There are many business situations that are too complex or large for a single organisation to tackle alone. In the past, this has led like-minded businesses to come together and form working groups, joint initiatives, or data sharing agreements.

Although cooperative, such initiatives aren’t necessarily critical to core business operations and tend to be finite engagements. New forms of corporate collaboration may provide an opportunity to deliver ongoing and direct benefits to businesses that face a common opportunity or threat.

An area ripe for increased competitive collaboration is the private health insurance (PHI) industry. According to APRA’s Private Health Insurance Quarterly Statistics more than 13 million people are covered by PHIs in Australia, which generates treatment claims of around $4.5 billion each quarter. Unfortunately, a small number of people and health service providers, engage in ‘improper claims’—potentially fraudulent behaviour. Especially concerning is that it can be very challenging to catch the behaviour, and figures around the cost of improper claims are hard to come by, as no one wants to admit they have been duped by fraudulent behaviours.

A paper prepared by the University of Minnesota for the FBI in 2012 suggested that 3-10% of health spending in the US is fraudulent. Therefore, it’s conservative to assume it’s a multi-million dollar problem here in Australia.

Tackling health fraud today

PHIs today use business intelligence tools to report red flag triggers for further investigation if a claim contains certain suspicious criteria. There are several reasons why the red-flag approach is problematic.

Firstly, it takes a lot of resources to review thousands of claims that have been ‘red-flagged’ every month. Sometimes an error is just a simple mistake with no nefarious motivation, yet every flag requires the same degree of investigation. In the health industry this is known as “pay and chase” and is a significant burden for health funds.

Secondly, and more importantly, a red-flag review can really only check claims against validated data, such as known medical history, billing or practitioner speciality. This means ‘left-field claims’—those which are legitimate at face value, but hide fraudulent practices—are significantly more difficult, if not impossible, to catch.

Finally, a single PHI cannot see across all the relevant industry data to allow for a comprehensive check. It is perfectly possible for fraudulent behaviour to move from PHI to PHI and the same fraud be perpetrated multiple times.

Local health insurers now realise they need to use technology to help anticipate and prevent fraud rather than chasing down fraudsters after the fact. They are looking to big data and analytics to find patterns and anomalies that deliver predictive insights, rather than just setting ‘red flags’ or screening for data conflicts.

The challenge

Developing larger data sets and improved analytical tools will require a greater deal of cooperation amongst insurers. Industry groups in Australia have taken some steps, however, there has been limited success with this approach.

The Operations of the Private Health Insurers report reveals that, unlike credit card issuers, or even Medicare, which can draw from a national transaction database, most of Australia’s 30-40 mid-size and small health insurers have less than 100,000 policy holders.

For a fiercely competitive industry, the prospect of such close collaboration might sit uneasily. Collaborative data would mean breaking out of these silos and actively pursuing collaboration with competitors, and using trusted external partners such as Civica—who currently provide the underlying technology that supports over 30% of PHI claims in the country—as facilitators.

The next steps

There are a number of initial steps that health insurers would need to take before they can replicate the success of government organisations, including establishing a governance system and roadmap for constructive and effective information exchange.

Another challenge will be to promote a shift in mind-set within organisations. Identification of which data to share will also be required, and there needs to be a desire to overcome any competitive, policy and regulatory issues that are identified.

Conclusion

Creating and analysing aggregated data sets and utilising predictive analytics could provide the Australian health insurance industry with a powerful tool for battling fraud.

The Australian finance sector has shared data for years to help prevent bad debt and prevent financial crime despite being in a highly competitive market. They realised they had more to gain than lose by getting into bed with their competitors.

Civica is hosting sessions to encourage such a debate in the Australian PHI sector which we are confident will lead to a new industry initiative where everyone will be winners, except of the course the fraudsters.