HDM Initiates Study of Relative Efficacy
of Random Sample Versus Claim-By-Claim Health Plan Audits

Healthcare Data Management, Inc. (HDM) has partnered with two academians from the Haub School of Business at St. Joseph’s University in Philadelphia to compare the efficacy of the company’s claim-by-claim method of auditing medical and prescription drug claims versus the random sample methodology that relies on a 200 to 400-claim sample derived from 100 percent of claims for the audit period.

The study was launched the week of June 22, 2009, using actual HDM client data and audit results.

“We’re not looking to pick a fight with actuaries who support audits based on random samples, said David McSweeney, HDM’s COO. “We’re simply trying to illustrate the differences that occur in results when you truly audit 100% of claims versus selecting a random sample from among 100 percent of claims. There can be huge swings in the number and dollar amount of errors caught, and we’re going to quantify that,” McSweeney added.

The study will be conducted by Professor Ronald K. Klimberg, Ph.D., M.S. of the Decision and System Sciences Department at Saint Joseph’s and George P. Sillup, Ph.D., M.S., Assistant Professor in the Department of Pharmaceutical Marketing and a Fellow of the University’s Pedro Arupe Center for Business Ethics.

Before joining the Saint Joseph’s full-time faculty in 2004, Dr. Sillup held various positions, including COO, over a 28-year career in the pharmaceutical and medical device industries.

Among many other accomplishments, Dr. Klimberg has published over 30 articles and delivered presentations in the U. S. and abroad in the areas of management science, information systems, statistics and operations.

Using HDM client data from multiple recent audits, Drs. Klimberg and Sillup will develop simulations of random audits, and then make comparisons to the original HDM audits.

Some of the questions Drs. Klimberg and Sillup have set out to address are:

• What is the magnitude of the sample error in the random audit simulations?
• What is the shape of the distribution of errors?
• What is the average cost per error missed?

“We think the study will have significant implications for self-insured plans that are being audited using the random sample method… particularly now, when there are so many cost escalating factors at play in the healthcare marketplace.

Drs. Klimberg and Sillup will release a white paper on the study in mid-September 2009. The paper will be available for download at www.hdminc.com and www.cfo.com.