Two scientists hired by the Public Health Service to prepare a report for Congress on lead poisoning in children have resigned in protest, contending that the federal agency plans to delete and dilute critical portions of their work.

The scientists, who resigned June 5, said in interviews yesterday that their draft report, which details the adverse health effects of lead at blood levels common to 17 percent of urban preschool children, suggested the need for more far-reaching and costly remedies than the Reagan administration is willing to consider.

They said a condensed version of a draft sent out last week for interagency review fails to present the national scope of an environmental problem once thought to be confined to poor, inner city dwellers and to detail the health consequences.

"No way in hell you can comprehend the complexity of this problem in a boiled down, very misleading, essentially neutral document," said Paul Mushak, senior author of the draft and adjunct professor of environmental pathology at the University of North Carolina School of Medicine.

"It's one of the most subtle, nastiest rewrites I've ever seen," said coauthor Annemarie F. Crocetti, retired associate professor of community medicine at New York Medical College.

Frank Mitchell, chief medical officer of the PHS' Agency for Toxic Substances and Disease Registry, said the 330-page draft has been cut to 46 pages to create a "readable, usable" document for Congress complete with all vital findings, such as the Mushak-Crocetti estimate of 2.4 million urban preschool children who have lead levels associated with growth and developmental problems.

"We're not suppressing anything," Mitchell said. He said the 46-page version dated June 5 is not in final form, although he expects no substantive changes. The excised charts, appendices and judgments of the two authors will be made available to Congress as "backup data," he said.

Mushak and Crocetti were hired by PHS in November to write the report, which Congress requested last year to analyze the childhood dangers of lead that have been cited in an increasing number of research papers in recent years. Mitchell's division of the PHS was instructed to coordinate the study.

The scientists, who distributed their draft twice for review by government and nongovernment experts, decided to resign a day after Mushak received a call June 4 from Mitchell "indicating that a decision had been made above his head to send a very abbreviated version" to Congress, Mushak said. "I didn't argue. I thought it was a lost cause," he added.

Mitchell refused to disclose the decision-making process behind the revision, saying only that "there was nothing sinister."

Mushak and Crocetti cited numerous omissions from their draft, including its "most important finding," which they said is that "there are no strata" of children who are "exempt from the risk of lead levels high enough to represent a potential adverse health impact." A large number of white, suburban, affluent youngsters are as vulnerable to dangers of lead as traditional victims in the inner city, the draft said.

Crocetti said the social pervasiveness of the problem was purposely emphasized so that members of Congress would realize its dimension and not be lulled into thinking that lead pollution simply plagues the poor.

Mitchell said the socio-economic scope of the problem "isn't missed at all" in the June 5 version.

None of the draft's broad socio-economic judgments appear in a copy of the June 5 text obtained by The Washington Post. A chart is included, giving geographic and financial breakdowns of the preschool children with worrisome levels of lead in their blood.

The scientists also were angered by the short shrift given to lead's dangers at the level of 15 micrograms per liter of blood absorbed by the estimated 17 percent of preschool children in 1984. Their draft described in great detail such ill effects as IQ and hearing loss, growth retardation and impaired hemoglobin formation.

Although the June 5 document includes a chart listing the adverse effects and mentions them in passing, it fails to describe them in the detail that the scientists believe illustrates the seriousness of the threat.

Mitchell said that the 17 percent estimate "speaks for itself" in portraying lead as a public health problem. He said Congress is not interested in "an in-depth explanation of every number. That's why we have backup data."

Another omission in the June 5 version was the 16-page bibliography and extensive references in the draft, which Crocetti said, gave the report professional credibility.

Mitchell said a paragraph had been inadvertently dropped from the June 5 document directing the reader to backup materials for references. It will be restored, he said.