
84.3% (54 PCDs), for the same sequence. CONCLUSIONS: For both approaches,
TCAs provided the least costly alternative, and better outcomes, for both for primary
and secondary effectiveness measures, effectively dominating SNRIs and ACs.
PSY17
DECISION-ANALYTIC MODELLING OF TREATMENT OPTIONS FOR
PATIENTS WITH SEVERE HAEMOPHILIA A AND INHIBITORS IN
GERMANY
Ehebergg D1, Berger K2, Schramm W3
1
IMS Health GmbH & Co. OHG, Munich, Germany, 2University Hospital of Munich, Munich,
Germany, 3University Hospital of Munich, München, Germany
OBJECTIVES: Development of a cost effectiveness analysis to evaluate treatment
options for haemophilia patients with inhibitors in Germany. METHODS: Costs and
outcomes of four different treatment regimens were compared by using an Excel-based
Markov model with a three month cycle length. Comparators are low- and high dose
protocols vs. strict on-demand treatment and to a mixture of both protocols after
conducting a risk assessment. Time horizon: 18 years. Success rates of the different
ITI-protocols, rates of anamnestic response, bleeding event rates and success rates for
the treatment of bleeds were obtained from the literature. Costs included the costs for
factor concentrates and hospitalization. Costs for one average patient were reported
from a third party payers’ perspective. Annually discounting rates were 3% for costs
and 5% for benefits. RESULTS: One average patient treated on-demand costs
€1,558,962.54 and has 181.92 minor and 2.43 severe bleeds. The cost for the same
patient treated with the high dose ITI-protocol is approximately double (€3,340,976.81),
but the patient has 2/3 fewer bleeds. The incremental cost per additional avoided
minor bleed compared to on-demand treatment is €14,816.77 for the high dose pro-
tocol, €9,716.50 for ITI after risk assessment and €8,938.84 for the low dose protocol.
The incremental costs per avoided bleed is €12,810.19 more for ITI after risk assess-
ment than for the low dose ITI protocol. When comparing the high dose ITI protocol
vs. ITI after risk assessment, each avoided bleed costs an extra €313,417.11. CON-
CLUSIONS: The results of the model suggest that use of risk assessment is the most
cost effective approach to ITI. Although data on rare diseases are scarce and hard to
compare, decision-analytic modelling is a valuable tool to weight the costs and con-
sequences of different treatment regimens for patients with severe haemophilia com-
plicated by inhibitors.
PSY18
ECONOMIC EVALUATION OF AN INTERNET BASED WEIGHT
MANAGEMENT PROGRAM
Rasu R1, Hunter C2, Maruska H3, Peterson A4, Foreyt J5
1
University of Missouri- Kansas City, Kansas City, MO, USA, 2National Institute of Diabetes &
Digestive & Kidney Diseases, NIH, Bethesda, MD, USA, 3University of Missouri Kansas City
School of Pharmacy, Kansas City, MO, USA, 4University of Texas Health Science Center at
San Antonio, San Antonio, TX, USA, 5Baylor College of Medicine, Houston, TX, USA
OBJECTIVES: To determine whether a Behavioral Internet Treatment (BIT) program
for weight management is a viable cost-effective option, compared to the usual
care in a diverse sample of overweight (average BMI  29 kg/m2) healthy adults (mean
age 34) of the United States Air Force (USAF). METHODS: A 2-group parallel-
randomized controlled trial (N  442) outcomes were considered for this economic
evaluation. Participants were randomly assigned into two groups 1) Usual Care (UC,
N  215) and 2) Usual Care plus Behavioral Internet Therapy (BIT, N  227). Individu-
als were evaluated twice, at baseline and at 6 months. Only statistically significant
primary outcome measures (changes in body weight, percent body weight changes and
waist circumferences) and a secondary outcome measure (Weight Efficacy Life Style
Questionnaire/WEL) were included in incremental cost effectiveness analysis (ICEA)
model. Costs of the intervention were computed by using the perspective of an agency
wanting to replicate the intervention. Intervention costs consisted of: recruiting and
training of telephone counselors, orientation, baseline and follow up appointments,
paperwork and measurement costs. One-way sensitivity analyses were performed to
measure the robustness of this ICEA model. RESULTS: The intervention produced a
total direct cost of $7,993.32 for BIT intervention with a cost of $35.21/BIT partici-
pants. Total staff time costs were $64.58/BIT participants. The overall cost for the
entire BIT intervention was $22,653.92 or $99.80/BIT participants. Intervention cost
was $52.52 for each additional kg of weight loss. Additionally, intervention costs were
$58.70 for each cm waist circumference loss. On the WEL subscales costs were $76.77
A133
for each additional point gained on the social pressure subscale; where increasing
scores indicate increased confidence in managing social pressures to eat. CONCLU-
SIONS: Although different forms of weight management program exist, the Behavioral
Internet Therapy program is found to be a cost-effective choice for weight manage-
ment. Internet Therapy program is costly initially, but may provide a cost-saving
long-term impact.
PSY19
COST-EFFECTIVENESS COMPARISON ANALYSIS OF ANTI-TNF AGENTS
WITH MULTIPLE INDICATIONS
Szkurhan AR
Dymaxium Inc, Toronto, ON, Canada
OBJECTIVES: A variety of anti-TNF agents have been approved for the treatment of
multiple autoimmune disorders. The objective of the current study is to compare the
cost-effectiveness of these biologics across each drug’s approved therapeutic indica-
tions. METHODS: A systematic literature review was conducted of journal articles
published during the period of 1998 to 2008 to identify all relevant cost-effectiveness
analyses investigating the following three multiple indication drugs: adalimumab,
etanercept, and infliximab. Specifically, for adalimumab cost-effectiveness results were
analyzed for rheumatoid arthritis (RA), ankylosing spondylitis (AS), psoriasis (PsO),
and Crohn’s disease (CD); for etanercept cost-effectiveness results were analyzed for
RA, AS, and PsO; and for infliximab cost-effectiveness results were analyzed for RA,
AS, PsO, psoriatic arthritis (PsA), CD, and ulcerative colitis (UC). For each of the
three therapies, the incremental cost-effectiveness ratios (ICERs) among their various
indications were compared. All costs were converted to 2007 USD using the Consumer
Price Index (CPI) and Purchasing Power Parities (PPP) for comparison. RESULTS: A
total of 25 cost-effective evaluations were included for comparative analysis. While
all studies found adalimumab, etanercept and infliximab to be cost-effective for at
least one indication (using a threshold of $50,000 – $100,000/QALY gained), ICER
values across approved indications varied greatly; for example, for infliximab, ICER
values as high as $290,000 were reported the treatment of RA, $75,000 for the treat-
ment of AS, $46,000 for the treatment of PsA, and $60,000 for the treatment of CD.
Reported ICER values for each indication also varied among studies as a result of
differences between comparators, analytic timeframes, and study perspectives. CON-
CLUSIONS: Based on published ICER values, the cost-effectiveness of multiple indica-
tion drugs varies across approved indications. From an economic standpoint, although
various biologic agents have approval for several indications, these drugs do not neces-
sarily offer a cost-effective solution across all approved indications.
PSY20
COST-EFFECTIVENESS OF PROPHYLAXIS AGAINST ON-DEMAND
TREATMENT IN BOYS WITH SEVERE HAEMOPHILIA IN IRAN
Ahmad Kia Daliri A1, Mamikhani J2
1
Ministry of Welfare and Social Security, Tehran, Tehran, Iran, 2Behinpuyan Hutan Research
Institute, Tehran, Tehran, Iran
OBJECTIVES: The aim of this study was to assess the incremental cost-effectiveness
of on-demand versus prophylactic haemophilia therapy in Iran from the third-party
payers’ perspective. METHODS: A retrospective chart review of 25 types A haemo-
philiacs that treated in 3 hemophilia treatment center was conducted for patients aged
0–9 years receiving one of two treatments: 1) prophylaxis with concentrate at clinic;
2) concentrate at clinic as on-demand. Fourteen boys receiving on-demand infusions
for bleeding events and 11 boys receiving infusions prophylaxis. Data were collected
from documents obtained from the hemophilia treatment centers during a period of
approximately 6 months. RESULTS: The patients receiving prophylactic treatment
had fewer bleeding events each month (mean, 0.26 vs. 2.74) but used more concentrate
(225.31 vs. 87.20 units / kg per month). Average cost per patient each month at pro-
phylaxis group was about 1.9 times higher than on-demand group. Compared with
on-demand infusion, prophylaxis costs 3201656 Rials (US $ 356) per bleeding event
prevented. CONCLUSIONS: Prophylactic care markedly reduces the number of bleed-
ing episodes, but at considerable cost.
PSY21
COST-EFFECTIVENESS OF PREGABALIN IN PATIENTS WITH
FIBROMYALGIA: A MEXICAN PERSPECTIVE
Arreola-Ornelas H1, Rosado-Buzzo AA2, García-Mollinedo MDL2, Dorantes-Aguilar J1,
Mould–Quevedo
Q J3, Davila-Loaiza G3
1
Fundación Mexicana para la Salud, Mexico City, Mexico, 2Links & Links S.A. de C. V, Mexico
City, Mexico, 3Pfizer Mexico, Mexico City, Mexico
OBJECTIVES: Patients with fibromyalgia (FM), a major cause of morbidity, use health
services extensively and have an estimated prevalence of 3.4% for women and 0.5%
for men. The aim of this study was to assess the cost-effectiveness of pregabalin in
the treatment of FM from the health care payer’s perspective. METHODS: We devel-
oped a three-state Markov model to simulate health and economic outcomes during
a time horizon of one-year (12-week cycles). The model includes several stages related
to functional disability (mild-pain, moderate-pain and severe-pain). Effectiveness was
assumed as the percentage of patients with no or mild pain (pain score 4) at the end
of the follow-up period using a Visual Analogue Scale(VAS). Transition probabilities
were obtained from a systematic review involving national and international published
trials. Comparators used in the assessment were amitriptyline (25–50 mg/day); prega-
balin (150–450 mg/day); fluoxetine (20–80 mg/day); duloxetine (60–120 mg/day);
gabapentin(300–1200 mg/day); tramadol(150–1300 mg/day)  acetaminophen(325–
2600 mg/day) and fluoxetine(20–80 mg/day)  amitriptyline(10–50 mg/day). Resource