New evidence about a cancer operation in women finds a higher death rate for the less invasive version, challenging standard practice and the “less is more” approach to treating cervical cancer.

The unexpected findings are prompting changes at some hospitals that perform radical hysterectomies for early-stage disease.

The more rigorous of the two studies was conducted at more than 30 sites in a dozen countries. It found women who had the less invasive surgery were four times more likely to see their cancer return compared to women who had traditional surgery. Death from cervical cancer occurred in 14 of 319 patients who had minimally invasive surgery and 2 of 312 patients who had open surgery.

Results were published online Wednesday by the New England Journal of Medicine.

Radical hysterectomy is standard treatment for women with early-stage cervical cancer. Rates are declining because of widespread screening. The number of operations has fallen, too, to several thousand a year in the United States. Some women with early-stage cervical cancer are choosing fertility-sparing techniques, treatments not included in the new research.

In both studies, researchers compared two methods for radical hysterectomy, an operation to remove the uterus, cervix and part of the vagina. The surgery costs around $9,000 to $12,000 with the minimally invasive version at the higher end.

Traditional surgery involves a cut in the lower abdomen. In a newer method, a surgeon makes small incisions for a camera and instruments. Patients recover faster, so laparoscopic surgery, which has been around for more than a decade, gained popularity despite a lack of rigorous long-term studies.

It’s not clear why it failed to measure up. Experts suspect there may be something about the tools or technique that spreads the cancer cells from the tumor to the abdominal cavity.

Some hospitals went back to traditional hysterectomy after the results were presented at a cancer meeting in March.

“We immediately as a department changed our practice and changed completely to the open approach,” said Dr. Pedro Ramirez of the University of Texas MD Anderson Cancer Center in Houston.

Ramirez led the more rigorous study, which randomly assigned 631 patients to one of two surgeries. After 4½ years, the rate of those still living without disease was 86 percent with less invasive surgery and 96 percent with traditional surgery.

The experiment was halted early last year when the higher death and cancer recurrence rates showed up. The original plan was to enroll 740 patients in the study, which was funded in part by surgical device maker Medtronic.

For 33-year-old Alicia Ackley, who was diagnosed with cervical cancer in July, the recommendation for traditional surgery came as a surprise, but she followed the advice of her doctor at MD Anderson after hearing about the research. Tests following her September operation show no signs of cancer.

“I’m very glad I went that route,” Ackley said. “The open hysterectomy got everything.”

The other study looked at 2,461 women with cervical cancer who had radical hysterectomies from 2010 through 2013. It found a 9.1 percent death rate after four years among women who got minimally invasive surgery compared to 5.3 percent for traditional surgery.

“We’re rethinking how we approach patients,” said study co-author Dr. Jason Wright of New York-Presbyterian Hospital. “There’s a lot of surprise around these findings.”

The research is “a great blow” to the technique and the findings are “alarming,” said Dr. Amanda Fader of Johns Hopkins Kimmel Cancer Center. She said Johns Hopkins Hospital in Baltimore has stopped doing less invasive hysterectomies for cervical cancer until there is more data.

While some patients with small tumors might do as well with minimally invasive surgery, “surgeons should proceed cautiously” and discuss the new information with patients, Fader wrote in an accompanying editorial.

AP video journalist Federica Narancio contributed to this report from Washington, D.C. The Associated Press Health & Science Department receives support from the Howard Hughes Medical Institute’s Department of Science Education. The AP is solely responsible for all content.