| Cutting fraud in health claims |
|
June 28, 2005
SOFTWARE developed in Australia for on-the-spot detection of suspicious claims to health insurers is being tested by 16 private health funds. Its developers say it could be used for detecting suspicious Medicare claims. The system was developed by the Capital Markets Cooperative Research Centre, an academic think-tank supported by private companies. The funds testing the package use the Hospital and Medical Benefits System, one of two systems used by health funds. CMCRC adapted research in data management, data mining and visualisation, developed for capital markets and widely used locally and overseas, to make it work in the health insurance sector. "People put claims into private health insurers and the question is whether they are valid," CMCRC chief executive and University of NSW finance professor Mike Aitken said. "At the moment, the way things work is that they get paid automatically unless there is some obvious reason for rejection, such as you have gone over your limit. "If there is any fraud they worry about that later." Early tests of the technology have identified 4 per cent of claims as suspect, but industry experts say it could be as high as 10 per cent. The Australian Health Insurance Commission estimates non-compliance through fraud and inappropriate practice is less than 1 per cent. Professor Aitken, from UNSW's Faculty of Commerce and Economics, said the software could deliver savings in excess of $800 million a year. "We can detect these things in real time and stop payment," he said. "We are not necessarily saying it is fraud, we are just allowing the insurance operator to ask questions about whether it should be paid." Contracts had been signed, for undisclosed amounts, with the NSW Teacher's Federation health fund and Lysaght Peoplecare, associated with BHP and BlueScope Steel, Professor Aitken said. Intellectual property for the technology has been spun out to form a separate company, called Dtecht, which is raising first-round venture capital of $1.5 million.
|