Children the winners in data linkage project

A University of Adelaide project to link information from multiple government databases could have major benefits for child health development and the delivery of support services.

For the past three years Professor John Lynch has been steadily navigating his way through an administrative jungle.

He's been liaising with multiple government agencies, talking with lawyers and ethics specialists, dealing with technology experts and lobbying leaders in health and education.

The subject - making better use of the masses of government data on child health and development.

From the time a child is conceived and during key stages of life, various federal and state agencies are collecting data.

Such information provides valuable insights into a child's progress and supports the delivery of government services. But there is one major flaw in the system - none of these separate packages of information is linked.

"Currently systems don't talk to each other because governments don't have integrated data systems," Professor Lynch said.

"Consequently, tracking the longer-term success or otherwise of new programs in public health and education, and assessing the impacts they are having on children, is very difficult.

"What we're doing is providing the links to identify groups of kids who aren't doing well, finding out how we can help them better prepare for school and giving them more life chances. That's really at the core of what this is all about
- and it's a very efficient way of doing it."

A trained epidemiologist and Professor of Public Health at the University of Adelaide, Professor Lynch began his work on the Early Childhood Data Linkage Project after being awarded funding through a prestigious National Health and Medical Research Council Australia Fellowship in 2009.

His project has now been given an additional $273,131 from the Federal Government's Partnerships for Better Health program plus $180,000 from State Government partners.

It's a whole-of-population study which captures more than 240,000 children up to the age of 13 born in South Australia since 1999.

Professor Lynch is also one of the leads on a mirror project in the Northern Territory which has a particular focus on indigenous communities and involves about 4000 children.

In total the project is linking together data on cohorts of children from 13 separate state and federal sources.

It covers the perinatal period, health facts about the mother and data collected through the Child and Family Health Programs, any hospitalisation or emergency admissions, and results of four-year-old health checks at kindergarten involving health, growth, vision and hearing.

Then it takes in data collected from the school readiness census for the Australian Early Development Index and information from the National Assessment Program - Literacy and Numeracy (NAPLAN) conducted in years 3, 5, 7, and 9 at school.

All these sources help develop a broad picture of each cohort covering health, social and emotional development, and cognitive, language and general knowledge.

A major appeal of the project is that it re-uses and adds value to publicly funded data which already exists.

"By linking the data we'll have a local evidence base to inform how health, child development and early education services direct their efforts to where they are needed most to have the greatest impact," he said.

"Ultimately it's about supporting children's healthy development which is important for their readiness to learn and academic achievement at school."

While it sounds a great outcome, Professor Lynch has had to step through a quagmire of legalities and ethical issues to reach this stage.

"The plan is that the database is updated and keeps growing each year, but it has been common that ethics approvals are time limited and data sets have often been destroyed once the original purpose has been filled," he said.

"We've been working hard with ethics panels to impress on them that is for the public good and that because we're adding so much value it doesn't make any sense to destroy them.

The fact that all data is de-identified has been instrumental in overcoming objections. The focus is on general patterns of development among child cohorts and not on individual progress.

"We're not interested in identifying individuals, that's not what this is all about."

A defining feature of the project is that it merges data from two levels of government, including agencies within SA Health and the State Department for Education and Child Development, and also the Federal Department of Education, Employment and Workplace Relations.

Gaining access to those databases takes time and considerable trust-building with the individual data custodians. The project team must also be mindful of the quality of the data which was not collected for research purposes.

"We've been receiving cooperation from the highest level from the Premier, Ministers and chief executives - they get it and are very supportive," Professor Lynch said.
"However, we've had a mixed bag of support further down which is why the past three years have been so important in building trust with custodians."

Professor Lynch now has a team of about 10 researchers working on the project, some located within the government agencies.

"Most of the data is in and we're hoping to get some analysis done and have some quick wins," he said. "The end goal is to try and create a national system - and that opens up all sorts of possibilities."