Integrating RFID Technology into a Drug Administration System

• Abstract
• Introduction
• The IDAS System
• Research Methodology
• Results
  • Selecting RFID Hardware
  • Attaching an RFID Tag to a Syringe
  • Reading an RFID Tag
  • Associating a DIN with an RFID Tag and Passing it to IDAS
• Outstanding Issues and Further Work
  • Undetected Syringes and Inadvertent Reads
  • Printing on RFID Tags
  • Electromagnetic Radiation and Interference
  • Reducing Anaesthetist Data Input into IDAS
• Conclusion
• References

Abstract
Errors in the administration of drugs to patients, along with other types of adverse drug events (ADEs), have proven to be very costly to the healthcare sector. One commonly proposed solution to the problem is the use of barcodes to uniquely identify both patients and drugs. The size of the ADE problem and the success of a limited number of barcode-based systems have prompted US health care regulators to endorse, or even mandate, their use. However, it has been suggested that limitations of barcodes adversely affect the usability of such systems, and that radio frequency identification (RFID) technology offers a more suitable alternative.

This paper documents the design and development of an extension to an existing, barcode-based, anaesthetic drug identification system, IDAS, which would allow the replacement of barcodes with RFID technology. The design is informed by a review of RFID application case studies, experiments with RFID hardware and observation of and interviews with anaesthetists. A demonstration of the prototype system suggests that RFID technology could functionally replace barcodes but that significant issues would remain to be addressed.

Introduction
Injuries to patients caused by errors in administering drugs are known in the healthcare sector as adverse drug events (ADEs). ADEs have proven to be very costly in terms of both human life and scarce healthcare resources. Bates, Cullen and Laird^{[ 1 ]} find that 7 percent of patients admitted to US hospitals experience an ADE. Wilson et al^{[ 2 ]} report that 8 percent of ADEs in Australian hospitals are fatal and 17 percent result in permanent disability. Recent research from New Zealand^{[ 3 ]} suggests around 2 percent of hospital admissions result in an ADE, with 12 percent of these resulting in death or permanent disability. Johnson and Bootman^{[ 4 ]} have estimated that ADEs were costing the US healthcare sector around US$80 billion annually in 1995.

Such figures have prompted regulatory bodies and patient safety lobby groups to call for the increased use of information systems (IS) that address the causes of ADEs. Electronic databases containing drug trial results and known drug interactions can reduce the chances of doctors ordering inappropriate medications. Computerised Physician Order Entry (CPOE) systems are intended to ensure that one doctor’s orders can be understood by other doctors, nurses and pharmacists.^{[ 5 ]} In some pharmacies, robots are being used to eliminate errors in dispensing drug doses.^{[ 6 ]} Bar Code Medication Administration (BCMA) systems can reduce the risk that patients are given the wrong medication by using barcodes to uniquely identify patients and drugs.^{[ 7 ]}

Clinical studies suggest that BCMA systems might prevent up to 58 percent of ADEs.^[7-10] Yet very few healthcare organisations have implemented them. Recent market research by healthcare IS vendor Exavera Technologies^{[ 11 ]} suggests that in the US users may number as few as 2 percent of the country’s 64,000 hospitals. However, this seems likely to change in the near future. The US Food and Drug Administration (FDA) recently mandated that all drugs supplied to US hospitals must include barcodes on their packaging.^[12] The US Joint Commission on Accreditation of Healthcare Organizations (JCAHO) has issued guidelines on the use of barcodes for unique identification of patients.^[13]

The leading proponents of BCMA in New Zealand include Merry, Webster and their colleagues from the University of Auckland’s Faculty of Medical and Health Sciences, Anaesthesiology Department. Barcoded drugs play a central role in their Injectable Drug Administration and Automated Anaesthetic Record System (IDAS).^[14] Simulated operations using IDAS and comparing it with traditional methods for administering anaesthetic drugs suggest that IDAS reduces the incidence of ADEs.^[15,16] However, they also identify some limitations of barcodes that may reduce the effectiveness of IDAS.

More recently, Merry and Webster have suggested radio frequency identification (RFID) technology as an alternative that addresses some of these issues.^[17]

This paper is a description of a constructivist research project undertaken to determine whether RFID technology can functionally replace barcodes in IDAS. The next section describes IDAS and its usability problems identified during simulated operations. This is followed by a brief discussion of the research methodology and the specific goals of the project. Next, the design and development of a prototype RFID-enabled IDAS is documented. Then, outstanding issues and possibilities for further work are highlighted. Finally, a summary and conclusions are presented.

The IDAS System
Several hospitals in Auckland are using IDAS to support the administration of anaesthetic drugs during operations. The system has been designed using first principles of patient safety. It provides procedures and tools for keeping the anaesthetic work area organised, for confirming drugs before administration and for keeping a record of drugs administered.

As illustrated in figure 1, during an operation an anaesthetist spends most of their time within a triangular area defined by one or more drug trolleys, an anaesthetic workstation and the patient.

Injectable drugs, some in ampoules and some in pre-filled syringes, are stored in the drug trolley. Under the IDAS system, the ampoules and syringes have a label attached showing the drug’s class, name and a barcode representing the Drug Identification Number (DIN). As figure 2 illustrates, the label is also colour-coded based on the drug’s class according to an international standard.^[18] The top of the drug trolley is used as a work surface, on which the anaesthetist prepares injectable drugs for the operation. Drugs kept in ampoules are "drawn up" into syringes, which then have a corresponding barcode label attached. Under the IDAS system, plastic trays are used to keep the drug trolley drawers and work surface organised. An example of the trays used on the work surface is shown in figure 3.

The anaesthetist controls infused drugs and monitors the patient’s condition via the anaesthetic workstation, shown in figure 4. Pulse, blood pressure and other vital signs are displayed on one monitor. A second monitor displays the IDAS software. This software allows an anaesthetist to access details on the operation and the patient, and to monitor the patient’s fluid balance, the drugs administered and other details of the anaesthetic. Drug injections are recorded by scanning the syringe’s barcode label. The IDAS software then displays the name of the drug and speaks it aloud, providing the anaesthetist with confirmation that they are about to administer the intended drug. If the anaesthetist has the incorrect drug he or she presses a key to indicate that the drug wasn’t administered. Otherwise the anaesthetist enters the dose administered.

The IDAS system has been compared with "traditional" anaesthetic practice through simulated operations.^[15,16] These have found that, while the IDAS system seems to reduce the incidence of ADEs, its effectiveness may be limited by usability issues. The major problem is an anaesthetist administering a drug without first scanning the syringe’s barcode label. Half of the trial participants did this on at least one occasion. Comments from the participants suggest that this was primarily due to simple forgetfulness. However, some comments were more directly related to problems using the barcode scanner. For instance, one participant didn’t use the scanner because of "technical difficulties".

As figure 4 illustrates, the barcode scanner used in the trials was a simple hand-held model. These require more effort from an operator than the bi-optic scanners common in supermarkets. IDAS requires the anaesthetist to hold the syringe in a certain way for its label to be scanned. He or she must move the syringe directly over the barcode scanner on the anaesthetic workstation, which may not be on the natural movement path from the drug trolley to the patient. This is reflected in the actions of some trial participants, who moved the scanner from the anaesthetic workstation to the drug trolley or to the operating table. The anaesthetist must also pause at the barcode scanner long enough for the barcode to be recognised.

Chassin^{[ 9 ]} finds that doctors are conditioned by their medical training to be self-reliant and to reject support systems. To encourage anaesthetists to use a system such as IDAS, the extra effort required should be minimised. RFID technology offers some opportunities to do this. RFID readers can sense RFID tags anywhere within range, even if they are obscured or moving. The anaesthetist can therefore hold the syringe in a natural way. While the tagged syringe must be moved somewhere near the RFID reader, it doesn’t need to be paused at a specific location. This allows the anaesthetist to move more naturally between the drug trolley and the patient.

Research Methodology
After reading of Merry et al’s interest in RFID, the researcher approached them with a proposal for a research project to determine whether RFID technology could functionally replace barcodes in the IDAS system. A constructivist methodology was suggested as most suitable for this research. Cornford and Smithson^[20] characterise constructivist research as:

Nunamaker, Chen and Purdin^[21] acknowledge that the research value of system development has been questioned. However, they go on to argue that it is a legitimate methodology for IS research. To ensure value from such research, Nunamaker et al propose an iterative multi-methodological approach. Theory building informs the development of prototype systems. These are evaluated through observation and experimentation, with the results informing a new cycle of theory building.

The short time available for this research meant that only the first part of this cycle would be possible. Theory building was based on a review of RFID application case studies, experiments with RFID hardware, and observing and interviewing anaesthetists. This informed the development of a prototype system. In order to show that RFID technology could functionally replace barcodes in IDAS, this prototype was required to allow the following actions:

1. Attach an RFID tag to a syringe. The RFID tag must show the anaesthetist the same information as the current barcode labels – drug class, name and colour coding.
2. Associate a DIN with an RFID tag.
3. Read a tagged syringe at some point between it being removed from the drug tray and being placed in an injection port.
4. Send a DIN to the IDAS software, activating the existing confirmation and recording function. The IDAS software itself could not be changed.

Results
Case studies of RFID applications were reviewed to determine which type of RFID tags and readers would be most suitable for use within the "anaesthetic triangle". Experiments were then conducted to find out how different methods of attaching tags to syringes affected their performance. Anaesthetists were observed during an operation and later interviewed to determine how RFID readers could best be deployed to fulfil requirement 3. Two small software tools were developed to perform requirements 2 and 4.

Selecting RFID Hardware
There is a small but growing body of case studies describing RFID applications in hospitals. The applications can be broadly classified into three types. Identification applications involve a single action at a single location. Identifying a staff member for access to a secure area is one example. Location-based applications perform continuous actions at a single location. For instance, an RFID-enabled "smart" medicine cabinet can provide a real-time inventory of the drugs it contains, recording removals and additions.^[22] Tracking applications use continuous actions at multiple locations. For instance, individual pieces of equipment may be tracked to prevent them being lost or stolen^[23] or staff and patients may be tracked to analyse workflow.^{[ 24 ]}

The case studies reviewed show that almost all of these applications use RFID tags and readers that operate at either 13.56 Mhz or the 900–915 MHz band. These are the two frequencies most widely endorsed in RFID-related standards. This makes them popular choices for RFID manufacturers, ensuring that there should be a steady supply at a reasonable cost.

Pappu, Singhal and Zoghi^[25] suggest that the 13.56 Mhz frequency is more common in identification and location applications, with 900 MHz typical in tracking applications. This reflects the different "read ranges" of the two frequencies. Readers that operate at 900 MHz can detect a tag over several metres, while 13.56 MHz readers typically have a range of one metre or less. The "anaesthetic triangle" only covers a couple of square meters, so the latter should be more suitable.

Another factor that distinguishes RFID frequencies is how the radio signals sent from the reader to the tag are affected by intervening objects. Walker et al^{[ 26 ]} point out that 13.56 MHz signals can be blocked or distorted by metal and 900 MHz signals are absorbed by liquids. While there may be some metal objects, such as surgical instruments, in the anaesthetic work area, the RFID tags in this application will always need to be read through syringes filled with liquid. Based on this, it was decided that the prototype system would use 13.56 MHz RFID hardware.

Attaching an RFID Tag to a Syringe
Experiments were carried out to determine how RFID tags could be best attached to syringes. The aim was to find the method that resulted in the longest read range, while still allowing the anaesthetist to handle the syringe comfortably. The three most common sizes of syringe were used: 60 mls (approximately 30 mm in diameter and 110 mm long), 10 mls (16 mm diameter, 80 mm long) and 5 mls (13 mm diameter, 60 mm long).

Three sizes of RFID tag were used: 80 x 50 mm, 50 x 50 mm and 40 x 25 mm. The larger the exposed area of an RFID tag, the more easily it can be read. A larger tag can be read at a greater range than a smaller tag, and a tag laid flat at a greater range than a folded or bent tag.^[27,28] The maximum read range would therefore be provided by attaching the 50 mm edge of the largest tag to the syringe, with the rest of the tag protruding out from the syringe.

Such large protrusions, however, would be impractical when placing syringes in close proximity on a drug tray. In addition, anaesthetists suggested that having the RFID tag wrapped around the syringe would be preferable. When using IDAS they currently have to "tent" barcode labels to use them on smaller syringes. The label is folded in half so that the printed barcode is flat enough to be scanned, and attached so that the barcode itself protrudes around 10 mm from the syringe. Even this relatively small protrusion can cause handling difficulties.

Table 1: Read ranges of 13.56 MHz RFID tags

Table 1 shows the maximum read range of the smallest and largest tags attached by various means to the three sizes of syringe. Results for the 50 x 50 mm tag are excluded as they fall predictably in between. The results show that the smallest tag could be read as easily as the largest tag when wrapped around a large syringe, and more easily when wrapped around medium and small syringes. These results led to the decision that the prototype system would use 40 x 25 mm tags wrapped around syringes.

Reading an RFID Tag
In the IDAS trials the barcode scanner was placed on the anaesthetic workstation. As already noted, this required the anaesthetists to pause while moving from the drug trolley to the patient and a few participants reduced this inconvenience by moving the barcode scanner on to the end of the operating table. This suggests that a more natural movement from drug trolley to patient would be facilitated by locating RFID readers at one, or both, these endpoints. As only a single reader was available for this prototype, one location had to be chosen. Both have strengths and weaknesses as the location for a reader.

An RFID reader located on the drug trolley would detect when a tagged syringe has been removed. This is the earliest opportunity for the anaesthetist to be warned that he or she has selected the wrong drug, allowing the error to be corrected with the least effort. However, some anaesthetists keep syringes in places other than the drug trolley, so a reader there would not necessarily detect all syringes.

Locating an RFID reader at the patient would detect all tagged syringes, regardless of where they had come from. But it has the disadvantage of being the last possible point for confirming that the correct drug has been selected. It is also possible for the electromagnetic radiation (EMR) emitted by an RFID reader to interfere with sensitive medical electronic devices such as pacemakers.^{[ 29 ]} In light of the IDAS focus on patient safety, it was decided that the prototype system would use an RFID reader placed on the drug trolley, under the drug tray.

Associating a DIN with an RFID Tag and Passing it to IDAS
A simple software application, Associator, was developed to associate a DIN with an RFID tag. It accepts an RFID tag number from a reader, and a DIN from a keyboard or a barcode scanner. The combination is written to a database record. If an RFID tag reader/writer is used, the DIN can also be written to the tag itself.

A second application, Operator, was developed to read a tag and retrieve the associated DIN. It polls the RFID reader under the drug tray every half-second, receiving a list of RFID tag numbers. When a tag number disappears from the list, this is taken as a sign that a syringe has been removed from the tray. Operator looks up the database populated by Associator and retrieves the corresponding DIN. The IDAS software expects input from a barcode scanner attached to a serial port. Operator’s output is therefore passed to IDAS by channelling through Eltima Software’s Virtual Serial Port Developer application to make it appear to the IDAS software as though the information comes from a serial port.

The prototype system was demonstrated to Merry and his colleagues at Auckland Hospital. They seemed confident that RFID technology could functionally replace barcodes in the IDAS system, and raised a number of issues, and additional functionality that they believed should be addressed in the next stage of development.

Outstanding Issues and Further Work
The prototype system has problems detecting syringes in certain conditions. Showing information from the existing barcode labels on RFID tags also proved difficult. The demonstration was not conducted in a "live" operating theatre so did not experience any EMR interference. However, this needs to be considered. Suggestions for additional functionality centred around reducing the need for anaesthetists to enter data into IDAS.

Undetected Syringes and Inadvertent Reads
The drug trays used in the IDAS system are 25cm wide. As table 1 indicates, an RFID reader placed under the middle of a tray would not be able to read small syringes placed on the edge of the tray. One solution is to rely on RFID readers at the patient. The disadvantages of this have already been noted above. Another solution is to use two RFID readers to cover a tray. However, this would require careful placement and synchronisation to avoid the "reader collision" problem.^[30] A third solution is to experiment with types of RFID reader other than the "pad" style used in the prototype. "Gate"- and "loop"-style readers offer a greater read range, but at greater expense.

The possibility of the opposite problem, inadvertent reading, was also raised. This might occur where the anaesthetist removes a tagged syringe from the reader’s range, but not to inject it. For example, the anaesthetist may be moving the syringe to another part of the drug tray, or placing it in the waste bin. Excessive inadvertent reading might in fact lower the usability of the system, as the anaesthetist would have to manually remove the supposed injections from the automated record. Determining the actual incidence of inadvertent reading would best be done through usability trials, with anaesthetists using the Operator application in some simulated operations. If it does present a problem, the "syringe movement" scenario could be countered by requiring that a syringe is removed for a certain length of time before it is considered removed. A "syringe disposal" could be correctly identified by having an additional reader located on the waste bin.

Printing on RFID Tags
As noted under Research Methodology, some information from the current barcode labels must be visible on any RFID tags that replace them. For the demonstration, the barcode labels were simply attached to the RFID labels, and the combined label attached to the syringe. However, to reduce the effort involved in preparing syringes it would be preferable to have a single label. Combination RFID-encoder / barcode printers are available^[31] but do not appear to be suitable as they cannot produce the colours required for the standard. There are label printing bureaux that may produce suitable labels, but their services are only economical in print runs of hundreds of thousands.

Electromagnetic Radiation and Interference
As noted in the section "Reading an RFID Tag", the EMR emitted by RFID readers may affect pacemakers. It is also uncertain how prolonged exposure to EMR might affect patients. There is a good deal of research on the effects of EMR on people. However, the results to date are not conclusive and tend to focus on EMR from mobile phones which operate at frequencies higher than 13.56 MHz. On the basis of documentation for the RFID reader used in the prototype, it appears that it does comply with the New Zealand standard.^[32]
The anaesthetists interviewed pointed out that in a "live" operating theatre, there are a number of devices known to produce high levels of EMR. The greatest risk appears to be surgical diathermy machines. Nominally, they operate in the 1MHz frequency, but emit so much power that they produce interference harmonics at much higher frequencies.^{[ 3 ]} This interference could drown out the relatively low-power transmissions of RFID readers. Experiments in a live theatre situation would be required to determine the actual effect.

Reducing Anaesthetist Data Input Into IDAS
As noted in the section "The IDAS System", IDAS currently requires the anaesthetist to provide manual input after scanning the syringe. If the audio and visual cues indicate that the wrong syringe has been picked up, the anaesthetist must "cancel" the drug administration. This is also required if the anaesthetist selects the correct syringe, but doesn’t go ahead with the injection. The need to manually cancel a drug administration could be removed by introducing a second RFID reader, near the patient. This reader would signal to IDAS when a syringe is brought near an injection port. If the syringe is detected leaving the drug tray and then returning, without being detected at the patient, then IDAS could automatically treat this as a cancellation. IDAS could treat in the same way the situation where the syringe is detected at the patient but too briefly for an injection to have been completed.

If the audio and visual cues indicate that the correct syringe has been picked up, the anaesthetist must enter the drug dose administered. By using the signals from RFID readers to activate other devices it may be possible for IDAS to automatically determine the dose. For instance, a camera may be positioned over the drug tray to capture images each time a syringe is placed on it. Knowing which syringe has just been placed, the new image could be compared with a previous image to calculate the change in the level of liquid in the syringe. Positioning the drug tray on a scale may allow similar processing based on changes in weight.

Conclusion
ADEs are costly, both in lives and scarce healthcare resources. BCMA systems, such as IDAS, are one solution to reducing the incidence and impact of ADEs. However, barcode scanning appears to disrupt the natural anaesthetic workflow, possibly compromising the effectiveness of BCMA. The prototype system developed in this constructivist research project demonstrates that RFID technology offers the potential to functionally replace barcodes with minimal disruption to natural workflow.

A review of RFID application case studies indicates that RFID tags and readers operating at 13.56 MHz appear to offer a read range and interference profile that makes them the most suitable for location-based systems such as BCMA. Experiments with different sized tags and syringes suggest that the best balance of read range and usability results from smaller RFID tags wrapped around syringes.

The development of the prototype system illustrates a number of challenges in working with RFID in the hospital environment. The appropriate choice, and carefully designed layout, of RFID readers is necessary to minimise both undetected tags and inadvertent reading. Some aspects of barcode labels are difficult to reproduce with RFID tags. And the effects of EMR from RFID readers, and on readers from other devices, must be considered.

References

1. Bates DW, Cullen DJ, Laird N. Incidence of adverse drug events and potential adverse drug events. J Am Med Assoc 1995;1995(274):29–34.
2. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Newby L, Hamilton J. The quality in Australian health care study. Med J Aust 1995;163:458–471.
3. Davis P, Lay-Yee R, Briant R, Ali W, Scott A, Schug S. Adverse events in New Zealand public hospitals II: preventability and clinical context. N Z Med J 2003;116(1183).
4. Johnson JA, Bootman JL. Drug-related morbidity and mortality: a cost of illness model. Arch Intern Med 1995;155:1949–1956.
5. Berger RG, Kichak JP. Computerised physician order entry: helpful or harmful? J Am Med Inform Assoc 2004;11(2):100–103.
6. Brooks A. Robot fills prescriptions at UHS. Retrieved 21 October 2004 from http://www.news.harvard.edu/gazette/2003/07.17/13-robot.html
7. Patterson ES, Cook RL, Render ML. Improving patient safety by identifying side effects from introducing bar coding in medication administration. J Am Med Inform Assoc 2002;9(5):540–553.
8. Jensen LS, Merry A, Webster C, Weller J, Larsson L. Evidence-based strategies for preventing drug administration errors during anaesthesia. Anaesthesia 2004;59:493–504.
9. Anderson JG, Jay SJ, Anderson M, Hunt TJ. Evaluating the capability of information technology to prevent adverse drug events: a computer simulation approach. J Am Med Inform Assoc 2002;9(5):479–490.
10. Anderson JG, Jay SJ, Anderson M, Hunt TJ. Evaluating the impact of information technology on medication errors: a simulation. J Am Med Inform Assoc 2003;10(3):292–293.
11. Collins J. RFID remedy for medical errors. Retrieved 1 June 1 2004 from http://www.rfidjournal.com/article/articleview/961/1/1/
12. Food & Drug Administration. 2002N-0204: bar code label requirements for human drug products and biological products; final rule. Retrieved 21 October 2004 from http://www.fda.gov/cber/rules/barcodelabel.pdf 
13. Joint Commission on Accreditation of Healthcare Organisations. National patient safety goals. Retrieved 21 October 2004 from www.jcaho.org
14. Merry A, Webster C, Mathew D. A new, safety-oriented, integrated drug administration and automated anesthesia record system. Anesth Analg 2001;93:385–390.
15. Webster C, Merry A, Gander PH, Mann NK. A prospective, randomised clinical evaluation of a new safety-orientated injectable drug administration system in comparison with conventional methods. Anaesthesia 2004;59:80-87.
16. Merry A, Webster C, Weller J, Henderson S, Robinson B. Evaluation in an anaesthetic simulator of a prototype of a new drug administration system designed to reduce error. Anaesthesia 2002;57:256–263.
17. Merry A, Webster C. Bar codes and the reduction of drug administration error in anesthesia . In press 2004.
18. Standards New Zealand. User-applied labels for use on syringes containing drugs used during anaesthesia (AS/NZS 4375:1996). Wellington: Standards New Zealand; 1996.
19. Chassin MR. Is healthcare ready for sigma six quality. Milbank Q 1998;76(4).
20. Cornford T, Smithson S. Project research in information systems. New York: Palgrave; 1996.
21. Nunamaker JF, Chen M, Purdin TPM. Systems development in information systems research. Journal of Management Information Systems 1991;7(3):89–106.
22. Collins J. Healthy RFID rivalry for hospitals. Retrieved 9 September 2004 from http://rfidjournal.com/article/view/1094.
23. Broder C. Hospitals wade into asset-tracking technology. Retrieved 30 October2004 from www.ihealthbeat.com
24. Anonymous. Smart healthcare 2003: Conference Review. Retrieved 1 November 2004 from http://www.idtechex.com/smarthealthcareusa/2003review.asp.
25. Pappu M, Singhal R, Zoghi B. RFID in hospitals: issues and solutions. Retrieved 21 October 2004 from http://www.rfidjournal.com/whitepapers/download/20
26. Walker J, Spivey Overby C, Mendelsohn T, Wilson CP. What you need to know about RFID in 2004. Retrieved 9 February 9 2004, from www.forrester.com
27. Foster PR, Burberry RA. Antenna problems in RFID systems. In: IEE Colloquium on RFID Technology; 25 October 1999 IEE Savoy Place, London. IEE; 1999. p1–5.
28. Chen SCQ, Thomas V. Optimization of inductive RFID technology. In: IEEE International Symposium on Electronics and the Environment Denver, Colorado, 7–9 May 2001. Denver, Colorado: IEEE; 2001. p.82–87.
29. Irnich W. Electronic security systems and active implantable medical devices. Journal of Pacing and Clinical Electrophysiology 2002;25(8):1235–1258.
30. Engels D, Sarma S. The reader collision problem. In: 2002 IEEE International Conference on Systems, Man and Cybernetics 6–9 October 2002,. Hammamet, Tunisia. IEEE; 2002.
31. Collins J. Label printer for the undecided. Retrieved 1 August 2004 from http://rfidjournal.com/article/articleview/998/1/1/
32. Ministry for the Environment, Ministry of Health. National guidelines for managing the effects of radiofrequency transmitters. Wellington, New Zealand: Ministry for the Environment; 2000.
33. Bassen H. Electromagnetic interference of medical devices and implications for patient safety. International Journal of Bioelectromagnetism 2002;4(2):169–172.