Annual Research Meeting (ARM)
Erasmus University Rotterdam
AcademyHealth, Orlando, 03Jun07
Evaluation of the Dutch Risk
Equalization system: are the insurers
confronted with predictable losses for
the chronically ill?
Wynand P.M.M. van de Ven
(vandeven@bmg.eur.nl)
Pieter J.A. Stam
Rene C.J.A. Van Vliet
Erasmus University Rotterdam
ARM Orlando 03Jun07
1
Erasmus University Rotterdam
Health Insurance Act: 01jan06
• Mandate for everyone in the Netherlands
to buy private health insurance;
• Standard benefits package;
• Selective contracting allowed;
• Open enrolment per product per insurer;
• Community rating per product per
insurer per province;
• Risk equalization.
ARM Orlando 03Jun07
2
Risk Equalization Fund (REF)
Erasmus University Rotterdam
Gov’t contribution
(18-)
REF
Income-related
contribution
50%
50%
Insured
REF-payment based
on risk adjusters
Insurer
premium (18+)
Two thirds of all households receive an income-related care allowance
(at most € 420 per person per year)
ARM Orlando 03Jun07
3
Erasmus University Rotterdam
Risk adjusters in the Dutch REF
Year
1992
1995
1997
2002
New risk adjuster
Age/gender
Region, yes/no employee, disability
Age/disability
Pharmacy-based Cost Groups (PCGs)
(13 PCGs and about 7% of population)
2004 Diagnostic Cost Groups (DCGs) (about 2% of pop)
yes/no self-employed
2007 Multiple PCGs allowed (co-morbidity);
New PCGs: mental health (3% pop.), cancer
and growth hormons
(20 PCGs and about 16% of population)
ARM Orlando 03Jun07
4
Erasmus University Rotterdam
Effects of selection
• Disincentive for insurers to be responsive to
the high-risk consumers and contract the
best quality care for them;
• Disincentive for providers to acquire the
best reputation for treating chronic diseases;
• Selection more profitable than efficiency;
• High premiums for high-risk patients;
• Instability in the insurance market.
ARM Orlando 03Jun07
5
Erasmus University Rotterdam
Objective & Research questions
Objective: evaluate the risk equalization system.
Research questions:
1. Are there identifiable subgroups of consumers
with predictable lossses?
2. If so: How large are these subgroups? And
how large are the predictable losses?
In particular we focus on subgroups of
persons with a chronic condition or with
above average utilization rates in previous
years.
ARM Orlando 03Jun07
6
Erasmus University Rotterdam
Method
• Data: all information in the files of a large
insurer (Agis) over the period 1998 – 2004,
combined with an individual health survey
(held in 2001); some 30,000 observations.
• Method: the Dutch 2007 risk adjusters are
applied to the 2004-data. By comparing the
predicted 2004-expenditures (based on the
2007 risk adjusters) with their actual 2004expenditures we calculated the average profits
and losses for many subgroups.
ARM Orlando 03Jun07
7
Results (costs and losses in euro)
Erasmus University Rotterdam
Subgroup 2001
Size
Costs Predictable
2004 losses 2004
Self-reported health status fair/poor
21.2%
3404
541
Worst score Physical functioning (SF36)
10.0%
4469
1140
Worst score Social functioning (SF-36)
10.0%
3190
649
Restricted in mobility (OECD-score)
14.9%
3740
653
Stroke, brain haemorrhage/ infarction
2.6%
4341
943
Myocardial infarction
3.3%
4755
789
Other serious heart disease
2.3%
4654
926
Some type of (malignant) cancer
4.8%
3440
689
ARM Orlando 03Jun07
8
Erasmus University Rotterdam
Results (costs and losses in euro)
Costs
2004
Predictable
losses 2004
Subgroup 2001
Size
High bloodpressure
15.2%
2961
342
Astma, chronic bronchitis, emphysema
8.1%
3182
460
3-6 self-reported conditions
22.3%
2848
333
7 or more self-reported conditions
2.9%
4833
1461
Prescribed drugs (self reported, 2
weeks)
48.2%
2597
220
Contact specialist (self reported, 1 year)
39.8%
2586
317
Hospitalization (self reported, 1 year)
7.5%
3611
1034
Home care (self reported, 1 year)
2.2%
4258
1152
ARM Orlando 03Jun07
9
Erasmus University Rotterdam
Results (costs and losses in euro)
Subgroup 1997 - 2001
Size
Costs
2004
In top-25% highest costs, in 3 of 5 years
5.9%
2537
238
In top-25% highest costs, in 4 of 5 years
4.5%
3240
304
In top-25% highest costs, in 5 of 5 years
8.2%
6131
1757
Hospitalization in 2 of the 5 years
4.7%
3613
728
Hospitalization in 3 of the 5 years
1.1%
6606
2030
Hospitalization in 4 of the 5 years
0.3%
11763
5933
Hospitalization in 5 of the 5 years
0.1%
14373
6453
ARM Orlando 03Jun07
10
Predictable
losses 2004
Erasmus University Rotterdam
Conclusions
1. Many subgroups, from <1% to 30% of population,
with predictable losses in the order of hundreds
to thousands euros per person per year.
2. Also predictable losses for subgroups of insured
whose disease is included as a risk adjuster in
the risk equalization formula (e.g. heart
problems, cancer, …).
3. Improvement of the risk equalization system
needs a high priority. Otherwise the
disadvantages due to risk selection may
outweigh the advantages of competition.
ARM Orlando 03Jun07
11
Erasmus University Rotterdam
New (potential) risk-adjusters
• Diagnostic information not only from
prior hospitalization, but from all prior
medical encounters (Diagnosis Treatment
Combinations, DTCs) expected to be
implemented in 2009;
• Multiyear-DCG’s;
• A better indicator of invalidity (or
functional heath status);
• Yes/no voluntary deductible;
• ……
ARM Orlando 03Jun07
12