A vast amount of data is generated in Finnish laboratories, and it plays a critical role in healthcare. Up to 70% of clinical decisions are based on laboratory test results.

In Finland, it is common practice that laboratory data flows smoothly between stakeholders. Data from different laboratory disciplines can be combined and compared, and historical data can be used for longitudinal analysis when needed. However, this level of interoperability is not self-evident in many countries.

What is everyday routine for us is still aspirational in many other countries, says Juha Högmander, Chief Product Officer at Mylab, a Tampere-based laboratory information systems company.

A unified data model as a foundation for effective use of data and AI

The smooth operation model in Finland is built on a well-designed IT solution. Approximately 80% of laboratories in Finland use Mylab’s My+ laboratory information system, which is based on a unique unified data model.

This means that data in laboratory information systems is standardized, i.e. interoperable across systems, and structured, meaning consistent in format and structure. This enables large-scale and efficient utilization of data.

Mylab has, together with its customers, built a shared understanding of the laboratory domain, diagnostics, and the data required in diagnostics for over 40 years. For years, the company has systematically developed a future-proof foundation for the efficient use of data.

In many organizations, poor data quality is a barrier to adopting AI. We do not have this problem. Going forward, we can effectively leverage AI in the analysis of laboratory data. Over the past ten years, we have collected data in accordance with a unified data model, in collaboration with major domestic customers, into our current next-generation My+ system,” says Högmander.

Thanks to the unified data model, we can also bring historical data from other systems into My+ and utilize it in the same way, he adds.

Enriching data enables more predictive healthcare

Due to the unified data model, Finland has excellent conditions for leveraging AI very effectively in diagnostics.

We are only at the beginning of this journey. With AI, diagnostics can be taken to an entirely new level, while significantly improving the effectiveness of healthcare. This is exactly the kind of meaningful use of AI,” says Högmander.

In the future, laboratory data can be enriched, making patient care more predictive. AI can, for example, predict the onset of type 2 diabetes by combining data such as the development of a patient’s blood glucose levels and weight. It can also detect cancer-indicative findings from imaging studies, even at the cellular level.

At the population level, a unified data model offers significant opportunities. When data is available from nearly the entire population, it becomes possible to model the development of diseases or pandemics.

In addition to patient care, data also supports the operational management of laboratories.

The data can be used to monitor and manage the day-to-day operations of laboratories; for example, it shows how quickly laboratory results are reported. Similarly, it is possible to anticipate laboratories’ workload and resource requirements: if, for example, the number of positive influenza results begins to rise, we can prepare in advance for a surge in patient numbers, Högmander explains. I believe that, going forward, the role of laboratories in healthcare will become more prominent and stronger than it is today. 

A unique model with global potential

How is it done elsewhere? In many countries, hospital laboratories may use up to eight different laboratory information systems, with separate systems for each specialty, such as pathology and microbiology. In addition, various imaging and testing devices introduce their own systems into the mix.

It is similar to a manufacturing company operating with multiple ERP systems, along with the systems that feed data into them, Högmander compares.

Even if the systems communicate with each other, utilizing the data effectively is difficult, as each system has its own way of collecting data and there is no unified data model. The challenge exists at the ecosystem level—individual healthcare providers cannot realistically solve or change such a situation alone,” he continues.

IT architecture also becomes extremely heavy and costly over its lifecycle when multiple fragmented systems are maintained and developed, and their data cannot be utilized across systems.

Especially if different regions and cities use different systems, efficient and large-scale use of data is simply not possible,” Högmander states.

In addition to Finland, Mylab operates in Sweden and Denmark.

We see that the Finnish model could benefit many other countries that are currently underutilizing their laboratory data,” he concludes.