A testing laboratory some of the nation's leading generic drug companies hired to help win government approval for their products used seriously flawed scientific methods and procedures, according to a Food and Drug Administration inspection report released yesterday.

The FDA examined 13 of the hundreds of drug studies conducted by BioDecision Laboratories over the last 18 years. In 11 of them, "some or all the submitted data has been called into question based on such practices as manipulation of control data, selective reporting of sample values, improper calculation of results, etc.," the FDA said.

BioDecision officials said the Pittsburgh-based company re-tested all but one of the drugs mentioned in the report and found no unsafe drugs were approved because of testing errors. The FDA said it found no "evidence of fraud, conspiracy and conflict of interest" at BioDecision.

The FDA report is the latest in a list of problems faced by the generic drug industry, which was shaken last summer by allegations of shoddy manufacturing practices and corruption.

The flaws at BioDecision uncovered by the inspection also led to calls in Congress for closer FDA scrutiny of the generic industry.

In a letter to the FDA, Rep. John D. Dingell (D-Mich.) -- whose Energy and Commerce subcommittee on oversight and investigations spearheaded last year's generic industry investigation -- called for the agency to withhold approvals of all drugs that were tested by BioDecision.

"Given the impact on the generic industry from corruption found to date, this latest revelation of impropriety is truly calamitous," Dingell wrote.

BioDecision is one of several commercial laboratories in the drug business that perform what are known as bioequivalence tests. When generic drug firms produce their copycat, low-cost version of a brand name drug, they generally send it to a company like BioDecision for validation.

The laboratory tests the generic product against the brand name it is copying, measuring whether the new version performs as effectively as the original. FDA approval is contingent on the results.

What the inspection revealed is that BioDecision, which performs about 20 percent of the generic industry's bioequivalence testing, violated basic scientific practice and operations in 15 areas. Among the problems identified by the FDA:

In one instance, the generic drug being evaluated and the drug it was being compared to were mixed up, making it impossible to interpret the results of studies.

Laboratory procedures did not follow approved guidelines, data was often arbitrarily and inexplicably rejected and methods were changed without evaluation or authorization.

Simple calculations of the dosages of drugs being tested were incorrect, throwing final results into question.

White correction fluid was used to obliterate and change original results of drug tests.

Personnel were poorly trained, "resulting in significant undetected errors in data."

"There are a couple of things here that are really worrying," said Arthur Kippe, senior director of pharmacy and scientific affairs at the American Pharmaceutical Association, a professional assocation for druggists. "If this company can't make a simple calculation and figure out something like the dilution factor, then you have to wonder what else they can't do . . . . This company didn't even have a quality assurance officer in place until 1988. Give me a break. That's just unconscionable."

Officials of BioDecision said they have responded to each complaint in the FDA report and corrective measures are underway.

"We view the entire process as very constructive and have taken it extremely seriously," said BioDecision President Robert Lecher. "We don't take issue or disagree with any of the observations."