Medical Device Daily Contributing Writer

PARIS — Can radio frequency identification technology (RFID) become for hospitals what bar-coding is today for the local grocery store?

If only it were as simple as installing laser readers at the checkout. But a living, breathing, fast-paced hospital, where human lives hang in the balance, presents challenges to information technologies that engineers most likely never imagined back in the lab.

The problems multiply and become more critical when RFID is introduced into operating rooms (ORs) or intensive care units (ICUs).

Results from a two-year study of RFID at the Academic Medical Center (AMC) of the University of Amsterdam (the Netherlands) indicate good news for the accounting department but, at the same time, show that RFID technicians have large problems to solve before the technology is ready for prime time.

The Amsterdam project is literally on the "bleeding edge" of healthcare RFID. One of the three pilots tracked blood supplies used in surgical operations; a second pilot tracked patients and hospital personnel; a third tracked OR materials.

After two years of fine-tuning RFID equipment and its installation, the report says the tracking systems "were in successful operation all during the last quarter of 2006. No problems related to RFID interference with medical equipment were reported." A fuller report will be published later this year.

The preliminary report provided to Medical Device Daily offers a characteristically open Dutch discussion of problems encountered introducing digital inventory control to the OR and, critically, testing the influence or interference of RFID devices.

The Dutch Ministry of Health, Welfare and Sport requested the study to determine the added value of RFID applications in healthcare, the project partially funded by the Dutch Ministry of Economics Affairs, though the greater share of the 11.3 million project ($1.8 million) was funded by project partners with cash or in-kind donations of equipment and personnel.

Capgemini Netherland (Amsterdam) led the project, with its partners including Oracle (Redwood Shores, California), Intel (Santa Clara, California) and Geodan Mobile Solutions (Amsterdam).

The financial benefit for an average-sized hospital, calculated on the AMC experience, is estimated at 1380,000 ($550,000) annually, realized through lower stocks of materials, more efficient use of blood stocks and reduced administration. And the RFID system created detailed data analysis for workflows and processes, precisely measuring the time for patient waits, holding periods and transport of materials.

Three main benefits were quantified by the report: the optimization of materials flow, reduced waste in blood products and reduced paperwork in transfusion laboratory. Other qualitative benefits included improved patient safety through identification, increased insight into logistics and enabling activity-based costing for surgeries.

The preliminary report on the Amsterdam RFID experiment covers 178 surgeries, 122 surgical materials, 141 blood products and tracking of 801 healthcare professionals.

For the blood-tracking pilot, passive RFID tags — embedded chips read by antennas or handheld readers — were used for blood bags, allowing a specific product to be identified in storage or traced when used on return to recycling.

An active RFID tag, capable of sending a radio signal, was added to blood bags moving out of the transfusion lab to the OR or ICU in order to monitor blood temperature. If the temperature rose or fell beyond a preset threshold, an alert was triggered.

Active tags are removed once the blood bag is used. The active tag is placed in a "stop basket" that reads usage information. The passive tag remains on the bag to track its processing or disposal.

Physicians and other staff wore passive RFID badges and were tracked by function, not as individuals. This was a first lesson learned for winning acceptance among participants.

"Providing information about the pilots at the right moment stimulated cooperation and interest," the authors said.

The report acknowledges "significant discrepancies" in data captures during the course of the three different pilots. For example, in some cases doctors were tracked but not a corresponding patient for the procedure, accounting for the difference between surgeries and patients.

Time stamps included in data sets was another issue. Where hundreds of blood bags were tagged, for example, only the 141 in the study had time stamps.

Yet the data sets in the preliminary report are far below what might be expected for a two-year study in a major center. The actual number of observations, especially during the early stages of the pilot, was much larger than the data sets validated for the study findings and considered valid for drawing conclusions.

The non-valid data sets had components missing due to what the report calls, "ongoing fine-tuning of the infrastructure."

Almost 50 medical devices were tested, including ventilators, monitors, anesthesia apparatus, infusion pumps and, significantly, external pacemakers.

Another environmental issue affecting the IT infrastructure is the variations in electrical current running through hospital walls that initially need to be identified and modified for data capture for the passive tags operating at 868 MHz.

Dr. Pieter Bakker, of the department of medical oncology at AMC, said flatly in his presentation of results, "There was a total failure, a meltdown when the active RFID infrastructure was implemented and adjusted."

Although infrastructure and equipment issues were solved over time, problems specific to RFID still need to be addressed before an RFID system can be reliably, and safely, deployed in an OR or ICU.

In his conclusions, Bakker said that standard RFID tags utilizing EPC global Gen2 at 868 MHz "could not be used inside the OR, due to interference. In summary, most of the medical apparatus were susceptible to the signals from the fixed antennae of the RFID 868 MHz system," the report concludes. Distances varied from direct contact up to 6 meters.

A smaller number of systems were affected by RFID devices operating at 1356 MHz. The fewest problems for interference and susceptibility of medical devices were to signals from the RFID 125 KHz system. Distances where interference occurred varied from 5 cm up to 2 meters.

The RFID 125 KHz system was selected for the pilot project in the operating room of the hospital, though the report notes there are doubts about the system's effects on pacemakers.

"A rather surprising finding," said the report, "was that RFID frequencies could disturb the radio channel of a pacemaker programmer operating at the same frequency as the RFID system. After the removal of the RFID signal, most apparatus functioned normally. In a few cases, manual interruption was required. Some apparatus had to be switched off and on again to produce a return to normal functioning."

The preliminary report on RFID tentatively concludes that under real-use conditions, with careful placement of antennas and the correct power setting for the unit, there is no patient risk.

Oracle and Intel estimate the market for RFID in European healthcare for 2008 will be 1300 million ($420 million).