My wife wore a serious expression as she entered our living room. “I think my water just broke,” she said. My heart skipped, half in excitement and half in concern. Her due date had arrived, so it was time for the baby to come. But she wasn’t yet in labor, and contractions are supposed to be well underway before the gush that signals the rupturing of the amniotic sac.

Events, in other words, were out of order. Her nearly textbook-perfect pregnancy had taken an unexpected last-minute turn.

The medical dilemma we would soon face is far from rare. As many as one pregnancy in 10 — about 400,000 annually in the United States — are complicated by the expectant mother’s water breaking before labor begins. Doctors call it PROM, for prelabor (or premature) rupture of the membranes. (The condition has nothing to do with premature birth.) While it’s unclear why PROM happens to some women and not others, studies suggest that women who experience PROM once are likely to have it happen again.

Despite how common PROM is, there’s no universally accepted way to care for women who experience it. Childbirth professionals disagree, and practices can vary from one hospital to the next.

If the pregnancy is at term, as my wife’s was, obstetricians typically recommend prompt induction of labor, in which drugs are used to bring on contractions and hasten delivery. Any delay, they say, poses an unacceptable risk of infection. That’s because bacteria in the lower genital tract can migrate upward and enter the uterus, potentially threatening both mother and fetus, which is no longer sealed inside the amniotic sac’s protective envelope.


But other professionals, including many midwives, argue against universally inducing. In their view, induction can increase the need for subsequent interventions, potentially culminating in a Caesarean section. They maintain that many women with PROM can safely stay at home and wait for labor to begin naturally.

My wife and I had been oblivious to the simmering debate, but we didn’t remain that way for long. Two phone calls later, we’d learned that our obstetrician wanted us in the hospital as soon as we could make it, while the midwife-in-training we’d hired to be our doula, or personal birth assistant, urged us to stay put.

I’m a medical journalist. My wife is a doctor. Through our work, we have greater access than most people to the sprawling library of medical evidence that doctors use to guide their decisions. That August night in 2009 we felt sure that close examination of relevant studies would tell us the right move. So we pulled up trustworthy medical sites, punched in passwords, downloaded studies and guidelines, and began reading. But a definitive answer remained elusive.

Sometimes, medical science can’t address — or, at least, hasn’t addressed — the pressing questions patients can find themselves facing. What to do when PROM occurs at full term, we realized, was one such instance.

When obstetrician Aaron B. Caughey moved from Boston to California years ago, he was stunned that his new colleagues’ approach to PROM was diametrically at odds with the one familiar to him. In Boston, women who experienced PROM at term were routinely admitted to the hospital and induced. In San Francisco, they were often sent home to wait for labor to begin spontaneously.

Which approach was better, wondered Caughey, who is now the chair of obstetrics and gynecology at Oregon Health & Science University in Portland. What he and others widely recognize as the most important medical trial on PROM at term — a three-year-long international study that my wife and I pored over that summer night — had been published in 1996 in the New England Journal of Medicine. In it, researchers randomly assigned more than 2,500 women who were experiencing PROM to undergo immediate induction. About 2,500 similar women with PROM received so-called expectant management, in which labor was not induced unless medical complications arose or four days passed without contractions beginning on their own.

The trial’s leaders concluded that mothers were less likely to experience an infection or postpartum fever if they underwent immediate induction. A follow-up study found that the newborns were about twice as likely to suffer an infection if the mother hadn’t gone into active labor within 24 hours of her membranes rupturing as were infants whose mothers had gone into labor within 12 hours.

In other words, induction is better for mom and at least as good for the baby. Case closed, right?

Yes, says the American College of Obstetricians and Gynecologists. An ACOG bulletin intended to guide obstetricians’ clinical decisions states, “For women with premature rupture of the membranes (PROM) at term, labor should be induced . . . to reduce the risk of [maternal infection].”

But the American College of Nurse-Midwives espouses a different view. Its position statement on PROM says that women should be “allowed to select expectant management” if they meet certain criteria indicating they are at low risk for infection — and if precautions are taken to keep their risk low.

The nurse-midwives are among those who take issue with whether the 1996 trial’s conclusion is meaningful today, given improvements in obstetric care that have made expectant mothers safer than they once were. The trial itself contributed to some of those improvements.

For example, medical providers now know better than to needlessly probe women who are awaiting the onset of labor. In the trial, all women received a “baseline” vaginal exam, which involves inserting a gloved hand. More than 55 percent of the women in the expectant-management group received at least four such exams by the time they delivered, and some of them received eight or more.

In a follow-up study published a year after their initial paper, the University of Toronto-led researchers reported that women in the trial who received eight or more exams faced five times the infection risk of women who received no more than two. Doctors, midwives and nurses took note.

“There’s really no need,” said Michelle Collins, a spokeswoman for the nurse-midwives group, to do such exams before labor begins. “Once you’ve stuck your hand in there, you’ve introduced bacteria.”

The trial also underscored the dangers of group B streptococcus, or GBS, a kind of bacteria that can reside harmless in the genital tract of a healthy woman but can cause mischief during labor. Nowadays, health-care providers test pregnant women for GBS before they reach full term and administer antibiotics if the bacteria are present.

But in the 1990s, precautions surrounding GBS weren’t always taken.

Some GBS-positive women in the trial were assigned to expectant management — which obstetricians and nurse-midwives now agree is unwise — and a few didn’t even receive antibiotics. One baby in the expectant-management group died of GBS infection, a tragedy that today’s standard of care might have prevented.

Another rigorous trial on PROM may never be mounted, Caughey said.

That’s partly because the first trial produced, at its core, a reassuring finding: No matter what mothers do, the risks are fairly low. While Caughey did once treat a PROM-affected mother who nearly died of an infection after waiting too long at home, such infections are rare, he said, and can generally be treated with antibiotics.

To assess the importance of rapid induction, Caughey and several colleagues recently conducted a study in which they reviewed 20 years’ worth of records at University of California at San Francisco. They identified 3,841 cases in which women had experienced PROM at term, and they compared the rates of infection in those who delivered relatively rapidly after their membranes ruptured and those who took longer to deliver.

Infection rates began to rise in women within 12 hours if they hadn’t delivered by that time, the researchers found. That supports immediate induction as the best move from a purely medical standpoint, Caughey said.

But medical factors aren’t the only ones worth considering, he quickly added. Patient preference matters, too.

Part of the variation in practices between the East and West coasts, he said, reflects regional differences in patients’ and doctors’ tolerance for intervening in what are essentially healthy pregnancies. If an expectant mother is informed about the risks of sitting things out, he said, it’s her prerogative to do so.

On that point, Collins agreed. “Women are most satisified with their births when they feel empowered, [when] they have been part of the decision-making process,” the nurse-midwife said.

In the end, my wife and I decided to wait. She had a strong desire to use no medications for induction or pain relief unless absolutely necessary. Honoring her wishes, her obstetrician agreed to postpone induction — but only until 6 a.m. the following morning. With a plan in place, we went to bed.

At 1 a.m., my wife awoke to the first contractions of her labor. There would be no need to induce after all. Labor progressed quickly and smoothly and, as it happened, we arrived at the hospital right about 6 a.m. True to her goal, my wife used no medications in delivering our son at 8:55 that morning. Had we induced, that goal would have been out of reach.

Did we make the medically correct decision? Experts may continue to disagree on the answer. But mother and child are doing well. That’s enough for me.

Harder is the general manager of health and science at U.S. News & World Report.