Navigating healthcare services:

During my fellowship at the International Rescue Committee’s (IRC) Regional Innovation Hub in Jordan, I conducted a 1.5 month research project that looked at understanding the refugees experience both for seeking and receiving health care services. Those services are provided by the three different types of clinics: humanitarian, Ministry of Health, and private.

This project’s outcome was threefold;

1. Providing insights and a better understanding of how refugees navigate a fragmented healthcare system.

2. Testing the use of machine learning (K-means clustering) in synthesizing the data from field research.

3. Create and propose more structure and guiding processes to develop the community-led research initiative started by IRC’s innovation hub ‘Mahali’.

My roles:

Project Management,
Design Researcher,
Interview & workshop facilitation

Background & Research:

As of 2017, the latest data which is available, refugee households were spending an estimated 41% of their household income on health care, and cited cost as a primary barrier to care. Refugees reported primarily relying on government clinics (57%), and private clinics (38%). Only 14% of refugees reported going to NGO clinics (which provide free services) to receive care. These numbers have surely changed over 2018, as the MoH increased prices for Syrian health seekers. All the same, this seemingly represents a paradox-- if costs are a key barrier, why aren’t more health seekers going to free clinics?

The two main objectives of the research:

1. Have a better understanding of the health-seeking behaviors of refugees in Jordan especially after changes in the country policies.

2. Understand key definitions related to care from the perspective of refugees, including types of care, types of clinics, and definitions of quality all along the care-seeking journey.

Group work and facilitation:

An Important aspect of this research was working with the field researchers. They were previous participants from the community-led innovation hub ‘Mahali’, and that IRC started hiring as contracted staff. As field researchers they are the people facilitating focus groups, interviews, documenting, and reporting the results back to the designers on the team.

During my time working in this organizational structure, I found that there was an opportunity in getting the field researchers more involved in the whole process. As a result, I started creating guides that would be handed to the facilitators one week in advance, as well as have dry runs of the workshop, role-playing and evaluating the tools before conducting them with participants.

After my fellowship, the organization looked into continuing that work and looked into other training opportunities they can provide to the field researchers.

Synthesis:

Along with analyzing the data around the two main research questions, this research project also experimented through the use of cluster analysis methods (k-means clustering) during synthesis and the processing of raw data gathered from focus groups. In order to see

if using machine learning could speed up or support the process of synthesis during a humanitarian crisis.

Working with Steven the data vis. designer and researcher, we developed reporting systems that would allow for suitable data collection, as well as analyzing & cross matching the results that came out from the clustering synthesizes with standard synthesis results.

Steven visualized the insights and pain points of the system on an interactive map shown above.

TWO MAIN INSIGHTS FROM USING K-MEANS CLUSTERING FOR DESIGN RESEARCH SYNTHESIS:

• It can help validate and/or expand a designers synthesis outcomes.

• If you have more data, then it would save time. However, if you have less data it could take longer to synthesis which can defeat the purpose of using it as a tool.

1. REFINE THE EXPERIENCE/JOURNEY OF REFUGEES IN FREE CLINICS

e.g leveraging way-finding systems, line management, etc. Developing a guide on user experience at NGO clinics could be an impactful, yet achievable step in the right direction.

Research insights & Further exploration areas:

2. DEVELOPING DECENTRALIZED MODELS FOR DELIVERING HEALTH SERVICES

e.g building on existing health practices like visiting the pharmacy instead of going to the clinic - having doctors rooms at pharmacies

3. OPTIMIZE HEALTH SECTOR INFORMATION SHARING

While services may be coordinated from a supply perspective among the different agencies, these decisions are opaque to clients and do not take into account how they seek to access services.