The average annual pregnancy rate was 34 per 1,000 girls in the study — far below the national average of 158.5 for sexually active teens.
Doctors say the results show that when money is not a factor, teens will chose IUDs and hormone implants over less reliable methods such as birth control pills or condoms. Nearly three-fourths of teens in the study picked long-acting methods; only 5 per cent of U.S. teens use those now.
"When costs are removed, young people and families will use these effective methods," said Dr. Mary Ott of Indiana University. She had no role in the study but led an American Academy of Pediatrics' policy statement earlier this week urging long-acting contraception, a stance other doctor groups endorse.
She said the study shows that more can be done to curb teen pregnancy. "It's good for the teens and it's cost-effective for society" to offer free, effective birth control, Ott said.
One study participant, Marlice House, 25, of St. Louis, said she got an IUD in 2008 as a teen in the project after having tried two less reliable methods.
"I wanted to go to college and I didn't want to have a child at an early age," she said.
The study was led by Gina Secura and Dr. Jeffrey Peipert of Washington University in St. Louis. Results are in this week's New England Journal of Medicine.
Paying for birth control, especially IUDs, or intrauterine devices, has long been a contentious issue. In June, the Supreme Court sided with the company Hobby Lobby and said that employers with religious objections could opt out of a federal rule requiring insurers to cover contraceptives. The Obama administration has been looking at other ways to increase access to birth control.
The study involved a project in St. Louis that offered 1,400 girls their choice of methods but stressed the benefits of IUDs and hormone implants that last three to 10 years. The girls were at high risk of unintended pregnancy — nearly 500 were just 14 to 17 when they enrolled, and half of those young teens had already been pregnant and 18 per cent had had an abortion.
During follow-up between 2008 and 2013, the average annual birth rate was 19.4 per 1,000 teens in the study, versus 94 for sexually active teens in the general population. The comparable abortion rates were 9.7 versus 41.5.
Of the 56 pregnancies that occurred during the study, teens reported using no birth control at the time of conception in 25 cases. Only two with IUDs became pregnant. Thirteen said they had been using birth control pills and nine reported using condoms. Other methods accounted for the rest.
Researchers relied on teens to report pregnancies in surveys throughout the study, so some may have been missed. National teen pregnancy rates have been falling, but even so, the differences seen in the study "are substantial and of public health importance," the authors wrote.
The project was sponsored by the Susan Thompson Buffett Foundation, the National Institutes of Health and others. Two authors have financial ties to companies that make contraceptives.
The project also included about 7,500 older women. Similar results — lower pregnancy, birth and abortion rates — for the group as a whole were reported two years ago.
Marilynn Marchione can be followed at http://twitter.com/MMarchioneAPSuggest a correction