Transcript Waiver Systems for Government

Waiver Systems for GovernmentFinanced Health Care:
Lessons from Suriname and Jamaica
Ricardo Bitran, Ph.D.
Rodrigo Muñoz, M.S.
Ursula Giedion, M.S.
Bitran & Asociados
December 2003
1
Contents
• Part I: Suriname
• Part II: Jamaica
• Part III: Conclusions
2
Suriname
Part I
Evaluation of Suriname’s Ministry of
Social Affairs (MSA) Card System
3
Suriname
Study objectives
1. Compliance of MSA Card system with own policy
2. Type I and II errors in MSA Card system
3. Use of services and out-of-pocket expenses by
MSA Card beneficiaries
4. If necessary, explore alternative identification
mechanism based on “proxy means test”
4
Suriname
Description of MSA Card System
• Mission:
– Identify the poor and near-poor and subsidize their health care
– Pay health providers for services delivered to the MSA Card holders
Meets
Requirement
s (Poor or
Near Poor)
Get MSA Card. If they get care, MSA pays for it
MEANS
TEST
MSA Social
Worker
Do not meet
requirements
Do not get MSA Card. If they get care they, or
their insurer or employer, pay for it, not MSA
5
Suriname
Description of MSA Card System
• Official selection criterion and fees:
MSA Card system summary, in SF. and $ (period of Jan. 1999 to Sep. 2000)
MSA Card type
A: Poor
B: Near-poor
Administrative fees
Co-payment for
hospitalization (per
day)
Co-payment for
medicines at the
pharmacy
0 – 20,000
1,000
300
150
$
0 – 14.30
0.70
0.20
0.10
SF.
20,000 – 30,000
2,000
1,200
300
$
14.30 –21.40
1.40
0.90
0.20
Currency
Monthly income
range
SF.
– Income-based criterion to identify the poor & near-poor
– Small administrative fees
– Modest co-payments for hospitalizations and medicines
6
Suriname
Description of MSA Card System
• Other (informal) selection criteria:
 Household size
 Education level
 Medical condition
(chronic, disabled,
handicapped, elderly)
 Early adulthood
pregnancy
 Dwelling condition
 Female single
 Running water and
electricity
 Square meters per
household member
 Number of bedrooms
 Cooking fuel used
 Means of
transportation
 Distance to work
 Presence of a previous
card
 MSA staff knows that the income-based criterion has
flaws: they reach the poor more accurately with
informal socio-economic criteria
7
Suriname
1. Effectiveness of MSA Card System:
Compliance with own policy
Possession of MSA Card according to self-reported income
Number of households
reporting income
•
•
•
•
Has an MSA Card
Does not have an
MSA Card
Total
Below SF. 30,000
23
40
63
Above SF. 30,000
363
829
1,192
Total
386
869
1,255
6% of MSA Card holders declared income below SF.30,000
Among those with income below SF. 30,000, 37% held MSA Card
94% of MSA Card holders with income above SF. 30,000.
Limitations of these measurements:
– 5 percent of the 1,255 households declared an income below SF.30,000
– Income data from Household Budget Survey unreliable (poor correlation
with expenditure)
 Household survey: MSA officials use a higher implicit cut-off point. This
is consistent with the other selection criteria based on the socioeconomic assessment of applicants
8
Suriname
2. Effectiveness of MSA Card
System: Type I and II errors
Poor:
should
have a
card
• Type I error or
Under-coverage:
– Number of individuals
entitled to an MSA Card
who do not have one,
divided by the total number
of people entitled.
• Type II error or
Leakage:
– Number of MSA Card
holders that are not entitled,
divided by the total number
of MSA Card holders.
Has a
card
Does not
have a
card
NonPoor:
should
not have
a card
Ok
Type II:
Leakage
(94%)
Type I:
Undercoverage
(63%)
Ok
9
Suriname
2. Effectiveness of the MSA Card
System: Type I and II errors
Causes of error:
• Inappropriate classification criteria:
– Actual cut-off line of SF.30,000 is too low  it does not represent
the target population, i.e. the poor and near-poor.
• Some individuals are classified incorrectly:
– Income-based classification is difficult to implement reliably.
• Some individuals are never classified:
– Actual policy does not actively search for potential beneficiaries.
10
Suriname
2. Effectiveness of the MSA Card
System: Type I and II errors
The SF.30,000 set too low, thus cut-off line does not
identify the poor and near-poor. Researchers used
official Surinamese poverty line to define the target
population.
Total number of individuals and households from the HBS
Individuals
Households
Below the official poverty line
2,780
617
Above the official poverty line
2,177
638
Total
4,957
1,255
 More than half of the population lives under
the poverty line
11
Suriname
2. Effectiveness of the MSA Card
System: Type I and II errors
Health insurance among poor
households
Households with an MSA Card
Poor
64%
Not all MSA cards
belong to poor
households: leakage
Other insurance
37%
MSA Card
40%
Type II error
Type I error
Not all poor
households have
health insurance:
under-coverage
Non-poor
36%
No insurance
23%
141
households
Cards needed = Cards leaked
140
households
12
Suriname
2. Effectiveness of the MSA Card
System: Type I and II errors
What are the causes of the under-coverage?
 The data available do not permit clear determination of the causes
However, the data suggest that
a main cause is that some
individuals are never classified.
Percent of individuals without MSA Card
(five lower deciles)
30
25
 Actual policy does not
actively search for potential
beneficiaries
20
15
10
5
0
I
II
III
IV
V
13
Suriname
3. Effectiveness of MSA Card System:
Access to health services
Kind of
insurance
coverage
Ambulatory care
Health problem
during last 30
days
Took no action
regarding health
problem
20
16
None
11
MSA
Card
0
20
40
0
99
99
46
17
7
26
78
67
8
24
26
72
67
4
10
20
0
20
Received care
61
38
5
29
27
Sought formal
care only
28
17
23
20
Other
Sought informal
care only
40
0
50
100
0
50
99
99
100
98
Hospitalized
during the last
year
3
1
Nonpoor
7
5
Poor
11
9
100
0
10
20
Chronic illnesses
Pap-smear test
during the last
year (women)
Children under 3
w/ immunizations
up to date
25
None
MSA
Card
26
86
75
29
24
82
77
Sought informal
care for chronic
illness
36
33
5
67
34
Took no action
regarding
chronic illness
6
93
19
Other
Prevalence of
chronic illnesses
11
21
24
20
7
14
11
43
37
30
14
7
Sought formal
care for chronic
illness
25
26
Nonpoor
63
73
16
16
Poor
59
59
14
0
20
40
0
50
100
0
10
20
0
20
40
0
20
40
0
50
100
Suriname
3. Effectiveness of MSA Card System:
Financial protection
Kind of
insurance
coverage
Mean
expenditure in
$ (over the
relevant
population)
21
None
Other
136
10
9
Non-poor
Poor
4
13
MSA
Card 4
19
38
10
Non-poor
Poor
20
40
0
100
0
Non-poor
Poor
7
Non-poor
Poor
4
4
1
0
Non-poor
Poor
2
1
5
10
20
3
5
4
3
5
10
8
3
0
200
11
7
2
Non-poor
Poor
3
2
7
MSA
Card
10
6
31
6
None
Other
10
11
0
Percentage of total
household
consumption
expenditure
Chronic illness care
in the formal sector
(monthly)
Hospitalizations
(yearly)
Ambulatory care in
the formal sector
(monthly)
2
10
20
0
5
10
15
Statistically significant
variables
Initial list of variables
Suriname
4. Exploration of alternative eligibility
criteria
–
–
–
–
–
–
–
–
–
–
–
Number of members under 15 years of age
Cooking fuel
Condition of the dwelling
Presence of electricity
Presence of telephone
Presence of toilet inside the dwelling
Construction material of the dwelling
Ownership or mortgage of the dwelling
Company water inside/outside the dwelling
Dwelling surface per capita
Number of bedrooms per capita
R² = 0.267
16
Suriname
4. Exploration of alternative eligibility
criteria
Type I error
Under-coverage
Type II error
Leakage
Measured from HBS
1999/2000
23
36
Predicted with proxy
means test
22
28
Includes only the error
caused by wrong classification
 total under-coverage would
be slightly higher
Slight reduction in leakage
17
Suriname
Conclusions
• With time, MSA staff has reduced classification errors by:
– Using an implicit higher income cut-off point
– Adopting informal criteria based on socio-economic status of applicants
• MSA beneficiaries tend to be poorer  System is progressive
• Under-coverage (poor people with no insurance): 23 percent
–
–
–
–
Caused mainly by lack of policies that actively search and screen the poor
Also caused by inadequate income cut-off line and informal selection criteria
Good performance in comparison with other countries
Coverage (poor people with MSA Card): 40 percent
• Leakage (MSA Card holders above the poverty line): 36 percent
– Caused by errors during the screening process
– Standard performance in comparison to other countries
• The good news: Needed cards = Leaked cards
18
Suriname
Conclusions
• MSA Card increases utilization of health services by
the poor:
– Use of services by poor MSA Card holders approaches that of the
non-poor, whereas the use by the uninsured poor is much lower:
Card promotes equity in delivery
– Adverse selection is also observed
• MSA Card reduces the financial burden of the poor:
– Proportion of income allocated to health expenditure by MSA Card
holders approaches that of the non-poor, whereas the uninsured
poor spend much more: Card promotes equity in financing
• Available proxy means test:
– Would not improve (may worsen) under-coverage
– Would only slightly reduce leakage
– Higher cut-off line would reduce both errors, but more cards
needed
19
Jamaica
Part II
Assesment of User Fee Program (UFP)
in Jamaica
20
Jamaica
Background
• Strengthening of user fee program (UFP) in late 1980s
and 1990s
– User fee revenue from a low 1% in 1986 to 10% in year 2000
• Concerns that UFP creates access problems leading to:
– Non-attendance at primary care (e.g. family planning) and other
outpatient clinics (e.g. diabetes, hypertension)
– Early hospital discharges
– Increasing hospital re-admission rates
– Non-attendance for elective surgery
– Denial of access to birth and death registration data at hospitals
21
Jamaica
Study Objectives
1. Examine policy and operational frameworks of UFP
2. Evaluate data on service use before and after UFP
3. Get views of key stakeholders on performance of
UFP
4. Do ‘cost-benefit’ and ‘equity’ analysis of UFP
5. Recommend actions to re-design or expand UFP
22
Jamaica
Review of “history” of UFP and
exemption schemes
• 1993 User Fee Schedule
–
–
–
–
–
Different fees for public and private patients
Different fees for patients with or without insurance
Different fees for hospitals and health centers
Fixed fees
List of waived people + inability to pay
• 1999 User Fee Schedule
–
–
–
–
More detailed specification of services
Patients with insurance treated as private patients
Fees increase
Similar waiver scheme
23
Jamaica
Review of “history” of UFP and
exemption schemes
• Unusually large number
of user fee studies in
Jamaica over last 20
years.
• Consensus that
collection of fees has
increased steadily and
substantially, though
problems in collection
detected (collection is
only 60-65% of amount
billed, Lefranc & Lalta
(2001)).
User fee collection as a percentage of total public
health expenditure in selected countries
14
12
10
8
6
4
2
0
SVG 1996
Zimbabwe
1995
Jamaica
2001
Chile 1995
Ghana
1999
Thailand
2000
• Recognized that utilization of health services dropped
especially in public health facilities previous to 1994
24
Jamaica
Review of “history” of UFP and
exemption schemes
• Drop of utilization in public facilities due to many factors
– Low quality of government health services
– Adoption of fees
• Several studies identify equity problem in access. For example:
– Wagstaff (1998): horizontal equity problems in service provision,
explained by lack of financial protection for the poor.
– Lefranc & Lalta (2001): equity in access problems as income is
shown to be a strong predictor of service use.
• Consensus that the waiver policy not working well, leading to
basic problems shared by countries implementing waivers to
protect the vulnerable:
– Lack of a strong national waiver policy, leading providers establish
their own waiver criteria
– Substantial administrative burden to assess waiver eligibility
25
Jamaica
Study results based on previous surveys:
Hospital Information System (HIS)
• Results:
Change in utilization of services in
public health facilities 1991-2002
– Total utilization of
health services:
10.0
5.0
Total health
centers visits
0.0
-5.0
Total services
public facilities
-10.0
-15.0
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
 “M”-shaped: pickup
after historical low
figures in 1994 and
renewed drop after new
increase of user fees in
1999.
Total hospital
services
15.0
%
 Total decline 14.4% for
the period 1994-2002
20.0
Year
Increases in user fees
26
• Results:
– Primary care services:
 Maternal child health
services seem less
elastic to price changes
than curative OP
(outpatient) visits and
casualty visits
 Substantial drop of
service utilization in
1999
Curative visits
2500
2000
Maternal/child
and other
visits
Total health
centers visits
1500
1000
500
0
19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
 Total decline 34% for
the period 1994-2002
Utilization of services in health centers
1990-2002
Thousands
Jamaica
Study results from previous surveys:
Hospital Information System (HIS)
Year
Introduction of user fees
in health centers
27
Percentage of those reporting illness/injury
in last 4 weeks, 1992-2001 (Quintile 1 and 5)
Days of illness/injury related impairment ,
1992-2001 (Quintile 1 and 5)
Quintile 1
AVG
2001
2000
1999
1998
1997
1996
1995
Quintile 5
1992
AVG
2001
2000
1999
1998
1997
1996
1995
1994
Quintile 5
4.8
1994
Quintile 1
1993
6.8
10,0
9,0
8,0
7,0
6,0
5,0
4,0
3,0
2,0
1,0
0,0
1993
18,0
16,0
14,0
12,0
10,0
8,0
6,0
4,0
2,0
0,0
1992
No substantial differences in
self-reported health status
Illness-related impairment
higher among the lowest
quintiles
Percentage of those ill seeking care, 19922001 (Quintile 1 and 5)
80,0
70,0
60,0
50,0
40,0
30,0
20,0
10,0
0,0
Quintile 1
AVG
2001
2000
1999
1998
1997
1996
1995
1994
1993
Quintile 5
1992
Jamaica
Study results based on previous surveys:
Jamaican Survey of Living Conditions (JSLC)
Lower propensity to seek care among
the poor (steady decline since 1999)
28
• Use of public/private facilities
Quintile 1
40,0
Quintile 5
20,0
AVG
2001
2000
1999
1998
1997
1996
1995
1994
0,0
28.5%
Use of public services by those seeking
care (Quintile 1), 1992-2001
38.7%
80,0
60,0
40,0
Medical Care
Medication
AVG
2001
2000
1999
1998
1997
0,0
1996
20,0
1995
– Public sector used mainly by the poorest
– For hospitalizations, the poorest only use
hospital services whereas the majority buys
medication in the private sector
– These data suggest existence of qualityrelated problems in public sector
100,0
1994
 2001:
60,0
1993
 1992:
80,0
1993
– General trend for this period: towards
increase in proportion of people using public
facilities
Public sector use by those seeking care,
1992-2001 (Quintile 1 and 5)
1992
– Jamaicans make strong use of private
services. Even among the poorest, on
average only one-half of those seeking care
went to public facilities
1992
Jamaica
Study results based on previous surveys:
Jamaican Survey of Living Conditions (JSLC)
Hospitalization
JSLC 1992-2001
29
80,0
60,0
40,0
20,0
Medical Care
Medication
2001
2000
1999
1998
1997
1996
0,0
1995
– For hospitalizations, public
facilities are used
extensively
100,0
1994
– They rarely use public
facilities when seeking
medical ambulatory care or
medications
Use of public services by those seeking
care (Quintile 5), 1992-2001
1993
• Use of public/private
facilities by the wealthiest
(Q5)
1992
Jamaica
Study results based on previous surveys:
Jamaican Survey of Living Conditions (JSLC)
Hospitalization
JSLC 1992-2001
30
• Cost of services in the
public/private sector
All Patients: Mean Patient Expenditure
in Public and Private Facilities
1992-2001 (Real $ 1990)
200
80
Public
40
AVG
2001
2000
1999
1998
1997
1996
1995
1994
1993
0
Quintile 1: Mean Patient Expenditure
in Public and Private Facilities
1992-2001, (Real $ 1990)
120
100
80
60
40
20
0
Private
AVG
2001
2000
1999
1998
1997
1996
1995
Public
1994
– For the poorest, the mean
expenditure on visits in private
sector dropped substantially after
1999
Private
120
1992
– On average, Jamaicans spend three
times more in private sector than in
public sector
160
1993
– Overall, the mean cost for visits in
the public health sector has
increased from 5 J$ to 63J$ in year
2001 (Real 1990 $)
1992
Jamaica
Study results based on previous surveys:
Jamaican Survey of Living Conditions (JSLC)
31
Jamaica
Study results based on surveys collected
for this study
• Survey Design
– Applied to patients, front line workers and directors of health
facilities
– 16 public health facilities surveyed
 2 hospitals and 2 health centers from each RHA
– 280 Patients interviewed
 231 outpatients and 49 inpatients
– 42 Frontline workers
 26 cashiers, 11 SAOs and 5 others
– 27 Managers
 13 administrative and 14 medical
32
Jamaica
Study results based on surveys collected
for this study
•
User charges tend to be regressive
–
–
•
•
•
In public facilities, most of poorest
Patients according to payment of services
pay upfront whereas 50% of
wealthiest either receive total or
120
Service free/full exemption
partial waiver or are granted a credit 100
for future payment
80
Payment of total or part of
% 60
Worrisome situation as rich use
fee in the future
predominantly more costly hospital 40
Part of fee at the point of
services Leakage of public
20
service
subsidies
0
Total fee at the point of
1
2
3
4
5 Total
83% of inpatients did not pay
service
charges
Quintiles
–
•
90% of patients in Q1 paid
65% of patients in Q5 paid
Most of inpatients in the sample
belong to quintiles 3,4 and 5.
8% of outpatients did not pay
charges
–
Most outpatients in sample from Q1
33
Jamaica
Study results based on surveys
collected for this study
• Patient attitudes:
–
Most of patients in Q1(86%) did not request waiver compared with 61% in Q5.
–
All waiver requests accepted and full or partial waiver granted.
–
More needs to be known to explain this situation. On the basis of other countries
experiences this might be due to one or a combination of the following reasons



Inadequate information
Stigma
Complicated and time consuming procedures to receive an exemption together with low fee
levels (“it is not worthwhile”)
Waivers: Patients’ behavior and system response
Did not ask for
Exemption
Partial
exemption --full refused --full
exemption -- Full exemption
Quintiles
payment made payment made payment made or service free
1
86
0
2
2
2
74
0
2
10
3
78
0
0
4
4
74
0
4
4
5
61
0
0
8
Total
75
0
2
6
Source: Outpatients and inpatients survey, August 2003.
34
Jamaica
Study results based on surveys
collected for this study
• Coping with payment of
health care costs:
–
–
Most patients reported to have
been in situation of not having
funds to pay for their health care
Of those under financial pressure:
 Borrowing from the
extended family network
seems to be the principal
way of coping
 One-fourth decided to delay
treatment showing a
problem of access to health
services.
Patient coping under financial pressure
Frequency % of total
Never under financial pressure
Have been under financial pressure
Total
79
179
258
31
69
100
Of those under pressure
Borrow
Delay treatment/stay at home
Still come for treatment try best to pay
Use home remedies
Borrow or delay treatment/stay at home
Go to a cheaper public facility
Savings
Pray/God provides/other
69
46
35
13
4
3
2
7
39
26
20
7
2
2
1
4
179
100
Total
35
Jamaica
Study results based on surveys
collected for this study
• Financial aspects
– Between 1998 and 2001
budget cuts were substituted
by user fee increases
– This implies a shift of financial
responsibility from the public
sector to patients without a
simultaneous quality
improvement
12.000
10.000
8.000
6.000
4.000
MOH current
budget
User fee
collection
2.000
0
1.000
900
800
700
600
500
400
300
200
100
0
2002/03
2001/02
2000/01
1999/2000
1998/99
1997/98
– In this situation, utilization of
services will drop (as has been
observed) as there is only a
price and no quality effect.
Total budget and fee collection (real $J 2003)
36
Jamaica
Conclusions and recommendations
• Institutions should have incentives to collect fees
• Price definition in 1999 Gazette defies basic economic design criteria
• User fees to be applied uniformly across country
• Fee collection evaluation and monitoring tools in health facilities
• Administrative cost of UFP is 8 percent of fee collection. However, data
on UFP staff and salaries is still not clear.
• The good news: the institutional implementation of the UFP is in
accordance to internationally accepted guidelines.
• Institutions to have incentives to grant waivers and to collect fees
• Government should set aside special fund for waiver reimbursement
• Users unaware of waiver rules
• Waivers to cover the neediest
37
Conclusions
Part III
Conclusions
38
Conclusions
Conclusions
1.
2.
3.
4.
5.
6.
7.
8.
9.
Pricing policy to be consistent with health policy objectives
Tendency to define user fee and waiver systems loosely
Designing & implementing systems of waivers: easier said than done
Successful waiver systems must be accompanied by appropriate
incentive schemes
Chief incentive: provider to be reimbursed for forgone income from
waivers
Propensity to abuse waiver systems by non-poor is substantial: leakage
remains a threat
Waivers to cover the neediest: fine tuning of targeting systems seems
possible mainly through individual means testing
Widespread education of the poor about their rights is essential
Evidence that waiver systems can improve equity in delivery and in
financing
39