New York- With the help of their doctors, women planning to become pregnant should take an inventory of the medications they take, researchers from Canada advise.

In a study, they found that many pregnant women still take medications long known to cause birth defects.

Some medications with known fetal risk, such as drugs that control epilepsy, are essential during pregnancy, Dr. Anick Berard, at the University of Montreal in Quebec, noted in an email correspondence to Reuters Health.

Other medications, such as those that treat severe acne, anxiety and psychiatric drugs, antibiotics, and many drugs prescribed for heart disease and medical conditions, "can and should be avoided," according to Berard.

Women should understand the side effects of any drug they are taking -- especially drugs treating a chronic condition -- and plan pregnancies to avoid or minimize risks such drugs pose to babies, Berard added.

For the 5 years between January 1998 and the last day of 2002, Berard and colleagues analyzed the prescriptions filled by pregnant women for drugs available at the time and known to pose fetal risks.

Their report, in BJOG: An International Journal of Obstetrics and Gynecology, shows 56 percent of 109,344 pregnant women filled at least one medication prescription. A total of 6.3 percent (6,871 women) did so for at least one medication known to pose a risk to the fetus.

"These pregnancies were associated with an elevated number of (pregnancy terminations) and babies born with major (birth defects) in comparison with the expected numbers in the population," they note.

Specifically, terminations occurred in 47 percent of the pregnancies exposed to drugs with known fetal risks. Six percent of these pregnancies ended in miscarriage.

By contrast, in the much larger non-exposed group about 36 percent of the pregnancies had been terminated and fewer than 5 percent ended in miscarriage.

Berard's team further identified birth defects in 8.2 percent of 2,842 infants exposed to risky drugs during gestation and available for assessment, compared with 7.1 percent of the 59,287 infants not exposed. This is "a statistically significant difference," they note.

They emphasize, however, that it cannot be concluded that the drug exposure caused the birth defects. These pregnancies may have also been exposed to other harmful agents or maternal health conditions, they point out.