Everyone within the health care community recognizes that the traditional ‘pay-and-chase’ model for fraud cost reduction has some serious limitations. 

Recovery of the lost funds is in some cases a mere single digit percentage point (with some exceptions – but not many) of the moneys paid.  And the effort and difficulty involved in prosecution is significant drain on time, money and resources.  Also many of the larger, conclusive fraud schemes are criminal in nature (organized crime), so the parties or money has a tendency to disappear. 

It would appear that the allure of ‘pre-payment’ fraud detection would address all of the problems.  However, there are a few catches.  Health care claims are far from real-time; there is no comparison to the banking world.  Firstly, the claims are not received when the service is provided, and most health care systems do not allow for the access to their systems of record, requiring a day to extract the data to a data warehouse. But that is only the start.  What about the requirement to pay ‘none suspicious’ claims in a reasonable time coupled with the number of potentially fraudulent claims and the limited number of investigators?  There is also the issue with the time required to triage and investigate a potentially fraudulent claim – more and more claims accumulate in the queue. 

It all translates to a very large potential bottle neck of what may be fraudulent claims.   This is by no means to say that pre-pay is not a great money saver, and by far one of the best means to reduce fraudulent claims costs, only that the Holy Grail of pre-pay must be approached in a reasonable fashion with reasonable expectations. One method that seems to get significant results is to scrape the cream of the most likely and most expensive fraud off the top and drop that into the pre-pay bucket to limit the number of claims to those that are most likely to be fraudulent and bring the most value to the organization.