Professional Documents
Culture Documents
S tudent’s
CORNER
ABSTRACT The purpose of this paper is to examine medication errors in Australian hospitals
from a nursing perspective.This paper provides a detailed review of past and
current literature to examine prevalence rates and risk factors for medication
errors. It also describes unintended injury or harm, also known as adverse events,
associated with a medication error, and proposes strategies to counteract them.
According to the literature, approximately two percent of patients in an acute
hospital setting will experience a medication error.Administration of medications
has remained a high risk task for registered nurses, and for this reason, the
strategies to reduce error discussed in this paper will be predominantly nursing
focussed.The occurrence of a medication error is a significant issue in the health
Key Words care setting, and although the relative risk of an individual error occurring
nursing; is low, the rate of error remains perceivably high in clinical practice today.
medication With continual technological and pharmaceutical advances, the medication
error; adverse
administration environment is becoming increasingly complex. Given the nature
CN
event; drug;
patient safety; of medication therapy, it is near impossible to avoid all error, human and
risk otherwise.
Received 7 February 2008 Accepted 11 February 2009
INTRODUCTION
n the clinical setting, all too frequent errors
I in the administration of medications are a
persistent problem. As registered nurses have
the responsibility of the final contact with the
JENNIFER EVANS medication prior to administration, they are
Undergraduate student often held accountable for subsequent adverse
Deakin University –
Burwood Campus
events. Administering a medication is the nurs-
Melbourne VIC, Australia ing task that carries the highest risk, and the
hoped that the rigidity of this view is rapidly hospitals were due to omissions of medication,
disappearing, even though the prevalence of and 20 percent were as a result of overdosing a
medication errors remains perceivably high in patient. Administering the wrong medication
clinical practice today. accounted for ten percent of errors made in
hospitals and incorrect labelling was reported
PREVALENCE OF MEDICATION to be responsible for less than five percent
ERRORS IN AUSTRALIA (Hodgkinson 2006).
Given the time-critical, delicate and variable In terms of patient safety, the NCC MERP
nature of health care, all health professionals (2002) stated that there is no acceptable occur-
must understand that every patient provides an rence rate of medication errors. The rate of
opportunity for a serious mistake to be made error can be influenced by many factors which
(Wolf et al. 2000). In the clinical environment can increase the likelihood of an error occurring
in Australia, the prevalence of medication errors in specific situations.
is of serious concern (Joanna Briggs Institute
2006).The importance of the statistics outlined RISK FACTORS LEADING TO
below is highlighted by Fogarty and McKeon MEDICATION ERRORS
(2006) who identify adverse events due to med- Mayo and Duncan (2004) surveyed 983 reg-
ication errors as a leading cause of unintentional istered nurses working in acute care hospitals
patient harm within an acute care environment. to determine their perception of medication
The literature shows a reasonably wide range errors.They reported that illegible hand writing
of recorded rates of adverse drug events, rang- of prescriptions, distraction and fatigue were
ing from one to four percent of hospital admis- the highest contributing factors to medication
sions. Specifically, clinical indicators show a one errors. A British study undertaken by Fry and
percent chance of adverse drug event occur- Darcy (2007) identified the most common con-
ring, whereas the medical record reviews have tributors to error as: unclear prescriptions,
shown two to four percent of admissions are high workload, and a busy schedule with many
associated with an adverse medication event pressures. Hand and Barber (2000: 558) identi-
(Runciman et al. 2003). fied ‘personal’, (such as a high level of stress),
In 1995, The Quality in Australian Health ‘contextual’ (such as a noisy ward with many
Care Study was conducted and it remains the patients) and ‘knowledge base’ as the three im-
largest nationally representative study to date portant categories of human error which con-
(ACSQH 2004). Using the data contained in the tributed to medication errors.
study, it can be calculated that approximately As similarly defined by many, human error
two percent of patients will experience a med- involves the failure of intended actions to elicit
ication error, which is of similar magnitude to a desired response, causing deviation from what
the data presented above, from Runciman et al. is right or true (Reason 1997; Hansen 2006).
(2003). Kleinpell (2001) suggests that many of the fac-
Errors can occur at any step in the medica- tors leading to medication errors are human
tion process (from prescribing to administra- related, such as inexperienced or temporary
tion), and an important aspect to consider when staff, caused by a knowledge deficit or inac-
examining the prevalence of medication errors curate documentation. Distractions and other
is the type of error, or at which phase the error impeding factors such as fatigue, stress and not
actually occurred. Data from the Australian focusing on the task can also have very detri-
Incident Monitoring System shows that over a mental effects on the patients’ well being and
quarter of medication incidents occurring in are known to contribute significantly to human
cations being administered, nurses are also 2001). This punitive environment can result in
required to comprehend the application of the adverse outcomes for the nurse, and indeed the
medication administration procedure. Mathe- patient, as damaged confidence can often result
matical ability is an essential skill for many in changes to clinical practise and competence.
aspects of a nursing career, yet a number of More importantly, it fails to address the under-
studies have identified insufficient levels of lying cause of the error such as the work condi-
accuracy (Polifroni, McNulty & Alllchin 2003; tions which may have contributed.
Bruce & Wong 2001). Performing a medication
calculation requires both arithmetic skills and ADVERSE EVENTS RELATED TO
conceptual skills, including being able to con- MEDICATION ERRORS
textualise the calculation with relevance to the Adverse drug events have been estimated to
patient, medication or clinical scenario (Rainboth cost the Australian public $350 million annually
& DeMasi 2006; Bliss-Holtz 1994).This can be a (Hodgkinson 2006). It has been calculated that
time-consuming, difficult, and anxiety provoking adverse drug events consume approximately
task for nursing students and registered nurses one percent of the total sum spent on health
(Pozehl 1996; Bayne & Bindler 1988). care nation wide (Runciman et al. 2003). Along-
Over the past decade, the health sector has side this, each year in Australia, it is estimated
seen many technological advances, resulting in that general practitioners manage up to 400,000
changes to nursing practice. Although the skill adverse drug events, many of which could
is far from obsolete, there has been a decrease be related to medication errors (Joanna Briggs
in the frequency of medication calculations Institute 2006). Many medication errors can
required of a nurse. The increased reliance on end in tragedy for the patient and family, and
technology has resulted in less demand for even the health professionals and associated
nursing staff to execute medication calculations. agency (Wolf et al. 2000).
This reduced clinical exposure could potentially In 1995, the Quality in Australian Health
lead to a loss of proficiency (Rainboth & DeMasi Care Study reviewed medical records of over
2006; Bindler & Bayne 1991). In 2005 a report 14,000 patients in hospitals across Australia.
published in Finland indicated that nurses were More than half of all occurring medication
aware of their own substandard medication cal- errors were considered preventable. In 77.1
culation abilities (Grandell-Niemi et al.). Prior percent of cases, the resultant disability had
research demonstrating the ability of nurses to resolved within a year, and 13.7 percent of cases
accurately predict their own calculation skills resulted in a long term disability. Medication
supports this assertion (Bayne & Bindler). This errors can sometimes cause fatality, reported as
lack of discrepancy between perceived and actu- occurring in nearly five percent of cases (Wil-
al ability has been apparent for nearly 15 years, son et al 1995).
with potentially incompetent nurses practising Depending on the clinical situation, medica-
medication calculations on a daily basis. tion errors can have a broad range of effects,
Although human error is accountable for from going unnoticed to causing death. For
some medication errors, it is believed that this example, a patient with an Aspirin allergy was
is only a small percentage (Henneman & Gawl- administered Indocin and developed a non-fatal
inski 2004). Unfortunately, because medication respiratory arrest. This event may have been
errors attributable to human fallibility are some- preventable if allergies and contraindicated con-
what easier to recognise, misplaced emphasis ditions were checked first (Hartigan-Go 2006).
contributes to a culture of blame within health- Serumbus,Wolf, and Yongblood (2001) report-
care organisations (Institute of Medicine [IOM] ed a paediatric patient who received an injec-
the anticoagulant, Warfarin. This might have sionals to be near perfect appears to have creat-
been prevented if the five rights were vigilantly ed an atmosphere in which mistakes are per-
followed. Much research has been conducted in ceived as intolerable and unacceptable, resulting
this area and it mostly shows that checking these in a degree of dishonesty. Given the potential
five rights helps to ensure safe and effective ad- significance of retribution, (ranging from loss of
ministration of a medication (Wolf et al. 2000). job and self-confidence, to an increased level of
Conflicting results have also been obtained, vigilance when conducting future medication
showing that errors still occur when nurses administrations), many professionals choose to
adhere to the five rights (Smetzer 2000). As in conceal the error (Serumbus et al. 2001). The
most clinical scenarios, there is room for human lack of reporting of errors is a key issue within
error and many human factors such as poor the health care system that needs to be addressed.
handwriting or lack of experience, can con- Wolf et al (2000) suggest that health care
tribute to an inaccurate verification of a right. workers need a better understanding of the
Healthcare facilities that do not employ the use personal impact a medication error can have.
of identification wrist bands (such as aged care Appreciating the anxiety and guilt associated
hostels) can have problems with accurate verifi- with the reporting of a medication error would
cation of identity as verbally questioning the enable them to develop appropriate support in
individual could be tainted by poor hearing or a personal and clinical context. This increased
confusion. Therefore, the five rights on their awareness would also need to incorporate a
own are not enough to prevent high rates of change in perspective, viewing medication errors
medication error. as indicators for continuous quality improve-
Some literature now refers to the sixth right ment, rather than an individual’s failure. In
of medication administration. Wilson and Di- order to do this, medication errors need to
Vito-Thomas (2004), propose the sixth right: be dealt with in a constructive manner. This
right response. This sixth right is intended to includes exploring underlying causes, further
focus on the evaluation of the effectiveness of education or training if required, and support
the medication, and reiterates the importance or professional help to optimise coping mecha-
of taking responsibility for the administration of nisms (Schelbred & Nord 2007).
medications. In clinical scenarios involving pain While the importance of a comprehensive
control, anticoagulant and antiarrhythmic thera- incident reporting scheme is well recognised in
pies, the importance of assessing patient res- the nursing community, many reporting schemes
ponse is evident.This proposed sixth right is not that have been implemented do not elicit very
applicable in all administration situations; how- high levels of response, thus rendering them
ever, it is important that the medication process inadequate (Cullen et al. 1995). If all medica-
does not end at administration. tion errors and potential medication errors are
Many medication errors are not reported reported, the system has the ability to identify
and thus are undetected, so only a minority of and correct problems, sometimes before they
causes can be identified (Anderson & Webster occur (Anderson & Webster 2001). It is also
2001).The reason many nurses and other health essential that near-misses are reported, as near
professionals do not report some medication misses occur more frequently than actual ad-
errors or near errors, is fear of professional and verse events. The NCC MERP (2002) suggest
personal retributions, which could lead to a the following three factors which can affect the
damaged career, reputation or confidence (Wolf reporting of a medication error: a non-punitive
et al. 2000). culture in the health organisation can result in
The social pressure for health care profes- increased reporting, the varied definitions
port for clinical decisions (Huffstutler,Wyatt & has now been implemented, bringing with it a
Wright 2002). As demonstrated by Greenfield form of safer communication, with a better
(2007), providing immediate access to health understanding of the prescribing process and
information and data at the bedside increases strategies to minimise patient harm (ACSQH
the accuracy, speed and safety of nursing care, 2004; 2005).
decreasing the incidence of medication errors. A strategy that would involve many dis-
The potential of PDA use to improve medica- ciplines within healthcare is the avoidance of
tion safety depends on the sufficiency of the unsafe abbreviations (Abushaiqa et al. 2007).
information available. To be considered suffi- Abbreviations that would be deemed unsafe are
cient, the source must contain specific, accurate those which provide the opportunity for error
and detailed information, able to meet the to occur through misinterpretation of medica-
needs of each clinical decision involved in the tion name, dosage or units. Unsafe abbreviations
process of medication use. One recommended are not only misunderstood, but can also appear
change to further enhance medication safety to mean something completely unintended
was the inclusion of specific product details when combined with other surrounding words,
(such as colour, markings and size) in PDA data numerals or symbols (New South Wales Thera-
sources (Galt et al. 2005). peutic Advisory Group [NSW TAG] 2007).
In 2003, the IOM published a report which Meadows (2003) for example, described a patient
identified the use of information technology as who died as a result of an order for 20 units of
one of five core areas in which students and insulin being abbreviated to ‘20U’. The U was
working health professionals must develop and mistaken as a zero and subsequently the patient
maintain proficiency. All clinicians need to was injected with 200 units. In order to pro-
understand the limitations of technology and mote adequate patient safety, universal and
continue to exercise vigilance in all patient unambiguous symbols and abbreviations need to
duties.Whilst computerised technology has the be used consistently (NSW TAG 2006).
potential to improve the level of patient safety, The ACSQH (2002) suggest the risk of error
it is also possible that new types of errors will in medication administration from a ward stock
arise. Clinicians are more likely to ignore exces- supply is 15–20 percent, a much higher inci-
sive warnings produced by PDAs or other such dence compared to when individual patient
technology, which may lead to an adverse event stocks are used (five to eight percent).This evi-
occurring (Gaddis, Holt & Woods 2002). The dence suggests that it is a safer option to have
reliance on computer systems may also lead to a pharmacy staff measure and dispense individual
reduction in human vigilance, resulting in nurs- patient doses. This would also reduce nursing
es being less aware of medication characteristics workload, as medication preparation and ad-
and possibly making an error they would not ministration have been reported to consume
have made otherwise (NHMRC 2006). nearly 40 percent of nursing shift time (Armi-
Health ministers across Australia supported tage & Knapman 2003). However, giving the
the idea of a national drug chart across all Aus- extra responsibility of measuring individual
tralian hospitals as a way to reduce medication doses to pharmacists may act as transference of
errors related to documentation (Leach 2006). the problem, rather than a long term solution.
A study conducted by Leach in 2006 showed a Mathematical proficiency has been identified
marked reduction in documentation errors (of as an area requiring improvement at the tertiary
allergies, dose and frequency) and reduced pre- level (Rainboth & DeMasi 2006), which in-
scription errors when 31 hospitals used a stan- dicates the need for better mathematical or
dardised drug chart. This national drug chart numeracy training prior to commencing an
patient, it is important that all errors are ade- Healthcare [ACSQH] (2002) Second National
quately recognised and specific nursing strate- Report on Patient Safety: Improving Medication
gies are implemented to reduce medication Safety, Australian Commission on Safety and
error rates. Although nurses bear a lot of res- Quality in Healthcare, Canberra.
ponsibility and are frequently targeted as the Australian Commission on Safety and Quality in
primary cause of the error, it must be realised Healthcare [ACSQH] (2004) Charting the
that in the majority of cases, the error can be Safety and Quality of Healthcare in Australia,
attributed to ‘the systems, rather than the indi- ACSQH, accessed at http://www.safetyand
vidual’ (Beyers 2000: 56; Anderson & Webster quality. org/internet/safety/publishing.nsf
2001). /Content/charting on August 29 2007.
Many strategies have already been imple- Australian Commission on Safety and Quality in
mented in Australia and many are nursing tar- Healthcare [ACSQH] (2005) National
geted. For example adequate incident reporting Inpatient Medication Chart (NIMC) –Version E,
schemes (Anderson & Webster 2001), enhanced ACSQH, accessed at http://www.health.gov
mathematical preparation (Rainboth & DeMasi .au/internet/safety/publishing.nsf/Content
2006) and a nation-wide common drug chart /80A0EF37F281A8D7CA25718F000CCC2
(Leach 2006). The strategies that have been F/$File/chverse.pdf on November 20, 2008.
implemented targeting the reduction medica- Australian Nursing Federation [ANF] (2005)
tion error rate have been quite successful, and it National Consensus Statement on Nurses and Life-
is envisaged that they will continue to do so. long Learning, Canberra, Australia.
However, due to the nature of nursing and med- Barr D (1955) Hazards of modern diagnosis and
ical care – both human and technological, the therapy – the price we pay, JAMA 159(15):
risk of error can theoretically never reach zero 1452–1456.
(Leape 1995b). Bates D, Leape L, Cullen D, Laird N, Petersen L,
Teich J, Burdick E, Hickey M, Kleefield S, Shea
Acknowledgement B,Vander M and Seger D (1998) Effect of
Thank you to Malcolm Elliott and Jane Walker computerized physician order entry and a team
for their continual support and assistance with intervention on prevention of serious medica-
editing. tion errors, JAMA 280(15): 1311–1316.
Bayne T and Bindler R (1988) Medication calcu-
References lation skills of registered nurses, The Journal
Abushaiqa M, Zaran F, Bach D, Smolarek R, and of Continuing Education in Nursing 19(6):
Farber M (2007) Educational interventions 258–262.
to reduce use of unsafe abbreviations, Ameri- Benner P, Sheets V, Uris P, Malloch K, Schwed K
can Journal of Health Systems Pharmacy 64(11): and Jamison D (2002) Individual, practice
1170–1173. and system causes of errors in nursing: a
Anderson D and Webster C (2001) A systems taxonomy, Journal of Nursing Administration
approach to the reduction of medication 32(10): 509–523.
error on the hospital ward, Journal of Advanced Beyers M (2000) About a four-point plan to
Nursing 35(1): 34–41. reduce medication errors, Nursing Management
Armitage G and Knapman H (2003) Adverse 31(2): 56.
events in drug administration: a literature Billings C and Woods D (2001) Human error in
review, Journal of Nursing Management 11(2): perspective: the patient safety movement,
130–140. Postgraduate Medicine Online 109(1).
Australian Commission on Safety and Quality in Bindler R and Bayne T (1991) Medication