Background: Error prevention in healthcare is paramount. Beyond introducing computerised provider order entry (CPOE) and automated medication dispensing systems, patient identification systems have also been proposed in an effort to increase patient safety while reducing misidentification. Electronic patient records are thus integrated into closed-loop systems. In Switzerland, there are no systematic data available on misidentification rates in hospitals. This study sought to evaluate the nature and frequency of the misidentifications prevented using a computerised, multipurpose, workflow-integrated patient identification system as part of a fully electronic patient record. We analysed actions comprising phlebotomy, drug administration, and transfusion.
Methods: Over a 30-month period, 24 879 in-patient stays in a public hospital were analysed concerning identification system usage and prevented misidentification, focusing on phlebotomy, transfusion, and drug administration areas. All identification checks were performed using the same device and software.
Results: The results of 38 199 bedside checks revealed low misidentification rates and moderate system usage within a non-mandatory setting. Clear differences were noted between specific tasks. In this setting, a total of 457 misidentifications were prevented by the system.
Conclusion: Misidentifications are not common, generally occurring in less than 2% of bedside actions. The absolute numbers are, however, cause for concern and thus merit significant preventive efforts. While implementing a multipurpose system can increase nurse acceptance, this is associated with additional workload pertaining to identification checks, which proves an issue. Increasing patient safety by introducing a wristband identification system, for example, thus appears worthwhile, although it must be designed as to respect the socio-technical environment.
Key words: patient safety, patient identification system, medical errors, quality assurance, misidentification, wristband
There are decades’ worth of studies proving the existence of and analysing medical errors. Following the publication of several milestone publications [1–4], awareness concerning errors in patient trajectories significantly increased in the early 2000s [5, 6]. As patients are in a vulnerable condition per se, any critical event or error may have deleterious consequences. It is thus crucial to prevent all error in the healthcare setting.
Introducing electronic support into healthcare processes has been associated with a decrease in both the opportunity for error and actual error rates . The medication process – one of the most common, fragmented, and interdisciplinary workflows in all healthcare, prone to a wide variety of errors – is especially well-studied. Traditionally, computerised provider order entry (CPOE) has been advocated in order to vastly reduce prescription, documentation, and translation error [8–11]. In addition, electronic systems for drug dispensing and administration have been shown to improve both quality of care and patient safety [12–17]. Most medication errors have been shown to occur in two specific areas, namely drug prescription and administration . Consequently, several publications have advocated the introduction of so-called closed-loop systems, where the whole process, beginning with prescription and ending at the bedside identity check, is conducted on electronic devices . Computerised bedside systems (e.g., during medication administration), and thus closed-loop systems, are still very rare in Swiss hospitals, however. In addition, while there are vast data available on errors occurring in the medication process, far fewer studies have been published on electronic systems used for other diagnostic or therapeutic interventions (e.g., sampling errors, labelling errors, wrong blood in tubes, or wrong-site surgery) [20–25]. These tasks may not be less prone to errors and the serious potential consequences than those of the medication process.
Although computerised systems offer opportunities in terms of patient safety, the unintended consequences of introducing information technology are also well known [26–30] and deserve great attention, given that changing the socio-technical  environment can give rise to new errors . CPOE itself does not guarantee success, and many paper-based prescriptions are still made out, with electronic systems often insufficiently used. According to Radley et al. , only 39–46% of prescriptions (depending on hospital size) are digitally made out, even when CPOE is implemented, revealing how difficult it is to reach high acceptance and high usage. Consequently, its implementation has to be carefully planned and monitored [34, 35]. Given the fragmented und multifaceted tasks inherent in this domain, especially considering nursing activities, it seems reasonable to reduce the number of applications and devices as much as possible in an effort to facilitate efficient work flow. Little is known so far as to how best to integrate identification systems into (nursing) workflows, and most studies to date have been focused on just one specific kind of identity check (e.g., either medication or transfusion).
Misidentification (e.g., administration of the wrong medication to a patient, mistransfusions, or wrong-site surgery) may exert a particularly high impact on patient safety, thereby representing a special burden on nurses and physicians alike. Although nearly all hospitals practice positive patient identification prior to conducting any type of intervention on a patient, few use computerised systems to prevent misidentification in the last steps of many healthcare actions. Wristband technology, integrated into the workflow, represents a possibility to reduce bedside errors. As to the underreporting of misidentifications, studies have found that many more near-misses have been detected in centres using barcode-based identification systems , emphasising the relevance of both actual underreporting and prevention.
Although numerous investigations into misidentification rates have been conducted in primarily paper-based environments, very little data has been published on misidentification rates or what misidentifications are actually prevented by using computerised systems within European hospitals, with no data at all available for Switzerland. It must also be mentioned that, while several publications dealing with this area are based on direct observation or reporting system results (e.g., critical incident reporting systems or CIRS), these papers do not specifically address misidentification. Direct observations are prone to human error too, and underreporting of critical or adverse events has been widely suggested [37–40].
This quantitative study sought to evaluate misidentification rates detected in a closed-loop system, occurring not only in the area of medication administration but also blood sampling/phlebotomy and blood component transfusions. As only anonymous patient data was analysed, no formal approval of the institutional ethics committee was necessary.
The local setting was a public, 300-bed primary and secondary acute care hospital in Thun, Switzerland. The establishment has three major departments (Internal Medicine, Surgery and Orthopaedics, as well as Gynaecology and Obstetrics), admitting 16 000 in-patients annually. For this analysis, the Gynaecology and Obstetrics department was completely excluded, given that the patient identification system was not used in this department during the analysed period. Only in-patients were analysed. First and foremost, usage of the system was not mandatory. The nursing management of the internal medicine department decided to additionally employ the system for medication administration, whereas the surgical department voted to use it for blood sampling and transfusions.
The identification system (IDEF-IS), built between 2006 and 2008 as a home-grown solution to prevent misidentifications of any kind, was fully integrated into the complete electronic patient record. The system includes identification data of the personnel (radio-frequency identification [RFID] chip on the ID badge), patients (RFID wristband), and materials (barcodes on blood sampling tubes, blood products and medication). While system usage was not mandatory, the identification check could be used at each nurse’s discretion. This was for the following reasons: (a) positive patient identification without computerised systems had been used for decades, yet with the exclusion of wristbands; (b) the system was assumed to create significant additional workload – especially during medication administration; (c) (digital) wristband technology has never been either mandatory or part of routine workflows in Swiss hospitals.
The mapping between the electronic patient record and the patient’s wristband (RFID, 13.56 MHz, ISO 15693, PDC Corporation) was performed at the patient’s bedside by the nurse in charge every time a patient was admitted to the designated ward for the first time. The unique RFID identifier served as the only link between both the wristband and the electronic patient record. There was no human-readable information on the wristband (thus avoiding optical identification only), and no data stored on the wristband’s chip.
Barcodes (gs1 standardised for drugs and interleaved 2 of 5 for blood samples) based on digital orders (electronic patient record, Phoenix, and CompuGroupMedical) were printed, then adhered to specimen tubes or medicine bags during preparation. Specimen tubes were directly labelled, based on the order, while medication orders were transmitted via HL7 message to an automated medication distribution system (AMDS) (Pyxis Medstation 3500, Carefusion/Becton Dickinson). Using this ward-based computerised distribution system, single drugs or bags containing multiple drugs to be administered at a specified time (e.g., at the beginning of a shift) were barcoded by the nurses. Distribution and labelling were strictly based on a computerised order.
Blood products were identified via the imprinted International Society of Blood Transfusions (ISBT) barcode registered in the blood bank. None were relabelled. The barcoded identification information was transferred via HL7 interfaces from the blood bank to the clinical information system, and thus to the identification information system.
A handheld scanner (Socket PDA) with a custom Windows Mobile application was used in order to check the different identifiers and their associated resources at the patient’s bedside. Nurses and physicians logged into the handheld application by scanning their personal staff badge. Patients and resources were scanned using the device’s combined RFID/barcode scanner. The software verified correspondences between the patient and resource (e.g., medication, blood sample or transfusion product), producing three different result types: ‘OK’ for a match between patient and resource within the due date; ‘Out of time’ for a match between patient and resource – though too early or too late with regard to the due date; ‘No Match’ for a mismatch between patient and resource. The results were colour-coded green (OK), yellow (Out of time), and red (No Match). With every match request, the intended action and matching result were separately stored in the database, in real time and independently of any further action by nurses.
In the event of a match confirmed by the nurses, additional actions within the clinical information system were automatically triggered (e.g., status changes in orders; transmission of HL7 messages to the laboratory information system) in order to improve workflow integration. Since the automated medication distribution system is not a unit-dose environment, a medication application check was not typically performed for every single pill/drug, but rather for a ‘drug portion’ in a bag labelled with a barcode. The electronically-controlled dispensing process prevented mixing up medications among patients and distributing medicines without orders. Hence, this simplification process during drug dispensing, as well as the bedside checks, did not decrease patient safety. It was the nurse’s decision to label either a single drug (e.g., one specific antibiotic) or a complete portion bag. For this reason, the analysed medication administration figures cannot be correlated with single drug orders over the same period, given that the number of drug orders far exceeded the number of bedside checks.
Study design and period
For this observational study conducted in the field, every bedside identity check and its result was statistically analysed over a 30-month period, from January 2013 until June 2015. Test subjects were excluded from the analysis. Within the departments under study, no other patient exclusions were carried out. As the analysis focused on misidentifications, only ‘No Match’ results were assessed. With respect to phlebotomy and transfusions, a valid correlation between identity checks and the corresponding order was possible in order to calculate usage. Regarding medication administration, the same correlation had to be estimated based on pooled medication prescriptions, given that there is no 1:1 correlation between planned medication administrations and barcodes (as no unit-dose system was in place, as mentioned above).
All calculations and statistical analyses were performed using SPSS Statistics 24 for Windows.
As shown in table 1, 24 879 in-patient stays were analysed during the 30-month period between January 2013 and June 2015. Distribution according to each department (Surgery and Internal Medicine) has been outlined. Overall, this (non-mandatory) identification system was used in 20.5% of phlebotomy and 32.7% of transfusion actions. The surgical department mainly employed the identification system for transfusions, and less frequently for phlebotomies. A similar distribution was revealed in the internal medicine department, where drug administration was additionally cross-checked. The inter-departmental differences concerning relative usage were not statistically significant. Monthly usage rates indicated relative stability over time. Concerning medication administration, only figures for single-drug orders as the reference parameter have been indicated, though they cannot be directly correlated, given the aforementioned reasons. As outlined in the Methodology section, no unit-dose system was in place, and consequently no valid statement concerning usage rates for medication could be made. This was because the nurses, with exception to cases of single-drug orders, decided independently whether to label a single drug or pooled drugs to be administered at the same time.
A total of 457 misidentifications were prevented by the identification system (table 2). The misidentification rates were very similar between departments, varying from 0.8 (medication administration) to 1.9% (phlebotomy). As medication administration proved more frequent than phlebotomy, and more common by far than transfusion, the absolute numbers of misidentification were highest for medication administration, yet with the lowest percentage of misidentifications. Of note is that medication administration was cross-checked only within the internal medicine department.
|Table 1: Patients, orders, and system usage. The table shows the numbers of analysed patients, orders, and wristband-based actions over the 30-month period. Personal digital assistants (PDA) were used during the computerized bedside identity check. As the absolute usage strongly depended on the number of patients in a given month, relative numbers concerning the median monthly usage are indicated.|
|Surgery & Orthopaedics||Internal Medicine||p-Value||Total|
|In-patient stays||15 381||9498||24 879|
|Average length of stay (days)||5.2 ± 3.2||6.7 ± 4.0||5.5 ± 4.6|
|Orders (%)||39 397 (100)||30 829 (100)||70 226 (100)|
|Orders/month||1271 ± 186||1063 ± 271||2334 ± 252|
|PDA-controlled (%)||9168 (23.9)||5181 (16.9)||0.01||14 349 (20.5)|
|Median PDA usage in %/month (IQR#)||23.9 (7)||16.9 (4)||0.85||20.5 (7)|
|Orders (%)||2762 (100)||1994 (100)||4756 (100)|
|Orders/month||89 ± 25||57 ± 27||71 ± 31|
|PDA-controlled (%)||951 (34.5)||606 (30.4)||0.76||1557 (32.7)|
|Median PDA usage in %/month (IQR#)||34.4 (12.9)||30.4 (11.8)||0.66||32.7 (11.8)|
|Orders (single drugs)||–||244 694||244 694|
|Estimated medication bags* (%)||–||224 617 (100)||224 617 (100)|
|PDA-controlled (%)||–||22 293 (9.9)||22 293 (9.9)|
|Median PDA usage in %/month (IQR#)||–||9.9 (3.9)||9.9 (3.9)|
|# IQR: Interquartile Range* There is no clear correlation between single-drug orders or single-drug administrations and printed barcodes (no unit-dose system in use) possible. The number of medication bags was estimated based on the usual pooling of drugs (e.g. morning shift)|
|Table 2: Observed and prevented misidentifications. Table 2 indicates the PDA (personal digital assistant)-controlled actions and the observed and prevented numbers of misidentifications in phlebotomy, transfusion, and medication administration, respectively. Medication administration was only performed in the internal medicine department.|
|Surgery & Orthopaedics||Internal Medicine||p-value||Total|
|PDA-controlled (%)||9168 (100)||5181 (100)||14 349 (100)|
|Misidentifications (%)||151 (1.7)||92 (1.9)||0.85||243 (1.8)|
|Monthly misidentification rate in % (95% CI#)||1.7 (1.4–2.0)||1.9 (1.5–2.3)||0.85||1.8 (1.6–2.0)|
|PDA-controlled (%)||951 (100)||606 (100)||1557 (100)|
|Misidentifications (%)||17 (1.8)||9 (1.5)||0.71||26 (1.7)|
|Monthly misidentification rate in % (95% CI#)||1.8 (0.8–2.2)||1.5 (0.3–2.3)||0.76||1.7 (0.8–2.1)|
|PDA-controlled (%)||--||22 293 (100)||22 293 (100)|
|Misidentifications (%)||--||188 (0.8)||188 (0.8)|
|Monthly misidentification rate in % (95% CI#)||--||0.9 (0.7–1.0)||0.9 (0.7–1.0)|
|PDA-controlled||10 119||28 080||38 199|
|Misidentifications (%)||168 (1.7)||289 (1.0)||457 (1.2)|
|Estimated medication bags* (%)||–||224 617 (100)||224 617 (100)|
|PDA-controlled (%)||–||22 293 (9.9)||22 293 (9.9)|
|Median PDA usage in %/month (IQR#)||–||9.9 (3.9)||9.9 (3.9)|
|# Confidence Interval|
This observational study investigated misidentification error rates and their eventual prevention by use of a computerised, multipurpose identification system fully integrated into nursing workflows. Our results concerning medication administration, blood sampling, and blood transfusions revealed relatively small percentages of misidentifications within a total of 38 199 bedside checks. Misidentifications occurred in 0.8% (medication administration) to 1.9% (phlebotomy) of patient-related actions, and could be prevented. Given the high workload and fragmented tasks inherent to in-patient care, this percentage appears quite low. However, the absolute numbers of preventable misidentifications proved rather concerning. With the observed system usage, 457 misidentifications were prevented over the study period. Assuming a 100% system usage for all three action types in the two departments studied, approximately 3395 misidentifications would occur annually, based on our observations, with 1.2% misidentification rates overall (table 2). Being able to detect these events thus strikes us as extremely worthwhile. On the other hand, performing identification checks was shown to induce a considerable increase in workload. This issue may be the reason behind the low system usage, especially with respect to medication administration, in a non-mandatory setting. Although medication administration is often pooled within the day (e.g., once per round), these actions prove very common over all shifts, though sometimes limited to one single drug. Furthermore, in our setting, a high safety level was ensured, and perceived by nurses, because all steps preceding drug administration had already been electronically monitored and checked by CPOE and AMDS. Additional tasks such as identification checks were deemed time-consuming, and thus tended to be skipped, although healthcare professionals do often agree on the relevance of positive patient identification.
The usage figures collected generally varied between 20.5% (phlebotomy) and 32.7% (transfusions). Similar percentages were reported for traditional paper-based settings using wristbands to manually identify patients. Franklin et al.  reported that patient identity was not checked in 82.6% of cases, a percentage very similar to that of our non-mandatory setting. Monthly usage rates in our non-mandatory environment proved very stable over the study period, especially when taking into account typical patient fluctuations, thus emphasising the validity of the observed results. Clear differences in the system’s usage were noticeable for specific tasks. As expected, identification checks were mostly performed for blood transfusion, but less frequently in the phlebotomy setting (p-value: 0.11, not outlined in table 2), though clear usage frequencies could not be calculated for medication administration. Given that nurses could apply barcodes on either pooled drugs (e.g., one barcode on a plastic bag containing the entire morning’s medication) or single drugs (e.g., one dose of a given antibiotic), the total identification number to check concerning medication could only be estimated, and has been outlined as such in the Results section. The different system usages could reflect the differing risk potential as assessed by nurses with respect to specific tasks, and thus the presumed impact of a misidentification. It seems reasonable to consider that transfusion misidentifications pose the highest risk. Interestingly, phlebotomy does not appear to be judged as important as transfusion, despite phlebotomy activities forming the basis for blood product compatibility tests.
As we can see from the research and numerous study results dating back over 20 years, medical errors remain an important issue. Be it medication errors, wrong-site surgery or transfusion errors, patient misidentification is among the most dreaded mistakes to occur, having a potentially huge impact on patient safety and outcome, in addition to the grave emotional and financial consequences. Transforming traditional and mostly paper-based workflows into computerised settings has been promoted for years, and carries both advantages and disadvantages. Electronic patient records, including CPOE, and automated medication dispensing systems are integrated into so-called closed-loop systems by adding identification systems, mostly based on patient wristbands. Consequently, every step pertaining to the processing of orders and actions is controlled and digitally documented. It seems reasonable to assume that such a system would help efforts to do the right thing at the right time for the right patient. While closing the gap in bedside identification issues, wristband-based identification checks should be able to prevent errors pertaining to this last step, and their usage has thus been promoted for several years now. Alternatives are scarce, and the efficacy of human-performed verification like double-checks appears limited, estimated at roughly 80% .
Few hospitals comprehensively use wristband technology for any action, and there is very little published data on success factors concerning their implementation and usability. Moreover, most of these studies were performed in quite different settings from Swiss hospitals. Additionally, wristband usage without computerised systems has proven prone to error. Many related errors seem to be primarily due to human factors . In some cases, accurate and consistent wristband placement, the cornerstone of patient identification, was shown to be lacking , with the wristband either not accessible, such as in the perioperative setting , or new, causing safety hazards due to either wristband usage  or missing standardisation . A significant percentage (15.3%) of errors were shown to be technology-related . While the time-consuming nature of the identification process obviously hinders aims of widespread usage, its purpose has nevertheless garnered wide support. The low usage could be further aggravated by the fact that each nurse individually perceives only a very low incidence of misidentification. With all these opportunities and obstacles in mind, we should emphasise the system’s high-value implementation process, seamless integration into any practitioner’s workflow, speed in terms of data transfer, and ease of use when combined with a polyvalent architecture. The latter appears especially relevant, given that we should perform identity checks not only for drugs and blood samples, but also for avoiding wrong-site surgery, as well as transmission of nosocomial infections. In terms of safety, we should therefore search for a single solution for all necessary identity checks.
This study’s strengths lie in its long observation period, coupled with the high numbers of identification checks included, and the data collection originating from a Swiss hospital, unique to the literature and including over 38 000 identification checks for analysis. The results were uniform across the analysed period, suggesting high data validity. Owing to the polyvalent architecture, misidentifications could be detected in not only one but three different and relevant areas: medication administration, blood sampling, and transfusion. The study setting did not depend on direct observation techniques or reporting systems, but rather reflected the actual usage and identity check results in a real-world setting. In addition, it is of note that there was no experimental study setting, the nurses being unaware of the data collection. Moreover, usage of the system could be measured as a distinct correlation among all planned procedures, especially with respect to blood sampling and transfusion.
The study had several limitations. First, and most obviously, making the systems’ use optional could have influenced the results. As the nurses had the choice whether to use the system or not, they could have chosen to use it only when they had low workload, for example, or only in situations where they felt particularly concerned or unsure. The results could thus have under- or overestimated the real number of misidentifications. However, the misidentification rate proved quite constant, and the variability among different wards and clinics rather low. Furthermore, the monthly usage rates were stable, pointing towards a somewhat routine use. Secondly, the system was not used in an ambulatory setting or emergency medicine. These areas have been previously shown prone to misidentifications due to the high workload and pace, and therefore their respective figures may differ. Thirdly, there were no similarities in system usage according to task. The system was far less used in medication administration, for example. With only approximately 10% of medication administration cross-checked, we may have significantly underestimated misidentification occurrence in this particular setting. Additionally, with respect to orally-administered drugs, single pills were not checked but rather bags containing, for example, the entire morning’s orders of oral drugs for a given patient, though automated dispensing was used.
Based on the primarily positive results from several studies, Khammarnia et al.  have recommended the widespread use of wristband systems, as they are believed safer than the previously used manual workflow. With these recommendations in mind and their implications for nursing workflow, it seems wise to implement comprehensive systems to be used for any identification process in a given institution. In this way, mistransfusions, WBIT errors, wrong-sitesurgery or medication administration errors could be prevented with a single, easy-to-use, and fully-integrated identification system. Though only a few hospitals in Switzerland currently work with a closed-loop system, bedside identification tests are deemed, at least to some extent, to be an isolated action, and an automated medication distribution system, for example, is thus not a mandatory prerequisite to digitally control bedside actions. With these circumstances in mind, we can assume that our results could be easily extrapolated to other Swiss hospitals with similar settings and thus help design a multipurpose and easy-to-use identification system for other settings.
Increasing the system’s net benefit – especially for nurses – by developing computerised systems perfectly adapted to the environment might likewise increase utilisation, and thus patient safety.
Preventing medical errors is crucial. Only within a closed-loop setting, as in the described setting, can the whole process, from the medication order to the patient’s bedside, be computerised, and errors thus prevented. Misidentification per se is a rare event, occurring in only 1 to 2% of cases. As the perceived misidentification incidence appears low to each individual healthcare worker, the system’s design and its perfect fit into workflows are of particular relevance, encouraging caregivers to use it often and over the long term. In conclusion, although widespread use of identification systems, such as the wristband, can effectively prevent errors in Swiss hospitals, they must be better designed for more than just one purpose, and be implemented with great care.
1 Allan EL, Barker KN. Fundamentals of medication error research. Am J Hosp Pharm. 1990 Mar;47(3):555–71. Review.
2 Leape LL, Brennan TA, Laird N, Lawthers AG, Localio AR, Barnes BA, Hebert L. The nature of adverse events in hospitalized patients. Results of the Harvard Medical Practice Study II. N Engl J Med. 1991;324:377–84.
3 Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG. Incidence of adverse events and negligence in hospitalized patients. Results of the Harvard Medical Practice Study I. N Engl J Med. 1991;324;370–76.
4 Bates DW, Cullen DJ, Laird N, Petersen LA, Small SD, Servi D, Laffel G, Sweitzer BJ, Shea BF, Hallisey R. Incidence of adverse drug events and potential adverse drug events. Implications for prevention. ADE Prevention Study Group. JAMA. 1995 Jul 5;274(1):29–34.
5 Aspden P, Wolcott J, Bootman J, editors. Institute of Medicine. Preventing medication errors: quality chasm series. Washington DC: National Academy Press; 2006.
6 Kohn LT, Corrigan JM, Donaldson MS. To err is human: building a safer health system. Washington DC: National Academic Press; 2000.
7 Bates DW, Cohen M, Leape LL, Overhage M, Shabot MM, Sheridan T. Reducing the frequency of errors in medicine using information technology. J Am Med Inform Assoc. 2001;8(4):398–9.
8 Bates DW, Teich JM, Lee J, Seger D, Kuperman GJ, Ma’Luf N, Boyle D, Leape LL. The impact of computerized physician order entry on medication error prevention. JAMIA. 1999;6:313–21.
9 Mekhjian HS, Kumar RR, Kuehn L, Bentley TD, Teater B, Thomas A, Payne B, Ahmad A. Immediate benefits realized following implementation of physician order entry at an academic medical center. J Am Med Inform Assoc. 2002;9(5):529–9.
10 Ammenwerth E, Schnell-Inderst P, Machan C, Siebert U. The effect of electronic prescribing on medication errors and adverse drug events: a systematic review. J Am Med Inform Assoc. 2008;15(5):585–600
11 Keers RN, Williams SD, Cooke J, Ashcroft DM. Causes of medication administration errors in hospitals: a systematic review of quantitative and qualitative evidence. Drug Saf. 2013;36(11):1045–67.
12 Borel JM, Rascati KL. Effect of an automated, nursing unit-based drug-dispensing device on medication errors. Am J Health Syst Pharm. 1995;52(17):1875–9.
13 DeYoung JL, Vanderkooi ME, Barletta JF. Effect of bar-code-assisted medication administration on medication error rates in an adult medical intensive care unit. Am J Health Syst Pharm. 2009;66(12):1110–5.
14 Helmons PJ, Wargel LN, Daniels CE. Effect of bar-code-assisted medication administration on medication administration errors and accuracy in multiple patient care areas. Am J Health Syst Pharm. 2009;66(13):1202–10.
15 Young J, Slebodnik M, Sands L. Bar code technology and medication administration error. J Patient Saf. 2010;6(2):115–20.
16 Wulff K, Cummings GG, Marck P, Yurtseven O. Medication administration technologies and patient safety: a mixed-method systematic review. J Adv Nurs. 2011;67(10):2080–95.
17 Poon EG, Keohane CA, Yoon CS, Ditmore M, Bane A, Levtzion-Korach O, Moniz T, Rothschild JM, Kachalia AB, Hayes J, Churchill WW, Lipsitz S, Whittemore AD, Bates DW, Gandhi TK. Effect of bar-code technology on the safety of medication administration. N Engl J Med. 2010;362(18):1698–707.
18 Krähenbühl-Melcher A, Schlienger R, Lampert M, Haschke M, Drewe J, Krähenbühl S. Drug-related problems in hospitals: a review of the recent literature. Drug Saf. 2007;30(5):379–407.
19 Franklin BD, O’Grady K, Donyai P, Jacklin A, Barber N. The impact of a closed-loop electronic prescribing and administration system on prescribing errors, administration errors and staff time: a before-and-after study. Qual Saf Health Care. 2007;16(4):279–84.
20 Morrison AP, Tanasijevic MJ, Goonan EM, Lobo MM, Bates MM, Lipsitz SR, Bates DW, Melanson SE. Reduction in specimen labeling errors after implementation of a positive patient identification system in phlebotomy. Am J Clin Pathol. 2010;133(6):870–7.
21 Ansari S, Szallasi A. ‘Wrong blood in tube’: solutions for a persistent problem. Vox Sang. 2011;100(3):298–302.
22 Francis DL, Prabhakar S, Sanderson SO. A quality initiative to decrease pathology specimen-labeling errors using radiofrequency identification in a high-volume endoscopy center. Am J Gastroenterol. 2009;104(4):972–5.
23 Ohsaka A, Abe K, Ohsawa T, Miyake N, Sugita S, Tojima I. A computer-assisted transfusion management system and changed transfusion practices contribute to appropriate management of blood components. Transfusion. 2008;48(8):1730–8.
24 Phillips SC, Saysana M, Worley S, Hain PD. Reduction in pediatric identification band errors: a quality collaborative. Pediatrics. 2012;129(6):e1587–93.
25 Burrows JM, Callum JL, Belo S, Etchells E, Leeksma A. Variable pre-transfusion patient identification practices exist in the perioperative setting. Can J Anaesth. 2009;56(12):901–7.
26 Patterson ES, Cook RI, Render ML. Improving patient safety by identifying side effects from introducing bar-coding in medication administration. J Am Med Inform Assoc. 2002;9(5):540–53.
27 Koppel R, Metlay JP, Cohen A, Abaluck B, Localio AR, Kimmel SE, Strom BL. Role of computerized physician order entry systems in facilitating medication errors. JAMA. 2005;293(10):1197–203
28 Han YY, Carcillo JA, Venkataraman ST, Clark RS, Watson RS, Nguyen TC, Bayir H, Orr RA. Unexpected increased mortality after implementation of a commercially sold computerized physician order entry system. Pediatrics. 2005;116(6):1506–12.
29 McDonald CJ. Computerization can create safety hazards: a bar-coding near miss. Ann Intern Med. 2006;144(7):510–6.
30 Cochran GL, Jones KJ, Brockman J, Skinner A, Hicks RW. Errors prevented by and associated with bar-code medication administration systems. Joint Commission Journal on Quality and Patient Safety / Joint Commission Resources. 2007;33(5):293–301, 245.
31 Berg M, Aarts J, van der Lei J. ICT in health care: sociotechnical approaches. Methods Inf Med. 2003;42(4):297–301.
32 Raban MZ, Westbrook JI. Are interventions to reduce interruptions and errors during medication administration effective? A systematic review. BMJ Qual Saf. 2014;23(5):414–21.
33 Radley DC, Wasserman MR, Olsho LE, Shoemaker SJ, Spranca MD, Bradshaw B. Reduction in medication errors in hospitals due to adoption of computerized provider order entry systems. J Am Med Inform Assoc. 2013;20(3):470–6.
34 Adelman JS, Kalkut GE, Schechter CB, Weiss JM, Berger MA, Reissman SH, Cohen HW, Lorenzen SJ, Burack DA, Southern WN. Understanding and preventing wrong-patient electronic orders: a randomized controlled trial. J Am Med Inform Assoc. 2013;20(2):305–10.
35 FitzHenry F, Peterson JF, Arrieta M, Waitman LR, Schildcrout JS, Miller RA. Medication administration discrepancies persist despite electronic ordering. J Am Med Inform Assoc. 2007;14:756–64.
36 Nuttall GA, Abenstein JP, Stubbs JR, Santrach P, Ereth MH, Johnson PM, Douglas E, Oliver WC Jr. Computerized bar code-based blood identification systems and near-miss transfusion episodes and transfusion errors. Mayo Clin Proc. 2013;88(4):354–9.
37 Varey A, Tinegate H, Robertson J, Watson D, Iqbal A. Factors predisposing to wrong blood in tube incidents: a year’s experience in the North East of England. Transfus Med. 2013;23(5):321–5.
38 Gonzalez-Gonzalez C, Lopez-Gonzalez E, Herdeiro MT, Figueiras A. Strategies to improve adverse drug reaction reporting: a critical and systematic review. Drug Saf. 2013;36:317–28.
39 Pirmohamed M, Breckenridge AM, Kitteringham NR, Park BK. Adverse drug reactions. BMJ. 1998;316(7140):1295–8.
40 Fujihara H, Yamada C, Furumaki H, Nagai S, Shibata H, Ishizuka K, Watanabe H, Kaneko M, Adachi M, Takeshita A. Evaluation of the in-hospital hemovigilance by introduction of the information technology-based system. Transfusion. 2015;55(12):2898–904.
41 Facchinetti NJ, Campbell GM, Jones DP. Evaluating dispensing error detection rates in a hospital pharmacy. Med Care. 1999;37(1):39–43.
42 Ohsaka A, Abe K, Ohsawa T, Miyake N, Sugita S, Tojima I. A computer-assisted transfusion management system and changed transfusion practices contribute to appropriate management of blood components. Transfusion. 2008;48(8):1730–8.
43 Mayor S. Hospitals must standardise patients’ wristbands to reduce risk of wrong care. BMJ. 2007;335(7611):118.
44 Khammarnia M, Kassani A, Eslahi M. The efficacy of patients’ wristband bar-code on prevention of medical errors: a meta-analysis study. Appl Clin Inform. 2015;6(4):716–27.
Spital Thun, Spital STS AG, Thun, Switzerland
The author would like to thank Ueli Dummermuth, Mathias Fahrni, and Andrea Baumgartner for their long-lasting and ongoing technical support.
No potential conflict of interest relevant to this article was reposted.
Dr. med. Marc Oertle
Spital STS AG
Published under the copyright license
“Attribution – Non-Commercial – NoDerivatives 4.0”.
No commercial reuse without permission.