A preliminary study of dioxin in everyday paper products shows that five of seven items sampled pose a slightly higher risk of cancer than the Environmental Protection Agency considers acceptable, according to a report released yesterday.

The assessment of risk, based on hypothetical assumptions and written by the consulting firm of Arthur D. Little Inc., "does not warrant any immediate action," said Bill Waugh, of the EPA's Office of Toxic Substances, because the conservative assumptions may have overstated the extent of exposure to the toxic chemical.

Still, said Waugh, the findings of elevated cancer risks in coffee filters, uncoated paper plates, disposable diapers, tampons and paper towels "raise some concern" and underline the need for in-depth government investigation.

Last month, the paper industry revealed that trace amounts of dioxin were found in a wide range of its products. The EPA, which had turned up small quantities of dioxin in the sludge, effluent and pulp of paper mills, had anticipated the product findings early last year and commissioned the assessment.

Seven common items were evaluated based on assumptions about the frequency of product use, dioxin levels and the rate of dioxin's migration from the product and human absorption.

The projected risks of dioxin in the products were compared to the EPA's standard of no more than one additional cancer for every one million people.

Only coffee filters showed a substantially higher risk. Assuming that the consumer used one filter daily for 50 years, that each filter contained 10 parts per trillion of dioxin (a level commonly found in pulp in the mill study) and that half of that dose was absorbed by the coffee drinker, then the risks of cancer would rise to one additional case for every 10,000 people, according to the report.

The plates, diapers, tampons and towels showed lifetime risks ranging from 1.4 to 6.6 extra cases of cancer per one million people.

Dinner napkins and uncoated sheets of paper showed risks below the EPA standard.

Waugh said that since the consultant probably overestimated the assumptions of the study levels, it is "highly unlikely that the consumer actually experiences risks" at the level cited by Little.