Tuesday, September 05, 2017
A Valuable Review Of The Potentialities Of The Use Of Health Data For System And Clinical improvement Purposes.
This useful review appeared last week:
30 August 2017
I’d just crumpled onto the couch at home when an unknown number flashed up on my mobile. “It’s Christian Nøhr calling from Denmark,” said a warm voice down the line. I did the awkward phone-to-shoulder contortion as I scrambled for a pen, notepad and dictaphone.
I wasn’t expecting this call.
My half-hour chat with Professor Nøhr, a health informatics expert from Denmark’s Aalborg University, opened a small window into what I’ve decided to call “data heaven”.
Until very recently, Australian researchers have had to beg federal, state and territory governments for linked health data and could wait years to gain access.
In one instance, cited by the Productivity Commission in its March 2017 report, there was a five-year delay in the supply of vital data that went on to show an increased cancer risk for young people undergoing CT scans.
“Had [the] study been approved sooner, and been able to proceed at an earlier date… we would have had results sooner, with potential benefits in terms of improved guidelines for CT usage, lesser exposures and fewer cancers,” one researcher said.
Contrast this to the Danish system, where GPs, specialists, pharmacists and hospitals all interact using a central database, which has been in place since 2010. Each patient’s journey through the system can be tracked in detail using a unique ID number, and there is a clearly defined process through which researchers can seek ethics approval and access de-identified data.
Denmark’s primary care portal, Sundhed.dk, allows 98% of GPs to communicate directly with other healthcare professionals. Patients cannot receive treatment outside the system and are unable to edit their medical information. But they do have the ability to tag specific medicines as “private”, which means doctors can be fined for not seeking consent prior to accessing the information (unless it’s an emergency situation).
The most remarkable aspect of the Danish system – and the most alien from an Australian perspective – is the broadbased acceptance of data collection and monitoring by government. “We expect the public system to take good care of our data”, said Professor Nøhr. By comparison, lack of trust in data sharing arrangements is “choking” the use of health data in Australia, according to the Productivity Commission.
This is unfortunate because linked, de-identified medical records are the Holy Grail for health researchers. Retrospective analysis of millions of anonymised data points can reveal a whole galaxy of patterns, without the expense and hassle of a clinical trial.
Once granted access to government-held medical data, researchers can view real-life information about the long-term outcomes of patients taking particular medications. And linking that information to social variables can yield an even richer understanding about how the health of individuals is affected by education, employment, housing, or income. The health system can be made vastly more efficient using the insights from data analytics.
But while other countries are increasingly using health data in clever ways, Australia is being left behind. “This is a global phenomenon and Australia, to its detriment, is not yet participating,” the commission’s report said.
So what are Scandinavia’s secrets to safely unleashing the power of health data, and what’s stopping Australia from doing the same?
To find out, I emailed peak bodies for health research in Norway and Sweden and asked them to spill the beans. Both of these countries were listed as top nations for health data sharing and linkage by the OECD’s 2015 report, along with Denmark, Finland, Canada, Czech Republic, Israel, Korea, New Zealand, Singapore, and the UK.
Birgitta Lindelius, a spokesperson for the National Board of Health and Welfare in Stockholm, Sweden, said: “One reason is probably our long tradition when it comes to registers and the personal identity number (PIN) that is an important part of the registers. The PIN gives us the possibility to link data.”
Statistics of diseases and surgical treatment of patients have been published for more than 100 years in Sweden, regulated by the Act for Health Data Registers. County councils are, by law, obliged to send information on the care given to the National Board of Health and Welfare.
Data are disclosed for research purposes up to 400 times every year by means of an exception in the Public Access to Information and Secrecy Act. “We believe there is a big trust among both researchers and citizens in general in this system,” said Ms Lindelius.
Norway has a similar tale. Using the 11-digit ID number given to all citizens, data from different registries can be linked together. “This provides valuable knowledge about the occurrence of disease, the use of healthcare services and results from public health measures,” Marta Ebbing, the director of Health Registries at the Norwegian Institute of Public Health, said.
Denmark introduced its central person register number in 1968. This ID is used for everything in relation to national and local government, including healthcare, drivers’ licences, bank accounts, public education and employment. “It gives a unique opportunity to do research, especially within epidemiology where you can actually trace many, many different bits of information about the individual from birth to death,” said Professor Nøhr.
In fact, all OECD countries with a strong health data linkage use a single number to identify individuals and link data sets together. New Zealand has a national health index number, South Korea has a resident registration number, Israel has an ID number and the UK has an NHS number.
By comparison, there was a furore over the Hawke government’s proposed Australia card in the 1980s. Even today, Australian health data can’t be linked using the Medicare number because people can have multiple cards (or multiple people can have the same Medicare number). Instead, researchers use statistical techniques to piece together health data sets in Australia as best they can.
“Researchers might take the first three letters of your first name and surname, your date of birth and something else and then creating a hash key out of it and that then becomes unique to you,” Dr Christopher Pearce, the vice-president of the Australasian College of Health Informatics, said.
There is vastly more here:
The bottom line, to me, in all this is that we need to do a much better job of communicating the benefits of data linkage and use and establish a trusted framework for this to happen.
As I put it to a correspondent the issue is this:
"To point out the obvious they (the countries cited) are all small homogeneous countries with high levels of trust in Government.
We (Australia) are much bigger, more diverse and less trusting!"
To me it is these issues that need to be addressed first. It will also need dramatic improvement in the levels of trust in Government and the ADHA specifically!
Posted by Dr David G More MB PhD at Tuesday, September 05, 2017