Thursday, June 26, 2008


RFID tags and medical equipment

There’s been a study in the Journal of the American Medical Association that’s made the news in the last couple of days, showing that RFID tags (those little devices — or sometimes not so little ones — that are used for inventory control and such) interfere with medical equipment [see USA Today, New Scientist, and the summary at the JAMA site].

The researchers set up readers and tags near active equipment, and recorded whether the equipment was adversely affected, and at what range. From New Scientist (where they give more details than USA Today does):

His team tested two RFID systems and dozens of medical devices, none of which were connected to patients. The researchers tested devices within meters of an RFID reader broadcasting a radio signal to nearby tag. The strength of the radio signal, though high, was not unrealistic.

If the signal caused a device to break down, the researchers moved the device away until the interference stopped. If the initial test turned up negative, van Lieshout’s team inched the device closer.

In 123 tests of 41 different pieces of equipment, the devices malfunctioned 34 times — 22 of the glitches were deemed serious enough to affect patients. For instance, eight of nine syringe pumps, which trickle medicine to patients, failed completely when exposed to an RFID field, anywhere between a centimetre and a meter away from a transmitter.

This sounds bad, and, indeed, it makes it clear that we can’t yet use RFID tags for some of the purposes that have been proposed. But it’s easy to overreact, so let’s make one thing clear: passive RFID devices are not a hazard when they’re not being read.

“Passive” RFID devices are called that because they don’t transmit anything on their own. They have no internal power source, and get their power from the electromagnetic field generated by the reader. So they only transmit when they’re within range of a reader. The range, here, depends upon the size of the coils in both the reader and the RFID tag, but the small tags that you’ll find inside CD or DVD cases, for example, or that are installed in employee badges (and passports), can usually only be read at distances of a few inches, and certainly no more than a couple of feet, even with a powerful receiver.

When they’re not being read, they emit no electromagnetic radiation, and can cause no electromagnetic interference. So it’s important to note, here, that the tests done in the study were done with a live reader — the IDs were being read at the time, so there was electromagnetic radiation generated by both the readers and the RFID devices.

Since passive tags are not a danger in the absence of a reader, they can safely be used in cases where the reading is done away from patients (for inventory control, or personnel access, for example), and there’s no reason to worry about preventing the tags from being brough into hospitals. If, say, medicine bottles are tagged, the presence of a bottle in someone’s room is not a worry. But if someone was planning to install a reader at the patient’s bed to warn that the wrong drug was just brought within range... that could be a problem.

That means that until the problem is dealt with, we have to be careful about what we use the tags for — specifically, about when we read them. It’s still OK to use RFID tags to check meds, identify patients, or, as the New Scientist article mentions, detect surgical equipment that might have been left inside a patient. We just can’t read the tags in close proximity to sensitive equipment that’s in use. That’s important to understand, but it’s not as limiting as one might think from reading the articles.

Update, 15 July: The NY Times picked this up today, and gets it completely wrong. The whole article blames the tags, and never mentions the readers once. For example:

But a new study suggests that the tags may pose a danger when they are kept in places like intensive care units: the signals they give off can interfere with the working of critical care equipment.
Sigh. They do have a pointer in the sidebar to the JAMA summary, which never mentions the tags themselves (it refers to “RFID systems”).

Update, 16 July: I've corresponded with the NY Times writer, Eric Nagourney, and he's checking with the study's authors and has suggested a clarifying change to the online article that I think corrects the misunderstanding nicely.

1 comment:

Anonymous said...

RFID tag readers may broadcast one watt. Cell phones broadcast up to 10 watts. I guess that is why you are not supposed to use the phones in hospitals.

However, a "normal" RFID tag with a 10 cm antenna can be read from 3 meters. Cutting the power in the reader back by a factor 100 reduces the read range a factor 10. So all you have to do is make sure that tag readers in sensitive locations transmit little power, and make people hold the tags close to the reader.

Rod Hodgson