Dear Mr. Hiary,

Thank you for sharing this information about Immunify. I’m sorry I missed the webinar, but I’ve reviewed the original UNICEF RFP and the public case study on the ISEET website. Congratulations on developing and launching such a significant platform.

I’d happy to share some perspectives informed by my work on CCE data interoperability and the integration of CCE data with supply chain systems. I’m currently contributing to the Cold Chain Data Interchange (CCDX) program in Nigeria, led by the National Primary Healthcare Development Agency (NPHCDA), which is extending the country’s supply chain systems to incorporate CCE administrative, performance, and status data, alongside other eLMIS enhancements.

The architecture is not yet fully settled, but there is strong alignment in many areas, and a working prototype already connects CCE temperature and performance data to OpenLMIS. The prototype builds on the OpenLMIS Cold Chain Equipment Management module, which the program will extend.

As countries move from periodic EVM data collection to continuous system data, I see five key considerations:

Master data management

Without a designated, authoritative system of record for health facility data, fragmentation is inevitable: every system and program that needs facility data builds its own copy. Your case study makes this point well — a second place to maintain the facility list is the fastest route to failure for a new system. Many countries are starting from exactly that position, with multiple competing facility sources and no clear system of record. The same applies to other critical master data, such as CCE inventory. When it is spread across systems, consumers cannot tell which records are authoritative or who owns them. No single eLMIS or CCE inventory system can resolve these conflicts; a new platform simply risks becoming one more copy. The obstacle is clarity and governance, not software.

Countries should establish one authoritative system of record per master data domain (health facilities, CCE inventory, etc.), each with a named owner and defined scope of authority. An eLMIS platform might be authoritative for CCE inventory but would rarely be authoritative for facility data. Policy should require systems to consume master data from the relevant system of record rather than create duplicate repositories. Systems of record must be online, with published APIs and low-friction access for authorized systems and programs. Programs should also establish feedback loops for surfacing and resolving discrepancies between authoritative records and field-observed conditions.

Standards and tools for CCE data management are improving

The EMS standards describe how data will be recorded on new WHO-prequalified refrigerators and freezers, including proposed standards for interoperable CCE data delivery from EMS/RTMD suppliers to countries. Having a common data dictionary and data format for CCE performance and administrative data will make it easier for countries to connect CCE data to national decision-making systems, regardless of fridge make/model. I encourage countries and technical partners to familiarize themselves with the details of the EMS data definitions and structure.  

I would also encourage countries to become familiar with data collection tools such as the Varo mobile app, which let technicians and health workers create detailed cold chain status reports — with facility and digital temperature data — for national or sub-national aggregation. As EMS appliances become more prevalent in the field, I expect the number of mobile tools for cold chain reporting/analysis to increase.

Linking loggers to appliances to facilities

The difficulty of connecting data from CCE loggers to CCE appliances to health facilities should not be underestimated. Nigeria is fortunate to have high-quality facility and CCE inventory data, refined over years and multiple collection efforts. Even so, establishing and maintaining the linkage between these three data entities cannot be fully automated; low-confidence matches must be escalated for human review. Geospatial query capabilities in the health facility registry could make it easier to link geolocated reports (e.g., CCE reports collected on mobile phones with geopositioning) to the correct facility.

CCE temperature data resolution

RTMD and new EMS devices produce high-resolution streams of temperature and performance data, typically at 10–15 minute intervals. A combined eLMIS/equipment management system (probably) does not need all of it. Functional status and daily KPIs cover most operational use cases; the full-resolution stream matters mainly for remote diagnostics by technicians and for anticipating spare-parts needs. It’s not clear to me how Immunify receives CCE data, but it appears to have reached the same conclusion, storing aggregated status and KPIs at regular intervals (daily?) rather than the raw stream. That raises the question of where ingestion, aggregation, and (if needed) full-resolution storage should live, which I take up below.

Interoperability layer

A unified platform like Immunify still has boundaries that require integration with other systems: CCE data ingestion across suppliers, normalization and enrichment of RTMD alerts, logger-to-appliance-to-facility matching, synchronization with systems of record, and KPI calculations. Every one of these is likely to differ in the next country. If that logic is built into the platform itself, adapting it elsewhere is an engineering effort that could limit the platform’s portability.

Countries should consider adopting an interoperability / systems mediation layer to hold this country-specific logic, keeping the platform itself clean and decoupled. For CCE data specifically, the interoperability layer receives data from all suppliers and sources, applies the country’s rules, and delivers the appropriate derivatives to each downstream system — daily KPIs to the eLMIS, for example. This is consistent with the WHO Digital transformation handbook for health supply chain architecture. Whether the layer should also facilitate long-term storage of the full-resolution CCE data streams depends on priorities: if remote diagnostics or predictive maintenance are genuine goals, it is a sound investment; if not, aggregating at intake and discarding the raw data is the simpler path. My organization, Phoenix GH, has published some expanded thoughts on this topic at https://docs.2to8.cc/cce-data-interop/.

I’m curious how these thoughts resonate with your experiences with Immunify. Please let me know if you have questions or would like to expand on any of these topics in the coming days.

Thank you again for sharing these details on Immunify. I am personally interested in anything you can share about the data schema for CCE inventory and performance/functional status in the platform. I see that Immunify has been described in a few settings as open source, but I can’t seem to find a public repository. Is the source available to inform efforts in other countries? The work in Nigeria might benefit from a peek at certain implementation details and I would appreciate anything you can share here.

Warmest regards,
Benson Miller
Principal
Phoenix GH