Impact Evaluation of Super Abbu: A Speech Based MNCH Platform in Pakistan

University of California (UC) Davis

Status

Terminated

Conditions

Maternal and Child Health

Treatments

Other: Super Abbu

Study type

Interventional

Funder types

Other

Identifiers

NCT04008199

1255578

Details and patient eligibility

About

The investigators have developed, piloted, and propose to expand and to experimentally evaluate a speech-based service that connects expectant fathers to doctors and to each other over a simple phone call. This service, dubbed Super Abbu (Super Dad in English), addresses the challenges faced by existing efforts in several new and important ways:

The service targets fathers. Currently, the entire public health infrastructure in Punjab is geared towards providing information to women.
It supplements LHWs by providing information between LHW visits.
It does so in a manner appropriate for those who are illiterate and do not have smartphones.
It does not require synchronous phone calls. Expectant fathers can leave questions to be answered. And they can access past questions and answers asked by themselves and other users.
Through experience developing, publicizing, and driving adoption of speech-based services, the investigators believe they can overcome the slow adoption rate faced by the Punjab Health Line. Principal investigator Raza previously developed and launched Baang in Punjab, an entertainment speech-based social service, similar to Reddit, which accumulated 42,500 calls by 1,550 users within 71 days of its launch. These users listened to 393,448 speech posts.

Full description

Problem: Pakistan has the highest neonatal mortality rate in the world (44.2 per 1000 live births), accounting for 7% of the world's neonatal deaths, and among the worst maternal mortality rate in the region (178 per 100,000 live births). These indicators remain high despite the best efforts of the government and NGOs to leverage modern practices and technology to improve maternal, newborn, and child health (MNCH). To this end, the province of Punjab, Pakistan has hired 48,000 Lady Health Workers (LHWs), has introduced multiple information communication technology (ICT)-based health monitoring systems, and has launched the Punjab Health Line staffed by 150 doctors 24/7. LHW effectiveness has been limited, however, given the size of Punjab (100+ million people). LHWs are expected to visit nine homes per day and in 40 minutes cover a long syllabus and perform several tasks. Similarly, high deployment costs, low smartphone penetration, and low literacy rates in rural areas have limited the effectiveness of ICT-based interventions. The Punjab Health Line has been ineffective for more mundane reasons-its adoption has been very slow despite extensive publicity.

Research goals and expected outcomes: Super Abbu itself will generate useful information on frequently asked questions for public health professionals. In addition, the investigators seek to evaluate the effectiveness of Super Abbu at reducing maternal and infant mortality rates, and key drivers of these rates, in Punjab, Pakistan. In order to do so, the service must first be expanded to accommodate a new base of users. The investigators will then conduct a careful, large-scale randomized control trial of the service to understand its impacts on drivers of maternal and infant mortality rates, such as the number of antenatal visits, tetanus injections administered before birth, iron supplements consumed, and breastfeeding during neonatal days, and attitudes about vaccination, breastfeeding, and delivery.

Impact on research fields: Current speech-based health services in the developing world either depend on the functionality of a smartphone, involve synchronous contact with a healthcare provider, only target women, are costly, or are any combination of these. Super Abbu pushes the envelope. In addition, a careful evaluation of the service conducted by an interdisciplinary team would impact future research in public health, in computer science/language technologies, as well as in development economics.

Enrollment

276 patients

Sex

All

Ages

18+ years old

Volunteers

Accepts Healthy Volunteers

Inclusion and exclusion criteria

Primary study subjects The target study population are households in Punjab, Pakistan, with a pregnant woman with a male spouse. The researchers will identify households that fit these criteria village-by-village, through community-based Lady Health Workers (LHWs) that, among other things, are charged with knowing which women are pregnant in her catchment area. LHWs are appointed by Punjab's Department of Health at the village level, and are supervised by Lady Health Supervisors (LHSs) at the sub-district level, who are in turn supervised by district coordinators and assistant district coordinators at the district level. The research team has relationships with LHSs and district coordinators in several districts in Punjab through prior work and it is these districts from which the study sample will be selected.

From all of the selected districts, the researchers will draw a random sample of 40 LHWs, stratified on sub-district. The team will then work with each of these 40 LHWs to identify all of the households in their catchment area that meet the study requirements. Note that these LHWs will share location and pregnancy status of households with the researchers, with permission of their supervisors, which the researchers have verified does not break any confidentiality rules within Pakistan, Punjab, and/or Punjab's Department of Health. It is discussed below how the researchers will maintain confidentiality of the research subjects once they are in the sample.

The researchers expect each LHW to have 1000 households in her catchment area on average. It is expected that four percent of households will have a pregnant woman at any given time. This implies that each LHW will have forty pregnant women in her catchment area. The researchers thus expect 1600 pregnant women across the sample catchment areas.

Once all of these 1600 households have been identified across 40 LHW's catchment areas, the researchers will randomly select 20 catchment areas for "treatment" and 20 to be "control". While all 1600 households will then be administered identical baseline and endline surveys, as described below, spouses in households in "treatment" catchments will also be invited to join Super Abbu at the conclusion of their baseline survey. What Super Abbu entails is described below.

Before enrolling households into the study during the baseline survey, the research team will verify that the household does in fact have a pregnant woman with a male spouse. The woman and her spouse will then be enrolled into the study. The only additional criteria will be that the subject is an adult who gives consent, where adult in this setting is 18 years or older.

There will be no other inclusion or exclusion criteria for primary study subjects.

Preliminary study subjects The researchers will conduct a preliminary study with approximately 100 subjects prior to beginning the primary study, to pilot the service and to test different versions of Super Abbu. These participants will be selected following the same procedure as with the primary study subjects, with one exception---they will all be invited to join Super Abbu. They will also be in a geographically distinct location from the primary study participants.

Focus group subjects The research team will also conduct focus groups prior to the primary study. Separate focus groups will be conducted with: gynecologists, LHWs, pregnant mothers, expectant fathers, fathers that were invited to be preliminary study subjects, fathers that were invited to be primary study subjects who were also invited to Super Abbu, and fathers that were invited to be primary study subjects who were not invited to Super Abbu. In all cases, the researchers will only include adults (18 years or older) who give consent. Selection criteria will be that subjects must match the group for which each given focus group is being conducted, with no additional criteria. Groups of 1-10, identified ad-hoc based on referrals, will be invited in-person to join a focus group at a central location. The researchers expect to conduct two focus groups per group, except four for preliminary study subjects, for a total of 16 focus groups and up to 160 participants. Note while the researchers expect discussion to differ based on the background and perspectives of those being focus grouped they plan to ask very similar questions of the group and, as such, only have one consent form for all focus group participants.