Have hysterectomies gotten riskier for some of the hundreds of thousands of U.S. women who undergo these procedures each year?

Researchers are trying to answer that question now that gynecologists have mostly stopped removing women's uteruses with the help of a tissue-slicing device called an electric morcellator.

Introduced in 1993, the device enables uterine tissue to be removed through tiny abdominal incisions. This reduces the complications and recovery time of traditional surgeries, which require large horizonal or vertical belly incisions. But, as the Food and Drug Administration warned last year, morcellation can also spread and worsen a rare, hidden uterine cancer called leiomyosarcoma.

That's what happened in October 2013 to Amy Reed, 42, a Philadelphia-area anesthesiologist and mother of six. She and her husband, Jefferson University Hospital heart surgeon Hooman Noorchashm, have since pushed for a total ban on the device.

Last week, a University of Michigan study suggested that curtailing its use has had downsides, as gynecological medical groups had predicted. The study found that after the FDA's April 2014 warning, there was a shift from small-incision to open-abdomen hysterectomies, and the rates of major complications increased from 2.2 percent to 2.8 percent. Hospital readmissions - defined as having to return within 30 days of discharge - also increased, from 3.2 percent to 4.2 percent.

However, the analysis has limitations. The researchers had no data on power-morcellation usage, only on whether the uterus was removed vaginally, through large incisions, or through small incisions (not all laparoscopic hysterectomies use a morcellator).

And the analysis of 15,300 Michigan hysterectomies covered such a short time span - the 15 months before the FDA warning, and the eight months afterward - that many findings were hard to interpret. In particular, the rise in abdominal hysterectomies and related complications such as infections, blood clots, and readmissions were not statistically significant, meaning the small increases could be due to chance.

Even so, obstetrician-gynecologist John Harris and coauthors said patients should be warned.

"The additional risks associated with changes in surgical practice over time, possibly due to a decline in the use of morcellation, must be discussed with patients in order to provide comprehensive informed consent," they wrote in the American Journal of Obstetrics and Gynecology.

In May, the same journal published a small poll of minimally invasive surgeons that found less use of morcellators and more use of larger incisions - but the survey had no data on complications.

Hal C. Lawrence, executive vice president of the American College of Obstetricians and Gynecologists (ACOG), said he wasn't surprised by the Michigan findings.

"It has been well-documented that minimally invasive surgery decreases the complication rate, especially for patients with obesity, diabetes, or renal disease," he said. "We had discussed that with the FDA."

Arnold P. Advincula, president of the minimally invasive surgeons' group, the American Association of Gynecological Laparoscopists, was also unsurprised by the Michigan trends - even though Columbia University, where he practices, hasn't seen such changes in practice pattern or complications.

"Although we don't use morcellation, we have come up with tissue-extraction methods to offer" minimally invasive operations, he said.

Evaluating the fallout of the morcellator storm is difficult because the prevalence of morcellation before Reed and Noorchashm sounded alarms, as well as the risk of morcellating a leiomyosarcoma, remain unclear.

Experts estimate that about 400,000 women a year now undergo hysterectomies, about 40 percent to treat heavy bleeding caused by benign uterine fibroids. That's far fewer than the almost 600,000 who underwent the procedure annually more than a decade ago, but it remains the second-most-common surgery among American women, behind C-sections.

Preoperative tests can't distinguish fibroids from leiomyosarcomas, which are rare - diagnosed in fewer than one in 100,000 women a year, studies show - but ferocious. Even when the malignancy is found while confined to the uterus, about half of patients die within five years, according to National Cancer Institute data. (Reed, who had intensive treatment for her stage-four disease, has had two recurrences.)

ACOG has estimated that 11 percent of all hysterectomies involved morcellation before the controversy, but rigorous data are lacking.

As for morcellating a leiomyosarcoma, estimates vary widely. The FDA reviewed 18 studies and concluded that the risk could be as high as one in 350 women undergoing hysterectomy. But a review of 133 studies led by Wisconsin researchers - and presented to the FDA last year - estimated the risk to be one in 2,000 to one in 8,300 surgeries.

"I do believe it needs to be relooked at," Advincula said, to explain "how you can have such discrepant numbers."