NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
Can a Decision-Making Nudge Improve Birth Outcomes?: A Pilot Randomized Experiment in Nairobi
Initial registration date
December 01, 2014
December 01, 2014 2:10 PM EST
Harvard School of Public Health
Other Primary Investigator(s)
Harvard School of Public Health
Additional Trial Information
Expecting women and their partners in urban Africa face a wide variety of choices of where to deliver babies. Many public and private delivery options exist, but reliable information about the quality or cost of facilities is lacking. Women do not always coordinate with their families to save enough money for delivery. Furthermore, labor often begins at an unplanned time or place, leaving women without appropriate transport, supplies, and necessary childcare for other children. These factors often lead to late arrival at whatever facility is closest and affordable or, in some cases, to home delivery. Preliminary research in Nairobi revealed that nearly half the time women do not end up delivering where they intended. The central idea behind our project is that behavioral barriers to deciding where to deliver represent a significant component of delays in reaching timely, high quality maternal and newborn health care. We will conduct a pilot study in Nairobi to test simple interventions addressing behavioral and structural barriers to arriving safely at a high quality delivery facility. We will relieve behavioral barriers by encouraging pre-commitments about where to deliver and structural barriers by providing cash transfers to help offset the costs of delivery (including transport). Our hypothesis is that encouraging expecting women to pre-commit to a delivery location will reduce delays in reaching the facility and increase the quality of the facility used.
Cohen, Jessica and Margaret McConnell. 2014. "Can a Decision-Making Nudge Improve Birth Outcomes?: A Pilot Randomized Experiment in Nairobi." AEA RCT Registry. December 01.
This project is a pilot randomized control trial to be conducted in peri-urban Nairobi, Kenya. In our experiment, women in their 5th-7th month of pregnancy will be randomized into one of three treatment arms: 1) a treatment group that receives an unconditional cash transfer at midline, 2) a treatment group that receives an unconditional cash transfer at midline and a conditional cash transfer during endline if they go to the facility they pre-committed to during their 8th month of pregnancy/midline visit, or 3) a control group that receives standard care. We will survey all study participants at baseline (during their 5-7th month of pregnancy), mid-line (during their 8th month of pregnancy) and endline (2-4 weeks after delivery). Recruitment into the study will involve several methods including listings from community health workers, snowball sampling, community information days and fliers.
Intervention Start Date
Intervention End Date
Primary Outcomes (end points)
Category 1: Intentions = Outcomes?
1.1a Yes/No: Did respondent deliver at facility that she said was most likely at midline?
1.1b Yes/No: Did respondent deliver at facility that she said was most wanted at midline?
“Attempt to deliver” means woman called or arrived at facility for delivery (but may have been referred elsewhere)
1.1c Yes/No: Did respondent attempt to deliver at facility that she said was most likely at midline?
1.1c Yes/No: Did respondent attempt to deliver at facility that she said was most wanted at midline?
Category 2: Birth Planning Primary Outcomes
2.1 Days before EDD that final decision about where to deliver was made 2.2 Number of facilities ruled out between baseline and midline 2.3 Days before EDD that resp began to think about where they wanted to deliver.
2.4 Days before EDD that first talked to spouse about delivery location 2.5 How much had household saved toward delivery (at endline)
2.6 Yes/No: has transport plan for delivery at midline
2.7 Yes/No: has childcare plan for delivery at midline (among those with other children) Secondary Outcomes
2.8 Total saved at midline (spouse + self)
2.9 Accuracy of perceptions of cost of delivery at each facility at midline
2.10 Accuracy of perceptions of transport time at midline
2.11 Amount plan to save for delivery at midline
2.12 Yes/No: Talked to spouse about where to deliver? (midline)
2.13 To what extent did you agree with spouse about where delivering on day of delivery (endline)
2.14 Yes/no: Woman is part of decision about where to deliver (i.e. answers that either she alone or she and partner jointly made decision about where to deliver)
Category 3: Delays
Primary Outcomes 3.1 Amount dilated at first exam 3.2 Time between first contractions and arrival at facility
3.3 Time waited between arriving at facility and talking to a health professional
3.4 Time between arrival and first exam
3.6 Time between contractions at time of departure to facility
3.7 Yes/No: woman feels she arrived when she wanted
3.8 Any complication during delivery 3.9 Any complication for mom postpartum
Category 4: Quality of care (Note: intervention could affect quality of ANC and of delivery facility)
4.1 Number of ANC visits at facility where respondent ultimately delivered
4.2 Days before EDD that high risk was identified (if ever identified as high risk after baseline survey)
4.3 Yes/No: Ever identified as high risk (after baseline survey)
4.3a: Yes/No: If identified as high risk after baseline survey, woman knows why
4.4 Yes/No: Did woman have emergency referral (while in labor)?
4.5a Number of blood pressure checks at ANC after baseline
4.5b Number of times had belly measured at ANC after baseline
4.5c Number of times had urine tested at ANC after baseline
4.5d Number of times had respiration/breathing measured at ANC after baseline
4.5e Number of times had fetal heart rate measured at ANC after baseline
4.5f Number of times weighed at ANC after baseline
4.5g Yes/No: Ever got Tetanus at ANC after baseline
4.5h Yes/No: Ever got iron supplements at ANC after baseline
4.5i Yes/No: Ever got vit A supplements at ANC after baseline
4.5j Yes/No: Ever left ANC with questions about preg that didn’t get answered?
4.5k Yes/No: Ever left ANC with medical advice you didn’t understand?
5. Delivery facility:
5.1 Overall quality on 1/10 scale
5.2 Yes/No: Would you deliver there again?
5.3 Yes/No: Would you recommend this place to friends? 5.4 Index of secondary delivery facility all quality outcomes Secondary Outcomes:
5.5 Yes/No: Running water at facility
5.6 Yes/No: Have own bed on day of delivery
5.7 Satisfaction with experience
5.8 Yes/No: Examiner wore gloves
5.9 Yes/No: Other patient sharing bed
5.10 Yes/No: Open 24 hours a day
5.11 Competence/Knowledge of workers
5.12 Friendliness of workers
5.13 Respect of workers
5.14 Communication of workers
5.15 Cleanliness of facility
5.16 Rating of availability of drugs, supplies, equipment
5.17 Yes/No: Partner allowed in the room
Predictions about heterogeneity:
1. The conditional transfer will be most effective for people who score high on the procrastination scale at baseline.
2. The conditional transfer will be most effective for people who made a last minute decision in previous deliveries.
3. The unconditional transfer will be most effective for women who have low bargaining power at baseline. 4. The unconditional transfer will be most effective for women who have low income. 5. Both treatments will be more effective for women with low medical risk than for women with high medical risk.
Primary Outcomes (explanation)
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
600 women in their 5th-7th month of pregnancy will be randomized into one of three treatment arms (with n=200 in each): 1) a treatment group that receives an unconditional cash transfer at midline, 2) a treatment group that receives an unconditional cash transfer at midline and a conditional cash transfer during endline if they go to the facility they pre-committed to during their 8th month of pregnancy/midline visit, or 3) a control group that receives standard care. We will survey all study participants at baseline (during their 5-7th month of pregnancy), mid-line (during their 8th month of pregnancy) and endline (2-4 weeks after delivery). Recruitment into the study will involve several methods including listings from community health workers, snowball sampling, community information days and fliers.
Experimental Design Details
Random assignment done in office with random number generator
Individual level random assignment. Random assignment stratified by neighborhood and by gestational month at enrollment.
Was the treatment clustered?
Sample size: planned number of clusters
Sample size: planned number of observations
Sample size (or number of clusters) by treatment arms
200 women in control group
200 women in unconditional transfer group
200 women in conditional transfer group
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
INSTITUTIONAL REVIEW BOARDS (IRBs)
Harvard School of Public Health Office of Human Research Adiminstration
IRB Approval Date
IRB Approval Number
Post Trial Information
Is the intervention completed?
Is data collection complete?