My thanks to Healthcare Informatics for providing such a good foil -- their article on Nine Tech Trends -- for a number of posts that I was planning to write. Here's the first, on Bar Coding and RFID.

We've all seen bar codes at the grocery or on UPS and FedEx
packages. Bar codes encode data into a machine readable format,
eliminating potential human error like transposing numbers. RFID (radio
frequency identification) encodes data electronically using radio
frequency (RF) rather than light to read data from a distance. The
distance varies, passive RFID (like the things they stick on clothes or
books) have no power and thus limited range. A passive RFID system has
readers that transmit RF signals at sufficient power to "excite" the
passive tag, causing it to transmit its encoded data. The more powerful
the RF signal from a passive RFID reader, the greater the distance
between the tag and the reader. Passive RFID can be problematic in some
areas due to potential interference with medical devices. Active RFID
includes battery power in the tag. Powered tags can communicate over
greater ranges at lower power than passive tags. Active RFID tag
batteries typically last a year or two.

Active RFID will be used for patient and asset tracking, with bar
codes used for meds administration and inventory control. Bar codes
require line of sight and manual (and sometimes repeated) scanning.
RFID simply automates the process with sensors that "grab" the tag
electronically. As prices fall and tagging (with either bar codes or
RFID) becomes more pervasive, the market will naturally transition to
the greater automation offered by RFID. Usually priced $25 to $20
bucks, some active RFID tags are starting to fall below $10 per tag.

Active RFID has huge potential as a tool to improve patient flow. By
tracking patients, assets and staff, patient management software can
orchestrate, track and report on the movement of patients through the
care delivery process providing an unprecedented level of automation
and data for fine tuning patient flow. Known as Indoor Positioning
Systems, these products are being integrated into all the patient flow
software apps.

Actual implementations mentioned in the article include Vocera, Radianse and Awarix. Parco also has an installation, as does PanGo and Ekahau.
With the exception of Radianse, who has more than a dozen hospital
customers, IPS vendors mostly have just initial installations. Vocera
is a communications solution using VoIP (voice over IP); their system
can tell you the access point (AP) that one of their devices is
associated with, and depending on AP density and location, could
provide a general location. Vocera has also integrated with Radianse.

Hospitals are particularly challenging for an IPS. Because of the
variety of materials used for interior walls, shielding used in some
areas, and the many electromechanical devices, an IPS that works great
in a warehouse or factory may not work in a hospital -- best not be the
first one to try.

Vendors with 802.11 based technologies will tell you that leveraging existing infrastructue is better.
In reality the only infrastructure at issue is the IPS receiver or AP
-- everything else is common infrastructue. An IPS based on 802.11
provides accuracy in the 3 to 4 meter range. These types of systems
usually require increased AP density, i.e., adding more APs. Access
points placed initially for optimal coverage may not be in a good place
to determine location. The tags for 802.11 can also be larger and more
expensive ($80 to $200) than those based on other technologies.

As one vendor put it, "A combination of
interoperability, infrastructure, cost-effectiveness and workflow
integration will determine success for vendors." This is really the
bottom line, regardless of vendor technology or implementation.

3/10/05 UPDATE: Please note the last sentence of paragraphy 5 above,
that Vocera has integrated with with Radianse, is incorrect. While a
number of vendors have integrated both vendors into their solutions,
Vocera and Radianse have not done any direct integration themselves.
Thanks to Dan for pointing out the error.

UPDATE: Tuomo Rutanen of Ekahau responds in the comments below.
While I've never seen any data that tracks the need for incremental APs
for indoor positioning, it seems logical that at least some sites would
not need any additional APs to gain sufficient positioning resolution. He also notes that active RFID tags have come down to $60 and are continuing to fall in price.