effect of bar-code-assisted medication administration on medication administration errors and...

Upload: alex-pieces

Post on 28-Feb-2018

213 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/25/2019 Effect of Bar-code-Assisted Medication Administration on Medication Administration Errors and Accuracy in Multiple

    1/9

    PRACTICE REPORTS Bar-code-assisted medication administration

    1202 Am J Health-Syst PharmVol 66 Jul 1, 2009

    P R A C T I C E R E P O R T S

    Effect of bar-code-assisted medicationadministration on medication administration errors

    and accuracy in multiple patient care areasPIETERJ. HELMONS, LINDSAYN. WARGEL, ANDCHARLESE. DANIELS

    PIETER J. HELMONS, PHARM.D., is Pharmacoeconomics Special-ist, Department of Pharmacy, University of California San Diego(UCSD) Medical Center, San Diego; at the time of this study he was aPostdoctoral Fellow, UCSD Skaggs School of Pharmacy and Pharma-ceutical Sciences, La Jolla. LINDSAYN. WARGEL, PHARM.D.,is ClinicalPharmacist, UCSD Medical Center; at the time of this study shewas Postgraduate Year 1 Resident, Department of Pharmacy, UCSDMedical Center. CHARLESE. DANIELS, PH.D.,is Professor of ClinicalPharmacy and Associate Dean for Clinical Affairs, UCSD SkaggsSchool of Pharmacy and Pharmaceutical Sciences, and Pharmacist-In-Chief, Department of Pharmacy, UCSD Medical Center.

    Address correspondence to Dr. Helmons at the University of Cali-fornia San Diego Medical Center, 200 West Arbor Drive, Mail Code

    8765, San Diego, CA 92103-8765 ([email protected]).The assistance of Evelyn Hoevatanakul, Pharm.D., Kimberly Lam,

    Pharm.D., Sanaz Farhadian, Pharm.D., Sharon Siu, Pharm.D., VivianLuo, Pharm.D., and Warren Tong, Pharm.D., who performed a largenumber of the observations, is acknowledged. Also, this study wouldnot have been possible without the dedication and commitment ofthe nurse managers and nursing staff to improve medication safetyon the units participating in this study.

    The authors have declared no potential conflicts of interest.

    Copyright 2009, American Society of Health-System Pharma-cists, Inc. All rights reserved. 1079-2082/09/0701-1202$06.00.

    DOI 10.2146/ajhp080357

    Purpose. The effect of a commercially

    available bar-code-assisted medication

    administration (BCMA) technology on six

    indicators of medication administration

    accuracy and nine types of medication ad-

    ministration errors in distinct patient care

    areas were studied.

    Methods. This prospective, before-and-

    after, observational study was conducted in

    two medicalsurgical units, one medical in-

    tensive care unit (ICU), and one surgical ICU

    of a 386-bed academic teaching hospital.

    Nursing staff were observed administering

    medications one month before and threemonths after implementation of BCMA

    technology. Observations were conducted

    by two pharmacists and four pharmacy stu-

    dents on weekdays and weekends. Medica-

    tion administration accuracy was measured

    using the accuracy indicator of the Califor-

    nia Nursing Outcomes Coalition.

    Results. The majority of medication ad-

    ministrations occurred during the 9 a.m.

    medication round. After BCMA imple-

    mentation in the medicalsurgical units,

    improved adherence to patient identifi-

    cation policies was observed, but more

    distractions of the nursing staff occurred

    and the medications administered were

    less frequently explained to the patient.

    Although an increase in wrong-time errors

    was observed in the medicalsurgical units,

    the total number of medication errors did

    not change. When wrong-time errors were

    excluded, the rate of medication errors de-

    creased by 58%. In the ICUs, the charting of

    medication administration improved after

    BCMA implementation, but total medica-

    tion errors and wrong-time errors did not

    change.Conclusion. Implementing BCMA technol-

    ogy decreased medication administration

    errors in medicalsurgical units but not

    in ICUs when time errors were excluded.

    BCMA technology affected different types

    of medication administration errors in dif-

    ferent patient care areas.

    Index terms: Codes; Drug administration;

    Errors, medication; Hospitals; Quality assur-

    ance; Technology

    Am J Health-Syst Pharm.2009; 66:1202-10

    Since the publication of To ErrIs Human,1 two other reportsissued by the Institute of Medi-

    cine have addressed the importanceof preventing medication errors.2,3The administration of medicationis one of the most errorprone stepsof the medication-use process, with34% of all errors originating in thisphase.4 In addition, less than 2%of medication administration er-rors are intercepted at the patient

    bedside. Bar-code-assisted medica-tion administration (BCMA) wasdeveloped as an additional safetybarrier between the nurse and thepatient if a medication error reachesthe patients bedside. This technol-ogy assists the nurse in confirmingthe patients identity and confirmsthe appropriate identity, dose, time,and form of the medication. Thenumber of hospitals using BCMAis increasing; in 2002, only 5% ofhospitals with 300399 staffed beds

  • 7/25/2019 Effect of Bar-code-Assisted Medication Administration on Medication Administration Errors and Accuracy in Multiple

    2/9

    PRACTICE REPORTS Bar-code-assisted medication administration

    1203Am J Health-Syst PharmVol 66 Jul 1, 2009

    had implemented this technology,compared with 40.2% in 2007.5

    Organizations such as the Food andDrug Administration, the HealthcareInformation and Management Sys-

    tems Society, and the American So-ciety of Health-System Pharmacistshave urged the adoption of BCMA,and the Institute for Safe MedicationPractices considers BCMA a maturetechnology.6 However, only a fewstudies have investigated the effect ofBCMA on medication error rates.7-11These studies were conducted in dif-ferent settings, limiting the externalvalidity of the results. Some studieswere conducted in only one or two

    general care areas (medicalsurgicaland cardiac telemetry units),8,9with-out implementing a computerizedprescriber-order-entry (CPOE) sys-tem9or using an institution-specificBCMA system.7,9 Other studies fo-cused on warning and alert data asa surrogate marker for certain typesof medication administration errorsprevented by the BCMA system.10,11

    The evaluation of BCMA shouldnot be limited to its effect on theprevention of medication admin-istration errors, as implementingBCMA has important implicationsfor nursing workflow as well. An av-erage of 25% of nursing time is spenton medication-related activities,12and the error-prevention potentialof BCMA is greatly diminished ifthe technology does not support

    nursing workflow. Already, numer-ous work-around strategies afterimplementation of BCMA have beendescribed,13-16 illustrating the needto assess medication administration

    accuracy in addition to medicationadministration errors after BCMAimplementation. Direct observa-tion of medication administrationis the most efficient and practicalmedication-error detection methodand produces valid and reliable re-sults.17-19 The objective of this studywas to determine the effect of a com-mercially available BCMA system onmedication administration accuracyand medication administration er-

    rors on two patient care areas (gen-eral and intensive care) in a highlycomputerized setting using a vali-dated observation method.

    Methods

    This was a prospective, before-and-after, observational study. Dataon all outcome measures were col-lected one month before and threemonths after BCMA implementation.Observations were scheduled on bothweekdays and weekends. The medica-tion administrations scheduled for 9a.m. were the focus of this study, as themajority of medications on the studyunits were administered at that time.

    Setting. This study was conductedin two medicalsurgical units (22and 26 beds) and two intensive careunits (ICUs) (one 13-bed medical

    ICU and one 20-bed surgical ICU) ofa 386-bed academic teaching hospitalin southern California. The maxi-mum nurse:patient ratios were 1:4 onthe medicalsurgical units, 1:2 on the

    medical ICU, and 1:2 on the surgicalICU. Additional study unit charac-teristics during the preinterventionand postintervention periods aresummarized in Table 1.

    CPOE was bidirectionally inter-faced with the pharmacy informa-tion system, eliminating the tran-scription of medication orders bythe pharmacy. Pharmacists serviceto the medicalsurgical units con-sists of continuous centralized order

    validation and the daily presence ofa clinical pharmacist on the units.Specialized clinical pharmacists arestationed on the medical and surgicalICUs daily. Medication dispensing isfacilitated by unit-based automateddispensing cabinets (ADCs). High-volume medication administrationtimes are 9 a.m., 12 p.m., 6 p.m., and9 p.m., with the majority of medica-tions administered at 9 a.m.

    Before BCMA implementation,the patient-specific medication ad-ministration record (MAR) wasprinted once daily and served as apaper reference for the medicationsto be given to patients and completedadministrations for that day. Thehospitals CPOE system had to beregularly checked for new or modi-fied medication orders. Any changes

    Table 1.

    Characteristics of Study Units Before and After Implementation of Bar-Code-Assisted MedicationAdministration (BCMA)

    Characteristic

    MedicalSurgical Unit 1

    Daily occupancy, % 82 86 78 81 84 94 97 93

    No. patients discharged 119 93 202 149 13 14 19 31

    Length of patient stay, days 4.6 5.8 4.5 5.1 7.3 15.4 9.0 11.5

    Career nurses, % 75 82 98 90 87 86 90 89

    Nurse vacancy rate, % 11 4 6a 8a 0 8 10 15

    Surgical IntensiveCare Unit

    BeforeBCMA

    AfterBCMA

    MedicalSurgical Unit 2

    Medical IntensiveCare Unit

    BeforeBCMA

    AfterBCMA

    BeforeBCMA

    AfterBCMA

    BeforeBCMA

    AfterBCMA

    aNegative values indicate a surplus in staffing.

  • 7/25/2019 Effect of Bar-code-Assisted Medication Administration on Medication Administration Errors and Accuracy in Multiple

    3/9

    PRACTICE REPORTS Bar-code-assisted medication administration

    1204 Am J Health-Syst PharmVol 66 Jul 1, 2009

    needed to be transcribed on thepaper MAR as this document wasused to retrieve medication from theADC.

    BCMA implementation. BCMA

    technology (Medication Administra-tion Check using Med Administra-tion Check, version 23.04.9, Siemens,Malvern, PA) was implemented inthe hospital from May 2007 to Feb-ruary 2008. Both medicalsurgicalunits went live in October 2007; thesurgical and medical ICUs followedin December 2007 and January 2008,respectively. BCMA was based onan electronic MAR (eMAR), acces-sible on computers throughout the

    hospital, including the medicationstorage room and each patient room.The BCMA system was integratedwith the pharmacy information sys-tem and interfaced with the CPOEsystem. This allowed the eMAR tobe automatically updated when newmedication orders were entered intothe CPOE system or when existingorders were modified. The BCMAsoftware displayed medications dueat a certain time in the active worklist, which was used to retrieve medi-cation from the ADC. In the patientsroom, the nurse used the bedsidecomputer to select the appropriateeMAR and confirmed the patientsidentity by scanning the bar code onthe patients wristband. By scanningthe bar code on each dosage form,the medication, dosage form, dose,and administration time were veri-fied using the patients eMAR. Staffwas trained on the new technologyby completing a mandatory train-

    ing program consisting of an onlinetraining module and hands-on train-ing sessions in the hospital units.

    Data collection. Medication ad-ministration accuracy was measuredusing the accuracy indicator of theCalifornia Nursing Outcomes Co-alition (CalNOC).20 CalNOC is thelargest ongoing statewide nursingquality database project in the na-tion and engages approximately 150hospitals in nursing quality database

    development, benchmarking, and re-search efforts. CalNOCs medicationadministration accuracy indicatorwas made available to other hospitalsin July 2006. The indicator contains

    a component to detect medicationadministration errors, which wasadapted from a nationally recognizedobservational medication-error-detection methodology previouslypublished.17,18,21An observed medica-tion administration error is definedas any discrepancy between the med-ication administered to the patientand the medication ordered on thepatients medical record.21The indi-cator also uses identical subclasses

    of medication administration errors(appendix). However, the CalNOCtool also contains six indicators ofmedication administration accu-racy that reflect error-prone processvariations. Some of these accuracyindicators have been proven effectivein other studies that focused on thequality of medication administra-tion22or work-around scenario afterthe implementation of BCMA.13,23

    In addition to collecting data forthese indicators, the duration of themedication administration processfor each patient was recorded. Toassess compliance with the BCMAtechnology, additional information(medication name, strength, route ofadministration, and reason for over-ride) was collected if BCMA was notor could not be used.

    Observation procedures. Twopharmacists and six pharmacy stu-dents were trained to unobtrusivelyperform the observations. Training

    consisted of studying the manual forthe CalNOC medication administra-tion accuracy indicator. Adequateknowledge of study procedures wasensured by attending a two-hourreview session of CalNOCs train-ing manual and study procedures,developed by one of the pharmacists.The data collection form provided byCalNOC was modified to allow fasterdata collection and to accommodatethe additional variables unique to

    this study. Usability of the data col-lection form and interrater reliabilitywere assessed during two pilot ob-servation sessions on one medicalsurgical unit. During the first pilot

    session, the two pharmacists simulta-neously observed one nurse, and thedata collected were used to ensureinterrater reliability between the twopharmacists. During the second pilotsession, two groups of three studentsand one pharmacist observed medi-cation administered by one nurse,with each group observing a differ-ent medicalsurgical unit. Interraterreliability between the students andthe pharmacist was ensured by com-

    paring the observation data of eachstudent with those of their peers.Before the study observation ses-

    sions began, the study team informedthe nurse managers and nursing staffof each unit of the studys purposeand methodology. Practical issues,such as the proposed observationschedule, situations that were ex-cluded from observations, and theinformed-consent procedure, werediscussed. To prevent interferencewith nursing workflow, a maximumof two observers could be assigned toeach study unit during the observa-tion sessions. However, a nurse wasaccompanied by only one observerduring the medication administra-tion round.

    Since nurses were the subjects ofour study, informed consent fromthe patient was not required bythe hospitals institutional reviewboard. After contacting the nurseat the beginning of the medication

    administration round, explaining thepurpose of the study, emphasizingthat none of the nurses personal in-formation was collected, and statingthat study participation was entirelyvoluntary, oral informed consent wasobtained from the nurse. Observerinteraction with the patient waslimited to explaining the nature ofthe study and the presence of the ob-server. If the patient was uncomfort-able with the presence of the observer

  • 7/25/2019 Effect of Bar-code-Assisted Medication Administration on Medication Administration Errors and Accuracy in Multiple

    4/9

    PRACTICE REPORTS Bar-code-assisted medication administration

    1205Am J Health-Syst PharmVol 66 Jul 1, 2009

    at the bedside, the observer left theroom, and no data were collected.Medication administrations duringemergencies (e.g., cardiopulmonaryresuscitation) were also excluded

    from this study. The observers wereinstructed to intervene if they wit-nessed actions of the nurse that couldlead to a medication administrationerror.

    Observers arrived on the nursingunit approximately one hour beforethe scheduled medication admin-istration time, as nursing staff wereallowed to administer medicationone hour before to one hour after thescheduled administration time. The

    observation period started when anurse entered the medication roomand began retrieving medicationfrom the ADC. The identity, strength,and dose of the medication removedfrom the ADC were recorded by theobserver. The route, infusion rate(when applicable), and medicationadministration accuracy indicatorswere assessed at the patient bedside.After completion of the medicationadministration, the observer re-turned to the medication room andfollowed additional nurses until themedication administration periodon the unit was completed.

    Medication administration er-rors were assessed by comparing theobserved medication administeredto the medication intended for thatpatient. Before the intervention, theintended medication was determinedby photocopying the paper MARof each observed patient, as well asretrieving the medication data in the

    electronic medical record. After theintervention, the medication data inthe patients electronic medical re-cord was interfaced with the eMAR,resulting in a continuously updatedeMAR. Thus, after the interven-tion, the intended medication wasderived from the information in theeMAR only. The rate of medicationadministration errors was calculatedby dividing the number of errorsby the total opportunities for error

    (OEs). OEs were defined as the sumof observed administrations andomitted medications. As wrong-timeerrors were generally considered lesssevere than other errors,24overall re-

    sults were reported as total errors anderrors excluding wrong-time errors.Data analysis. In the medical

    surgical units,the number of obser-vations needed to adequately powerthis study was based on the results ofa similar study investigating the effectof BCMA on medication errors in asimilar patient care area.7Assuminga similar baseline error rate of 6.3%,an a of 0.05, and a power of 80%,at least 654 medication administra-

    tions had to be observed before andafter BCMA implementation on bothmedicalsurgical units combinedto detect a similar 54% decrease inmedication administration errors.

    Depending on the type of medi-cation errors, medication admin-istration error rates in an intensivecare setting using observationalmethodologies varied between 6.6%and 54%.25-27 Two of these threestudies were conducted in EuropeanICUs.26,27 Therefore, the medicationerror rates found by Kopp et al.25were used in our sample size calcula-tion, as the setting of this study wassimilar to ours. Assuming a similarbaseline error rate of 20%, an a of0.05, and a power of 80%, at least 262medication administrations had tobe observed before and after BCMAimplementation on both ICUs com-bined to detect an expected 54% de-crease in medication administrationerrors.

    Data were initially entered intospreadsheets (Excel, Microsoft Corp.,Redmond, WA) for initial analysisand summary statistics. Stata 10(StataCorp LP, College Station, TX)was used to calculate power andconduct additional statistical tests.For nominal data, chi-square analysiswas used; if five or fewer data pointswere analyzed, Fishers exact test wasused. Continuous data were analyzedusing the unpaired ttest. Apvalue of

    less than 0.05 was considered statisti-cally significant.

    Results

    Observation characteristics.The

    preintervention and postinterven-tion characteristics are summarizedin Table 2. On the medicalsurgicalunits, all observation characteris-tics were similar before and afterBCMA implementation except fora slight increase in topical medica-tion administrations after BCMA.On the ICUs, fewer subcutaneousadministrations but more infusionsby i.v. minibag were observed afterBCMA implementation. Also, more

    observations were conducted duringthe 9 a.m. medication administrationround postimplementation, result-ing in fewer observations duringthe 6 p.m. and 9 p.m. medicationadministration rounds. After the in-tervention, fewer observations wereconducted during weekends.

    Medication administration ac-curacy. Baseline medication admin-istration accuracy was higher in themedicalsurgical units comparedwith the ICUs. On the medicalsurgical units, three accuracy indica-tors changed after the introductionof BCMA; improved compliancewith checking patient identity afterBCMA implementation was offset bymore distractions and interruptionsand less explanation of the medica-tion to the patient (Table 3). Thesethree indicators did not change in theICUs. However, implementation ofBCMA resulted in improved chartingand labeling of medications adminis-

    tered in the ICUs.Medication administration er-

    rors. The baseline medication errorrate was 10.7% and 12.6% on themedicalsurgical units and ICUs,respectively, or 8.0% and 11.0%,respectively, if wrong-time errorswere excluded. Although the totalerror rates on the medicalsurgicalunits did not significantly decreaseand an increase in time errors afterBCMA implementation was noted,

  • 7/25/2019 Effect of Bar-code-Assisted Medication Administration on Medication Administration Errors and Accuracy in Multiple

    5/9

    PRACTICE REPORTS Bar-code-assisted medication administration

    1206 Am J Health-Syst PharmVol 66 Jul 1, 2009

    Table 2.

    Observation Characteristics of Study Units Before and AfterImplementation of Bar-Code-Assisted Medication Administration

    (BCMA)

    Variable

    MedicalSurgicalUnitsa

    BeforeBCMA

    AfterBCMA

    Intensive Care Unitsa

    BeforeBCMA

    AfterBCMA

    aUnless otherwise noted, pnot significant.bOE = opportunities for error, defined as the sum of observed administrations and omitted medications.cp= 0.046.dp= 0.001ep= 0.042fExamples include rectal, intraocular, and intranasal routes.gp< 0.0001.hp= 0.006.

    Total no. OEb 888 697 374 394

    Median (range) OE per patient 5 (114) 5 (116) 5 (111) 4 (114)

    No. (%) OE per administration

    route

    Oral 736 (82.9) 581 (83.4) 255 (68.2) 261 (66.2)

    Subcutaneous 60 (6.8) 35 (5.0) 35 (9.4) 22 (5.6)c

    I.V. minibag 34 (3.8) 19 (2.7) 38 (10.2) 72 (18.3)d

    I.V. bolus dose 18 (2.0) 19 (2.7) 30 (8.0) 22 (5.6)

    I.V. large-volume

    parenteral 5 (0.6) 0 1 (0.3) 0

    Intramuscular 1 (0.1) 1 (0.1) 0 0

    Topical 21 (2.4) 29 (4.2)e 5 (1.3) 8 (2.0)

    Miscellaneousf 0 0 10 (2.7) 9 (2.3)

    No. (%) observations

    9 a.m. 839 (94.5) 651 (93.4) 329 (88.0) 389 (98.7)g

    12 p.m. 2 (0.2) 0 0 0

    6 p.m. 47 (5.3) 46 (6.6) 32 (8.6) 5 (1.3)g

    9 p.m. 0 0 13 (3.5) 0g

    Weekend observations, % 6.2 7.2 6.7 2.5h

    Median (range) duration of

    medication administration,

    min 10 (130) 10 (150) 12 (158) 13.5 (153)

    the error rate excluding wrong-timeerrors decreased by almost 58% af-ter BCMA implementation on theseunits (Figure 1). Substantially feweromitted medications and a decrease

    in the number of medications thatwere unavailable at the time ofadministration contributed to thiseffect.

    In contrast, no differences werefound for the overall error rate(12.6% before BCMA and 13.5% af-ter BCMA), the error rate excludingwrong-time errors (11.0% beforeBCMA and 9.9% after BCMA), andthe error types after BCMA imple-mentation (Figure 2) on the ICUs.

    Discussion

    A general medication administra-tion accuracy and error-assessmenttool, designed for use in multiplehospitals and in different care areas,

    was used to investigate the effectsof a commercially available BCMAsystem on two medicalsurgicalunits and two ICUs. This tool wasspecifically developed to comparemedication administration accuracyamong hospitals with different lev-els of automation. Therefore, thistool contains general medicationadministration accuracy indicatorsand multiple medication administra-tion error categories. However, the

    effect of bar-coding technology onmedication administration accuracyis reflected in only a limited numberof accuracy indicators (i.e., improvedchecking of patient identity, im-

    proved charting after administration,availability of MAR at the patientsbedside while administering medica-tion). BCMA technology specificallyaims to decrease the following errortypes: unauthorized drug, wrongform, wrong dose, wrong route, extradose, and omission. The tool usedin this study will only show a de-crease in medication errors if a largenumber of these error types werepresent at baseline, which explains

    the difference in the rates of medica-tion administration errors betweenmedicalsurgical units and ICUsafter BCMA implementation. Omis-sion was the predominant error typeon the medicalsurgical units. As aresult, the 58% reduction of totalerrors excluding wrong-time errorson the medicalsurgical units canbe largely explained by the decreasein errors of omission. Few errors ofomission were detected in the ICUsat baseline. Even though these errorsdecreased by 50% (from six to threeerrors), the low prevalence of thistype and other types of errors sus-ceptible to improvement by BCMAtechnology resulted in a nonsignifi-cant decrease in the total number ofmedication administration errors.

    On the medicalsurgical units,the number of wrong-time errorsincreased after BCMA implementa-tion. It is unlikely that the observedincrease in these errors was a result of

    a longer duration of the medicationadministration round after BCMAimplementation, as the medianduration of medication administra-tion on these units did not changeafter BCMA implementation andvaried widely per patient (Table 2).Similarly, the number of wrong-timeerrors observed did not change afterBCMA implementation in the ICUs.These findings are consistent withresults of a recent study that detected

  • 7/25/2019 Effect of Bar-code-Assisted Medication Administration on Medication Administration Errors and Accuracy in Multiple

    6/9

    PRACTICE REPORTS Bar-code-assisted medication administration

    1207Am J Health-Syst PharmVol 66 Jul 1, 2009

    differences in nursing time spenton medication administration afterBCMA implementation, which alsofailed to show a difference.28

    The number of medications notavailable on the medicalsurgicalunits at the time of administrationdecreased by 61%. A possible expla-nation for this is the implementationof a new hospitalwide ADC refillpolicy. However, this seems unlikely,as the new policy resulted in fewerdaily refills of the ADCs, which theo-retically could lead to more unavail-ability errors. Also, this was a hos-pitalwide policy change, and similardecreases on the ICUs would havebeen observed. More-likely expla-nations are changes in pharmacyprocurement practices and thorough

    checks of bar-code readability onarrival of new inventory in the phar-macy as a result of the BCMA imple-mentation. Differences in the typesof medications used in the ICUsversus the medicalsurgical unitscould explain the differences betweenthe number of wrong-time errors inthese patient care areas.

    Nurses were often distracted dur-ing medication administration: oneof every six and almost one of every

    Table 3.

    Accuracy of Medication Administration Before and After Implementation of Bar-Code-Assisted

    Medication Administration (BCMA)

    Indicator

    MedicalSurgical Units

    Distraction or interruption during

    medication administration

    Two forms of identification not checked

    Medication not explained to patientc

    Medication charted immediately after

    administration

    Medication not labeled at patient bedside

    Medication not compared to MAR before

    administration

    BeforeBCMA After BCMA

    BeforeBCMA After BCMA

    127/822 (15.5) 169/670 (25.2)b 104/352 (29.5) 113/373 (30.3)

    110/822 (13.4) 46/667 (6.9)b 104/353 (29.5) 90/372 (24.2)

    88/808 (10.9) 93/623 (14.9)d 53/170 (31.2) 50/155 (32.3)

    74/825 (9.0) 56/668 (8.4) 86/352 (24.4) 25/371 (6.7)b

    17/838 (2.0) 7/666 (1.1) 25/353 (7.1) 12/370 (3.2)e

    4/837 (0.5) 9/686 (1.3) 3/365 (0.8) 0/393 (0)

    a% observations = no. observed indicator/no. possible occurrence of the indicator (e.g., medications could only be explained to a conscious patient).

    bp< 0.0001.cMAR = medication administration record. Medication is considered adequately explained if at least the name of the medication is mentioned to a conscious patient.dp= 0.045.ep= 0.026.

    No. (%) Observationsa

    Intensive Care Units

    Figure 1.Total errors in the medicalsurgical units and intensive care units (ICUs) beforeand after bar-code-assisted medication administration (BCMA) implementation. Errorrates were calculated by dividing the number of errors by the total opportunities for error(observed administrations plus omitted medications). Numbers in bars indicate absolutenumbers of errors, * indicates p< 0.05, and *** indicates p< 0.0001.

    Before BCMA

    MedicalSurgical Units

    15

    14

    13

    12

    11

    10

    9

    8

    7

    6

    5

    4

    3

    2

    1

    0

    After BCMA Before BCMA After BCMA

    ICUs

    ErrorRate(%)

    Wrong-time errors

    Other errors

    24

    71

    33

    24

    6

    41

    14

    39

    *

    ***

    three medication administrationswere interrupted on the medicalsurgical units and ICUs, respectively.After BCMA implementation, the

    number of interruptions on themedicalsurgical units increased. It isunlikely that this was due to BCMAimplementation, as there were no

  • 7/25/2019 Effect of Bar-code-Assisted Medication Administration on Medication Administration Errors and Accuracy in Multiple

    7/9

    PRACTICE REPORTS Bar-code-assisted medication administration

    1208 Am J Health-Syst PharmVol 66 Jul 1, 2009

    obvious differences between thetwo observation periods that couldexplain this. Although decreasingthe number of interruptions and dis-tractions should always be a priorityduring medication administration,they will never be fully eliminated(especially in critical care areas). Itis therefore reassuring that BCMA is

    now in place to prevent medicationadministration errors resulting fromthese interruptions.

    In the medicalsurgical units,BCMA implementation resulted inincreased compliance with the hos-pitals policy of checking a patientsidentity; no difference was observedin the ICUs. Current policy requirestwo forms of identification to bechecked (orally confirming the pa-tients name and scanning the bar

    Figure 2.Types of errors excluding wrong-time errors before and after bar-code-assistedmedication administration (BCMA) implementation. Error rates were calculated by divid-ing the number of errors by the total opportunities for error (observed administrationsplus omitted medications). Numbers in bars indicate absolute numbers of errors, * indi-cates p< 0.05, and *** indicates p< 0.0001.

    Before BCMA

    MedicalSurgical Units

    12

    11

    10

    9

    8

    7

    6

    5

    4

    3

    2

    1

    0

    After BCMA*** Before BCMA After BCMA

    ICUs

    ErrorRate(%)

    Unauthorized

    Wrong technique

    Wrong dose

    Drug not available

    Omission

    5

    31

    34

    4

    12

    7

    15

    8

    12

    6

    13

    9

    14

    3

    *

    ***

    code on the patients wristband). Inthe ICUs, baseline compliance withthis policy remained low after BCMAimplementation. One reason couldbe that most patients in the ICUs areunconscious, so oral verification ofthe patients identity is impossible.The second method, visually check-ing the patient name and medical

    record number on the wristbandand scanning the wristband, wasoften not performed, as nurses wereassigned to the same patient for theentire day. This accuracy indicatoris probably less suitable for ICU set-tings, where the nurse:patient ratiois very high. However, checking twoforms of identification is one of theJoint Commissions National PatientSafety Goals29 and failing to checkpatient identity has led to patient

    harm, even after BCMA had beenimplemented.30

    After BCMA implementation,medications were less frequentlyexplained to the patients on the

    medicalsurgical units. This findingwarrants further investigation.In the ICUs, charting compliance

    after medication administrationgreatly improved. Baseline chartingcompliance was low in the ICUs; asBCMA technology is specifically de-signed to facilitate this process, thisindicator was expected to improve.

    CalNOCs medication admin-istration accuracy tool allowed forthe monitoring of medication ac-

    curacy and errors in different careareas, using indicators of multipleerror-prone steps of the medica-tion administration process. Thisis important, as implementationof BCMA technology can result inimprovements in one error-proneprocess but have unintended conse-quences on others.13,16,23,30

    This study had several limitations.The observational methodology hasbeen criticized for causing alteredbehavior of the observed subject (theHawthorne effect). This effect hasbeen shown to be negligible if theobservers are experienced, objective,unobtrusive, and nonjudgmental.17In this study, almost half of the ob-servations were done by fourth-yearpharmacy students who could beconsidered nonexperienced observ-ers. However, the expected change inbehavior of the nursing staff wouldresult in improved medication ad-ministration accuracy and fewer er-

    rors, as nursing staff were aware of thepurpose of the study. Also, we did notmatch the route of administration ofthe preintervention observations tothe postintervention observations,as this proved to be very impractical.As a result, the distribution of theobserved routes of administrationdiffered between the preinterventionand postintervention periods (Table2). In addition, the majority of obser-vations of medication administration

  • 7/25/2019 Effect of Bar-code-Assisted Medication Administration on Medication Administration Errors and Accuracy in Multiple

    8/9

    PRACTICE REPORTS Bar-code-assisted medication administration

    1209Am J Health-Syst PharmVol 66 Jul 1, 2009

    occurred during the 9 a.m. medica-tion round. Further, the observer whoconducted 50% of the observationswas also responsible for most of thedata entry. However, it is unlikely that

    this biased the results, as data integritywas ensured by using an automateddata-checking tool developed byCalNOC. Further, errors were as-sessed by the observers immediatelyafter each observation and not byindependent researchers. In addition,we did not assess the severity of theadministration errors detected butbased error assessment on the rigiderror definitions of CalNOC. Finally,the postintervention assessment was

    conducted three months after BCMAimplementation, which could beconsidered too short. However, otherstudies evaluating the effects of tech-nology on health care have also useda three-month implementation pe-riod.31-33We ensured appropriate useof BCMA technology by assessingscanning compliance during the post-implementation observations. Thecompliance rates were 89% on themedicalsurgical units and 94% onthe ICUs, similar to compliance ratesfound in other studies investigatingBCMA technology.7,16

    Despite these potential limita-tions, the error rates found weresimilar to those found in other stud-ies using similar methodologies anddefinitions of medication adminis-tration errors. Paoletti et al.7 founda preimplementation error rate of6.3% on a 36-bed medicalsurgicalunit, excluding wrong-time errors.Another study investigating the effect

    of a closed-loop electronic prescrib-ing and administration system onadministration errors in a 28-bedgeneral surgical unit found an errorrate of 8.6% if wrong-time errorswere excluded.8Although baseline er-ror rates in the ICUs were lower thanexpected, this study was adequatelypowered to detect a 50% decrease inerrors after BCMA implementation.

    This is the first study showing ma-jor differences in the effect of BCMA

    technology on medication admin-istration accuracy and errors in dif-ferent patient care areas. Recently,similar results in medication admin-istration errors and improvements in

    accuracy were reported in a subset ofseven California hospitals, using theCalNOC medication accuracy indi-cator.34 These improvements wereachieved by adherence to protocolsand increased auditing. Their resultsand the effects shown in this studyemphasize that implementing BCMAforces organizations to take a closerlook at the whole medication ad-ministration process. ImplementingBCMA technology has been shown

    to be a cost-effective intervention

    28

    and makes empirical sense. However,other hospitals using the CalNOCmethodology have found that im-proving current systems by adheringto protocols and educating staff cangenerate similar results.34

    Conclusion

    Implementing BCMA technologydecreased medication administra-tion errors in medicalsurgical unitsbut not in ICUs when time errorswere excluded. BCMA technology af-fected different types of medicationadministration errors in differentpatient care areas.

    References1. Kohn LT, Corrigan JM, Donaldson MS,

    eds. To err is human: building a saferhealth system. Washington, DC: NationalAcademy Press; 1999.

    2. Committee on Quality of Health Carein America. Crossing the quality chasm:a new health system for the 21st cen-tury. Washington, DC: National AcademyPress; 2001.

    3. Aspden P, Wolcott J, Bootman JL et al.,eds. Preventing medication errors: qual-ity chasm series. Washington, DC: Na-tional Academy Press; 2007.

    4. Bates DW, Cullen DJ, Laird N et al. Inci-dence of adverse drug events and poten-tial adverse drug events. Implications forprevention. JAMA. 1995; 274:29-34.

    5. Pedersen CA, Gumpper KF. ASHP na-tional survey on informatics: assessmentof the adoption and use of pharmacyinformatics in U.S. hospitals2007.Am J Health-Syst Pharm. 2007; 65:2244-64.

    6. Institute for Safe Medication Practices.A call to action: safeguard drug admin-

    istration within 2 years! www.ismp.org/newsletters/acutecare/articles/whitepaperbarcodding.asp?ptr=y (accessed2008 Apr 17).

    7. Paoletti RD, Suess TM, Lesko MG et al.Using bar-code technology and medica-tion observation methodology for safer

    medication administration.Am J Health-Syst Pharm.2007; 64:536-43.

    8. Franklin BD, OGrady K, Donyai P et al.The impact of a closed-loop electronicprescribing and administration systemon prescribing errors, administrationerrors and staff time: a before-and-afterstudy. Qual Saf Health Care.2007; 16:279-84.

    9. Poon EG, Keohane C, Featherstone E etal. Impact of barcode medication admin-istration technology on how nurses spendtheir time on clinical care. AMIA AnnuSymp Proc.2006:1065.

    10. Sakowski J, Leonard T, Colburn S et al.Using a bar-coded medication admin-

    istration system to prevent medicationerrors in a community hospital network.Am J Health-Syst Pharm. 2005; 62:2619-25.

    11. Johnson CL, Carlson RA, Tucker CL et al.Using BCMA software to improve patientsafety in Veterans Administration medi-cal centers. J Healthc Inf Manag. 2002;16:46-51.

    12. Keohane CA, Bane AD, Featherstone Eet al. Quantifying nursing workflow inmedication administration. J Nurs Adm.2008; 38:19-26.

    13. Patterson ES, Rogers ML, Chapman RJ etal. Compliance with intended use of barcode medication administration in acuteand long-term care: an observationalstudy. Hum Factors.2006; 48:15-22.

    14. Cochran GL, Jones KJ, Brockman J et al.Errors prevented by and associated withbar-code medication administration sys-tems. Jt Comm J Qual Patient Saf. 2007;33:293-301.

    15. Carayon PP, Wetterneck TB, Hundt ASet al. Evaluation of nurse interactionwith bar code medication administrationtechnology in the work environment. JPatient Saf.2007; 3:34-42.

    16. Koppel R, Wetterneck T, Telles JL et al.Workarounds to barcode medication ad-ministration systems: their occurrences,causes, and threats to patient safety. JAm Med Inform Assoc. 2008; 15:M2616.

    Abstract.17. Barker KN, Flynn EA, Pepper GA. Obser-

    vation method of detecting medicationerrors. Am J Health-Syst Pharm. 2002;59:2314-6.

    18. Flynn EA, Barker KN, Pepper GA et al.Comparison of methods for detectingmedication errors in 36 hospitals andskilled-nursing facilities.Am J Health-SystPharm. 2002; 59:436-46.

    19. Dean B, Barber N. Validity and reliabilityof observational methods for studyingmedication administration errors. Am JHealth-Syst Pharm.2001; 58:54-9.

    20. Aydin CE, Bolton LB, Donaldson N et al.Creating and analyzing a statewide nurs-

  • 7/25/2019 Effect of Bar-code-Assisted Medication Administration on Medication Administration Errors and Accuracy in Multiple

    9/9

    PRACTICE REPORTS Bar-code-assisted medication administration

    1210 Am J Health-Syst PharmVol 66 Jul 1, 2009

    ing quality measurement database.J NursScholarsh. 2004; 36:371-8.

    21. Barker KN, Flynn EA, Pepper GA et al.Medication errors observed in 36 healthcare facilities. Arch Intern Med. 2002;162:1897-903.

    22. Greengold NL, Shane R, Schneider P

    et al. The impact of dedicated medi-cation nurses on the medication ad-ministration error rate: a randomizedcontrolled trial. Arch Intern Med. 2003;163:2359-67.

    23. Patterson ES, Cook RI, Render ML.Improving patient safety by identifyingside effects from introducing bar codingin medication administration. J Am MedInform Assoc. 2002; 9:540-53.

    24. Allan EL, Barker KN. Fundamentals ofmedication error research. Am J HospPharm. 1990; 47:555-71.

    25. Kopp BJ, Erstad BL, Allen ME et al. Medi-cation errors and adverse drug events inan intensive care unit: direct observation

    approach for detection. Crit Care Med.2006; 34:415-25.26. Van den Bemt PM, Fijn R, van der Voort

    PH et al. Frequency and determinantsof drug administration errors in theintensive care unit. Crit Care Med. 2002;30:846-50.

    27. Tissot E, Cornette C, Demoly P et al.

    Medication errors at the administrationstage in an intensive care unit. IntensiveCare Med. 1999; 25:353-9.

    28. Maviglia SM, Yoo JY, Franz C et al. Cost-benefit analysis of a hospital pharmacybar code solution.Arch Intern Med.2007;167:788-94.

    29. Joint Commission. 2008 Joint Commis-sion National Patient Safety Goals. www.jcipat ientsafety.org/257 64/ (acce ssed2008 May 6).

    30. McDonald CJ. Computerization can cre-ate safety hazards: a bar-coding near miss.Ann Intern Med.2006; 144:510-6.

    31. James BC, Weingart SN, Wilson RM etal. Making it easy to do it right. N Engl JMed.2001; 345:991-3.

    32. Bates DW, Cohen M, Leape LL et al. Re-ducing the frequency of errors in medi-cine using information technology.J AmMed Inform Assoc.2001; 8:299-308.

    33. Lin L, Vicente KJ, Doyle DJ. Patient safety,potential adverse drug events, and medi-

    cal device design: a human factors engi-neering approach.J Biomed Inform. 2001;34:274-84.

    34. Medical News Today. UCSF pro-gram achieves over 56% reduction inmedication administration error. www.medicalnewstoday.com/articles/102068.php (accessed 2008 May 7).

    AppendixVariables of the medication

    administration accuracy tool developed

    by the California Nursing Outcomes

    Coalition20

    Medication administration errors

    1. Unauthorized drug

    2. Wrong dose3. Wrong form4. Wrong route5. Wrong technique6. Extra dose7. Omission8. Wrong time9. Drug not available

    Medication administration accuracy indicators

    1. Medication is not compared to medi-cation administration record (MAR) beforeadministration

    2. Distraction or interruption of the nurseduring medication administration

    3. Medication is not labeled at patientbedside

    4. Two forms of patient identification arenot checked

    5. Medication is not explained to the pa-tient during administration

    6. Medication is charted on the MAR orelectronic MAR immediately after administration