Transcript Oral Presentation 1

Rapid Penetration of COX2 Inhibitors in
Non-Steroidal Antiinflammatory Drug Market:
an Implication to Hospital Cost Containment Policy
Supon Limwattananon, MPHM, PhD *
Chulaporn Limwattananon, MPharm, MSc, PhD *
Supasit Pannarunothai, MD, PhD **
* Faculty of Pharmaceutical Sciences, Khon Kaen University
** Center for Health Equity Monitoring, Naresuan University
- Thailand
Cyclo-Oxygenase-2 (COX2) Inhibitors
In Thailand,
Celecoxib and Rofecoxib have been available since 1999,
each by a sole pharmaceutical company
“single-source product”
Report from MOPH-provincial hospitals (N =41*, Year 2002),
Spending for COX2 inhibitors
Total acquisition costs
42.9 million Baht
Share of top 50 high cost drugs
6.2%
Ranking
• Celecoxib
1st (in secondary care hospitals*)
3rd (in tertiary care hospitals*)
Objectives
1. To examine variations in hospital NSAID expenditures
as related to the use of COX2 inhibitors
2. To assess patterns of drug channeling for
COX2 inhibitors
Study Population
Settings: 18 provincial hospitals in 4 regions of Thailand
(secondary and tertiary acute care settings)
Sample: 1,558,633 prescriptions for oral NSAID solid forms
rendered to. ambulatory patients in 4 health insurance schemes
• Civil
Servant Medical Benefit Scheme-CSMBS
• Social Security Scheme-SSS
• Low-Income Card & Universal Health Care
Coverage-LIC/UC schemes
• Rest of the population-ROP
Time periods: Fiscal years 2000-2002
Study Design & Analysis
Retrospective, secondary analysis of electronic databases
of hospital drug use
Statistical analysis *
• For drug expenditures: a generalized linear model (GLM)
• For propensity of drug use: logistic regression analysis
• Control for the underlying differences in drug use patterns due to
• patient demographics (age groups and sex)
• years of drug use (and interaction with health insurance schemes)
• hospital settings
(proxy for variations in prescribing practice styles)
Utilization and Expenditures
All Types of NSAIDs
Prescriptions
Daily doses
Thai Baht
Year 2000
484,452
4,944,285
23,205,944
Year 2001
549,366
5,658,362
34,257,243
Annual growth
from Year 2000
(13.4%)
(14.4%)
Year 2002
538,517
5,260,404
Annual growth
from Year 2000
(11.2%)
(6.4%)
(47.6%)
37,991,221
(63.7%)
Daily Doses by Types of NSAID
Days
1,500,000
1,400,000
1,300,000
1,200,000
1,100,000
1,000,000
900,000
800,000
700,000
600,000
500,000
400,000
300,000
200,000
100,000
0
5.2%
10.0%
0.7%
COX2 inhibitors
Other NSAID-NED
Meloxicam
02
420
02
Q
TR
320
02
Q
TR
220
02
Q
TR
120
01
Q
TR
420
01
Q
TR
320
01
Q
TR
220
01
Q
TR
120
00
Q
TR
420
00
Q
TR
320
00
Q
TR
220
00
Other NSAID-ED
Q
TR
120
Q
TR
8.2%
Expenditures by types of NSAID
Baht
10,000,000
9,000,000
8,000,000
7,000,000
COX2 inhibitors
6,000,000
5,000,000
4,000,000
3,000,000
2,000,000
33.9%
6.5%
46.5%
52.1%
Other NSAID-NED
Meloxicam
Other NSAID-ED
1,000,000
Q
TR
120
00
Q
TR
220
00
Q
TR
320
00
Q
TR
420
00
Q
TR
120
01
Q
TR
220
01
Q
TR
320
01
Q
TR
420
01
Q
TR
120
02
Q
TR
220
02
Q
TR
320
02
Q
TR
420
02
0
Factors Affecting NSAID Expenditures per Capita
(Competing Models)
Model with interaction terms
Coefficienta
P value
Main effect model
Coefficienta
P value
COX2 inhibitors
2.486
< 0.001
2.488
< 0.001
Age 36 – 49 years b
0.368
< 0.001
0.370
< 0.001
0.798
< 0.001
0.805
< 0.001
- 0.158
< 0.001
- 0.158
< 0.001
0.864
< 0.001
0.847
< 0.001
Age 50+ years
b
Male
CSMBS c
LIC/UC c
-
0.001
0.954
- 0.053
< 0.001
ROP c
-
0.022
0.188
- 0.084
< 0.001
Year 2001 d
- 0.035
0.065
0.038
< 0.001
Year 2002 d
0.186
< 0.001
0.025
0.002
CSMBS x Year 2001
0.083
0.002
CSMBS x Year 2002
- 0.093
< 0.001
LIC/UC x Year 2001
0.123
< 0.001
LIC/UC x Year 2002
- 0.205
< 0.001
ROP x Year2001
0.070
0.002
ROP x Year2002
- 0.249
< 0.001
a Based
on generalized linear model (GLM) using log link, gamma distribution , adjusted for hospital indicators
b
Age of 18-35 years as the reference category
c
SSS as the reference category
d Year
2000 as the reference category
Effects on Difference in NSAID Expenditure
% difference a
95% CI
COX2 inhibitors
vs. other NSAID
1,101.2%
1,056.5 to 1,147.6%
Age 36-49 years
44.5%
42.3 to 46.7%
122.0%
118.5 to 125.6%
-14.6%
- 15.7 to -13.5%
vs. 18-35 years
Age 50+ years
vs. 18-35 years
Male
vs. Female
a
% difference due to an indicator variable = exp(Coefficient) - 1
Effects on Difference in NSAID Expenditure
(Trends for Each Scheme)
% difference a
Year 2001 vs.
LIC/UC
SSS
ROP
CSMBS
9.2%
-3.4%
3.5%
4.9%
-1.9%
20.4%
-6.1%
9.7%
Year 2000
Year 2002 vs.
Year 2000
a
% difference due to an indicator variable = exp(Coefficient) - 1
Based on GLM with interaction of schemes and years
Effects on Difference in NSAID Expenditure
(Comparison between Schemes for Each Year)
% difference a
Year 2000
CSMBS vs. SSS
Year 2001
Year 2002
137.2%
157.7%
116.1%
ROP vs. SSS
-2.2%
4.9%
-23.7%
LIC/UC vs. SSS
-0.1%
13.0%
-18.6%
a
% difference due to an indicator variable = exp(Coefficient) - 1
Based on GLM with interaction of schemes and years
Propensity to Receive COX2 Inhibitors
(Competing Models)
Model with interaction terms
Coefficienta
P value
Age 36 – 49 years b
Age 50+ years
0.619
< 0.001
0.617
< 0.001
1.267
< 0.001
1.270
< 0.001
- 0.302
< 0.001
- 0.301
< 0.001
2.279
< 0.001
2.434
< 0.001
b
Male
Main effect model
Coefficienta
P value
CSMBS c
LIC/UC c
-
0.845
< 0.001
- 0.585
< 0.001
ROP c
-
0.407
< 0.001
0.178
< 0.001
1.105
< 0.001
1.200
< 0.001
1.145
< 0.001
1.512
< 0.001
CSMBS x Year 2001
- 0.009
0.936
CSMBS x Year 2002
0.303
0.003
LIC/UC x Year 2001
0.367
0.009
LIC/UC x Year 2002
0.285
0.038
ROP x Year2001
0.461
< 0.001
ROP x Year2002
0.853
< 0.001
Year 2001
d
Year 2002 d
a Based
b
on logistic regression analysis, adjusted for hospital indicators
Age of 18-35 years as the reference category
SSS as the reference category
d Year 2000 as the reference category
c
Odds Ratio of Receiving COX2 Inhibitors
(Comparison between Schemes for Each Year)
Odds Ratio a
Year 2000
CSMBS vs. LIC/UC
Year 2001
Year 2002
22.74
15.62
23.14
CSMBS vs. SSS
9.77
9.68
13.22
ROP vs. LIC/UC
1.55
1.70
2.73
LIC/UC vs. SSS
0.43
0.62
0.57
a
Based on logistic regression model with interaction of schemes and years
Odds of Receiving COX2 Inhibitors
Year 2000
Year 2001
Year 2002
-1
-1.5
Odds* (in log scale)
-2
-2.5
-3
CSMBS
-3.5
-4
-4.5
-5
SSS
-5.5
ROP
-6
LIC/UC
* Odds = exp(constant+bAge+bGender+bScheme+bYear+bSchemexYear+bHosp)
Conclusion
• Growth in NSAID expenditures was largely driven by
rapid penetration of the expensive COX2 inhibitors.
• The prime target for the patent-protected, single-source drugs
was patients covered by fee-for-service scheme like CSMBS.
• To contain hospital drug costs, a generic substitution for
COX2 inhibitors is unfeasible due to market exclusivity nature.
• Therapeutic substitution with the multi-source NSAID is
a viable alternative in curbing the expenditure growth.