Nearly seven months pregnant with twin girls, Natalie DeStefano felt huge but happy as she waddled into her obstetrician's office in late August.
She was shocked when her doctor, Robert Debbs, finished his weekly exam and gently explained that he needed to admit her to Pennsylvania Hospital. She was in labor -- even though she couldn't feel the contractions -- and he was going to try to stop it.
This was the crisis DeStefano and her husband, John, had hoped to avoid.
The 29-year-old professionals from Mays Landing, N.J., 50 miles southeast of Philadelphia, were facing one of the most controversial, sometimes heartbreaking, problems in obstetrics: Despite many medical breakthroughs, no one has figured out how to prevent premature birth.
No matter what doctors prescribe -- bed rest, weekly checkups, home monitoring machines, antibiotics, labor-suppressing drugs, hospitalization, stitching the birth canal closed -- almost half the women who go into labor too early, deliver too early.
Of course, prematurity is far less dangerous than it used to be, thanks to high-tech nurturing in intensive-care nurseries.
Still, the couple realized that if the twins were born at that point -- nine weeks early, barely 3 pounds each -- their health could have been compromised for life.
Normally, a woman delivers at 38 to 42 weeks of pregnancy. But one in nine babies is born prematurely -- before 37 weeks.
The rate has been creeping up since the federal government began tracking it almost two decades ago. While some of this increase can be attributed to more multiple births resulting from assisted reproductive technology and more induction of labor for reasons of convenience or health, the rest is unexplained.
Very premature babies -- born before 32 weeks -- face risks of respiratory and digestive problems, brain hemorrhages, and developmental delays; about 5 percent die in their first year.
In most cases, preterm births are caused by preterm labor -- the kind DeStefano experienced. But the cause of most preterm labor remains a mystery.
Contractions alone are not necessarily dangerous or painful; many pregnant women have several per hour. The danger comes when contractions involve irreversible changes -- when the birth canal (the cervix) opens, or the woman's "water breaks."
"We don't know what causes the contractions or the cervical dilation. We don't understand this process at all," said Judith Maloni, a nursing professor at Case Western Reserve University in Cleveland, who has researched the problem.
Doctors do know that smoking, certain infections, uterine abnormalities and other factors increase the likelihood of preterm labor. And they have developed tests to help monitor high-risk women.
The dilemma is how to manage a woman when, for example, an amniotic membrane protein test reveals delivery may be imminent.
Usually, the first thing doctors prescribe is bed rest.
"Unfortunately, there is no well-done clinical trial that shows it's beneficial," said Debbs.
"Pregnant women get pretty uncomfortable, so they naturally slow down," Maloni said. "But lying in bed all the time, there's just no evidence it helps."
Like many modern couples, DeStefano, a family law attorney, and her husband, a school administrator, postponed pregnancy so they could establish careers and buy a house -- only to have fertility problems when they tried to have a baby six years into their marriage.
DeStefano had three miscarriages over two years. After the second, they consulted Debbs, a high-risk pregnancy specialist at Pennsylvania Hospital.
It turned out DeStefano had a coagulation problem, easily corrected by a blood thinner. She also had eggs with fragile shells that were being penetrated by too many sperm. In effect, her eggs were being overfertilized.
Conception had to be rigged in a lab dish under a microscope, where each egg was injected with a single sperm from her husband. Three resulting embryos were placed in her womb in February. Two took hold.
Carrying twins put DeStefano at high risk of early labor. Although twins and other multiples make up less than 3 percent of all births, they are 13 percent of preterm births.
By her 23rd week, in mid-July -- with 40 extra pounds on her 5-foot, 2-inch frame -- Debbs recommended bed rest, telling DeStefano to stay off her feet except for meals and trips to the bathroom.
At 29 weeks -- nearly seven months -- Debbs, while doing a vaginal ultrasound, saw that DeStefano's cervix had begun to open, and ordered her into the hospital.
Before she even put on a hospital gown, she was given a shot of betamethasone, a steroid to hasten the twins' lung maturation. Steroids need at least 48 hours to work.
While steroids don't prevent premature birth, they have dramatically reduced the incidence of life-threatening respiratory distress and brain hemorrhages in "preemies."
Next, DeStefano got a shot of terbutaline, one of the tocolytic drugs used to arrest premature labor. When the drugs were introduced about 30 years ago, doctors hoped they would reduce the incidence of preterm birth. They haven't. But they do buy time for the steroids to work.
DeStefano's labor subsided and, after three days of treatment, Debbs was satisfied the crisis had passed.
But what if DeStefano went home to Mays Landing -- an hour from doctors, monitors and delivery rooms -- and slipped back into labor without knowing it?
Debbs tried to reassure her. He explained she would be on a low daily dose of terbutaline, and would use a home uterine activity monitor that would allow a nurse miles away to assess her contractions.
Subject of controversy
Home monitoring is a controversial subject in the medical community.
One large study found women with monitoring services had no better outcomes than those who got a weekly call from a nurse.
A panel of public-health authorities called the U.S. Preventive Services Task Force, its Canadian counterpart, and the American College of Obstetricians and Gynecologists recommend against home monitoring.
Yet the Food and Drug Administration has approved a monitoring device, and thousands of women swear by it.
The only company that sells the home-monitoring service -- Matria Healthcare Inc. of Marietta, Ga. -- cites research that shows monitoring helps prolong pregnancies, thus reducing preterm babies' time in intensive care, which can cost $2,000 a day.
When Destefano went home from the hospital and crawled back into bed, she was leery of the $80-a-day monitor that Debbs had persuaded her health plan to pay for.
Soon, though, she saw the machine as a lifeline. Each morning and night -- and whenever she was worried -- she strapped a sensor onto her abdomen for an hour. The information was transmitted to Matria's regional office outside Philadelphia, and within 30 minutes, a Matria nurse called to discuss the results.
The week the Olympics started, in mid-September, DeStefano felt she had won a gold medal in pregnancy. She had reached 32 weeks, a threshold in terms of the twins' development. The weekly ultrasound showed they each weighed about 4 pounds and were flexing the muscles they would use for breathing.
The next day, at 5 a.m., her water broke.
DeStefano's nine-hour labor and delivery was arduous. After the first twin was delivered vaginally, the second twin's heartbeat slowed alarmingly -- almost forcing an emergency caesarean section. But Debbs managed to pull the second baby out with a suction device.
Gabriella debuted at 2:17 p.m., just shy of 5 pounds. Isabella emerged 36 minutes later, weighing 3 pounds, 14 ounces.
Although the twins breathed on their own, they spent three weeks in intensive care overcoming relatively minor problems, including jaundice and unstable body temperatures.
On Oct. 6, they went home. In the final analysis, the twins were born early, despite medical interventions. But did those interventions save them from being born far earlier and at greater risk?
DeStefano, the elated parent, said, "Absolutely. It was like a miracle that I got those three extra weeks."
Debbs, the scientist, said, "It's impossible to know."