Data collected at the point of care and transmitted to where it can be effectively consumed has the potential to inform policy and improve the way healthcare is delivered. Data sharing is key to this process and interoperability between systems that collect and share this data is at the heart of the health service improvement cycle. Simply put, interoperability is the ability to share information between systems but this definition is where the simplicity ends.

It is well known that Kenya is deploying electronic health record (EHR) systems and EHR vendors often use the label - ‘HL7 compliant’ to indicate the system’s ability to exchange information with other systems. HL7, the shortened acronym for Health Level 7 refers to the international standard for sharing health information. There are other data sharing standards such as FHIR, RIM, CDA, DICOM etc. though not as commonly referred to, also specify how particular data elements can logically be transmitted.

International standards that guide the flow of health information from and to health information systems are abundant but widespread interoperability is still yet to be achieved. Why?

In Kenya, one possible reason is that implementation of health information systems is still new and interoperability has not been of key concern. But this is changing rapidly. As it is now, that a system is ‘HL7 compliant’, as they often need to be, seems to be adequate for system purchasers to tick a box during procurement.

Even if they wanted to test the sending and receiving of data, they may be unable to. They would need access to the technical architecture of target systems, databases schemas and source code. This is not straightforward and legal issues around intellectual property of proprietary systems they'd like to share data with often come to the fore. These factors slow or completely stop this intention.

Getting to the point where there can be seamless exchange of data between differing systems in a form where it can be sensibly reconstructed has proved difficult across the globe and Kenya is not unique. The reasons are many and seamless sharing of data still seems to be some distance away.

In Kenya, it doesn’t help when users are unclear with whom they are sharing their data with and for what purpose. To DHIS2 they say! But what about the health record officers? What will they do? This is a significant part of their work is it not?

It is apparent that achieving interoperability is also about the human issues, ethics and data governance as it is about technology. Things have to be looked at in entirety. Many questions around what it would take to sensibly and coherently share any information electronically needs to first be answered.

To share which data with who? Are any interfaces or middleware needed to act as bridges between these systems required? How much would they cost and are these solutions cost effective? These questions mostly just represent the technological aspects. There are still legal, ethical and data governance questions that also need to be looked into.

In an ideal world, health data can be sent instantaneously and seamlessly to national level systems or to other facilities in cases of patient referral. This data should make sense and be accurate where it is received. Most countries including Kenya are still far from this ideal world but a lot of progress is being made and truly open source solutions have helped significantly as they allow access to a systems architecture, source code, databases etc with minimal legal encumbrances.

Currently, the Kenyan Ministry of Health and partners are initiating discussions to look at some of these issues. This is definitely a good step and even though it will take some work, Kenya will eventually get there.