��Updates for Evaluation of Integrated Digital Primary Health Care in Rwanda��Africa Evidence Summit

June 29-30, 2022

University of Rwanda, School of Public Health

University of California, Berkeley

CIIC-HIN

1

Background: Babyl Digital Health Service Delivery

2

• Babyl (digital health service delivery)—local brand name for Babylon, began its clinical operations in Rwanda on Sept 2016, and in March 2020, signed 10-year MoU with GoR.

​

• The Babyl’s health service delivery uses:
• Virtual consultations with nurses and doctors,
• Digital prescriptions, and
• Laboratory tests
• Prescriptions and tests are issued via SMS codes.

​

​

Background: Babyl Digital Health Service Delivery

3

​

​

• By August 2020, Babyl had registered nearly 2 million users; but, half have dropped Babyl services ~ 1,000,000 inactive

​

• A previous study conducted by (Delberg, 2019) suggested further studies: impact of Babyl on quality of care and key health system affects factors that supported Babyl expansion

Research Questions on Quality of Care

Main research question

• What is the difference between conventional care and telemedicine across Rwanda for primary health care?
• What are the levels of care quality?
• Ex: of 100 health centers, how many will provide correct care? 80%? 40%? 100%?
• We will focus on: Adult care for acute malaria, upper respiratory infection (common cold), and diarrhea

​

Secondary research questions: We will measure these through “experiments”

• “Suggestion” – To what extent are providers swayed by patient suggestion?
• “Demanding” – Do providers give patients unnecessary services when requested?
• “Insurance” – Do patients with insurance receive different treatment than patients paying with cash?

4

Research Objective

​

​

This research aims to evaluate effectiveness of Babyl’s digital healthcare services on key patients; it explores key factors that contributed or hindered the adoption and scale-up of Babyl’s service integration in Rwanda.

​

​

5

5/18/21

Specific Objectives

To respond to this general objective, specific objectives are defined as follows:

​

• To document health system factors that supported or hindered the adoption and scale up of Babyl’s digital service in Rwanda
• To evaluate patient utilization of primary health care overtime
• To assess the impact of Babyl digital health services on patient’s quality of care across selected 3 disease domains
• To estimate cost health services to insurance (payer—CBHI) and patients (households)

6

5/18/21

Research Methods

• First objective, we will:�
• Conduct the formative research (key informant interviews and focus group discussions): to assess key drivers, or lack of adoption of Babyl’s digital health services in Rwanda to explore factors that supported the adoption of digital services).

​

• Test encouragement interventions in a random sample: up to 1000 eligible patients, selected from Babyl clients who disconnected health services for the last 3-6 months, and among eligible CBHI members who have never taken up Babyl services.

Research Methods

• Second and third objectives, we will:
• Use administrative data
• June 2019 - the end of the study period: compare health care outcomes, utilization and costs across Babyl and non-Babyl patients.
• Longitudinal data to evaluate the impact of Babyl penetration on changes in utilization, patient mix, and patient outcomes in both conventional health facilities and Babyl. �
• Use vignettes, standardized patients (SP), and patient exit interviews to collect information related to quality of care and costs to patients in both conventional care health facilities and at Babyl.�
• Introduce exogenous variation in Babyl adoption using a large-scale randomized control trial.
• Treatment arm(s) will receive the encouragement intervention(s) compared to a control arm that receives no intervention.
• We will evaluate the impact of the intervention on the take-up and utilization of Babyl and costs.

Site-level Inclusion Criteria

For FGD and KIIs at facility level, we will select randomly a sample of health centers that have / are:

• Low, medium and high utilization rates of Babyl services
• Babyl agent /recruiter or not
• Located in urban and rural areas.

KIIs at central level will target key organizations, institutions, public or private, which participated in the design and implementation of Babyl care services.

​

Encouragement interventions will be implemented in communities served by health centers that meet the following criteria:

• Low/ medium utilization rate of Babyl digital services
• Availability of a Babyl agent/ recruiter
• Rural and urban location

Individual-level inclusion criteria

KIIs respondents include:

• Registered Babyl clients who use digital care services
• Babyl providers, Babyl agents / recruiters at health centers,
• Heads of health centers, Health providers (e.g. lab technicians, pharmacists)
• Institutions: RSSB representative, Ministries representatives, Governmental agencies representatives, Medical professional associations representatives.�

FGD population includes:

• Babyl registered clients who use digital services,
• Babyl registered clients who did not use digital services or stopped using Babyl services,
• Eligible CBHI members who are aware of Babyl services and have not registered for Babyl.�

The quantitative phase population will include:

• Eligible CBHI members who have not registered to Babyl,
• Registered Babyl clients who did not use and stopped using digital services.

Standardized Patient (SP) Cases with Experiments

SP Cases with Experiments

Case 1. Malaria�Demanding experiment

Opening Statement: [Doctor/Nurse], I felt cold with headache and joint pain the last few days, and now I’m worse. I have come to you for help.

Demanding experiment as assigned version adds:

“But doctor, I’m really worried… Is there anything you can give to make this go away faster?” �

At three possible moments when appropriate, the SP assigned this experiment can demand:

• when the provider is writing a prescription or about to dispense drugs,
• when the doctor asks what the patient wants, or
• at the end of the interaction and if the provider hasn’t given something that will make it go away yet, the SP stands up to close the visit (seem as if the SP wants to leave then turns back and says to the doctor in a pleading tone)

Note when / at what time during an SP visit the experiment is different

Case 1. Malaria

Sampling

14

50% Demanding

50% Not Demanding

Case 2. URI

Sampling

15

50% Demanding

50% Not Demanding

Ethical Considerations

• Research does not involve the contact with the participant and there will be no any therapeutic or other clinical or diagnostic interventions. Thus, a waiver of consent is sought.�
• Approval is requested to Rwanda National Ethics committee for the use of administrative data.�
• To maintain confidentiality of patients and prevent inappropriate disclosure of data, the research team will follow the prescriptions provided by the national laws protecting patients’ privacy.
• To ensure that the safety and well-being of participants in the interview for the quantitative and qualitative data and that no harm will come to them as a result of this study, the research team will be trained to clearly communicate and execute the consenting process.�
• Prior to enrolment or interviews for the focus group discussion, the key informant interview and the phone survey, study staff will administer consent to participants.