A new device that could make life easier for America's 16 million diabetics is available in Europe.

But don't ask your American doctor about the Diasensor; chances are he or she has never heard of it. That's because the Food and Drug Administration has managed to throttle four years of effort by the product's makers to get government approval and bring this new device to market.The Diasensor is described by its makers as the world's first "non-invasive" glucose monitor. It was developed for diabetics who must prick their fingers at least once a day to help monitor blood-sugar levels. If a diabetic's blood sugar becomes too unbalanced, the result can be blindness, kidney disease, loss of limbs or even death.

But rather than using a needle, the Diasensor uses an infrared beam of light to determine the body's glucose level. The machine's makers - Pittsburgh-based Biocontrol Inc. - readily admit that their product doesn't work on all diabetics. But it does work on some, a claim they say is backed by clinical evidence.

So far, they've had a hard time convincing the FDA. And Bio-con-trol officials we spoke to believe that's because the expert review panel that studied the application was biased from the start.

Biocontrol spokesman Jack Nard, who has waged a one-man war against the FDA for more than two years, believes the panel rejected his company's product because the panelists themselves had built-in conflicts of interest that corrupted their evaluation of the Diasensor. He cites one panelist who was employed at the time by a direct competitor of Biocontrol.

The FDA denies any bias in its decisionmaking and says Bio-con-trol's application was rejected because company officials couldn't prove the product worked.

Nevertheless, FDA official Bruce Burlington acknowledged in congressional testimony that he allowed the late Dr. Julio Santiago to serve on the advisory panel even though he held a paid position with a company that produces finger-prick equipment. Santiago led the panel's opposition to Biocontrol, at one point referring to the Diasensor as "the frigging machine." Witnesses also describe Santiago as making disparaging comments about Biocontrol stock owners before the meeting.

Nard's allegations have found receptive ears within the congressional subcommittee that oversees the FDA, which has been fighting since 1995 to reform the agency. As a committee source told our associate Kathryn Wallace, the FDA generally doesn't waste time with panel reviews for machines that don't work.

"This case has the markings of a set-up to push a preordained decision," the source told us.

Nard says that the problems began the moment blueprints of the Diasensor were first sent to the FDA in 1994. The plans languished well beyond the agency's 180-day deadline. Then Biocontrol was told three times to perform more studies and resubmit data.

The agency completely altered the topic list for the review panel the night before the review. The company tried to keep up with the shift in agenda and laboriously compiled information for the panelists - data Nard claims the reviewers never saw.

But even if all the reviewers received Biocontrol's data, chances are the results would have been the same. All but a small minority of the 17 panelists lacked the expertise to comment on this machine, congressional sources say, which is not uncommon in FDA reviews.

"Panels are too vulnerable to manipulation," one congressional investigator told us. Though three House bills are pending to hasten the approval process for experimental treatments, chances are slim any will become law this year.

United Feature Syndicate Inc.