The Reagan administration yesterday formally proposed new regulations aimed at speeding up approval of new prescription drugs.

One proposal would require drug companies seeking marketing approval from the Food and Drug Administration systematically to update safety information on their products. This stemmed from recent evidence that there were numerous unreported deaths and injuries in users of Oraflex, an arthritis medicine pulled from the market 3 1/2 months after the FDA released it for sale.

Perhaps the most controversial proposal is one that would allow the FDA to approve a new drug solely on the basis of data from human testing done abroad. Rep. L. H. Fountain (D-N.C.) questioned the proposal at an Aug. 3 hearing, citing several FDA documents that revealed serious problems with testing in other countries, particularly those that bar agency medical officers from checking test data.

The proposed rules are being published in today's Federal Register. The public has 60 days to comment before final action is taken.

The proposals "represent the most significant reform of the new-drug approval process since the 1960s," Richard S. Schweiker, secretary of Health and Human Services, and FDA Commissioner Arthur Hull Hayes Jr. said in a joint statement yesterday.

Schweiker had announced the proposals, in general terms, to a pharmaceutical industry meeting June 23, estimating "that 200,000 persons will benefit each year from significant new drugs, which will become available sooner than under present rules." Yesterday, however, he and Hayes said simply that the new proposals "will benefit thousands of Americans."

The FDA has acknowledged the proposals would not have a major impact on most new drugs that may be therapeutic breakthroughs, because these are already on a fast regulatory track. As for drug trials done abroad, Schweiker said in June that "existing policy makes it virtually impossible to approve [new-drug marketing applications] based solely on foreign testing."

Then at the Aug. 3 hearing of Fountain's House Government Operations subcommittee, FDA officials testified that they would continue to audit all major U.S. human studies offered in support of new-drug applications while indicating that they would try to assess foreign clinical trials only if they had good cause for concern about the validity of the data.

The results of this, Fountain and his staff suggested, could be a "dual standard" based on an implied assumption that foreign studies are more reliable than U.S. studies, and also an incentive to drug companies to do more research abroad.

Fountain produced FDA documents involving Arlidin, an existing drug that USV Pharmaceutical Corp. wants the agency to approve for a new use: treating symptoms of mental deterioration in the elderly. USV gave the agency data from studies done in Oak Forest, Ill., by Dr. Dodda B. Rao and in Ottawa, Canada, by Dr. Stanley E. Goldstein.

FDA inspectors found serious problems with the Illinois study, and then went to Ottawa in November 1980 to inspect Goldstein's records. But, FDA supervisory investigator David R. Foltz said in an internal memo, Goldstein turned the inspectors away, citing confidentiality provisions in an Ontario law.

Fountain also produced reports on FDA inspections in Mexico. One report said that a Mexican researcher for Ovcon-35, an oral contraceptive, claimed to have discarded records of a favorable study that was used in promotional materials by the manufacturer, the Mead Johnson subsidiary of Bristol-Myers. Another report said that neither "primary data nor original documentation" were found to support claims in three different studies of Stadol, a Bristol pain-killer.

The FDA's Smith said Friday, "If we have doubts or some problem we can't resolve," the agency can require "a confirming study" to be done in the United States.

By contrast, Dr. John O. Nestor, who had responsibility at FDA for the Arlidin application until he retired a year ago, termed the controversial proposal a bad idea. "We always had to be very careful about accepting foreign data because we did not have the means of checking on the validity of foreign data that we have in checking U.S. data," he said.