CHARLOTTE, North Carolina – Having built a molecular diagnostics program at William Beaumont Hospital (Royal Oak, Michigan), Frederick Kiechle, MD, PhD, noted that Laboratory Corporation of America (LabCorp; Burlington, North Carolina) had recently released financials on its operations, as one of the first companies to enter genomic testing.

"Folks, we are looking at a cash cow," he said of LabCorp.

He began the program at the hospital (which had 1,061 beds, 114,790 emergency room visits and 7,663 employees in 2004) in 1991. Kiechle, who is about to take a new job in Hollywood, Florida, is the past chairman of the department of clinical pathology and past medical director of Beaumont Reference Laboratory (BRL).

"It is an enormously profitable [operation]," Kiechle told the audience at the CLMA (Wayne, Pennsylvania) ThinkLab conference here, where he was speaking about how to build a financially viable molecular diagnostics program. The William Beaumont program has since been selected as one of 35 centers of excellence by Roche (Indianapolis).

When he began the program, Kiechle said he was given $200,000 to buy equipment for the lab and a goal of supporting 10% to 20% of the annual growth of BRL. He also was charged with generating revenue-cycle metrics to monitor financial data for BRL and the hospital, and generating enough money for future capital equipment needs.

The hospital wanted to introduce tests based on spending.

"I told [hospital] administrators this program won't be profitable for three years," Kiechle said. In reality, the program actually lost money for more than three years.

The molecular diagnostics lab began with the technique known as Southern blotting, which takes three to four days of turnaround time. But in 1993, the hospital signed a polymerase chain reaction (PCR) license with Roche to incorporate that method.

Kiechle, who also was responsible for reducing the number of "send-out tests," or tests sent to other labs, said that from 2003 to 2004, send-out tests decreased from 40,893 to 28,101, a 31.3% decline.

Molecular tests per year went from 387 in 1992 to 43,824 in 2004, he said.

One of the measures that the hospital undertook was an effort in what Kiechle called "up-selling," or a conference for "stakeholders," primarily physicians, who ultimately are the ones who order such tests. It also was an educational effort, because surprisingly, he said, sometimes physicians don't order tests simply because they aren't aware that such tests are available.

The hospital now is in its 14th year of hosting the conference.

The lab also undertook an effort to analyze the utilization of its outreach laboratory, measured by spending. In other words, he said, the hospital wanted to find out "which doctors were ordering which tests."

"You can play a lot of games with these numbers, but without them you can't do anything," he said of the data he shared with the audience.

One of the challenges of operating molecular labs are the patents on genes, which he said increased 62% in the U.S. from 1996 to 2001, which affects licensing required for DNA testing.

At the end of 2001, Kiechle said there were 18,174 patents on human DNA sequences and 365 patents on single-nucleotide polymorphisms (SNPs).

One of his favorite quotes on this issue comes from Nature Biotechnology, Thomas SM, et al.: "Lest we forget, the patent system was established to protect and promote inventions, not investments." Patenting of DNA has turned this thought "upside down," Kiechle said.

Among the patent problems he referred to, as pointed out in a 2005 article in Science, are that patent examiners spend only about 18 hours per patent reviewed. Also, he disagreed with the fact that patent examiners are financially rewarded for quickly pushing patents through the patent office.

He described the impact of gene patents on diagnostic testing as "major."

In a survey of 132 of 211 molecular lab directors, 25% said they discontinued performing genetic tests and 53% did not develop new clinical genetic tests.

In sharing information regarding the estimated revenue/profit of the molecular program at the hospital, Kiechle said that the lab studied 26 tests. Of that number, six tests lost $200,000, and 20 tests made about $1.95 million for a margin of 58.5%.

If a company had that kind of margin, he said, "I'd buy stock in it."