Impact Evaluation of the Demand-Side Financing Program for Maternal Health in Bangladesh

1
Impact Evaluation of the Demand-Side Financing
Program for Maternal Health in Bangladesh

• Laurel Hatt, Ha Nguyen, Nancy Sloan, Sara Miner,
  Obiko Magvanjav, Asha Sharma, Jamil Chowdhury,
  Rezwana Chowdhury, Dipika Paul, Mursaleena Islam,
  and Hong Wang

2
Bangladesh context

• High maternal mortality ratio (MMR) 320 maternal
  deaths per 100,000 live births
• 85 of deliveries at home, few with medical
  provider
• Millennium Development Goal 5 Cut MMR to 143 by
  2015

3
Bangladesh pilot DSF program

• Aims to rapidly increase utilization of maternal
  health (MH) services via
• Vouchers for free antenatal (ANC), delivery,
  emergency, and postnatal care (PNC), and for
  transportation
• Cash transfers (Tk. 2000 30) and gift bags if
  women deliver with skilled birth attendant at
  home or in facility
• Cash incentives for providers/field workers for
  registering women and providing MH services
  seed fund for facilities
• Eligibility local residents only pregnant for
  1st or 2nd time used family planning prior to
  2nd pregnancy means testing in some areas
• Started in 2007 in 21 subdistricts now in 45

4
Evaluation objectives

• Evaluate DSF program impacts on
• Use of skilled providers at delivery, C-section,
  ANC, and PNC
• Out-of-pocket expenditures on maternal health
  services
• Provider skills, knowledge facility quality
• Assess operations of the DSF program

ANC visit by voucher recipient
5
Methods

• Lack of useable baseline data
• 21 DSF subdistricts matched with 21 control
  subdistricts
• Data sources
• Household survey (N2,208) of women who delivered
  in prior six months
• 10 focus group discussions with women
• 16 facility assessments
• 51 provider interviews
• 90 key informant interviews at subdistrict and
  central levels
• Administrative reports
• Descriptive and multivariate analyses

6
Findings

• New mothers line up to receive cash stipends and
  gift boxes

7
receiving ANC significantly higher in voucher
areas
8
delivering with a skilled provider
significantly higher in voucher areas
9
More home deliveries attended by qualified birth
attendants
10
with PNC visit significantly higher in voucher
areas
11
OOP expenditure significantly lower in voucher
areas
12
Multivariate analyses confirm impact

• Cross-sectional regression
• Compared to living in a control subdistrict,
  residence in a DSF subdistrict increased the
  likelihood of delivery with skilled provider by
  42 to 50 percentage points
• Using pre- and post-DSF birth history data
  from recently delivered women
• Difference-in-differences and fixed effects
  analyses also confirmed program impacts

13
Conclusions Strong DSF impacts

• Women in DSF areas were significantly more likely
  to have at least 3 ANC visits (55) than women in
  the control areas (34).
• The DSF program is significantly associated with
  higher rates of delivery with qualified providers
  -- more than twice as high in DSF (64) as in
  control subdistricts (27).
• No significant differences in C-section rates
  (10 vs. 9)
• Women in DSF areas spent significantly less on
  pregnancy delivery care than women in control
  areas

14
Challenges remain

• Big upsurge in patient volumes is overwhelming
• Shortages of human resources, supplies and
  medicines
• Limited targeting
• Leakage to higher-parity women (11 of
  recipients)
• Little income targeting in means-tested areas
• Delays in funds disbursements are a serious
  problem
• Impact of incentives (to providers and consumers)
  diminished if not connected with desired
  behaviors
• Cost and sustainability concerns

15
Acknowledgements

• GTZ funded the original study, which was
  implemented by Abt Associates Inc. (USA) and its
  partner RTM International (Bangladesh)
• Special thanks to the Bangladesh Ministry of
  Health and Family Welfare, Health Economics Unit,
  WHO and DSF program office for their support
• http//www.abtassociates.com/HSRsymposium