chapter  14
14 Pages

Administration of Intravenous Medication: Process Variation across Hospital Wards ����������������������������������������������������������������������

Intravenous drug administration is a more risk-prone clinical process than almost any other procedure taking place in hospital settings (Kaushal et al� 2001, Gonzales 2010)� There are numerous strategies that can contribute to reducing the risk for medication errors and patient harm in the hospital drug administration processes (AAMI 2010, EU 2015)� These strategies include good clinical practices, adequate training and optimal use of technologies� While literature reviews suggest that computerised infusion devices, such as smart pumps (also called smart infusion pumps or intelligent infusion devices), can contribute to the risk reduction of medication errors in healthcare settings, data supporting this manner of risk reduction in the field remain limited (Black et al� 2011, Ohashi et al� 2014)�

Computerised patient infusion devices, infusion pumps, include features for preventing medication administration errors� According to earlier studies, over 90% of intravenous medications involve some type of an error (Husch et  al� 2005)� In Finland, a study by Ruuhilehto and her associates (2011) showed that 51% of 64,405

Background ������������������������������������������������������������������������������������������������������������ 211 Smart Pumps: Use and Challenges ������������������������������������������������������������������������ 212 Reporting Medication Errors’ Adverse Events: Reporting Technology ���������������� 213 Methods ������������������������������������������������������������������������������������������������������������������ 214

Setting ��������������������������������������������������������������������������������������������������������������� 214 Design ��������������������������������������������������������������������������������������������������������������� 214 Data Collection and Analysis ���������������������������������������������������������������������������� 215

Results �������������������������������������������������������������������������������������������������������������������� 217 Discussion ��������������������������������������������������������������������������������������������������������������220 Conclusion ������������������������������������������������������������������������������������������������������������� 221 References �������������������������������������������������������������������������������������������������������������� 221

web-based incident reports were concerned with medication, and the most common incidents were errors in documenting, dispensing and administering medications� Valentin and colleagues (2009) found out that the administration of parenteral medication was a vulnerable area in patients’ safety in intensive care� Their results were based on data from 113 participating units from all over the world (27 countries) and illustrated that this problem could not be attributed to suboptimal care in a few individual units but, instead, represented a common pattern� The most frequent errors were related to wrong time of administration and missed medication, followed by wrong dose, wrong drug and wrong route of administration� The authors stated that most medication errors occurred in routine care situation, not in extraordinary situations (Valentin et al� 2009)�

Medication safety in hospitals is dependent on the successful execution of a complex system of scores of individual tasks, namely prescribing, preparing, dispensing and transcribing a medication and monitoring the patient’s response� Many of these tasks lend themselves to technological devices� Patient safety is a matter of major concern� Emerging technologies, such as smart pumps, can diminish medication errors (Ohashi et al� 2013) as well as standardise and improve clinical practice, resulting in the subsequent benefits for patients (Rothschild et al� 2005, ManriqueRodríguez et al� 2014, Mason et al� 2014)�

Härkänen (2014) demonstrated that adverse outcomes due to medication-related factors were common� The use of combination of methods revealed information that was more diverse than what was previously known regarding medication-related problems in hospital setting and that can be used to increase safety in the medication process (Härkänen 2014)� In incident reports and observational data, the administration errors were the most common, followed by documenting errors, while in Global Trigger Tool data, the prescribing errors were the most frequently apparent (Härkänen 2014)� Technology may provide the key for reducing medical errors� Technological applications such as electronic health records (EHR), barcoding technology for medication administration and smart infusion pumps are widely held as providing solutions for patient safety (Bates and Gawande 2003, Wulff et al� 2011, Ohashi et al� 2014)�

This chapter presents the findings of a study conducted in a Finnish tertiary hospital in 2014� The findings are a part of a developing process based on an international collaboration with David W� Bates and Brigham and Women’s Hospital, Boston� The design section includes a description of the development project, and the results section will report on the preliminary results of a more extensive project (before-after study)� The research questions focus on the following issues: what are the frequencies and types of intravenous medication errors with infusion pumps, what are the frequencies and types of intravenous medication errors within a medication process and how much variability is there by frequency and type in the settings�

The term ‘smart pump’ was originally coined by the Institute for Safe Medication Practices in the United States to describe an infusion pump with an inbuilt drug library that contains correct parameters for all the medications to be delivered by the

pump (ISMP 2010, Quinn 2011)� Smart pump technology has progressed over the past 5 years (Ohashi et al� 2014), and the built-in safety features of the smart pump technology provide an additional double-check system for medication administration (Carayon 2010, Mason et al� 2014)� The proper use of the smart pump technology yields benefits that include enhanced workflow for nurses and error reduction in medication administration (Carayon et al� 2010)�

Technology provides error reduction capabilities in medication via programmed dose limit alerts with audio-visual feedback to staff regarding erroneous orders, improper dose calculations or programming errors� Ohashi and colleagues noted in their literature review lower compliance rates of using smart pumps, overriding soft alerts, non-intercepted errors and the possibility of using the wrong library� Rothschild et  al� (2005) evaluated a very early version of the pumps and found that smart intravenous pumps with decision support capabilities had the capacity to intercept many dangerous medication errors� Manrique-Rodrigues et  al� (2014) demonstrated that the implementation of smart pumps proved effective in preventing infusion-related programming errors from reaching patients� A study in a German hospital showed that smart pumps prevented potentially dangerous overdoses� In a before-after study on patient-controlled analgesia, the use of a smart pump resulted in a significant 22% decrease in adverse drug events recorded by an automated surveillance system, and voluntary report events also decreased significantly by 72% (Kastrup et al� 2012)�

The introduction of smart pump technology provides a technology-based final check of an infusion rate that creates a safe environment for care� Modern infusion devices are much more than mere pumps and are currently often regarded as part of the organisation’s information technology capability� However, only a small reduction in dosing errors has been found, while a greater reduction occurred in pumprelated errors� Nine out the 10 post-intervention pump programming errors occurred because users did not use the pump software correctly (Adachi 2005, Ohashi et al� 2014)� Previous studies have shown poor caregivers compliance with the drug library and users frequently ignoring the drug library when selecting a drug (Ohashi et al� 2014)� Creation of a safe and effective customised drug library is essential for the proper utilisation of smart pumps� A drug library should include at least all highalert drugs with standard concentrations as well as soft and hard stops to various dosage limits� Drug libraries must also be maintained and updated constantly� Wireless communication technology in an organisation’s infrastructure allows easier adjustment or updating of drug libraries, which otherwise would require manually updating each pump separately (ISMP 2010)�

The World Health Organization has presented Draft Guidelines for Adverse Event Reporting and Learning System, and a European Council Recommendation (2009/C 151/01) has also advised European countries on the establishment and revision of reporting systems (EU 2009)� The European Council Recommendation mentions that there are big differences between reporting systems in the EU

Member States� There are both mandatory and voluntary reporting systems in the Member States, and variety in the types of incidents that can be reported varies� However, a broad definition allows the reporting of any concerns, including near misses and ‘no harm’ incidents providing a rich resource for learning and systems improvement (EU 2009)�

The HaiPro system is intended for reporting patient safety incidents on the organisational level in Finland� The main properties of HaiPro are anonymity, confidentiality and freedom from sanctions� The HaiPro approach incorporates a system model that takes into consideration the features of natural human behaviour and the pathway of diverse events development� The local incident reporting system is meant to prevent adverse events of treatment through the improvement of operational procedures� At present, data are collected only at the level of the organisation and not sent forward to or nor aggregated or analysed on regional and national level (Doupi 2009)� In this study, reports from the HaiPro system illustrated one perspective of the administration of intravenous medication in the hospital, and the wards were studied�

The study was conducted in a Finnish tertiary hospital, which has 800 beds and provides specialised medical care to 860,000 inhabitants� Around 90,000 in patients are treated in the hospital annually (PSSHP 2015)� Computerised physician order entry has been in use in the hospital since 2009, and a barcoding system is used in its pharmacy� Smart pumps were in use in the intensive care unit and cardiac care unit, and a drug library was customised for each unit� The studied wards included a coronary care unit, an adult intensive care unit, an oncology unit and a cardiothoracic surgical ward� The pilot data were collected at a maternity ward, where the appropriateness of the questions was also tested for Finnish hospital�

This is a descriptive retrospective pre-post study conducted in three phases over the course of 36 months in total (2014-2016, Figure 14�1) based on a similar study implemented in the United States (Bates 2012)� During the first year, the quality and quantity of reported adverse events were screened in order to acquire baseline data concerning safety incidents (including HaiPro) in the wards� This was followed by an observational study where a multidisciplinary team of investigators prospectively compared the actual medication, its dose and the infusion rate of the infusion pump with the prescription in the medical record� Subsequently, during the second year (2015), these results were analysed� A consensus process including a face-to-face meeting with users and decision-makers took place in order to evaluate the types of events and to develop an intervention� After a run-in period, the effectiveness of intervention will be tested during the third year of the study (2016) to produce a set of evidence-based recommendations�

This study was approved by the University of Eastern Finland Committee on Research Ethics in 2014� Participants in the study were voluntary and aimed at patients aged 18 years or older�

The observation data were collected by four researchers� The observation data collection is based on methods employed by Husch and colleagues (2005)� Key elements, required to capture all kinds of intravenous medication errors, were identified using methods by Ohashi and colleagues (2013)� In assigned inpatient units, using a ‘prospective point prevalence’ approach, all data from intravenous medications at patient bedside were collected with the standardised form� Observers compared the infusing medication dose and infusion rate on the pump with the prescribed medication, the dose and rate in the medical record� All orders were obtained from both EHR and handwritten paper-based medical records, and all intravenous fluids were classified as medication� The availability of correct patient identification band and name verification was recorded for each patient� Labelling of the infusing medication, according to medication policy, was assessed�

Observations on the intravenous medication administration processes were made over 4 days, 8 hours per day in March 2014 in each of the four units� All epidural, patient-controlled analgesia and general use of intravenous infusion pumps on inpatient care units were included in the investigation� To capture data, two to four observers (registered nurses and pharmacists) went to a ward and conducted observation on patients� The observer entered a patient room; he or she introduced himself or herself and explained the purpose of the study� In order to confirm the presence of an error (Table 14�1), the observers had to agree that an error had been made� If such an error was identified, that had the potential to cause harm, the staff nurse caring for this particular patient was discreetly informed about the incident so that he/she could correct it�

The main study outcomes will be the intravenous medication error rate and the serious medication error rate� All errors were classified for every administered medication� Serious errors were rated by NCC MERP index (National Coordinating Council for Medication Error Reporting and Prevention) (Table 14�2)�

FIGURE 14.1 The study phases based on Bates’ study protocol, 2012�

TABLE 14.2 The Severity of Medication Errors according to NCC MERP Harm Index

TABLE 14.1 Medication Error Types in the Study

All factors collected in each participating ward included the report of patient safety incidents at the organisational level (HaiPro), the institutional policies and procedures around medication administrations� The data were analysed and then presented at each ward� The meetings were hosted by the multidisciplinary team� Written memorandums from each meeting were analysed by content analysis, and these data will be utilised in the recommendations�

The preliminary data were used as material in a meeting with physicians, nurses, pharmacists and a patient safety manager� In the meeting, data of the use of intravenous infusion pumps and medication administration workflow were reviewed� Based on the data and the summary of the meeting, the multidisciplinary team developed an intervention that may improve the safety of intravenous drug administration at the wards�

The HaiPro reports from between 2011 and 2013 indicated that adverse events had increased in each unit� In particular, the incidence of near misses has increased, which illustrates that the preventative meaning of reporting has been emphasised� Most of the incidents were related to medication or information management� The majority of the reported incidents consisted of detected errors; recently, reports on near misses had increased (Figures 14�2 and 14�3)�

During the data collection period, 194 inpatients in four inpatient wards were included in the study, and 492 medication procedures were observed (Table 14�3)� These consisted of 355 intravenous medication infusions and 137 fluids infusions� Out of these, infusion pumps were used in 56% of cases�

FIGURE 14.2 Reported incidents of detected-and near-miss medication and intravenous medication-related errors (2011-2013) in the hospital�

The use of an identification band is imperative to ensure that the right patient receives the right care� Information on the identification band has to correspond with details recorded on the patient’s medicine chart, and this should be confirmed whenever medications are administered� Remarkable variety between wards was found in the first phase of this study (Table 14�4)� Misidentification of patients and administering medicines intended for another patient to these patients are real dangers�

Stating patients’ allergies is key hospital policy� One aim of the EHR is to improve recording patients’ allergies and raise awareness of high-risk products among staff�

TABLE 14.3 Study Participants and Type of Medication

FIGURE 14.3 Reported incidents of detected-and near-miss medication and intravenous medication-related errors, 2011-2013, in the studied wards (excluding the pilot ward)�

The lack of information on allergies was a surprising finding in the study (Table 14�4)� Violations of medication policies regarding labelling were the most frequent error types� The details of the information missing on the labels varied between wards (Table 14�5)� The most common error was the lack of information on the label of the drug to be infused� There were also the differences in the way in which medications were prescribed, and medication orders were documented� Regardless of the fact that in some wards, all prescriptions were available in the EHR, the wards used several overlapping documentation forms and a paper-based form to manage medication information�

Each error in observation data was rated using the NCC MERP harm index� Data collection in the four study units further indicated that errors were rated as A-B, indicating that there was potential for an error (A) or an error occurred but did not reach the patient (B)� Indeed, in spite of the EHR system, several overlapping documentation forms were in use, and medication processes varied significantly across the units�

In summary, the work processes at each studied ward differed, even though the organisation’s policy was well known� There was a lot of variety in the use of identification bands between wards, and allergies were infrequently documented� Missing allergy information was recognised as a risk in the EHR� According to the organisation policy, medication and fluid prescriptions should be documented in the EHR system� A lot of room for development was found in connection to the labels of intravenous medication, because their contents varied between wards� The use of medical technology, such as infusion pumps, smart pumps and drug library of devices, should be increased� The staff was highly interested in improving their competences in using technology�

TABLE 14.4 Differences between Wards in the Use of Identification Bands and Documentation of Allergies

TABLE 14.5 Variety between Wards in Labelling Status

The results of the first phase of this intervention study show the significance for the hospital to improve patient safety� There was a positive development in the voluntary reporting of patient incidents, and especially reporting near-miss incidents had increased every year� These illustrate that the preventative importance of reporting had been emphasised� Members of staff were very interested in investing in their competence in using technology and were aware of the possibilities provided by technology�

In this study, the most common error was the lack of information on medication labels thus increasing the risk of administering wrong intravenous medication to patients� Previous studies have also demonstrated similar tendencies in medication errors (Keohane et  al� 2005, Ohashi et  al� 2013)� Documentation of allergy was the information found missing most frequently� The allergies of only 56% of the patients had been recorded in the EHR system� Differences in identifying patients in the studied wards were a further significant finding� Through the course of this study, such violation errors of hospital policy were found, which could potentially put patients at risk�

Traditional infusion pumps were in common use� Smart pumps were used mainly in cardiac care and intensive care units in this study� There was clearly a need to develop the use of technology in patient’s care� However, it has already been demonstrated that until barcode pumps are integrated with other systems within the medication administration process, the role of smarts pumps in enhancing patient safety will be limited (Trbovich et al� 2010, Ohashi et al� 2014)� Regardless of the technology, careful planning of care processes will ensure proper patient identification prior to any medical intervention and provide safer care with significantly fewer errors (Upton and Quinn 2013)� In the future, more interest must be focused on the work processes� The work processes at each studied ward differed even though the organisation policy was well known�

The main strength of an observation and medical chart review evaluation is that the observers directly perceived the use of medical devices in clinical environments� The advantages of observation include obtaining data on authentic surroundings (Elias and Moss 2011, Sinivuo et al� 2012)� The most remarkable preliminary result of this study was the high variability in the medication processes in the different studied units� Several types of errors were detected related to intravenous medication administration� These may increase the risk for administering wrong intravenous medications� The discussions in the first face-to-face meeting with the staff from the wards were productive� The staff was aware of the hospital’s instructions concerning intravenous medication, but each ward had and followed their own procedures� The tendency of nurses to ignore safety software in infusion devices where manufacturers have made it optional to use these is one of the obstacles for the research efforts attempting to demonstrate and quantify an increase in patient safety (Rothschild et  al� 2005)� The temptation to follow so-called workarounds must be resisted if potential benefits to patients and health professionals are to be realised (McAlearney et al� 2007)�

In the meetings with wards staff, the significance of the problem was recognised, and deeper commitment to hospital policy was stated� Overlapping documentation and ward-specific systems were highlighted, and there were lively discussions on

the recommendations� Collaboration in the hospital was emphasised� It is the prime responsibility of the prescriber to ensure that the patient’s information is checked for allergies before any prescriptions are written� Other staff responsible for dispensing, administrating and monitoring medicines can help prevent potential incidents by identifying the patient’s allergy status�

The results of discussion with the staff have condensed into four recommendations: (1) the identification band should be used according to the organisation policy; (2) allergy status should be documented for all patients, including the lack of an allergy (‘No known medicine allergy’); (3) medication and fluid orders should be prescribed on the EHR system according to the organisation policy and (4) the label of intravenous medication has to be filled as instructed�

There are limitations to our study� The small sample size of the studied wards limits the generalisability of the results� The study focused only on one tertiary hospital� However, as they are in use all over healthcare institutions, the safe use of infusion pumps is a shared challenge for healthcare organisations�

The first observations on intravenous medication errors identified the key issues related to the use of infusion pumps and smart pumps� This will support to the development of strategies to allow improving the prevention of intravenous medication errors� The utilisation of staff meetings will increase commitment to develop strategies for improving the prevention of intravenous medication errors� Support gained from hospital management and their commitment to improve patient safety was a facilitator for this study�

The most important preliminary result of this first observation study was the high variability in the medication processes in the studied wards� The hospital policy concerning intravenous medication has a central role in improving patient safety� The staff appeared to be well motivated to develop the medication processes and their own competence in using technology� Future challenges associated with the hospital policy especially to intravenous medication and increasing the use of medical technologies in patient care and in healthcare in general�

AAMI, 2010-last update, Infusing patients safely [Homepage of The Association for the Advancement of Medical Instrumentation (AAMI)], [Online]� Available at: http:// s3�amazonaws�com/rdcms-aami/files/production/public/FileDownloads/Summits/ AAMI_FDA_Summit_Report�pdf (accessed 20 May 2015)�

Adachi, W� and Lodolce, A�E�, 2005� Use of failure mode and effects analysis in improving the safety of i�v� drug administration� American Journal of Health-System Pharmacy, 62, 917-920�

Bates, D�W�, 2012� Research proposal� A national study of intravenous medication errors: Understanding how to improve intravenous safety with smart pumps� Brigham and Women’s Hospital, Boston, MA (unprinted)�

Bates, D�W� and Gawande, A�A�, 2003� Improving safety with information technology� The New England Journal of Medicine, 248, 2526-2534�

Black, A�D�, Car, J�, Pagliari, C�, Cresswell, K�, Bokum, T�, Mckinstry, B�, Procter, R�, Majeed, A� and Sheikh, A�, 2011� The impact of eHealth on the quality and safety of health care: A systematic overview� PLoS Medicine, 8, e1000387�

Carayon, P�, Hundt, A�S� and Wetterneck, T�B�, 2010� Nurses’ acceptance of Smart IV pump technology�International Journal of Medical Informatics, 79(6), 401-411�

Doupi, P�, 2009� National Reporting Systems for Patient Safety Incidents. A Review of the Situation in Europe� Jyväskylä, Finland: National Institute for Health and Welfare�

Elias, B�L� and Moss, J�A�, 2011� Smart pump technology� What we have learned� Computers, Informatics, Nursing, 29, 184-190�

EU, 2009, Council Recommendation of 9 June 2009 on patient safety, including the prevention and control of healthcare associated infections (2009/C 151/01) [Homepage of European Commission], [Online]� Available at: http://ec�europa�eu/health/patient_ safety/docs/council_2009_en�pdf (accessed 12 August 2014)�

EU, 2015-last update, Report on the public consultation on patient safety and quality of care [Homepage of European Commission, Patient Safety and Quality of Care working group], [Online]� Available at: http://ec�europa�eu/health/patient_safety/docs/pasq_ public_consultation_report�pdf (accessed 12 August 2015)�

Gonzales, K�, 2010� Medication administration errors and the pediatric population: A systematic search of the literature� Journal of Pediatric Nursing, 25, 555-565�

Härkänen, M�, 2014� Medication-related adverse outcomes and contributing factors among hospital patients an analysis using hospital’s incident reports, the global trigger tool method, and observations with record reviews, The Faculty of Health Sciences, University of Eastern Finland, Finland�

Husch, M�, Sullivan, C�, Rooney, D�, Barnard, C�, Fotis, M�, Clarke, J� and Noskin, G�, 2005� Insights from the sharp end of intravenous medication errors: Implications for infusion pump technology� Quality and Safety in Health Care, 14, 80-86�

ISMP, 2010� Multiple Intravenous Infusions Phase 1a: Situation Scan Summary Report. Toronto, Canada: ISMP�

Kastrup, M�, Balzer, F�, Volk, T� and Spies, C�, 2012� Analysis of event logs from syringe pumps a retrospective pilot study to assess possible effects of syringe pumps on safety in a University Hospital Critical Care Unit in Germany� Drug Safety, 35(7), 563-574�

Kaushal, R�, Bates, D�W�, Landrigan, C�, Mckenna, K�J�, Clapp, M�D�, Federico, F� and Goldmann, D�A�, 2001� Medication errors and adverse drug events in pediatric inpatients� The Journal of the American Medical Association, 285, 2114-2120�

Keohane, C�A�, Hayes, J�, Saniuk, C�, Rothschild, J�M� and Bates, D�W�, 2005� Intravenous medication safety and smart infusion systems� Lesson learned and future opportunities� Journal of Infusion Nursing, 28, 321-328�

Manrique-Rodríguez, S�, Sánchez-Galindo, A�C�, López-Herce, J�, Calleja-Hernández, M�Á, Martínez-Martínez, F�, Iglesias-Peinado, I�, Carrillo-Álvarez, Á, Sanjurjo-Sáez, M� and Fernández-Llamazares, C�M�, 2014� Implementing smart pump technology in a pediatric intensive care unit: A cost-effective approach� International Journal of Medical Informatics, 83(2), 99-105�

Mason, J�J�, Roberts-Turner, R� and Hinds, P�S�, 2014� Patient safety, error reduction, and pediatric nurses’ perceptions of smart pump technology� Journal of Pediatric Nursing, 29, 143-151�

McAlearney, A�S�, Vrontos, J�, Schneider, P�J�, Curran, C�R�, Czerwinski, B�S�, Pedersen, C�A�, 2007� Strategic work-arounds to accommodate new technology: The case of smart pumps in hospital care� Journal of Patient Safety, 3, 75-81�

NCC MERP, 2001� NCC MERP index [Homepage of National Coordinating Council for Medication Error Reporting and Prevention], [Online]� Available at: http://www� nccmerp�org (accessed 29 January 2014)�

Ohashi, K�, Dalleur, O�, Dykes, P�C� and Bates, D�W�, 2014� Benefits and risks of using smart pumps to reduce medication error rates: A systematic review� Drug Safety, 37, 1011-1020�

Ohashi, K�, Dykes, P�, Mcintosh, K�, Buckley, E�, Wien, M� and Bates D�W�, 2013� Evaluation of intravenous medication Errors with smart infusion pumps in an Academic Medical Center, AMIA, Ed� In: Proceedings of the AMIA 2013 Annual Symposium, 16-20 November 2013, Washington, DC: AMIA, pp� 1089-1098�

PSSHP, 2015 last update, Kuopio University Hospital [Homepage of University Hospital], [Online]� Available at: https://www�psshp�fi/web/en/ (accessed 12 August 2015)�

Quinn, C�, 2011� Smart practice: The introduction of a dose error reduction system� British Journal of Nursing, 20, S20-S25�

Rothschild, J�M�, Keohane, C�A�, Cook, F�E�, Orav, J�, Burdick, E�, Thompson, S�, Hayes, J� and Bates, D�W�, 2005� A controlled trial of smart infusion pumps to improve medication safety in critically ill patients� Critical Care Medicine, 33, 533-540�

Sinivuo, R�, Koivula, M� and Kylmä, J�, 2012� Observation as a data collection methods in clinical context (in Finnish)� Journal of Nursing Science, 24, 291-301�

Trbovich, P�L�, Pinkney, S�, Cafazzo, J�A�, Easty, A�C�, 2010� The impact of traditional and smart pump infusion technology on nurse medication administration performance in a simulated inpatient unit� Quality and Safety in Health Care, 19, 430-434�

Upton, D� and Quinn, C�, 2013� Smart pumps – Good for nurses as well as patients� British Journal of Nursing, 22, 4-8�

Valentin, A�, Capuzzo, M�, Guidet, B�, Moreno, R�, Metnitz, B�, Bauer, P� and Metnitz, P�, 2009� Errors in administration of parenteral drugs in intensive care units: Multinational prospective study� British Medical Journal, 338, b814�

Wulff, K�, Cummings, G�, Marck, P� and Yurtseven, O�, 2011� Medication administration technologies and patient safety: A mixed-method systematic review� Journal of Advanced Nursing, 67, 2080-2095�