Mikael Hagstrom, vice-president of SAS, a global leader in data analysis software, says Australia isn't alone. Health care worldwide lags the banking, telco, retail and other sectors in the application of intelligence tools. He claims advances in digital imaging, diagnostic tools, real-time electronic medical records and the ability to mine patient information from data streamed live from remote monitoring systems will "transform healthcare delivery".
It can happen now, he adds, simply by using information that's already available.
"When we look at most national e-health systems being rolled out, it's redoing what we've already tried, namely forcing nurses ... to retype patient data into a different storage environment, with all the associated errors and infrastructure costs," he says.
"Ninety per cent of available data has been created in the past two years. We don't need new ways of creating data. We need to use what's there, including the unstructured data contained in medical practitioners' ... notes."
Hagstrom is presenting a series of forums for senior executives in Australia and New Zealand next week. He says the state of the healthcare sector worldwide hurts him, because the potential for improvements is not being realised.
When the World Health Organization reported that 30 per cent of healthcare spending was wasted on wrong diagnoses and wrong hospital treatment, SAS began a pilot program with Sweden's Karolinska Institute.
"When we looked at what those figures meant for Stockholm, a city of two million inhabitants, well, that means two deaths a day and 100 injuries on top of the budget costs." Hagstrom says the three-year study helped to cut such medical errors by half because of access to real information.
"We simply looked at using what was there and that was the unstructured text data from doctors," he says. "We used text-mining techniques to let doctors look at what fellow physicians had done in similar situations.
"The doctors were very open to that. They could see their colleagues' decisions and the outcomes, and they found (this) cut their error rate in half.
"Yet it's a very simple thing to do, technology-wise."
Hagstrom says harnessing patient data will facilitate planning and establish "which treatment options work, and which don't. We need to make our voices heard on this. Advanced analytics is needed more than anything else in healthcare right now. It will improve care and reduce wrong diagnoses and treatments."
He says data quality is crucial and argues against relying on already heavily loaded nursing staff for data entry.
"We've proven human decision-making is very error-prone. Data-driven decision-making is not. So we can move from where people act on gut instincts or experience because they have no alternative, to being able to make decisions based on evidence. That's a powerful proposition."
---
Autor(en)/Author(s): Karen Dearne
Quelle/Source: The Australian, 21.07.2012
Bitte besuchen Sie/Please visit:
