Primary Care Provider Supply and Patient Outcomes

Overview

The evidence on how primary care provider supply and skill relates to patient outcomes is limited and inconclusive. The issue of skill is a particularly important one in low-income countries where the shortage of skilled medical professionals has led to greater use of the task-shifting model in which medical tasks are redistributed from highly skilled health workers to mid-level providers who receive less training. In this large-scale cluster-randomized trial, the investigators randomly select primary health care facilities to receive a highly skilled provider (a doctor), a mid-level health provider, or no additional providers (the control group). The investigators study the effect of this intervention on patient outcomes. Embedded within this trial is another experimental intervention in which pregnant women residing in communities served by the primary health care facilities are assigned to receive a cash transfer conditional on using antenatal, delivery and postnatal care. The investigators study the effect of the cash transfer on health care utilization and on maternal and infant outcomes.

180 primary health care facilities in Nigeria are participating in this study. In one arm of this trial, 60 of the participating facilities are randomly assigned a physician (in addition to existing clinic staff); in a second arm, 60 facilities are randomly assigned an additional mid-level provider (of similar training to existing providers); and the third arm consists of facilities that receive no additional providers (the control group). Assigned providers will practice in those facilities for approximately one year. Stratifying by treatment arm, the investigators further randomize 10-15 enumeration areas within each facility catchment (service) area to either a Conditional Cash Transfer (CCT) arm or to a control arm. The cash transfer is intended for currently pregnant women in those enumeration areas and is conditioned upon attending four antenatal care visits, giving birth in the facility, and attending at least one postnatal visit. To measure the effects of the interventions the investigators randomly sample 15 households in each facility catchment area and enroll all currently pregnant women within the study enumeration areas. Women at an advanced stage of pregnancy - third trimester - are excluded because such women will not be exposed to the intervention for the majority of their pregnancy. All household members are interviewed at baseline, midline (6 months later) and endline (12 months later) to collect information about health-seeking behavior and health status. Pregnant women are also interviewed at baseline and approximately 1-3 months after birth to collect information about health care utilization and birth outcomes. Additionally the investigators collect data from the participating health facilities and providers. Using a combination of provider surveys, clinical vignettes, direct clinical observation, and patient exit interviews the investigators examine the effects of the intervention on service delivery and quality.