Whether a woman who has had a baby by Caesarean section gets a "repeat" operation at the end of her next pregnancy depends more on whether she is a patient in a for-profit hospital than on her medical condition, according to a new study.

Despite national guidelines stating that most women who have had a Caesarean section can safely have a vaginal birth following subsequent pregnancies, the greater profitability of the surgery appears to be keeping the rate of "repeat" Caesarean sections high, according to a study of California births that appears in today's issue of the Journal of the American Medical Association.

One out of four babies in the United States is now born by Caesarean section, a proportion that is four times higher than it was 20 years ago. Thirty-six percent of Caesarean sections are "repeat" operations, performed on women who have had the surgery with a previous pregnancy.

The risk of dying from a Caesarean section is two to four times higher than from a vaginal delivery, and the operation entails a longer hospital stay and a much greater risk of infection, said Sidney M. Wolfe of Public Citizen's Health Research Group, a consumer organization. Surgery is also more expensive, bringing a hospital up to $2,500 more per delivery than a vaginal birth, he said.

Obstetricians long believed cutting through the uterus greatly increased the risk that the organ could rupture during labor if a woman tried to have a vaginal delivery with a subsequent pregnancy. The rule was, "Once a Caesarean, always a Caesarean."

But in the last two decades, a number of studies have disproved that theory and have prompted national organizations, including the American College of Obstetricians and Gynecologists, to recommend that doctors allow women who have had a previous Caesarean section to try to deliver vaginally, unless there is some other medical reason for surgery.

In spite of such recommendations, only 12.6 percent of American women with previous Caesarean sections go on to have vaginal deliveries. The new study, by epidemiologist Randall S. Stafford of the University of California at Berkeley, suggests that economic factors may partly explain doctors' apparent reluctance to follow the guidelines.

Stafford studied 45,425 births that occurred during 1986 to California women who had previously had Caesarean sections. Overall, 10.9 percent of their subsequent deliveries were vaginal. He found that a woman's likelihood of having a vaginal delivery depended on what kind of hospital she entered and on whether she had health insurance.

Private, for-profit hospitals had the lowest rate of vaginal births in such patients: 4.9 percent. The University of California Hospital had the highest rate: 29.2 percent. Hospitals owned by Kaiser-Permanente, a health maintenance organization (HMO) whose doctors are given strong incentives to keep down medical costs, had the second highest rate of vaginal births, 19.8 percent. Private, nonprofit hospitals, which lack such incentives, had a rate of 8.2 percent.

Women in the study who had private health insurance were much more likely than uninsured women to have surgery. Only 8.1 percent of privately insured women had vaginal births, compared with 18 percent of women responsible for their own medical bills and 25.2 percent of those whose bills were covered by "indigent services."

Women whose babies were born at hospitals with more than 3,500 births per year were more than three times as likely to have a vaginal delivery as women whose babies were delivered at hospitals with less than 1,000 births per year. The rate of vaginal deliveries in women with a previous Caesarean section was also three times higher in teaching hospitals -- those where medical students and residents are trained -- than in non-teaching hospitals.

Wolfe said national data collected by his organization support the study's conclusion that profit motives can influence decisions about whether pregnant patients should have surgery.

"The decision-making as to what is best for the patient gets distorted by an economic factor that has nothing to do with the patient," he said.