Solving the Mystery of an Increase in Adverse Events

FROM THE EDITORS

The Government Accountability Office issued a report at the end of July showing that the number of adverse events reported to FDA for medical devices increased to 588,161 in 2008 from 189,450 in 2004. That number of unwanted medical events in patients associated with devices even surpasses the adverse events reported for drugs. What the report fails to explain is why there was such a big increase in device adverse-event reports. Perhaps the market has grown proportionately, or maybe devices are more complex, resulting in more adverse events per device. Other reasons could be that companies are reporting more properly or are even overreporting out of caution.

“There may be a combination of factors,” says Jeff Shapiro, a partner in the law firm of Hyman, Phelps & McNamara (Washington, DC). “One big factor is certainly that FDA is taking a harder line. If FDA decrees that adverse events need to be submitted for each recurring event, the number really jumps.” Shapiro says that he has seen just this circumstance with several clients.
It may also be that hospitals are reporting more. “Without additional data—for example, how many by manufacturers, how many by device type, etc.—there is really no way to be sure of the cause.” Shapiro points out at least one “spectacular example in recent years of a company getting into trouble for not reporting.” As reported in the August 2003 issue of MD&DI:
The [Guidant] Ancure Endograft system was approved for commercial distribution in the United States in September 1999, but was withdrawn temporarily from the market beginning on March 15, 2001. Although the company filed 172 medical device reports during that 19-month period, it has admitted that there were an additional 2628 MDRs that it failed to file.
Shapiro speculates that Guidant's mistake may have led other companies to err on the side of reporting more than they may have previously. And certainly there has been growth in the complexity of both the market and devices themselves, especially in terms of software developments. This, too, could be a factor in the dramatic increase in the number of reported adverse events.
“I am aware of two factors that undoubtedly contributed,” says Brad Thompson, of Epstein Becker & Green (Washington, DC). The first, he says, is probably due to an easily verifiable growth in the number of unit sales of medical devices over that period of time. “It's not just the number of types of devices sold, but actual unit increases for existing devices. With a bigger pool of devices out there, one would expect more adverse events,” he says, noting that much of that growth has occurred in higher-risk devices, which also have a higher incidence of adverse events.
But the second and much harder to gauge contributing factor, he says, is a growing conservatism among companies on how they make judgments about what is reportable and what is not. “There's no specific penalty for overreporting—although it's discouraged—and there is a more definite compliance risk to underreporting,” he says, “and so, more and more companies have been reporting more events.”
In other words, if an event is in the gray area, it's safest to report it. He points out that the terms overreporting and underreporting are “loaded terms that assume a bright line as to what's reportable.” In fact, the line isn't that bright, and thus quite a few events fall into this murky middle area, he says.
“With the enforcement actions we've seen against companies for failing to report, particularly a few in the early part of this decade, other companies are simply being more conservative in how they treat borderline cases,” i.e., reporting even gray area events. Although he says it is unlikely to account for all of the growth, he says it certainly accounts for some of it.
No matter the reason, the numbers literally speak volumes. A tougher regulatory climate coupled with more—and higher-risk—devices on the market seems to have led to a huge leap in adverse events being reported. Perhaps it's time to better define the gray area.

Sherrie Conroy for the Editors