Some critics familiar with Food and Drug Administration practices say a regulation issued in 1993, concerning a drug for hepatitis, directed clinical researchers to presume that a new drug is to blame anytime a patient gets sick or sicker during a study.

Yet careful analyses by the National Institutes of Health and the National Academy of Sciences found that the system of clinical safeguards then in place had worked effectively to identify and call proper attention to the hepatitis drug's side effects. Observers say the agency created an arbitrary and significantly lower regulatory threshold for stopping a clinical trial, making the entire drug-development process unnecessarily risk-averse, slower and more expensive.

• The DuPont-Merck Pharmaceutical Co. estimated that its reporting burden has doubled for each perspective drug it develops.

• Amgen Inc. described the substantial practical difficulties of estimating the expected incidence of death and serious adverse events that arise -- not from the drug, but from underlying disease or other medications.

• Medical researchers at the Johns Hopkins University's Center for Clinical Trials found that the new requirements cost $24 million more than the old regulations -- with no commensurate advantage to patients.

The FDA had already pushed the average cost of developing a drug from $359 million to more than $500 million between 1990 and 1993.