A number of physicians hired by drug companies to test drugs in humans for the prescription market have committed fraud and other serious abuses, the Food and Drug Administration told a Senate hearing yesterday.

Abuses ranged from misrepresentations made to women in labor to win their "informed consent" for use of experimental drugs in their newborn, to counting the dead as active participants in experiments, the FDA said.

Commissioner Donald Kennedy told the Senate health and scientific research subcommittee that some of the findings could be viewed as "horrible" or "inconceivable." Subcommittee Chairman Edward M. Kennedy (D-Mass.) termed them a "perversion of science."

Agency investigators made 26 inspections where they had reason for suspicion. They found abuses committed by 13 physician-testers. None was named because investigations are continuing. Criminal prosecution of two of the doctors is in the government pipeline.

The 13 doctors were found doing studies for 48 major pharmaceutical manufacturers, including such leaders as Hoffmann-La Roche, Bristol-Myers, the McNeil Laboratories susbidiary of Johnson & Johnson, and Endo Laboratories.

In testimony, officials of the four named companies said they had been unaware of abuses and acted swiftly to correct them on becoming aware. None claimed to have notified the FDA.

The companies had retained most of the doctors for five to seven years. One who had been hired over a 15-yeard period never before had been audited by the FDA, Dr. Michael Hensley, the agency's director of scientific investigations, told reporters.

The commissioner said that the audits were a sampling too preliminary and possibly too biased - because they originated in indications of wrongdoing - to permit a reliable judgment on the state of affairs in the entire industry and in the corps of 6,900 physician-testers, about 2,000 of whom are active in any given year.

Nonetheless, said Sen. Kennedy, the public is "at unacceptable risk," because of clear signals that the FDA unknowingly could be deciding on the basis of sloppy or fraudulent research whether a medicine is safe, effective and deserves to go on sale.

For various reasons, the 48 companies got no drug on the market with data from the 13 doctors, although in a few cases they were seeking FDA approval of a new use for an already-marketed product.

The commissioner said the audits found:

Reports experimental drug, never gave informed consent to being tested, didn't have the disease the drug was supposed to treat, or got dangerously excessive doses.

False claims to the companies that approval of a drug study had been given by an institutional review board (IRB), which the FDA requires wherever clincial investigations are done to review benefit/risk decisions and ethics.

Sen. Kennedy disclosed that after inspecting 100 IRBs, the FDA rated the consent to testing given by patients as inadequately "informed")

Use of multiple experimental drugs to the exclusion of proven therapy, making the patient "nothing but a guinea pig;" in some cases, a medicine that interfered with the experimental drug was administered simultaneously, making the data from the study "valueless."

The centerpiece of the four-hour hearing was testimony by the FDA's Hensley about six of the 13 physicians, whom he identified only by numbers. For each drug they claimed to test, one of the 48 companies paid them between $24,000 and $106,000 over a period of up to a year. Their reports invariably were optmistic, sometimes to the point of inciting suspicion within the sponsoring firm.

Physician No. 2 got the $106,000 for "testing" a drug that Bristol Laboratories hoped would stimulate the performance of the senile. The test site was a nursing home, somewhere in Florida, characterized by a health official as "rat-infested dump."

The administrator had been indicted for embezzling funds from patients, mainly brain-damaged alcoholics. The IRB proved to be a phony. The physician has disappeared. The medical license found among his effects was forged. His credentials, testified Bristol research chief Stanley T. Crooke, had seemed to be "ideal."

Doctor No. 10 claimed to Hensley that a friend was doing his work on a McNeil muscle relaxant. Half-way through the test, the FDA official read in a newspaper of the friend's death. But No. 10 continued to report the friend's results.

No. 9, a senior physician at an unidentified Veterans Administration hospital, tested a Roche sedative. Aided by VA rules, he concealed from the company the deaths - for undisclosed reasons - of two of the 11 participants. He told Roche all 11 got the drug; patient charts show only one did.

"I'm surprised to see this information myself," said Roche vice president John J. Burns. He said Roche "corrected" the physician's data; before submitting it to the FDA; Hensley said the firm, the world's largest drug company, "reinterpreted and changed the data."