Transcript Slide Library: Thomson Medstat Business Overview Slides

T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
The Effects of Statin Copayments and Statin
Adherence on Medical Care Utilization and
Expenditures
Teresa B. Gibson1, Ph.D.; Tami L. Mark1 Ph.D., MBA; Kirsten Axelsen2, MS; Onur
Baser1, Ph.D.; Joan A. Mackell2, Ph.D.; Heidi King2, MS; Kimberly A. McGuigan2,
Ph.D., MBA
1 Thomson Medstat, Ann Arbor, MI
2 Pfizer, Inc., New York, NY
American College of Cardiology, 55th Annual Scientific Session
March 14, 2006
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Presenter Disclosure Information
“The Effects of Statin Copayments and Statin Adherence on Medical
Care Utilization and Expenditures”
Disclosure Information...
The following relationships exist related to this presentation:
•
•
•
•
•
•
•
T.B. Gibson, Research Grants, Pfizer, Inc., Significant
T.L. Mark, Research Grants, Pfizer, Inc. Significant
K. Axelsen, Salary, Pfizer, Inc., Significant
O. Baser, Research Grants, Pfizer, Inc., Significant
J.A. Mackell, Salary, Pfizer, Inc., Significant
H. King, Salary, Pfizer, Inc., Significant
K.A. McGuigan, Salary, Pfizer, Inc., Significant
2
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Introduction
• HMG-CoA reductase inhibitor (“statin”) therapy is a widely accepted
treatment for patients with high cholesterol.
• Clinical trials report benefits such as reductions in mortality and
morbidity from statin therapy (National Cholesterol Education Program(NCEP)
Expert Panel 2002, and Simes et al. 2002).
• The extent of cardiovascular risk reduction can increase in proportion
to the amount of time on statin therapy (Simes et al. 2002)
3
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Introduction (continued)
• Prescription drug copayments have increased as employers and other plan
managers attempt to contain prescription drug costs. (Kaiser Family Foundation
and the Health Research and Educational Trust 2005)
• Cost-sharing is likely to continue to rise. Many firms intend to continue to
increase cost-sharing in the near future. (Kaiser Family Foundation and the Health
Research and Educational Trust 2005; PriceWaterhouseCoopers 2005)
• Higher prescription drug copayments may lead patients with chronic conditions
to reduce utilization of maintenance drugs (Bierman and Bell 2004; Gibson et al.
2005)
• Higher statin copayments are associated with a reduction in compliance for
new users of statins. Higher statin copayments and lower levels of statin
compliance are also related to lower levels of outcomes (e.g., LDL-C goal
attainment and hospitalization). (Goldman et al. 2005, Schultz et al. 2005)
4
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Study Aims
1. To estimate the effects of statin copayments on statin adherence for
continuing users of statins, and,
2. To estimate the effects of statin adherence on expenditures and
utilization
5
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Data Source
• MarketScan Commercial Claims and Encounter Database and
Medicare Supplemental and Coordination of Benefits Database for
services provided from January 1, 2000 through December 31, 2003.
– Contains the healthcare experience of individuals with employersponsored health care insurance and Medicare supplemental insurance
in the United States
– Includes enrollment information and inpatient, outpatient and pharmacy
claims
6
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Inclusion Criteria and Study Sample
• Inclusion Criteria:
–
–
–
–
18 years of age or older
Continuously-enrolled from 2000 through 2003
At least one statin prescription fill January 2001 through June 2001
No indication of pregnancy during the study time frame
• Study Sample Construction:
– Continuing users: Filled a statin prescription in 2000
– Each patient was followed July 2001 through December 2003
• n=93,296 patients
7
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Measures
• Adherence to Statin Therapy (July 2001 – December 2002)
– Medication Possession Ratio (MPR) calculated by assessing whether
statins were on-hand each day, % of days with statins on-hand
– Adherent if MPR > 80%
• Expenditures and Utilization (January - December 2003)
– Expenditures –
• Total (Medical plus prescription drug)
• Medical
• Prescription Drug
– Utilization (1/0 variables)
• Physician Office Visit
• Emergency Room Visit
• Hospitalization
• Coronary Heart Disease-related (CHD) Hospitalization
8
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Explanatory Variables
• Patient Cost-Sharing
– Statin cost sharing amount ($ 2003 per day)
– Office Visit cost sharing amount ($ 2003 per visit)
• Sociodemographic - Age, Gender, US Census Region, Urban Area, Household
Income and % with College Degree (by ZIP code via Census information)
• Health Plan Type – (e.g., HMO, PPO, POS, Comprehensive)
• Type of Provider (prior 12 months)
• Medication (prior 12 months) - Number of prescriptions, Any use of mail order
• Severity/Comorbidity (prior 12 months)
– Acute Myocardial Infarction, Angioplasty, Coronary Bypass Surgery, Chronic
Ischemic Heart Disease (IHD), Coronary Atherosclerosis, Other IHD, Hypertension
– Anxiety, Dementia, Depression
9
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Multivariate Analysis
Part 1: Adherence, Logistic Regression
• Pr(Adherencei|x) = F(0 + 1sociodemographici + 2plani + 3providerip
+ 4medicationip + 5severityip + 6comorbidityip + 7cost-sharingi)
Part 2: Utilization and Expenditures
• G(Expenditurei) = ln(0 + 1sociodemographici + 2plani + 3providerip
+ 4severityip + 5comorbidityip + 6E(Adherence))
• P(Utilizationi|x) = F(0 + 1sociodemographici + 2plani + 3providerip +
4severityip + 5comorbidityip + 6E(Adherence))
i is patient, p is a 12 month lag, F is the cumulative logistic function and G is
the gamma distribution
10
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Selected Characteristics
Characteristic
n=93,296
Female (%)
44.5
64.1  11.2
Age (y)
Insurance Plan Type (%)
Comprehensive
28.9
HMO
3.2
Capitated POS
33.0
Non-Capitated POS/EPO
6.4
Preferred Provider Organization
28.8
$0.40  0.20
Statin Copayment ($ per day)
11
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Selected Characteristics (continued)
Characteristic
n=93,296
Disease Severity/Comorbidity (%) (12
month lag)
AMI
1.6
Angina
18.9
CABG
0.9
Chronic IHD
5.0
Coronary Atherosclerosis
24.8
Other IHD
3.4
PCTA
0.6
12
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Measures of Adherence, Utilization and Expenditure
Measure
n=93,296
0.58  0.49
Adherence (July 2001-Dec 2002)
Utilization and Expenditures (2003)
Total Expenditures
$6,589.70  7893.24
Medical Expenditures
$3,513.53  6866.97
Prescription Drug Expenditures
$3,076.17  2598.15
Office Visits
0.93  0.26
ER Visits
0.22  0.41
Hospitalizations
0.14  0.35
CHD-Related Hospitalizations
0.05  0.23
13
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Effects of Copayments on Adherence
Female
Age (y <=65)
Age, y over 65
Adjusted Odds
Ratio 95% CI
(0.863, 0.914)
(1.025, 1.03)
(0.982, 0.988)
Office Visit
Copayment
(0.98, 0.988)
Statin
Copayment
(0.632, 0.735)
61
Predicted Probability of Adherence
Selected
Effects
60
59
58
57
56
55
54
Low
Medium
High
Copayment Copayment Copayment
($.19/day)
($.42)
($.63)
All p<.01
Higher copayments are associated with lower levels
of adherence
14
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Effects of Estimated Adherence on Expenditures
Expenditures
Coefficient
Standard Error
p-value
Total Expenditures
0.139
0.172
0.420
Medical
Expenditures
-0.602
0.291
0.039
Prescription Drug
Expenditures
0.985s
0.103
0.000
Higher levels of adherence are associated with lower medical expenditures,
higher prescription drug expenditures and no change in total expenditures
15
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Estimated 2003 Expenditures
Effects of Estimated Adherence on Expenditures
$7,000
$6,000
$5,000
$4,000
$3,000
$2,000
$1,000
$0
Total
Medical**
Non-Adherent
Prescription Drug***
Adherent
* p<.10, *** p<.01
16
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Effects of Estimated Adherence on Utilization
Coefficient
Adjusted Odds Ratio
95% CI
1.427**
(1.34, 12.958)
ER Visits
-3.315***
(0.017, 0.076)
Hospitalizations
-1.226***
(0.118, 0.731)
CHD-Related
Hospitalizations
-3.972***
(0.005, 0.075)
Office Visits
Higher levels of adherence are associated with an increased likelihood of an
office visit, and a decreased likelihood of ER visits, hospitalizations and
CHD-related hospitalizations
* p<.10,** p<.05, *** p<.01
17
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Effects of Estimated Adherence on Utilization
1
0.8
0.6
0.4
0.2
0
Physician Office
Visits**
ER Visits***
Hospitalization***
Non-Adherent
Adherent
CHD-Related
Hospitalization***
* p<.10,** p<.05, *** p<.01
18
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Limitations
• Administrative Data
• Continuously-enrolled population with employer-sponsored insurance
• Selection
• Sensitivity Analysis
19
Copyright 2006 Thomson Medstat
T H O M S O N
S C I E N T I F I C
&
H E AL T H C AR E
Summary and Conclusions
• In this large cohort of continuing users of statins enrolled in employersponsored plans, prescription drug copayments are a financial barrier
to statin adherence.
• Reducing patient cost-sharing for a maintenance drug regime may be
an effective intervention.
• In turn, statin adherence is related to higher prescription drug
expenditures and an offset in medical expenditures, but total
expenditures are not significantly different.
• Lower statin copayments are associated with higher levels of statin
adherence. Total costs may not change, but fewer negative events
(ER visits, hospitalizations and CHD-related hospitalizations) occur.
20
Copyright 2006 Thomson Medstat