literature review: results-based financing in maternal and neonatal health care

Results Based Financing of Maternal and Neonatal Health Care

in Low- and Lower-middle-Income

Countries

Presentation of the literature review Expert talk Eschborn – 12 december 2012

Anna Gorter, Por Ir, Bruno Meessen

Outline of the Presentation

• General introduction• Methods• Findings of 14 reviews• Potential negative or unintended side-effects of

RBF• Supply and demand side barriers to access of

maternal and neonatal care • Findings of 70 individual studies• Summary of findings

Why did RBF develop in health

Because:• current service provision does not meet public

expectations, huge gaps, poor do not receive basic health services, catastrophic health costs..

• Frustration with lack of results (among governments, donors, services providers, and clients alike)

• One of the options developed: linking payments to results

A wide range of approaches

• Different objectives and expected results (from narrow targets to broad transformation health system)

• Different reward recipients (public, private providers, clients, district or provincial health offices)

• Type and magnitude of rewards

• Proportion of financing paid for results compared to rest of funding

• Different ways of measuring (indicators)

Development RBF approaches• Vouchers (started early 60ties)• Conditional cash transfers – CCT (90ties)• Performance based contracting – PBC (late 90ties)• Health equity funds - HEF (since 2000)• Performance based financing – PBF (since 2000)• Results Based Budgeting – RBB (since 2000)• Combinations:– vouchers and CCTs (Cambodia, Bangladesh)– RBB and CCTs (Nepal)

• Distinction can be rather artificial– E.g. between PBF and PBC

Demand-side and Supply-side RBF and their effect on providers

• Demand-side Money goes to the client (i.e. conditional cash transfers or vouchers)

• Vouchers: “money follows the client” resulting in a strong effect on provider side

• Supply-side Money goes to the provider, but is often linked to number of clients, and hence has an effect on the demand side

Supply-side with a demand-side component

(focus on provider)

Government/donor/health

insurance entity to agencies/facilities “Contracting-out”

Results-Based Financing (RBF)

1

Performance-Based

Contracting (PBC)

Performance-Based

Financing (PBF)

Results-Based

Budgeting (RBB)

Demand-side with a supply-

side component (focus on

provider and client)

Vouchers and Health Equity Funds (HEF)

Demand-side with no supply-side component (focus on client)

Conditional Cash Transfers

(CCT)

2 3 4 5Government/

donor to public or private (not-for-profit) facilities

“Contracting-in”

Government to all MOH

administrative levels and public health facilities

Government / donor to clients and providers

Government / donor to clients

Difference in impact on behaviour

Results Based Financing ApproachesInfluence on

provider behaviour

Influence on client

behaviour

Supply-side, with a demand-side component

Performance-Based Contracting (PBC)

XX X

Performance-Based Financing (PBF)

XX X

Results-Based Budgeting (RBB)

XX X

Demand-side with a supply-side component

Health Equity Fund X XX

Vouchers X XXX

Demand-side, no supply-side component

Conditional Cash Transfers (CCT)

-- XXXX

Methods (1)

• Objective: – compile evidence on RBF of MNH care in LLMICs that will

help program managers and partners answer relevant questions for programming of GDC in health

– Specifically look at RBF programs that focus on providers or have a strong supply-side component

• Inclusion and exclusion criteria:– Relevant supply-side RBF approaches were included, such as

PBF, PBC and RBB – Demand-side RBF approach with strong effect on the

supply-side (vouchers, but not HEF)– Excluded CCTs, vouchers for bednets etc.

• 14 review papers and 70 individual studies

Methods (2)• Not a Cochrane type of review• But extensive review using techniques of a

systematic review such as – pre-defined evaluation criteria, – evaluation of studies` methodological quality

• Less rigorous but this precisely allowed to consider a greater number of papers reflecting the actual state of research efforts implemented so far

• However the rigorousness of evaluation technique was scored; and this was taken into account in the final analysis

3 outcome categories

• Quantity of services provided / number of services utilised

• Quality of the services and satisfaction by clients

• Targeting of the services / equity among clients

Scoring of strength of evidence

• Very low: e.g. descriptive study using stakeholder interviews and no before and after comparison with or without a control.

• Low: comparison of data obtained before and after the intervention, but no control

• Medium: comparison before-after with control or other sophisticated design controlling for confounding factors

• High: very good study design with rigorous control of confounding factors

Conclusion categories

• Robust evidence: if 4 or more rigorous studies1 found a positive effect, and none a negative effect

• Modest evidence: if 2 or 3 rigorous studies found a positive effect, and none a negative effect

• Insufficient evidence: 0 or 1 rigorous study found positive effect or 1 or more studies a negative effect

• Conflicting evidence: if 2 or more rigorous studies had findings in opposite directions

• No effect: if more than half of the rigorous studies found no effect

1. Studies with strenght of evidence being medium or high

Results of review of 14 review papers (1)

• Strength of evidence: 6 low, 6 medium, 2 high• wide range of RBF approaches being piloted/scaled • few robust studies from LLMICs• RBF can make a difference in terms of utilisation and

coverage of those health services which are incentivised

• evidence on the effects on service quality and maternal health outcomes is limited

• anecdotal evidence suggests potential undesirable effects, such as motivating unintended behaviours, gaming or fraud

Results review of 14 review papers (2)

• If carefully designed and implemented, RBF can complement other interventions to address supply and demand barriers to effective MNH

• However not well documented: – efficiency or cost/effectiveness – the long-term effect of RBF on providers’ behaviours

and sustainability• research will be needed to disentangle positive

and negative effects of RBF in order to analyse the overall impact on the health system

List of potential negative or unintended side-effects of RBF (1)

• focus on ‘contracted’ indicators can lead to:– crowding out of other services, adverse selection of patients,

focus on quantity rather than quality, over-servicing and moral hazard, cherry-picking / cream-skimming

• fraud or abuse: – 'gaming', corruption: falsification of documents,

counterfeiting of vouchers, collusion between providers and voucher bearers or voucher distributors, bribery and kickbacks to verification agencies or voucher management agencies, demoralisation

List of potential negative or unintended side-effects of RBF (2)

• motivating unintended behaviours including distortions: – ignoring important tasks that are not rewarded, irrational

behaviour to fulfil requirements, paper work instead of clinical work, bureaucratisation

• undermining goals and motives– crowding-out intrinsic motivation, unsustainable improvement of

services, dependency on financial incentives• creating inequity

– increasing inequity by rewarding providers and facilities that are in better position to reach targets, widening the resource gap between rich and poor

Findings from the 70 studies

Characteristics PBC PBF Voucher

RBB Total

Countries with RBF 7 8 11 4 25

Programs studied: 7 8 18 4 37

Number of studies 11 18 33 8 70

Strength of evidence Very low Low Medium High

-551

4

1031

4

13142

-71-

835234

Objectives, type of services, for whom and where

PBCN=7

PBFN=8

Voucher

N=18RBBN=4

TotalN=37

Overarching objectives programs Reduce maternal/neonatal mortality (MNM) No MNM, focus on other aspects SRH/child Increase quality and use of essential service

package with focus on MNCH Increase quality and use of essential service

package with no particular focus MNCH

--

6 1

--

7 1

135

- -

4-

- -

175

13 2

Target groups: Poor Specific: sex workers, adolescents All: poor and non-poor

7--

7-1

1224

3-1

2926

Scale: Particular geographical area(s) National scale (and state level in India)

6 1

26

135

-4

21 16

Context programsPBCN=7

PBFN=8

VoucherN=18

RBBN=4

TotalN=37

Region Africa Asia Latin America and

Caribbean

142

62-

3

123

22-

12205

Ownership Donor-initiated Gov.-initiated Donor-initiated, scaled

by Gov.

421

1-7

864

-4-

131014

Where implemented, who initiated

List of MNCH interventions

PBCN=7

PBFN=8

Voucher

N=18

RBBN=4

TotalN=37

Adolescents & pre-pregnancyFamily planning 4 6 6 - 16

Prevent and manage STI - 5 4 - 9

PregnancyAppropriate ANC package and other relevant interventions 7 7 13 - 27

ChildbirthInstitutional or skilled normal delivery, referral of complicated deliveries 7 7 13 4 31

PostnatalPostnatal check and care of mother and child 7 7 13 4 31

Supply side barriers to access of maternal and neonatal care

• Availability/geographical accessibility:– Location, unqualified health workers, no 24/7,

waiting times, equipment, drugs, late/no referral• Acceptability – Staff interpersonal skills, including trust, inability

for patients to know prices beforehand• Affordability– Costs services, informal payments

Demand side barriers to access of maternal and neonatal care

• Availability/geographical accessibility:– Distance, availability of transport, information on

health care services/providers, awareness of services, demand for services

• Acceptability – Households’ expectations, low self-esteem and little

assertiveness, community and cultural preferences, stigma, lack of health awareness

• Affordability– Household resources and willingness to pay,

opportunity costs, cash flow within society

How does RBF address supply barriers (1)

• Supply side availability (all schemes): – e.g. waiting time, readiness of the facility to provide

services (availability of drugs, supplies, equipment), and improved referral

• Supply side acceptability (all schemes): – e.g. staff interpersonal skills

• Supply side affordability: – subsidizing fees (vouchers, RBB), – indirectly incentivising providers, regulating service prices,

controlling informal payments (some PBF and PBC schemes)

– Several PBF accompanied by abolishment of user fees

How does RBF address demand barriers (2)• Demand side availability (most PBC,PBF, vouchers):

– e.g. mostly through the provision of information on health care services and providers (voucher distribution, outreach)

• Demand side acceptability (most PBC,PBF, vouchers): – e.g. outreach activities increase health awareness, help overcome

cultural barriers. Vouchers empower the holder improving self-esteem;

• Demand side affordability (only 12 schemes): – 1 RBB through a CCT, 1 PBC through community work, 1 PBF

through organisation of services such that opportunity costs were reduced, 9 vouchers paid for transport and food costs and 1 of these also provided a CCT

Number of individual RBF programs addressing particular barriers

Barrier PBCN=7

PBFN=8

Voucher

N=18

RBBN=4

TotalN=37

Supply-side barriers addressed:

Availability 7 8 18 4 37

Acceptability 7 8 18 4 37

Affordability 3 2 18 4 27

Demand-side barriers addressed:

Availability 5 7 18 1 31

Acceptability 5 7 18 1 31

Affordability 1 1 9 1 12

PBC: 7 programs, 11 research papers

Type of Effect # of programs

N=7

# ofstudies

N=11

% with effect (of number of studies investigated the issue)

% with positive effect (of those with effect)

% with rigorous study design (of those which investigated issue and had positive effect)

# of studies with rigorous design and positive

Quantity/ utilisation

7 8 75% (6)

100% 50% 3

Quality / satisfaction

2 2 100% (2)

100% 50% 1

Equity / targeting 2 3 100% (3)

100% 67% 2

PBF: 8 programs, 18 research papersType of Effect # of

programs

N=8

# of studies

N=18






6 14 64% (9)

100% 11% 1


7 8 100% (8)

100% 50% 4

Equity / targeting 3 4 100% (4)

75% 33% 1

Vouchers: 18 programs, 31 research papers


N=18

# of studies

N=33






15 22 100% 100% 45% 10


11 16 100% 100% 50% 8

Equity / targeting

13 17 100% 100% 53% 9

RBB, 4 programs, 8 research papers


N=4

# of studies

N=8






4 7 100% 100% 14% 1


1 1 0% - 0% -

Equity / targeting 2 2 100% 50% 0% 0

Summary evidence on effect of RBF

Type of EffectRobust evidence>3 studies

Modest evidence2-3 studies

Insufficient evidence<2 studies or no effect

# rigorous studies positive effect

PBC Quantity/utilisation X 3Quality / satisfaction X 1Equity / targeting X 2PBF Quantity/utilisation X 1Quality / satisfaction X 4Equity / targeting X 1Vouchers Quantity/utilisation X 10Quality / satisfaction X 8Equity / targeting X 9RBB Quantity/utilisation X 1Quality / satisfaction X -Equity / targeting X 0

Summary findings: our assessment (1)• Maternal and neonatal health services have been

a major area of application of the RBF logic• High creativity in addressing barriers,

implementers adapt the RBF strategy to local bottlenecks and priorities

• Effectiveness: • there is more evidence for some RBF strategies than

others, e.g. for vouchers• a lot of research being implemented, esp. on PBF

• Not well investigated: negative and unintended side-effects of RBF, sustainability

Summary findings: our assessment (2)

• Little attention to efficiency (cost/effectiveness)• Efficiency of a RBF scheme depends on the

design, the funding, the implementation…• and heterogeneity across schemes (objectives,

experiences, contexts, combinations of RBF strategies) will not ease the synthesis of the evidence

• also schemes are improved while implemented• RBF is not a goal per se – the evidence should be

put in the broader context of road to UHC

Thank you

literature review: results-based financing in maternal and neonatal health care

Health & Medicine