HealthDirect Australia keeping tabs on directory data quality using federated platform

Having centralised data directories, in theory, is an ideal solution for keeping all necessary data stored and located at a single location.

But for HealthDirect Australia, the national government-owned organisation that provides health information and advice to Australians and is the host of the national health services directory, the experience of relying on centralised data directories were quite the opposite.

According to HealthDirect Australia engineering manager Mark Paul, having centralised data directories that are reactive rather than proactive is not ideal when there is an aim to maintain the quality of critical healthcare data.

"Current healthcare directories are essentially centrally managed database applications, with its data updated via content management systems and call centres. This model is highly reactive and inefficient, mainly because of the high frequency of change of the data within the directory," he said, speaking during the Databricks Spark + AI Summit this week.  

"For example, healthcare services constantly change their operating hours, and the practitioners in healthcare services constantly move between healthcare services, so there's high data volatility. We need a more proactive way to deal with this problem."

To solve this problem, HealthDirect developed a federated data directory platform, which Paul described as a database that "basically maps multiple autonomous database systems into a single federated data store". These database systems included bringing together records from the Australian Prudential Regulation Authority through to Medicare.

He compared the process to constructing a "federated data puzzle".

"There is no aggregation happening here. Instead, it's the abstraction of multiple data stores. Then you've got the federated data platform, which is basically controlled aggregation to create gold standard data by using multiple autonomous origin data sources," he said.

Making the switch has since enabled the organisation to succeed in its data improvement strategy.

"We can look at things like who was the exact source who came up with this change. We can even trace it back to the raw source it originated from, or even trace it back to an external identifier, like a source-based JIRA ticket number, for example," Paul said.

In addition, the new system can capture "exceptions", which are used to help fix the data, and provide "warnings" to improve data quality, according to Paul.

"We also use this for end-to-end data flow visibility, so we know exactly what's happening to our data … for example, over the last couple of months we have captured more than 25 million issues that can be traced back to specific sources," he said. 

Related Coverage

Australia's NIB on digitally fuelled healthcare in a post-coronavirus world

The health insurance company's managing director said COVID-19 has 'drained a few moats around so many healthcare fortresses and traditions'.

Electronic prescriptions in Australia to be available at end of May

The fast-tracked rollout is one of the measures announced by the federal government amid the COVID-19 outbreak.

ADHA details My Health Record breach attempt

The Australian Digital Health Agency also said it's working with healthcare providers to ensure they are cyber resilient.

A majority of Australian public pathology labs now uploading results to My Health Record

Australian Digital Health Agency says nearly 43 million pathology reports have been uploaded to system.