Challenges in hospital reform Barbara McPake London School of Hygiene and Tropical Medicine.

Challenges in hospital reform

Barbara McPake

London School of Hygiene and Tropical Medicine

Why reform hospitals?

Over-funded or under-funded?Failure to play designated role in the

health system?Political importance, especially of national

referral hospitalsDominance of medical professional

interests in hospital policy, combined with inappropriate incentives

Why reform hospitals?

Over-crowding, staff attrition, waiting and waiting lists?

Long neglect in public health debateNecessary adaptation to greater emphasis

on PHCAdaptation to new technical opportunitiesHIV/AIDS

What are the options for hospital arrangements?

Markets/Private sector

Broader public sector

Core public sector

PCAB

Source: Harding and Preker: 1998• Budgetary unit• Autonomy• Corporatization• Privatization

Important dimensions of these options

Source: Harding and Preker, 1998

Decision rights

Market exposure

Residual claimant

Accountability

Social functions

Public purse

Unspecified, unfunded mandate

Vertical hierarchy Management autonomy

Direct budget allocation Nonbudgetary revenues

Private owner

Direct hierarchicalcontrol

Rules, regulations and contracts

Specified, funded andregulated

Budgetary unitB

AutonomisedunitA

CorporatizedunitC

PrivatizedunitP

Alternative interpretations

Contract incompleteness and opportunism Are strengthened incentives compatible with

public objectives for hospitals? Access for patients who do not offer surplus

generating possibilities? Quality

Hospital objective function Alternative possibilities to surplus – what

implications for hospital behaviour?Are market forces allowed to work?

Hospital reform in the UK (1991)

Hospitals could apply for trust status Board of directors Determine management structure and profile of services

(with some provisos) Directly accountable to centre Employ own staff and set employment terms and

conditions Income determined by contracts with health authorities,

GP fundholders and private sector (also introduced for non-trust hospitals)

Retain surplus for following year Constraints on prices, and borrowing on capital markets

Trend in cost per episode by hospital group

-150

-100

-50

0

50

100

150

200

1991/2 1992/3 1993/4

1st wave trust

2nd wave trust

3rd wave trust

DMU

Source: Soderlund et al., (1997)

Hospitals’ competitive strategies

Competitive pressures not allowed much rein

Environment implied little scope for competition for main DHA contracts

More energetic competition for ‘extra-contractual referrals’, GP fund-holder contracts, pay beds 1990-1 81,366 patients treated in pay-beds 1994-5 99,399 = +22%

Median waiting times to elective admission, 1994-5

Operation

Operation on coronary artery

Excision of breast

Operation on inguinal hernia

Prosthesis of lens

All patients with surgical operation

Pay beds (days)

13

7

13

17

10

All beds days

76

15

86

175

46

Source Williams (1997) in Keene et al., (2001)

Health sector reform in Zambia

Separate policy making and purchasing from service delivery Creation of ‘Central Board of Health’,

implementing agency of MoH Commission services from public and private

(PNFP) tertiary hospital, and district boards. Contracts negotiated each year: set out

services which district and hospital boards commit to provide for catchment population

Position of tertiary hospitals in reformed system

Ministry of Health

Central Board of Health

Autonomous Hospital Boards

Tertiary hospital

Contract

Direct management

‘Implementing agency’ of the MoH

Features of implementation

3 changes of leadership of MoH between 1994 and 2000, stop-go cycle

Structures put in place but not used ‘Block contracts’ did not evolve Difficulties in de-linking staff from PSC (but some direct

recruitment) Direct interference by MoH continued

Background to reform implementation was economic decline and shrinking resource availability for health sector

Financing

Cost sharing fees for essential package, cost recovery fees for additional services

Package for tertiary hospitals only defined 2001

Interpretation: cost recovery fees for ‘high-cost’, ‘fast-track’, ‘private’ wards and clinics

Zambian hospitals could not gain by competing for ordinary patients – this became their main strategy

Implications of financing strategies

Major use of increased managerial discretion at hospital level

‘High-cost’ fees quite considerable when ‘bundled’ for an episode of care eg. Cerebral malaria, adult – price bundle = Kw

11,000, low cost; 294,980, high cost at one government hospital

What are the implications of these fees for hospital behaviour and the services received by low cost users?

Cost and revenue by ward (1998, Kwacha)

Unit cost % cost = staff

Unit revenue Revenue: cost ratio

High cost wardsMed, Surg, MMed, Surg, FPaediatricsMaternityLow cost wardsMedicalSurgicalPaediatricsMaternity

31,46932,43224,346113,885

16,68111,5799,92185,162

56554969

71656271

12,46112,4629,86115,181

358921,101

0.40.380.410.13

0.000.000.000.24

NB DATA PROVISIONAL

Quality differences between the two services extend to clinical QoC components

05

1015202530354045

HIGH C

OST WARDS A B C D

LOW

COST W

ARDS E F G H

For example: % Items purchased from the list of drugs prescribedPublic hospital X

Clinical staff presence per patient on ward, hospital X.

0123456789

HIGH C

OST WARDS A B C D

LOW

COST W

ARDS E F G H

Clinical staff presence per patient on ward, Hospital Y

0

0.5

1

1.5

2

2.5

3

Indonesia

Major objective of hospital autonomy programme (‘Swadana’) was to encourage hospitals to recover costs

Hospitals granted Swadana status in waves Autonomy circumscribed, but less so than in Zambia Hospital directors have greater control over own-

generated resources Hospitals set fees except for ‘class III’ beds (for the

poor) Hospital can determine service pattern subject to class

III beds constituting a minimum of 50% Favourable financial environment – subsidies increasing

throughout period

Own revenue as % total income

Source: Bossert et al. Hospital autonomy in Indonesia, 1996

0

10

20

30

40

50

60

70

80

90

1991/2 1992/3 1993/4 1994/5 1995/6

RSUP Persahabatan

RSU Tegalyoso

RSU Pasar Rebo

RSUP dr. Kariadi

RUS Tangerang

Trends in fee levels: RSUP dr. Kariadi

Source: Bossert et al. Hospital autonomy in Indonesia, 1996

0

10

20

30

40

50

60

70

80

90

100

1991/2 1992/3 1993/4 1994/5 1995/6

VIP

Class I

Class II

Class III a

Class III b

Numbers of Class III beds

Source: Bossert et al. Hospital autonomy in Indonesia, 1996

% change 1991-1996

-25 -20 -15 -10 -5 0 5 10 15 20

RSUP Persahabatan

RSU Tegalyoso

RSU Pasar Rebo

RSUP dr. Kariadi

RUS Tangerang

Vertical managementDistrict management Provincial management

Room charges per unit cost per class

Hospital

RSUP Persahabatan

RSU Tegalyoso

Small private hospital

ClassVIP

1.11

0.66

3.58

I

0.49

0.90

2.21

II

0.31

0.94

1.86

IIIa

0.20

1.57

0.52

IIIb

0.10

0.63

-

Source: Bossert et al., 1996

Efficiency effects?

Bossert et al.: no marked change in LOS and BOR or

differences Swadana, non-Swadana Management systems deemed to have

improved Physician absenteeism reduced due to

incentive paymentsLieberman and Alkatiri:

Similar conclusions Also, increases in BOR in both types of hospital

Colombia

Law 100, 1993 – framework for national health insurance system Contributory regime: contribution = 12% income Subsidised regime: reduced rate on sliding scale for

those judged unable to pay Cross-subsidy from contributory to subsidised

regimeAll insured entitled to package of care defined

separately for contributory and subsidised members

Insurance market in Colombia

Insurance regulation liberalised – market opened to ‘EPS’ organisations to sell insurance packages and contract with networks of provider institutions

‘Equalisation fund’: Each EPS collects according to the national schedule, remits to equalisation fund and receives back standard sum per patient =1X for each member of compulsory regime =0.5X for each member of subsidised regime

Changed role of hospitals

Previously directly managed by state Secretariats of Health or mandatory insurance agencies (eg. CISS), or private

Now providing services on the basis of contracts with EPSs, and while non-insurance persists, state Secretariats of Health

Key differences from Zambia and Indonesia

Avoids user fees at the point of use No ‘two-tier’ or multi-tier charging systems

Redistributes entitlements to hospital services by enforcing cross-subsidies within insurance system

Remaining inequities in the differences between contributory and subsidised packages of care

Background to reform has been considerably increased health funding (cf. Zambia)

Colombian reforms – evidence of impact

Study of Bogota hospitals: Admission rates

Source: McPake et al. Is the Colombian health system reform improving the performance of public hospitals in Bogota? 2002

0

50

100

150

200Hospital 1

Hospital 2

Hospital 3

Hospital 4

Hospital 5

Bogota hospitals: bed occupancy rates

020406080

100120140

1990

1991

1992

1993

1994

1995

1996

1997

1998

Hospital 1

Hospital 2

Hospital 3

Hospital 4

Hospital 5

Source: McPake et al. Is the Colombian health system reform improving the performance of public hospitals in Bogota? 2002

Some evidence of growing activity and productivity

No evidence of falling patient satisfaction or quality

No evidence or expressed concerns about equity impacts

Comparison of World Bank and ECLAC data 1992 and 1997 indicates slightly increased progressivity of government subsidy over period

Equity?

Jaramillo (2002) Hospitals increased coverage from 35% of population in 1990 to 63% (MoH data, no basis given)

0

5

10

15

20

25

30

1 2 3 4 5

Quintile

% s

ha

re

Weighted average, 1992

1997

Source: 1992: Molina et al. in (1993) World Bank (1993) and ECLAC (1997)

Incidence of public expenditure 1992 and 1997

Conclusions for market forces model

Limited scope for optimism with respect to equity gains

Clearer evidence that incomplete contracts carry risks of disadvantage for particular patients

Gap for political rather than market forces to dominate – in UK and Zambia, constraints to competition may have exacerbated equity effects – ordinary patients’ business can be taken for granted

Colombian model may protect equity but may not be feasible in low income countries

Indications for policy

Recognise that impact depends on context and policy detail

Role to be played by competition needs to be clear and enabled in context of realistic political analysis

Incomplete contracts: increase completeness; apply regulation external to contracts; allow perverse incentives to prevail?