Building better health information systems

Increasing complexity and costs put the health system on an uncertain trajectory. In a context of workforce pressures and increased need for services, where will the next advances in care delivery come from? One answer is health information technology (HIT), long expected to be a key driver of improvements in patient care delivery and healthcare transformation.

But behind this optimistic view of technology lies a troubling reality: stories of negative outcomes from HIT implementation are legion. The Telfer School’s Craig Kuziemsky says communication issues, creation of new or more work, and even adverse events such as medical errors are so habitual at the implementation stage that some are left to wonder if HIT will ever deliver on its promise.

Fulfilling technology’s potential

While interoperability is the underlying basis of integrated healthcare delivery, most of the work to date has not dealt with the interoperability of the people and processes that engage with HIT. To address this gap, healthcare decision makers will need to pay a lot more attention to the business processes that underpin HIT, argue Professor Kuziemsky and his colleague Professor Liam Peyton of the School of Electrical Engineering and Computer Science.

The team developed a framework based on their two-year case study of a palliative care information system (PALS-IS) in Ottawa. They identified three categories of process interoperability issues that arose during system implementation: care delivery, clinical practice and administrative. The researchers also found that many of these issues emerged over time, and that developing solutions to some process interoperability issues subsequently led to other problems.  

Taking a wider lens on interoperability

A significant challenge was when process automation crosses user groups, like clinicians and administrators. Administrators, for example, were under increasing pressure to provide accountability and performance management data. Meeting those needs often increased the workload of front-line clinicians who had to collect the necessary data to support administrative process interoperability.  

These conflicting results demonstrate that HIT really is a disruptive technology. But, as Kuziemsky says, “our ability to be successful at implementing HIT will be defined by how well we manage the disruptions.” And despite trials in putting HIT into operation, “all our stakeholder groups saw great potential in how it could enhance clinical processes like patient management, or administrative processes like monitoring and reporting.” 

An interaction of technology, processes and people

These findings provide the first tangible framework focused solely on process interoperability, and offer HIT designers and vendors practical guidance on how to evaluate this aspect. “My advice to people or organizations implementing HIT would be to take these findings seriously,” says Kuziemsky. “Most healthcare IT models have had very little to say about process interoperability, as interoperability is more typically considered from a technical or a semantic point of view.

“But if HIT is really going to be a driver of healthcare transformation, we will need to design systems with a lot more sensitivity to the overall healthcare ecosystem where technology, processes and people interact.