AEA RC T Registr y

The American Economic Association's registry for randomized controlled trials

The objective of this intervention is to reduce attrition amongst midwives by providing them with retention incentives.

All midwives who are part of this intervention are employed by SURE-P in SURE-P Primary Health Centres (PHC). In order to promote retention, midwives that remain at their duty health facility for increments of three months are eligible to receive a package of retention incentives every three months. The incentives are cumulative, i.e. a midwife cannot qualify for the second three months’ incentive if she has not qualified for the first three months’, and so on.

The implementation of SURE-P is divided in two phases. Phase 1 includes the first 500 PHCs to begin implementing the program, while Phase 2 expands this to a further 500 PHCs. The implementation of SURE-P in both Phase 1 and Phase 2 PHCs continues through the end of 2015.

2013-12-01

2015-12-31

Primary outcomes:

-Whether midwives are still working 9 months after baseline in the same SURE-P PHC where they were working at baseline
-Duration of service since baseline in the same SURE-P PHC where they were working at baseline

Secondary outcomes:
-Maslach Burnout Inventory of Midwives
-Neonatal mortality collected through an extension of the sisterhood method
-Edinburgh maternal depression index
-An index comprising variables related to health care use of pregnant women, mothers and newborns (detailed in pre-analysis plan)
-Midwives’ locus of control, life satisfaction, happiness and trust in institutions and other people
-Midwives' future prospect regarding job and livelihood
-Conflict between midwives and non-incentivised workers; performance of non-incentivised workers.
-Midwives participation in volunteering activities
-Perception of social norms
-Number of midwives available in the health facility
-Midwives feeling of support and appreciation by their employer

See pre-analysis plan

In order to evaluate the relative effectiveness of the different packages of incentives, midwives are randomly assigned to three treatment groups, each of each receives a different incentives package and one control group. Full details are provided in the pre-analysis plan.

Computer-based randomisation

The unit of randomization is the cluster, which is comprised of SURE-P PHCs that share the same referral hospital. There are four PHCs in each cluster.

Yes

500 SURE-P PHCs, i.e. 125 clusters receive the intervention in Phase 1;

An additional 126 clusters receive the intervention in Phase 2.

A target sample size of 2,500 households was planned at baseline. The response rate to the household questionnaire was of 95% (2,384 households) at baseline.

2570 midwives, assuming the same number of midwives in Phase 1 and 2

63 clusters in each treatment arm, and 62 cluster in the control arm.

To detect an increase in retention rate from 0.6 to 0.7 at 80% power and 5% significance, we will need 125 clusters and a minimum of 9 midwives per cluster under the assumption that the intra-cluster correlation is 0.1.