T H O M S O N
T H O M S O N
H E A L T H C A R E
H E A L T H C A R E
Background
• Children represent over one-quarter of nonelderly enrollees in private
health plans in the US. (Medical Expenditure Panel Survey, 2005)
Prescription Drug Cost-Sharing Among
Commercially-Insured Children and Adults with
Chronic Illness
• Children are dependent upon parents (or legal guardians) to mediate
the health care delivery system on their behalf
– Child is principal (P), parent is agent (A)
– This interaction differs from many principal-agent (P-A) interactions:
• The Principal is largely incapable of managing/supervising the
Agent
• The Agent is assigned to the Principal by law/custom/birth
• The P-A contract is implicit, since legal minors cannot sign or
negotiate contracts
– Contract is not between child and parent, but parent and state
– Parents have an “implicit promise” to behave in the interests of the child
Teresa B. Gibson, PhD
Thomson Healthcare, Ann Arbor, MI
(Becker and Murphy, 1988; Munro, 2001)
Medstat • MercuryMD • Micromedex • PDR • Solucient
©2007 Thomson Healthcare. All rights reserved.
T H O M S O N
H E A L T H C A R E
T H O M S O N
H E A L T H C A R E
Background (continued)
Cost-Sharing
• Information asymmetry in health care
•
Adults and children in the same employer-based health plan typically
face the same levels of cost-sharing (e.g., copayments, coinsurance)
•
Most cost-sharing studies have focused on the price-responsiveness of
adults.
– Parent/Provider: Parents seek help from physician agents to help
determine a course of treatment
– Parent/Child: Children must communicate symptoms to parents
–
–
•
2
Few studies include children or report results separately for children.
Little evidence regarding price-responsiveness and chronic illness in children
Price elasticity for medical services is different for children and adults
(Newhouse 1981)
–
–
•
Children: price inelastic response for inpatient services, price elastic response for
outpatient services
Adults: price elastic response for both inpatient and outpatient services
Price elastic demand for antibiotics among children and adults (Foxman
1987)
•
Adoption of a 3-tier formulary from a 1-tier formulary medications in
children resulted in a decline in the rate of adoption of ADHD
medications but few changes in utilization for existing users (Huskamp
2005)
©2007 Thomson Healthcare. All rights reserved.
T H O M S O N
3
©2007 Thomson Healthcare. All rights reserved.
H E A L T H C A R E
T H O M S O N
H E A L T H C A R E
Study Aims
Data Source
•
• 2001 through 2003 MarketScan Commercial Claims and Encounters
database
To examine the effects of higher levels of prescription drug costsharing on children with chronic illness
– Analyze price-responsiveness for a single, common chronic illness,
persistent asthma, affecting both children and adults
– Is price important when providing health care to children with a common
chronic illness?
©2007 Thomson Healthcare. All rights reserved.
4
– Representing the health care experience of enrollees in employersponsored health plans in the US
5
©2007 Thomson Healthcare. All rights reserved.
6
1
T H O M S O N
H E A L T H C A R E
T H O M S O N
Study Sample
H E A L T H C A R E
Study Sample
• Patients with Persistent Asthma age 5-54 years
• Met HEDIS denominator criteria for persistent asthma (493.xx) in index
year (2001 or 2002)
29%
– Based on: inpatient use, ED use, outpatient use and/or asthma
prescription drug use
56,381 adults
22,985 children
(18-54 years)
(5-17 years)
• Continuously enrolled at least 24 months
• Index year/measurement year combinations (2001/2002 or 2002/2003
- Children and adults were enrolled
in the same set of employer-based
health plans
71%
-22.9% of children and 27.5% of
adults appear in both years
Children
©2007 Thomson Healthcare. All rights reserved.
T H O M S O N
Adults
7
©2007 Thomson Healthcare. All rights reserved.
H E A L T H C A R E
T H O M S O N
Measures
Explanatory Variables
1. Any asthma drug use (1=yes, 0=no)
• Patient Cost-Sharing
8
H E A L T H C A R E
– Asthma drug cost sharing amount (US$ 2003 per 30-day supply)
– Office Visit cost sharing amount (US$ 2003 per visit)
– At least one prescription in the measurement year (2002 or 2003) if
identified as having asthma in prior year
2. Count of asthma prescriptions (in 30-day equivalents) in 2002 or 2003
• Sociodemographic - Age, Female, US Census Region, Median Household
Income (by ZIP code via Census information), salaried/hourly
3. Count of prescriptions conditional on use (in 30-day equivalents) in
2002 or 2003
• Health Plan Type – (e.g., HMO, PPO, POS, Comprehensive)
• Pulmonologist visit (prior 12 months)
• Disease Prevalence/Comorbidity (prior 12 months)
–
–
–
–
Charleson Comorbidity Index
Stage of Asthma (Disease Staging)
Sinus infection, otitis media, migrane, bronchitis
Anxiety, SSRI use, Depression
• Time (index year 2001)
©2007 Thomson Healthcare. All rights reserved.
T H O M S O N
9
©2007 Thomson Healthcare. All rights reserved.
H E A L T H C A R E
T H O M S O N
Multivariate Analysis
Results: Selected Characteristics
• P(Any useit|xit) = F(γ0 + γ1sociodemographicit + γ2planit + γ3providerip +
γ4severityip + γ5comorbidityip + γ6 cost-sharingit)
Characteristic
– Panel data logit model for any asthma drug use
• P(Number of Rxit|xit) = G(β0 + β1sociodemographicit + β2planit +
β3providerit + β4severityip + β5comorbidityip + β6 cost-sharingit)
10
H E A L T H C A R E
Children
Adults
n= 22,985
n= 56,381
Age (y)
10.7 ± 3.7
46.3 ± 9.4
Charlson Index
0.52 ± 0.54
0.68 ± 0.97
$9.6 ± 4.3
$9.4 ± 4.7
Copayment
– Panel data poisson model for counts of prescription drugs
Asthma Copayment ($/30 day supply)
Measures
• where i is patient, t is measurement year, p is index year
Any asthma drug
76.6%
78.0%
• panel data
Count of asthma drugs
4.0 ± 4.6
4.9 ± 5.5
Count of asthma prescriptions
5.4 ± 4.6
6.6 ± 5.4
conditional on use
©2007 Thomson Healthcare. All rights reserved.
11
©2007 Thomson Healthcare. All rights reserved.
12
2
Results – Any Asthma Drug
Copayment
Adults
Children
n=56,381
n=22,958
-0.019***
0.001
(0.004)
(0.006)
0.024***
-0.088***
(0.002)
(0.006)
Pulmonologist
Visit (last 12
months)
0.397***
0.544***
(0.054)
(0.172)
Household
Income
0.008***
0.008***
(0.001)
(0.001)
Age
Predicted Probability of Any Use
Selected
Effects
0.8
H E A L T H C A R E
T H O M S O N
Selected
Effects
Copayment
0.75
Age
0.7
Pulmonologist
Visit (last 12
months)
0.65
Household
Income
0.6
*** p<.01
Adults***
Children
Mean Copay
$10 increase
©2007 Thomson Healthcare. All rights reserved.
T H O M S O N
Children
n=42,763
n=22,958
Copayment
-0.004***
0.002
(0.001)
(0.001)
Age
0.007***
-0.002***
(0.000)
(0.002)
Pulmonologist
Visit (last 12
months)
Household
Income
0.086***
0.114***
(0.010)
(0.035)
-.0001
0.001*
(0.000)
(0.001)
6.8
Number of Asthma Prescriptions
Adults
Children
n=56,381
n=22,958
-0.008***
0.002
(0.001)
(0.002)
0.011***
-0.040***
(0.001)
(0.002)
0.08***
0.136***
(0.011)
(0.040)
0.001***
0.002***
(0.000)
(0.001)
5.3
Effects of a $10 increase
in Copayment
4.8
4.3
3.8
3.3
2.8
2.3
1.8
1.3
0.8
Adults
***
Mean Copay
13
Children
$10 increase
©2007 Thomson Healthcare. All rights reserved.
H E A L T H C A R E
T H O M S O N
14
H E A L T H C A R E
Other Results
Effects of a $10 increase
in Copayment
• Family Dyads
– Adults (parents) with asthma who had children with asthma (n=2,644)
had were less price sensitive than adults without asthmatic children for
each of the three measures: any use, number of prescriptions contingent
on use, number of prescriptions
5.8
4.8
• Adults and children with asthma in both years
– Patients appearing in both years (adults, n=21,423 ; children n=7,187)
had a less elastic price response than the full sample
3.8
2.8
• Income
– Price effects did not vary by income.
• Children residing in lower income areas (< $38K) had the same
inelastic response as children residing in higher income areas
(> $64K)
1.8
0.8
Adults
*** p<.01, * p<.05
Adults
*** p<.01
Results – Number of Prescriptions, Conditional on Use
Selected
Effects
H E A L T H C A R E
Results – Number of Asthma Prescriptions
Effects of a $10 increase
in Copayment
Number of Asthma Prescriptions
T H O M S O N
***
Mean Copay
Children
$10 increase
©2007 Thomson Healthcare. All rights reserved.
T H O M S O N
15
©2007 Thomson Healthcare. All rights reserved.
H E A L T H C A R E
T H O M S O N
Limitations
16
H E A L T H C A R E
Conclusions
• Commercially-insured parents in employer-sponsored health plans
may err on the side of caution by providing medications to their
chronically-ill children
• Measure prescription fills, not actual consumption patterns
• Persistent asthma criteria
– Meeting the asthma criteria for 2 years may improve ability to select
patients most likely to have asthma-related utilization (Mosen 2005,
Weiss 2006)
• Sensitivity analysis requiring 2 years of asthma revealed no
difference in results
– Criteria based upon utilization, not pulmonary function
• Higher copayments for children with asthma may not affect the
utilization of prescription drugs, as parents may seek to act in the best
interests of their children.
• Prescription drug copayments may not impede care for chronically-ill
children but may create a financial burden for families
• Continuously-enrolled population with employer-sponsored insurance
– Higher income
©2007 Thomson Healthcare. All rights reserved.
17
©2007 Thomson Healthcare. All rights reserved.
18
3
T H O M S O N
H E A L T H C A R E
T H O M S O N
Other Considerations
H E A L T H C A R E
Other Considerations
Demand for Asthma Prescriptions
Price per
Prescription
D
• “Even altruistic parents have to consider the trade-off between their
consumption and the human capital of children” (Becker and Murphy, 1988,
For child asthmatics, demand
is inelastic.
Copay2
p. 5)
• The loss in buying power may introduce principal-agent conflicts within
the family
Is Q’ optimal?
– Choices between medications and other goods
– Trade-off between the welfare of the child and the welfare of the parent
– Particularly important for lower income families (Munro 2001)
Copay1
Q* Q’
• Are higher user fees (e.g, higher copayments) the most effective way
to manage consumption of maintenance medications in chronically-ill
children?
Quantity
Of Prescriptions
©2007 Thomson Healthcare. All rights reserved.
19
©2007 Thomson Healthcare. All rights reserved.
20
4