Published On Dec 15, 2016
The Food and Drug Administration tries a fresh approach to medical devices suspected of falling short.
If a surgical tool or other medical device doesn’t perform the way it is supposed to, what happens next? That road forward is getting an upgrade, says Nat Sims, an anesthesiologist and device pioneer at Massachusetts General Hospital.
The Food and Drug Administration, which is responsible for licensing and monitoring these devices, is looking to make the response smarter. That will happen through two avenues: more extensive training for device users, and better information about what went wrong, from both the hospitals and the device itself.
This new “active” strategy makes a significant improvement over the old system, says Sims, in which an underperforming device might stay in the system for far too long. Sims, who holds more than 10 patents for devices he has invented, has been working with the FDA on designing and implementing this new approach.
Q: How did you become involved with medical devices?
A: My dad was a clinician and basic science researcher, but he was also amazingly inventive, with hands-on woodworking and metalworking skills. So it was accepted in our household that if something broke, you fixed it. Or you made a better version. I suppose that has played into the way I approach clinical practice—if a system of care is not working, redesign it.
Q: How has the FDA normally handled medical device safety?
A: Under the law, it is a manufacturer’s responsibility to respond to problems with devices that come up in the field. If there is an issue, the company is supposed to respond to complaints by notifying customers with safety advisories, or with field actions that can include inspections, repairs, modifications, adjustments, relabeling or destruction. The FDA keeps an eye out to make sure that is happening, so its involvement has typically been focused on the manufacturers, not on the users.
Q: And what is wrong with that?
A: Well, for one thing, it makes reporting device malfunctions dependent on front-line clinicians. And that makes a lack of detail very likely, because of the time and effort it takes to file a report. Reports are required by law when a malfunction results in death or serious injury, so they do get filed, but they may not contain sufficient information to be helpful. A user writing a report might simply note “device didn’t work.”
Gaps in medical device problem reporting can slow down the whole process. There’s one notable example of a defective infusion pump—a device that delivers fluids and drugs to a patient—introduced in the 1990s. It would shut down without warning and act in other unexpected ways that might have been harmful to patients. But it took multiple recalls and other actions through the years until the FDA sent a letter to the manufacturer in 2010, ordering the company to recall and destroy approximately 200,000 pumps in use in the U.S. And that happened only after significant advocacy.
Q: What is the proposed new system?
A: The FDA has initiated a new “active” strategy, aiming for a flow of digital information about devices that relies less on people. Almost all digital medical devices can now collect and store data. Microprocessors monitor sensors and software in the device and can detect and log when something is wrong. The idea is, why not automatically transmit the logged data to the manufacturers? They can look for error trends in real time and determine whether a device is exhibiting early signs of a significant problem. It might lead to quicker field actions or recalls, and would also lead to smarter next generations of devices. That type of monitoring isn’t required yet, but that is the direction we’re heading.
Q: How have you become involved with this plan?
A: Sometimes the problem turns out not to be about the device, but how it’s being used. So the FDA also wants to raise the bar for human performance. Between 2013 and July 2016, for instance, the FDA received over 300 reports of problems involving syringe infusion pumps—which are used particularly in the care of newborns. I looked through these reports and noticed that it wasn’t just that the devices had issues, but also that people operating the devices didn’t always understand elements of how they work (the key “tips and tricks” to ensure maximum safety).
I led an MGH team that engaged with an FDA task force to look for solutions. And the MGH team made it our job specifically to address the workforce training part—making sure that clinicians have the training they need, in the language they understand, to make these devices work properly.
Q: How do you do that?
A: We’ve created interactive educational modules that teach how to use certain medical devices, starting with syringe infusion pumps that are so often used in the care of newborns and infants. These e-Learning modules are now free and publicly available at syringeinfusionsafety.org, so that every hospital can train critical staff members. We released them at about the same time as the FDA sent out a safety communication about syringe pump problems.
Here at MGH, intensive care unit nurses and anesthesia caregivers will take these online classes, to make sure that we’re maintaining “excellence every day.”
All of this progress on medical device safety has been a significant journey. I’m very optimistic about where it will lead. And if it succeeds, it will be because of new partnerships between hospitals, manufacturers and the FDA, working together.