How are countries linking person level data systems together in a National Data Repository?  

During the Data Use Community meeting on Wednesday, March 29th, we heard about the work in progress happening in Malawi and Haiti.

Malawi

Brett Onions from Luke International shared work taking place in Malawi with the goal of the Master Patient Index (MPI) to identify all patients uniquely and accurately within the country.

Overview:

  • The Demographics Data Exchange (DDE) is a Master Patient Index (MPI) for Malawi, currently being developed by EGPAF in Malawi. The system is designed to store demographic and clinical information for patients across Malawi, with a focus on providing offline access to the data in areas with limited connectivity. The DDE is being transitioned to the MoH in Malawi, with the goal of improving healthcare delivery and patient outcomes.

Architecture and Infrastructure:

  • The DDE system is designed with an offline first architecture. This allows for the system to be accessed even when internet connectivity is limited or not available.
  • It is spread out across 170 sites within Malawi.
  • Each site has a proxy that is responsible for maintaining the system’s uptime when the site is experiencing connectivity issues.
  • The system is designed to be scalable, with the ability to expand as needed to accommodate new users and data.
  • DDE is bespoke, built on top of Ruby on Rails. 
  • Data is maintained between sites in a two way sync. Each site is given a predefined number of IDs they distribute. 
  • FHIR standards.

Unique Patient ID and Health Passports:

  • The DDE system produces a pre-generated non-identifiable patient ID for every client. These IDs are physically printed as barcodes and placed as stickers on the patient's Health Passport, which is a booklet containing the client's medical history/medical record. 
  • This unique patient ID allows healthcare providers to quickly and easily identify patients and access their medical records, even if they are not connected to the DDE system.

Client Registration

  • The DDE system currently has approximately 6 million unique clients registered.
  • The system is designed to allow for easy and accurate registration of new clients, with patient demographic and clinical information being collected at the point of care.
  • Demographic data can be changed (such as if a client changes their name), when the change happens locally it is then propagated to the other facilities. 

Challenges

  • The biggest challenge faced in terms of scale is connectivity - there are many solutions to solve it but not every site has the same problem.

Haiti

Christina White and Piotr Mankowski of DIGI ITECH-UW shared work with Client Registry in Haiti, with the goal to improve the ability to provide high quality care to patients to move between services, facilities and networks. 

EMR Background

Implementation Details

  • Using OpenCR (Intrahealth) for the Client Registry
  • Using FHIR based Client Registry workflows based on PIXm and PDCm IHE Profiles
  • Utilizing biometrics and demographics for resolving patient identity and cross-reference across systems. The fingerprint system focuses on all patients, not just HIV.
  • Within the architecture, there is a local server that is implemented that searches both locally and nationally. If the internet is not available, it will only search locally until it is available to send the information nationally. 

Current Status

  • The project is currently in the pilot phase. A pre-pilot was completed at certain locations looking at how functionally it would work on the ground with infrastructure.
  • Support for patient search and patient registration
  • Support for continuum of care workflows (shared health record)


  • Enrollment starts with registration within the EMR. Since Haiti uses a "Test & Treat" approach, this means diagnosis and treatment are starting at the same time. Currently, 3 EMRs are collecting information for about 95% patients under ART. These records are gathered into a national individual-level repository. Fingerprints are being collected at ART facility level. 
  • At the local level, when a patient is being registered, when they are entering demographic information it will automatically query locally & nationally for matching information & then displays potential matches. You can continue the registration process or import a matching patient.

Challenges

  • Integration with national fingerprint registry
  • Definition and implementation of patient matching criteria
  • Integration with existing infrastructure
  • Initialization of Client Registry with import of existing Patients.

Matching

  • Matching is automatic via fingerprint with manual adjudication using Identifier, First Name, Last Name, DOB.
  • We're using SOP from 2022 displaying mother's name, emergency contact, phone number and place of birth on registration for manual adjudication.

  

  • No labels

This work is licensed under a Creative Commons Attribution 4.0 International License.