Effective Delivery of Health Products in Ghana

Our sample consists of 72 health facilities selected for inclusion based on a census conducted of all of the public and private health facilities (over 300) in one region of Ghana, which for confidentiality purposes shall remain unnamed.
Inclusion criteria for health centers in the study were: (1) having an antenatal care clinic (ANC); (2) being rural; (3) having no other healthcare facilities within a 2 km radius, no hospitals within a 5 km radius, and not more than one other ANC within a 5 km radius (this was to keep subsidy costs manageable by limiting the potential increase in ANC attendance in response to the program); (4) having no free LLIN program currently in place (very few had); (5) having at least two stores within a 2 km radius willing to participate in a voucher scheme whereby health centers distribute vouchers redeemable at the store (only 6% of clinics were excluded by this criterion); and, (6) being accessible for net deliveries (less than 2% were inaccessible).
We asked selected health facilities to participate in a bed net distribution program called "SALI" and sponsored by the NGO JPAL. The bednet distribution program targeted pregnant seeking prenatal care. Each prenatal care registrant was eligible to one free long-lasting insecticide treated net.
We randomized the following features of the distribution program:
1. Distribution mechanism: whether the subsidized product was distributed directly through health facilities, or indirectly through a voucher scheme wherein the clinic distributes vouchers that could be redeemed at local stores (Direct vs. Voucher). We also randomized whether vouchers could be redeemed in 1 or 2 stores (1-store vs. 2-store)
2. Staff monitoring: whether the health works were told the distribution program would be monitored through audits (Audit threat vs. No Audit threat)
3. Incentives: whether, for Direct Distribution clinics only, the health worker received compensation for implementing the distribution (Compensation vs. No Compensation)
4. Stock levels: whether, for Direct clinics only, the clinic receives a high or low level of net stock (High stock vs. Low stock)

2011-10-02

2012-04-09

- coverage among eligible recipients
- payment requests from eligible recipients for nets that should have been free
- incidence of leakage of nets/vouchers from the health facilities to ineligible populations
- stockouts at facilities

After stratifying the 72 facilities (6 strata: size based on number of ANC visits, not registrants, and location: close to border or not, versus remote or not), they were randomized into 12 equally-sized groups:

V1 -voucher with 1 shop
V2 -voucher with 2 shops
VA1 - voucher with audits & 1 shop
VA2 - voucher with audits & 2 shops
D -direct distribution
DP -direct distribution with payment for staff
DA -direct distribution with audits
DAP -direct distribution with audits and payment
DAPL -direct distribution with audits and payment and large -stock
DPL -direct distribution with payment and large -stock
DAL -direct distribution with audits and large -stock
DL -direct distribution with large-stock

Randomization by a pre-set algorithm (done in office by a computer).

The randomization was conducted at the facility level.

No

72 facilities

72 facilities

24 facilities sampled for indirect distribution ("voucher"), 48 facilities sampled for direct distribution
36 facilities sampled for audit treatment, 36 not sampled for audit treatment
Among 48 facilities sampled for direct distribution: 24 sampled for bonus flat pay
Among 48 facilities sampled for direct distribution: 24 sampled for large delivery
Among 24 facilities sampled for voucher: half sampled to have two shops were vouchers can be redeemed

EDHEP Analysis Plan Final.pdf

Distributing subsidized health products through existing health infrastructure could sub- stantially and cost-effectively improve health in sub-Saharan Africa. There is, however, widespread concern that poor governance – in particular, limited health worker accountability – seriously undermines the effectiveness of subsidy programs. We audit targeted bednet distribution pro- grams to quantify the extent of agency problems. We find that around 80% of the eligible receive the subsidy as intended, and up to 15% of subsidies are leaked to ineligible people. Supplementing the program with simple financial or monitoring incentives for health workers does not improve performance further and is thus not cost-effective in this context.

Dizon-Ross, Rebecca, Pascaline Dupas, and Jonathan Robinson. "Governance and the Effectiveness of Public Health Subsidies: Evidence from Ghana, Kenya and Uganda." Working Paper, Janauary 2017

https://www.povertyactionlab.org/sites/default/files/publications/444_Governance%26Effectiveness_PublicHealthSubsidies_Jan2017.pdf