Are robotic surgery complications underreported?

Although US hospitals have been quick to embrace surgical robot technology over the past decade, a “slapdash” system of reporting complications paints an unclear picture of its safety, according to Johns Hopkins researchers.

The Johns Hopkins team, led by Martin A. Makary, MD, MPH, found that, among the 1 million or so robotic surgeries performed since 2000, only 245 complications—including 71 deaths—were reported to the US Food and Drug Administration (FDA).¹ When an adverse event or device malfunction occurs, hospitals are required to report these incidents to the manufacturer, which in turn is required to report them to the FDA—but this reporting doesn’t always happen.

“The number reported is very low for any complex technology used over a million times,” says Dr. Makary, associate professor of surgery at the Johns Hopkins University School of Medicine. “Doctors and patients can’t properly evaluate safety when we have a haphazard system of collecting data that is not independent and not transparent. There may be some complications specific to the use of this device, but we can only learn about them if we accurately track outcomes.”

The use of the robot in surgery has skyrocketed. Between 2007 and 2011, for example, the number of procedures involving the robot increased by more than 400% in the United States and more than 300% internationally. At the end of 2011, there were 1,400 surgical robots installed in US hospitals, up from 800 just 4 years earlier.

Some incidents went unreported until the news media highlighted them
Dr. Makary and colleagues found several incidents reported in the national news media that were not reported to the FDA until after the stories appeared in the press, even though the incidents took place long before the media exposure. Dr. Makary says it’s likely that many other incidents go unreported, never to be captured by research like his or by the FDA.

“We need innovation in medicine and, in this country, we are tremendously good at introducing new technologies,” he says. “But we have to evaluate new technology properly so we don’t over-adopt—or under-adopt—important advances that could benefit patients.”

How the study was conducted
Makary and colleagues reviewed the FDA adverse events database from January 1, 2000, to August 1, 2012. They also searched legal judgments and adverse events using LexisNexis to scan news media, and PACER to scan court records. The cases then were cross-referenced to see if they matched. The investigators found that eight cases were not appropriately reported to the FDA, five of which were never reported and two of which were reported only after a story about them appeared in the press.

Complication rate was highest for hysterectomy
When investigators reviewed complications that were reported, the procedures most commonly associated with death were:

• gynecologic (22 of the 71 deaths)
• urologic (15 deaths)
• cardiothoracic (12 deaths).

The cause of death was most often excessive bleeding. In cases where patients survived, hysterectomy by far had the most complications (43% of injuries).

A call for standardized reporting
Dr. Makary contends that standardized reporting is needed for all adverse events related to robotic devices. One rare complication that occurs, he says, is that a surgeon can accidentally cut the aorta because the surgeon cannot feel its firmness. For reporting purposes, however, it’s unclear whether such an event is surgeon error or device-related error. The FDA currently collects only device-related errors.

Dr. Makary argues that errors such as inadvertent cutting of the aorta, although preventable with proper technique, should be tracked as device-related errors because they are more common with robotic surgery than with conventional surgery. Without better reporting standards, he says, these complications are less likely to be reported to the FDA at all. And if they go unreported, they cannot contribute to the understanding or identification of safety problems.

He suggests one solution: use of a database like the one maintained by the American College of Surgeons, in which independent nurses identify and track adverse events and complications of traditional operations.

Good information on robotic surgery is needed not only for research, but also to ensure that patients are fully informed about potential risks. Right now, Dr. Makary says, it’s too easy for a surgeon to claim that there are no additional risks related to robotic surgery because the evidence is nowhere to be found.

“Decisions should not be made based on the information in the FDA database,” he says. “We need to be able to give patients answers to their questions about safety and how much risk is associated with the robot. We have all suspected the answer has not been zero. We still don’t really know what the true answer is.”