rfid in the blood supply...

As indicated by the Institute of Medicine’s

revolutionary studies To Err is Human:

Building a Safer HealthSystem1, Crossing the

Quality Chasm: A New Health System for the 21st Centu-

ry2 and Preventing Medication Errors3, the high rate

of medical errors and significant gaps in the quality

of healthcare are great cause for concern. Annually,

an estimated 44,000 to 98,000 Americans die from

medical errors.1,4,5 Accordingly, reducing medical

errors and improving patient safety has been identi-

fied as a top-20 priority for national action in trans-

forming healthcare quality.6 Playing an integral role

in patient care, transfusion medicine represents a

great opportunity for safety and quality improve-

ments. Risk of mis-transfusion, or mis-match

between patient and blood, is more than 100 times

greater than HIV or HCV transmission from blood

transfusion7, making mis-transfusion the arguably

the most important and serious hazard of transfu-

sion.7,8,9,10,11 Mis-transfusion, which most frequently

occurs among surgical patients, typically results

from an error made during the bedside check, per-

formed manually using eye-readable information,

just prior to transfusion. Despite the availability of

new supportive technology, and indications that

repetitive task performance is enhanced by using

technology12, the methodology used for this check

has changed little over the past 50 years.

RFID in the Blood Supply ChainIncreasing Productivity, Quality and Patient Safety

By Lynne Briggs; Rodeina Davis; Alfonso Gutierrez; Matthew Kopetsky; Kassandra Young; and Raj Veeramani

KeYwoRDs

RFID, supply chain, process, technology, pain points.

ABstRAct

As part of an overall design of a new, standardized RFID-

enabled blood transfusion medicine supply chain, an

assessment was conducted for two hospitals: the University

of Iowa Hospital and Clinics (UIHC) and Mississippi Baptist

Health System (MBHS). The main objectives of the study

were to assess RFID technological and economic feasibility,

along with possible impacts to productivity, quality and

patient safety. A step-by-step process analysis focused on

the factors contributing to process “pain points” (errors,

inefficiency, product losses). A process re-engineering

exercise produced blueprints of RFID-enabled processes

to alleviate or eliminate those pain-points. In addition, an

innovative model quantifying the potential reduction in

adverse patient effects as a result of RFID implementation

was created, allowing improvement initiatives to focus on

process areas with the greatest potential impact to patient

safety. The study concluded that it is feasible to implement

RFID-enabled processes, with tangible improvements to

productivity and safety expected. Based on a comprehensive

cost/benefit model, it is estimated for a large hospital (UIHC)

to recover investment from implementation within two to

three years, while smaller hospitals may need longer to

realize ROI. More importantly, the study estimated that RFID

technology could reduce morbidity and mortality effects

substantially among patients receiving transfusions.

Cast study: RFId

54 jhIm n FALL 2009 n voLume 23 / Number 4 www.himss.org

Barcode technology, using a hand-held laser that requires line-of-sight orientation to read a flat surface bearing a code, has reduced mis-transfusion errors, yet lacks broad adoption due to technological limitations.13-16 The use of RFID technol-ogy, which requires no line of sight, to improve safety, quality and productivity in transfusion medicine, has been explored at several institutions.8,14,17,18,19,20 Until recently, no study has explored the use of RFID in the entire blood medicine supply chain—from blood collection, to manufacturing and distribu-tion, through patient transfusion.

This paper focuses on the RFID assess-ment work performed for the hospital side of the transfusion medicine supply chain. A description of the preliminary RFID assess-ment work performed at the blood center side of the supply chain is documented in “Track-ing blood products in blood centers using radio frequency identification: a comprehensive assessment”. 21 During 2006 and 2007, a multidisciplinary project team, includ-ing individuals from the Blood Centers Consortium—BloodCen-ter of Wisconsin; Carter BloodCare, Dallas; and Mississippi Blood Services, Jackson—with support from SysLogic Inc., Milwaukee, Wisc., and the University of Wisconsin RFID Lab, Madison, was assembled to analyze and redesign processes in the hospital por-tion of the blood transfusion medicine supply chain. This analysis and redesign was accomplished by examining practices at two institutions: the University of Iowa Hospital and Clinics (UIHC) and Mississippi Baptist Health System (MBHS). Hospital experts and blood transfusion practitioners, including transfusion service medical directors, managers and supervisors, aided in the analy-sis. Additionally, directors, managers and supervisors from phle-botomy and nursing participated in evaluation activities to ensure a well rounded understanding of current transfusion processes. Two institutions of varying size and transfusion practice were selected to ensure that the results of the evaluation and redesign could be generalized.

Many challenges were faced throughout the duration of the analysis. To establish common industry baselines, it was impor-tant to establish the process commonalities between the partici-pating hospitals, while also documenting the differences in trans-fusion processes. Allowing for variation in the process redesign was crucial. Furthermore, agreeing upon common terminology to be used in reference to the processes also was vital, as well as obtaining industry benchmark data relative to transfusion errors and outcomes.

PROCESS ANALYSIS AND REDESIGN

Methodology: A three-step approach was followed to conduct the process-oriented assessment of both hospitals’ transfusion services. The first step was to document all current processes and associated tasks relating to identification, data capture and tracking of blood products. The overarching goal of the “as-is”documentation phase was to create process maps that would be applicable to both institutions, while allowing for any variations in transfusion practices. Delphi-like consensus-building interac-tions with the various stakeholders was used to create the maps.

As a final step in the documentation of the as-is process flows, the process owners and top management at each of the participating hospitals validated all process maps.

The second step consisted of identifying and characterizing “process pain points,” or steps in the processes that could intro-duce errors, inefficiencies and/or non-value-added effort and delays. These pain points would ultimately serve as the basis for identifying and prioritizing how RFID could be applied to improve hospital operations. Using the process flows as a guide, process

owners identified the locations of all pain points, and ranked each pain point according to frequency of occurrence and magnitude of the consequences resulting from the related inefficiencies or mistakes. An agreed upon scoring mechanism was used. The pain points were then categorized and the scores summed by category. Finally, the team focused in the process checkpoints (steps where a check is done to confirm the correct identification and matching of patient, product, order specifications, etc.). Checkpoints were defined as required tasks designed to reduce/eliminate transfu-sion errors at key points in every process. Due to the often manual nature of these tasks today, checkpoints frequently present oppor-tunities for error or can be bypassed.

The third and final step was to actually redesign the transfu-sion-related processes applying RFID technology as appropriate to mitigate the previously identified pain points. The redesign effort also aimed to leverage potential gains in efficiency through the use of automation as the result of RFID implementation. The “to be” process design serves as the main input for technical spec-ifications of the IT components needed to support the overall sys-tem. Detailed data models, process modules and system-to-sys-tem interfaces are then derived from this analysis.

Results: A high-level process flow, encompassing all of the hospitals’ transfusion related processes, and 14 mid-level process flows, illustrating greater detail of the actual processes, were gen-erated. The top-level process flow is shown in Fig. 1. Fig. 2 illus-trates an example of a mid-level process. The process flows depict operations that are currently performed at both hospitals, with any variation occurring between the hospitals called out via black circles. A complete, formal record of all variation in transfusion practices between participating institutions was maintained. This allowed for the creation of process maps that were applicable to both institutions, while still accounting for variation.

Due to the limited benchmark data in the United States, the research was extended to other countries to determine more com-plete baseline data (Quebec Health & Social Services Blood Sys-tem, Canada).

The analyses revealed that several common and recurring pain points in the transfusion medicine processes exist. The fol-lowing pain points were determined to be the most important by

-

- - -

RFID technology has the potential to provide many robust benefits beyond a traditional financial analysis. Therefore, benefits such as patient safety should also be formally quantified.

www.himss.org voLume 23 / Number 4 n FALL 2009 n jhIm 55

the process team, as shown in Fig. 3. Pain points are categorized based on the severity of consequences that could result were there a failure at the corresponding checkpoint. Pain points with a diamond-shaped symbol denote an area of primary safety concern; pain points with a circular-shaped symbol correspond to a secondary safety concern; and pain points with a square-shaped symbol result in detriment to efficiency or workflow, but no bodily harm.

By examining these pain points, opportunities for redesign-ing processes and workflows with RFID to increase productiv-ity, safety and cost-effectiveness were identified. Workflows were redesigned as to address the identified pain points and errors leveraging the new control capabilities brought by RFID technol-ogy. Checkpoints were included in the redesign to address the previously mentioned pain points. As they did in the existing processes, checkpoints still aid in the confirmation of the correct identification and matching of patient, product and order speci-fications. However, the newly designed processes leverage the technology of RFID to perform the actual check. The checkpoints use different shaped symbols to denote which type of pain point is being addressed, using the same key as previously described for the pain points themselves.

The process redesign resulted in 16 RFID-enabled process maps for common hospital transfusion processes. Improve-ments over the current model were estimated and classified into three categories: safety, productivity and quality. These esti-mates were used as the input for the financial ROI calculations performed at the conclusion of the process assessment. Due to the commonality of processes and pain points in the industry, it is expected that the RFID-enabled process designs can serve as a blueprint for process re-engineering with RFID in the transfu-sion medicine industry.

Fig. 4 illustrates an example of an RFID-enabled process: sam-ple collection. In the new process, a nurse or phlebotomist first electronically receives the transfusion request form (TRF). Subse-quently, this individual prints a barcode label for the sample col-lection tube, carries the label to the patient room and confirms that the label matches the patient with a handheld dual-RFID/barcode reader. A sample is collected from the patient and placed in the barcode-labeled tube, and the tube is again scanned to confirm that the patient and the sample match. If the patient and sample match, an electronic signature is generated, automatically updat-ing the status of the process. Later, a blood bank employee scans the sample tube to confirm the presence of the electronic signa-

Fig. 1: top-Level Process Flow


ture. Finally, the electronic TRF is placed in the pending sample testing file, where it awaits the subsequent process, patient sam-ple testing. Notice that checkpoints, shown with diamond-shaped symbols labeled C1, C2 and C3, were included in the RFID-enabled process. The use of the diamond-shaped symbols indicates that these checkpoints address pain points of primary safety concern.

Fig. 5 provides another example of an RFID-enabled process—transfusion. In the new process, the nurse or blood bank assistant, equipped with a handheld dual-RFID/barcode reader, receives the RFID tagged blood product. The nurse or blood bank assistant can use this technology to electron-ically check that the blood product matches what was ordered for the patient, verify that a valid crossmatch has occurred, and later confirm when the transfusion has been completed. If the transfusion has to be delayed and the product can be stored for later use on the hospital floor, the nurse or blood bank assistant can easily confirm that the right product is retrieved when the patient is ready to be transfused. Again, the re-designed checkpoints use different shaped symbols to signify-ing the type of pain point addressed by each checkpoint.

In developing and analyzing RFID-enabled process flows, the process design team noted that the majority of processes were accomplished by executing common sub-processes. Common sub-processes were identified as system building blocks. By assembling the functional components, or building blocks, it is

possible to build the majority of the system’s functionality with standard, reusable components. This will minimize application development time and cost. The building blocks for RFID-enabled hospital processes are shown in Table 1.

This methodology proved efficient and successful in the assess-ment of RFID application in blood center operations. A major benefit of this approach is that it first sought to understand the

Fig. 2: example of a Mid-Level “As-is” Process Flow

Based on the comprehensive cost/benefit model, a large hospital like UIHC is estimated to recover investments from implementation within two to three years. By contrast, smaller hospitals may require additional time to recover costs.


causes of common challenges (the pain points) in processes and then developed solutions that meet the needs of the processes. By redesigning processes to overcome the pain points, this method-ology helps ensure that the proposed changes are relevant, useful and have a high probability of acceptance by the intended users. This methodology is summarized in Fig. 6.

Overall, it was concluded that RFID-enabled processes have the potential to significantly reduce many sources of errors and inef-ficiencies in hospital operations. Various identification processes relating to patients, collection tubes and TRFs can be accomplished more accurately and efficiently by using RFID. As an example, an alert can be sent to the caregiver when patient identifiers do not match. Also, through automation, the implementation of RFID-enabled processes can ensure that compatibility test results are checked. Furthermore, safety, productivity and process quality all stand to gain improvements with the implementation of RFID.

IMPACT ANALYSIS

The purpose of the impact analysis was to estimate the effect RFID would have on hospital operations in terms of productiv-ity, quality, safety, responsiveness and cost. The impact analysis consists of a return on investment (ROI) analysis and a return on

safety (ROS) analysis. Subject matter experts from each organiza-tion provided the input for both of these analyses.

Return on Investment: To create the ROI model, spread-sheets were employed with a separate worksheet for each of the six model components. Following is a list of the relevant elements and an explanation of their purpose:

The profile worksheet contains various statistics relevant to UIHC and MBHS, including the number of blood products trans-fused over the course of a year, existing process lead times and hos-pital staffing levels. These values were provided by the hospitals. Additionally, data relating more generally to small (less than 100 beds); medium (100 to 299 beds); and large (300 or more beds) hos-pitals was included in the profile section. This data was collected as a result of research and the input from subject matter experts.

Control: the control worksheet summarizes the actual cost/benefit analysis of the model. A user first selects which institu-tion he would like to see information for: UIHC, MBHS or more general estimates for a hospital based on size. Calculations may be executed with or without depreciation considerations.

Deployment costs: the control worksheet pulls costing informa-tion from the deployment costs worksheet. Here, costs relating to hardware, software, infrastructure, tags and development costs

1.

2.

3.

Fig. 3: Pain Points


are detailed. Cost data was provided by IT experts from the par-ticipating entities.

Depreciation costs: the depreciation worksheet is similar to the deployment cost worksheet. This worksheet factors for the effects of depreciation to applicable items. If the depreciation option is chosen, the model will use these values in the control sheet when calculating costs.

Productivity: the productivity worksheet computes the produc-tivity gain or loss for all of the processes. Changes in productivity were determined based on consultation and validation from the appropriate project team members.

Quality: the quality worksheet is similar to the productivity worksheet. Quality focus is on blood product units lost quantifica-tion. Again, quality improvements were estimated based on the consultation of project team members.

cost/Benefit Methodology. Several options must be select-ed in the spreadsheet before using the model, as this will drive costs, benefits, net present value and, ultimately, payback period. First, the institution (UIHC, MBHS, or another small, medium or large hospital) must be chosen. This allows for the appropriate scaling parameters to be used. Generally, institu-tion selection will have the greatest impact in terms of profile

4.

5.

6.

data. Next, an accounting method—either cash-based or depre-ciation—must be selected. For ROI, the main cost categories included RFID tags, RFID hardware, IT infrastructure, software and implementation. Loaded staff salaries, accounting for both base pay (hourly wage) and benefits (insurance, vacation, etc.) were determined. Loaded salary and time required for staff to complete transfusion processes was used to determine staffing costs. Salary data for UIHC and MBHS was provided by each institution, while national averages were used as estimates for small, medium or large hospitals. The model has the flexibility to adjust RFID deployment schedules (for example, implementing varying percentages of the RFID technology over a multi-year planning period) based on institution-specific implementation needs, as well as other financial parameters such as minimum internal rate of return expected for hospital investments.

Benefit categories were classified as productivity gain/loss and quality. Input parameters for both categories were col-lected by surveying process owners and technology experts. To determine expected gains in quality, a baseline estimate of errors as they relate to the 14 existing processes was created. Next, since experts in the field of transfusion medicine gen-erally agree that errors are widely underreported22, baseline

Fig. 4: example of an RFID-enabled Process: sample collection


error estimates were inflated by approximately 50 percent. The actual error rate was ultimately scaled to each institu-tion as a function of the number of transfusions performed per year.

Results. It is economically justifiable to invest in building infrastructure and implementing new processes for UIHC, as investment may be recovered in less than three years. Prelimi-nary estimates based on a comprehensive cost/benefit model shows that UIHC, with RFID-enabled processes, would realize benefits (net present value over a five-year planning horizon) of nearly $1 million, while incurring less than $820,000 in one-time initial investment costs and under $190,000 in recurring oper-ating costs. This means a break-even period of just under three years (approximately 30 months), when cumulative benefits exceed the cumulative cost and positive cash flow ensues. This calculation is the net result of tangible gains in productivity and quality vs. the cost of implementing and operating RFID-enabled processes from blood center order receipts, to transfusion order management and final transfusion delivery to the patient. The

cost estimate assumes an RFID-based software solution being available in the near future for hospital implementation. By con-trast, a longer payback period of 6.9 years would be required to economically justify the implementation of the RFID-enabled processes at MBHS. Over the five-year planning horizon, MBHS would realize over $400,000 in benefits, while incurring nearly $460,000 in costs. Unfortunately, these figures would result in a negative net present value (total cumulative benefits minus total costs) at that stated benchmark timeframe, requiring close to two more years to complete the payback. The financial model developed by the study team permitted analyzing the effect of hospital volumes in the expected payback timeframe. As it also was found in the preliminary study, a direct correlation of hospi-tal size with the time to recover the investment was observed. In other words, bigger hospitals will realize the benefits of automa-tion by RFID faster.

Refinement of estimates. The payback period in implement-ing RFID technology is largely correlated to the frequency of occurrence for each process step. Each of these frequencies is cur-

Fig. 5: example of an RFID-enabled Process: transfusion


rently estimated to occur for each unit transfused in a hospital. However, data is currently being collected from UIHC and MBHS to define the frequency of each process step based on the indi-vidual institutions process flows. For example, at some hospitals, process “3.0 Sample Collection” may occur once for each patient receiving a transfusion, whereas, process “9.0 Retest Product ABO/Rh” may occur once for each red cell unit transfused. Defin-ing the frequency of each process step more accurately will ulti-mately ensure more accurate estimations of payback period for individual hospitals.

Additional Benefits. RFID enablement allows for an increase in blood processing visibility throughout the hospital and the entire supply chain. In the future, real-time information shar-ing could improve visibility of hospital actual demand, incom-ing units from collection sites allowing improved throughput throughout the entire system, better scheduling, smoothing of production planning, improved inventory balancing at both blood centers and hospital blood banks, reduce the number of expired products and improved rush/expedited (stat) order management. An end-to-end RFID-enabled supply chain can improve lead time management through enhanced visibility of inventory levels and location (multi-site visibility). Lead time reduction, in turn, will help reduce costs while improving the hospital’s ability to meet transfusion demand.

Patient Return on safety. In addition to the ROI component, the team also attempted to measure and analyze the safety impact as a part of the overall investment justification. Analysis indicates that there is a potential to directly improve patient safety via the use of RFID. However, quantitative models for assessing the effects of specific errors and the corresponding outcomes of the adverse reactions experienced by patients as a result of those errors have previously not been developed in the blood transfusion field. In the analysis, the team explored a return on safety (ROS) model capable of quantifying decreases in both patient mortality and less severe morbidity effects as a result of error reduction achieved through implementation of RFID technology. The following is a work in progress.

High-level Approach. To quantify the patient morbidity and mortality level improvements resulting from RFID implemen-

tation in hospital blood transfusion pro-cesses, development of an accurate base-line model, representing the current safety status of blood transfusions in the United States, is required. Unfortunately, although the National Healthcare Safety Network is in the process of implementing a hemov-igilance (blood transfusion monitoring) system in the United States23, such data does not yet exist, and therefore, available sources (noted below) were analyzed in an effort to best represent the North Ameri-can transfusion industry.

Through an intensive survey of available hemovigilance literature, the project team was able to obtain the following data:

Incidence of hospital errors in transfusion (Quebec Health & Social Services Blood System).24-26

Incidence of manufacturing errors in transfusion (FDA).27

Incidence of adverse reactions to transfusion (SHOT28, Quebec Health & Social Services Blood System, & Canadian TTISS29)

Incidence of mortality resulting from transfusion (FDA30,31, SHOT, Canadian TTISS).

Consequently, a team of expert practitioners was relied upon to provide their experienced-based best estimates of the incidences of unavailable data. Since data regarding the correlation between errors and adverse reactions was unavailable, the team estimated these correlations by rating each as high, medium or low prob-ability. Additionally, morbidity data regarding the severity of each morbidity effect; the likelihood that the effect would occur given the corresponding adverse reaction; and the duration that each effect would last also was rated on a similar scale. Mathemati-cal non-linear programming models were then used to correlate these various sources of data to create baseline indices for both mortality and morbidity effects resulting from transfusion, and the impact of process changes with RFID. The team is currently analyzing their findings.

Preliminary Results. As noted, the model and results are still being evaluated, but preliminary findings indicate a reduction in errors resulting in patient morbidity and mortality.

CONCLUSION AND NEXT STEPS

The methodology and approach taken by the project team for assessing and justifying the adoption of a technology such as RFID, is applicable to the assessment of other new technolo-gies in healthcare, providing a sound foundation for a success-ful implementation and adoption. A detailed analysis of existing processes should be a project’s starting point, and a fundamen-tal step for laying a sound understanding of current deficiencies, leading to a firm foundation of potential technological solutions. A deep process analysis and re-design as well as a comprehen-sive understanding of the potential impact (financial and safety), should precede technological application design and adoption, not vice-versa. This planning prevents healthcare institutions

•

••

•

table 1: Building Blocks for RFID-enabled Processes


from incurring expensive redesign costs during late stages of implementation and experiencing unforeseen changes in related processes post-implementation. RFID technology has the poten-tial to provide many robust benefits beyond a traditional financial analysis. Therefore, benefits such as patient safety should also be formally quantified. The overall impact analysis, both financial and patient safety related, should be conducted as early in a tech-nological adoption project as possible.

The preliminary findings and evaluation of this study pres-ent a positive outlook to the impact RFID technology would have in the processes at the hospital portion of the blood trans-fusion medicine supply chain. This new supply chain model is deemed feasible and desirable. As a result of implementation, tangible potential improvements to productivity and safety are expected. Based on the comprehensive cost/benefit model, a large hospital like UIHC is estimated to recover investments from implementation within two to three years. By contrast, smaller hospitals may require additional time to recover costs. Perhaps more importantly, the study indicates that RFID tech-nology could reduce morbidity and mortality effects substan-

tially among patients receiving transfu-sions. It is important to note that specific estimates relating to increased patient safety are pending validation of the pro-posed safety model.

There are several immediate activities planned as a continuation of this project. First, technical specifications relating to system requirements will be integrated with an initiative already underway in the blood center portion of the supply chain. Subsequently, an actual prototype will be developed for testing the concept of leveraging RFID to improve transfu-sion-related processes. Additionally, a rigorous validation of the proposed safe-ty model is currently taking place. This effort includes a review process by several experts in the field of transfusion medi-cine and patient safety. Finally, the finan-cial model assumptions are to be validated and refined as results from the prototype

stage of the overall RFID in blood transfusion medicine initia-tive are evaluated. JHIM

Lynne Briggs is Director, IS Applications/RFID Study Project Manager, at the

BloodCenter of Wisconsin.

Rodeina Davis is Vice President and CIO/RFID Study Program Director, at the

BloodCenter of Wisconsin.

Alfonso Gutierrez is Director, UW RFID Lab, at the University of Wisconsin-

Madison.

Matthew Kopetsky is Project Assistant, UW RFID Lab, at the University of

Wisconsin-Madison.

Kassandra Young is Project Assistant, UW RFID Lab, at the University of

Wisconsin-Madison.

Raj Veeramani is Executive Director, UW E-Business Consortium, at the University

of Wisconsin-Madison.

Fig. 6: Process Design and Analysis Methodology

ReFeRences1. Institute of Medicine. To Err is Human: Building a Safer Health System. Washington, DC: National Academy Press;2000.

2. Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press;2001.

3. Institute of Medicine. Preventing Medication Errors. Washington, DC: National Academy Press;2006.

4. Thomas EJ, Studdert DM, Newhouse JP, Zbar BIW, Howard KM, Williams EJ, Brennan TA. Costs of medical inquiries in Utah and Colorado. Inquiry. 1999;36(3):255-264.

5. Thomas EJ, Studdert DM, Burstin HR, Orav EJ, Zeena T, Williams EJ, Howard KM, Weiler PC, Brennan TA. Incidence and types of adverse events and negligent care in Utah and Colorado. Med Care. 2000;38(3):261-271.

6. Institute of Medicine. Priority Areas for National Action: Transforming Health Care Quality. Washington, DC: National Academy Press;2003.

7. Dzik WH. Transfusion safety in the hospital. Transfusion. 2003;43:1190-99.

8. Dzik WH. New technology for transfusion safety. Br J Haematol. 2007;2: 181-90.

9. Minz PD. Nishot: on target, but there’s no magic bullet. Am J Clin Pathology. 2001;116(6):802-805.

10. Sazama K. Reports of 355 transfusion-associated deaths: 1976 through 1985. Transfusion. 1990;30:583-590.

11. Love EM, Soldan K. SHOT Annual Report 2000-2001. 2002;1-239.

12. Bates DW, Cohen M, Leape LL, Overhage JM, Shabot MM, Sheridan T. Reducing the frequency of errors in medicine using information technology. JAMIA. 2001;8:299-308.


13. Hopkins S. Advances in patient and specimen ID. Advance News Magazine. [serial online]. 2005;14(2). Available at: http://laboratory-manager.advanceweb.com. Accessed February 1, 2005.

14. Dzik S. Case Study: RFID in action–the Massachusetts General Hospital START Project. IDTechEx Ltd Web site. 2004. Available at: www.idtechex.com/smarthealthcareusa/3.asp. Accessed February 1, 2005.

15. Computerworld. German hospital uses RFID for patient identification. Computerworld Web site. 2005. Available at: www.computerworld.com. Accessed May 20, 2005.

16. PDC. U.S. Navy tracks wounded in Iraq with PDC’s RFID technology. Precision Dynamics Corp Web site. 2003. Available at: www.pdcorp.com/healthcare/rfid_militaryuse.html. Accessed May 20, 2003.

17. Lusky K. Adding RFID layer to blood safety loop. CAP Today [serial online]. 2005;45-52. Available at: www.cap.org/apps/docs/cap_today. Accessed July 2005.

18. Dalton J, Ippolito C, Poncet I, Rossini S. Using RFID technologies to reduce blood transfusion errors. White paper by Intel Corporation, Autentica, Cisco Systems and San Raffaele Hospital. 2005. Available at: http://www.intel.com/it/mobility-wireless/rfid3.htm?iid=search&. Accessed September 2005.

19. Knells R. Radio frequency identification (RFID): an experience in transfusion medicine. ISBT Science Series. 2006;1(1):238-241. Accessed December 2008.

20. Stegwee RA. Presentation at World of Health IT: Visibility of Persons, Materials, and Blood Products with RFID: Results from a Large-scale RFID Pilot in the Operating Room of the AMC. RFID in de zorg Web site. 2008. Available at: http://www.rfidzorg.nl/WoHIT%20RFID%20v071022.pdf. Accessed December 2008.

21. Davis R. Geiger B. Gutierrez A. Heaser J. Veeramani D. Tracking blood products in blood centres using radio frequency identification: a comprehensive assessment. Vox Sang [serial online]. 2009. Available at: http://www3.interscience.wiley.com/journal/122269598/abstract. Accessed May 21, 2009.

22. Linden J, Paul B, Dressler K. A report of 104 transfusion errors in New York State. Transfusion. 2003;32(7):601-606.

23. Biovigilance Component. National Healthcare Safety Network (NHSN). 2009. Available at: http://www.cdc.gov/nhsn/bio.html. Accessed May 20, 2009.

24. Robilliard P, Nawej KI. Les incidents et accidents transfusionnels signalés au système d’hémovigilance du Québec en 2004. Québec Santé et Services sociaux. 2004. Available at: http://msssa4.msss.gouv.qc.ca/santpub/sang.nsf/fb8ffe4fe179b35785256c30005dd024/50561a37902dcbbf85256eb4006fc7c9?OpenDocument. Accessed May 20, 2009.

25. Robilliard P, Nawej KI, Chapdelaine A. Les incidents et accidents transfusionnels signalés au système d’hémovigilance du Québec en 2005. Québec Santé et Services sociaux. 2005. Available at: http://msssa4.msss.gouv.qc.ca/santpub/sang.nsf/fb8ffe4fe179b35785256c30005dd024/50561a37902dcbbf85256eb4006fc7c9?OpenDocument. Accessed May 20, 2009.

26. Robilliard P, Nawej KI, Chapdelaine A. Les incidents et accidents transfusionnels signalés au système d’hémovigilance du Québec en 2006. Québec Santé et Services sociaux. 2006. Available at: http://msssa4.msss.gouv.qc.ca/santpub/sang.nsf/fb8ffe4fe179b35785256c30005dd024/50561a37902dcbbf85256eb4006fc7c9?OpenDocument. Accessed May 20, 2009.

27. Biological Product and HCT-P Deviation Reports. FDA-Centers for Biological Evaluation and Research. 2007. Available at: http://www.fda.gov/cber/biodev/reports.htm. Accessed May 20, 2009.

28. Serious Hazards of Transfusion. 1996-2007. Available at: http://www.shotuk.org/. Accessed May 20, 2009.

29. National Transfusion Transmitted Injury Surveillance System. Public Health Agency of Canada. 2009. Available at: http://www.phac-aspc.gc.ca/hcai-iamss/tti-it/. Accessed May 20, 2009.

30. Fatalities Reported to the FDA Following Collection and Transfusion: Annual Summary for Fiscal Year 2005-2006. U.S. Food and Drug Administration. 2007. Available at: http://www.fda.gov/cber/blood/fatal0506.htm. Accessed May 20, 2009.

31. Fatalities Reported to the FDA Following Collection and Transfusion: Annual Summary for Fiscal Year 2007. U.S. Food and Drug Administration. 2008. Available at: http://www.fda.gov/cber/blood/fatal07.htm. Accessed May 20, 2009.


rfid in the blood supply...

Documents