A House hearing yesterday unfolded a bizarre story of how poor coordination among federal agencies led the Environmental Protection Agency to rely on a heart specialist who was under investigation by the Food and Drug Administration and Veterans Administration when it tried to set an important new air-quality standard for carbon monoxide.

EPA's five-year quest for a revised standard was nearly derailed in March 1983 by a news report that the FDA had barred Dr. Wilbert Aronow, former chief of cardiovascular medicine at the VA Medical Center in Long Beach, Calif., from testing drugs in humans. The FDA said that Aronow had falsified test results and "repeatedly or deliberately" violated FDA rules for four years.

Yesterday, the House Energy and Commerce subcommittee on oversight and investigations revealed that in 1981 the VA had found that Aronow had received $72,351 for "consulting services" from two drug companies, but had reported that he had received no outside "remuneration."

Under VA regulations, according to testimony yesterday, Aronow was supposed to report the income and turn at least some of it over to a special fund for veterans. VA Administrator Harry N. Walters testified that none of the money has been returned.

Walters also told subcommittee Chairman John D. Dingell (D-Mich.) that the VA had referred the case three times to federal prosecutors. He said criminal prosecution was finally declined, partly on the ground that "administrative or civil remedies were available."

As told at the hearing, the carbon monoxide (CO) story began in 1971, when EPA set the current standard for emissions of the gas, which can have adverse effects on persons with heart disease. The agency relied on a study -- not Aronow's -- that has not been replicated, but has not been refuted.

Six years later, EPA began revising its earlier standard. The agency acknowledged that its study was not scientifically adequate and issued a so-called "criteria document" that relied heavily on Aronow's published CO studies.

In June of the same year, the FDA -- unbeknownst to EPA -- began investigating Aronow's drug research, and in June, 1980, it notified him of the allegations against him. A year later, an FDA official saw a Federal Register notice about a study Arnow was doing for EPA and wrote an EPA official about the drug investigation. The alert "was reportedly lost in the maze at EPA," Dingell said.

FDA also notified the VA when the drug investigation began in mid-1979 and again when Aronow, in October, 1982, signed a consent agreement curtailing his drug-testing activities.

As a result, a VA professional standards board had decided by April, 1980, that Aronow had falsified scientific data and recommended he do no more research on humans.

Yet "the VA deliberately failed to inform EPA of any of these events," Dingell charged.

Dingell said that after a March, 1983, Washington Post report had disclosed FDA's consent agreement with Aronow, EPA convened a panel that advised it not to rely on Aronow's data.

In support of the revised CO standard, EPA, in August, 1983, issued an addendum to the 4-year-old criteria document. Dingell pointed out that it contained nine references to Aronow's studies.

EPA Administrator William D. Ruckelshaus pledged to Dingell that he will rely on none of Aronow's data when the agency finishes revising the standard. He, Walters and FDA associate commissioner Stuart Nightingale also said that they would make major changes in their reviews -- including better coordination with each other.