Primary Care Provider Supply and Patient Outcomes: A Cluster Randomized Trial in Nigeria

How to cite

Okeke, Edward. 2017. "Primary Care Provider Supply and Patient Outcomes: A Cluster Randomized Trial in Nigeria." AEA RCT Registry. AEARCTR-0001958. https://doi.org/10.1257/rct.1958-2.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Edward Okeke*

* Primary investigator

Registry ID
AEARCTR-0001958
DOI
https://doi.org/10.1257/rct.1958-2.1
First registered
Last updated
Status
Completed
Trial dates
2017-03-01 – 2022-06-30
Countries
Nigeria (Africa)
JEL codes
I12, J24, O15
Registry record
http://www.socialscienceregistry.org/trials/1958

Abstract

Despite a multitude of programs aimed at improving supply of primary health care providers in underserved communities, there is little convincing micro-evidence about the relationship between primary care provider supply and outcomes. In this study we circumvent thorny identification issues by randomly assigning primary care providers to health facilities and communities. 180 primary health care facilities in Nigeria are participating in this study. Two-thirds of these facilities will be randomly assigned an extra primary care provider (in addition to existing providers) who will practice there for a year. The remaining one-third will maintain the status quo and serve as a control group. We further examine whether the 'quality' of the assigned provider matters for outcomes. A relevant and easily observable proxy of quality is whether or not the provider is a doctor. Within the group of facilities assigned to receive an additional provider, we randomly assign one-half to receive a doctor while the other half receive a mid-level provider similar in skill level to existing providers in the facility. Within this trial, we embed a maternal conditional cash transfer intervention in which pregnant women can receive up to $16 if they attend antenatal and postnatal care, and deliver in the study health facility. Approximately half of participating women are assigned to this cash transfer arm while the other half are assigned to a control arm. We study the effect of the cash transfer on maternal health-seeking behavior and on maternal and newborn outcomes.

Intervention

1. Primary Care Provider (PCP) Treatment: Selected primary health care (PHC) facilities will be randomly assigned an additional primary care provider (either a doctor or a mid-level provider)

2. Maternal conditional cash transfer (CCT) Treatment: Selected pregnant women will receive a cash transfer conditional on attending antenatal care, delivering in the health facility, and attending postnatal care

Experimental design

This is a cluster-randomized trial. 120 (out of 180) participating primary health care (PHC) facilities will be randomly assigned an additional primary care provider (60 will receive a doctor and 60 will receive a mid-level provider). The remaining 60 facilities will receive no additional providers and will serve as a control group. Within each facility catchment area, approximately half of the enrolled pregnant women will be randomly assigned to receive a cash transfer conditional on attending antenatal and postnatal care, and delivering in the study facility.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Randomization takes place at the cluster level. Participating primary health care facilities will be assigned to either the doctor arm, the mid-level provider arm, or a control arm with stratification by location. Within facility catchment areas we will randomly sample 10-15 enumeration areas and randomly assign half to either the CCT arm or the control arm.
Clusters
Provider intervention: 180 primary health care facilities CCT intervention: Approximately 1800 Enumeration Areas
Observations
2,700 households (15 per catchment area) and approximately 10,000 pregnant women (approximately 60 per catchment area)
Arms
Primary Care Provider: 60 facilities - doctor, 60 facilities - mid-level provider, and 60 facilities - control Maternal CCT: Approximately 5,000 women - CCT, and 5,000 women - control
Primary outcomes
Health care utilization, quality of care, physical health score, child weight and height, mortality and morbidity

Cited by

No citation data available for this registration yet.

Keywords