The EPCON engine: advancing interoperability

Updated: Oct 19

Health information systems can work together across different organisations to improve the effective delivery of healthcare for individuals and communities. Such interoperability is a goal worth pursuing. Sensitive data requires privacy and security, which can make sharing tricky. Nonetheless, access is vital: the inability to access health data when needed can cause significant harm.

EPCON's AI model

The benefits of an interoperable healthcare system include:

  • The ability to monitor the patient journey throughout the cascade of care across different organisations and systems.

  • The elimination of double data entry at various points of care.

  • The capacity of government service providers to access data that reveals health trends, disease hotspots and gaps in the provision of care.

  • The potential for decision-makers to create better models of care and patient experiences across healthcare.

EPCON’s predictive AI engine helps users achieve optimum interoperability. How does it do so?

Drawing data from various sources

In any region or country, a wide variety of systems are in place for data accumulation, from paper-based approaches to Excel spreadsheets. A key component of EPCON’s AI engine is its openness: the platform can pull information from any source, store it in a data lake, or centralised repository, and then generalise the data into a single format for AI model training.

If data is missing or inconsistent, a change management process might be required. In low resource settings, TB screening data is frequently captured via paper-based methods or a combination of paper-based and digital records. This means that the real-time and location aware data is lacking, which reduces the effectiveness of the predictive AI system’s mapping and modelling. To overcome this issue, EPCON collaborates closely with its in-country partners and creates low entry, low cost tools that enable the digitisation of screening and patient level data. Existing communication channels like WhatsApp and Telegram are customised to facilitate this process. They get tightly integrated into the state-of-the-art AI models for dynamic and interactive program support.

Ensuring output meets a wide variety of needs

An interoperable system guarantees a smooth integration towards a range of healthcare applications. The visualisation of output provided by the EPCON model can be viewed using EPCON's dashboards or can integrate into existing tools such as DHIS2 or other.

For example, country-level strategic planners can pull real-time data - such as the number of people screened or diagnosed for a particular disease - in raw format, visual dashboards, or via open source tools like Graphite. Healthcare workers, on the other hand, require information in a format similar to how they are working - for example, via common communication tools like WhatsApp, text messages or email.

In essence, EPCON has the capacity to display output on multiple interfaces and push data to various systems, depending on client needs and existing IT infrastructure.

Providing a holistic view

An interoperable system allows a region to move away from viewing diseases in silos. Within most countries, data on various diseases like TB, malaria, HIV, leprosy and other, are collected via different methods and existing systems. The EPCON AI platform can tap into these different tools, collating the relevant information and train a model before providing a holistic view across disease profiles. This makes it possible to analyse how diseases are interacting with each other in the various communities and hotspots, assess the estimated burden and risks, and learn how integrated services can be optimised subnational level.

Maintaining data security

Sharing health data can raise various concerns, such as maintaining patients’ privacy. EPCON is always cognisant of data sharing agreements, data security and privacy. This means that data is often anonymised so that it cannot be traced back to a given patient.

Healthcare systems need to be able to work together to improve the delivery of healthcare at an individual, community and national level. The EPCON AI model has proven its capacities for interoperability.