Sei sulla pagina 1di 7

ARTICLE

Comparison of
documentation of patient
reported adverse drug
reactions on both paper-
based medication charts and
electronic medication charts
at a New Zealand hospital
Wilson Shen, Bernice Wong, Jessica Yi Ping Chin, Michael Lee, Carolyn
Coulter, Rhiannon Braund

ABSTRACT
AIM: Known adverse drug reactions (ADRs) can have profound effects on disease states, as well as
prescribing practice. Therefore, the correct and complete documentation of each individual patient’s ADR
history, upon hospital admission, is important in optimising that individual patient’s pharmacotherapy.
This study investigated the documentation of ADRs at a tertiary New Zealand hospital, on both paper-
based medication charts and electronic medication charts to quantify both the number of ADRs patients
self-report, as well as the differences between recording of that information in electronic and paper-based
charting systems.
METHOD: Following ethical approval, inpatient medication charts on the general medical ward (electronic
prescribing), or the general surgical ward (paper-based medication charts) were viewed for documented
ADRs—as reported by each patient on admission. Consecutive patient charts (and electronic clinical
management system) were viewed until 50 patients from each ward, each with at least one documented
ADR, (in any of the information sources) were obtained. Patient demographic information, ADR history and
discrepancies between information sources were determined.
RESULTS: In both wards 114 patients were reviewed in order to find 50 patients with documented ADRs.
In the medical ward (electronic) 44 (90%) patients had discrepancies in ADR information between different
information sources and in the surgical ward (paper) this occurred in 49 (98%) patients.
CONCLUSION: A large number of patients self-report ADRs. Full documentation of patient reported ADRs
is required to adequately inform future prescribing decisions. Discrepancies between ADR information
recorded in different information systems exist, but information sharing between electronic and
non-electronic sources could be prioritised in order to allow full and complete information to be collected,
stored and utilised; and reduce the current inadequacies.

The correct documentation of a history of [for] therapy of disease”1–2 and can span
true immune-mediated allergic reactions to from mild annoyances, such as indigestion,
medications can play an important role in to medical emergencies including anaphy-
the decision making process of prescribing laxis. ADRs are often poorly understood by
for that patient’s current condition. ADRs patients leading to both over and under-re-
are defined by WHO as “a response to a porting, and are often misdiagnosed and
drug that is noxious and unintended which incompletely categorised by doctors and
occurs at doses normally used in man … other healthcare professionals, both which

92
NZMJ 28 October 2016, Vol 129 No 1444
ISSN 1175-8716 © NZMA
Te Hauora mō ngā Iwi Katoa
www.nzma.org.nz/journal
ARTICLE

can lead to inaccurate and incomplete contraindicated medication is prescribed,


documentation. Inadequate reporting and as well as detailing severe intolerances,
documentation of ADRs can lead to greater to aid appropriate prescribing. Within
morbidity and mortality, with increased many health care settings, electronic
patient suffering, predominantly due to systems are changing the way medica-
limited prescribing choices potentially tions are prescribed, administered and
leading to poor therapeutic responses.3 dispensed. While illegibility errors have
Deviations from first line treatments and been almost eliminated by these systems
international guidelines, largely due to there are still issues surrounding the
possible adverse reactions, have been shown completeness of information held, including
to lead to suboptimal therapy, additional ADR reporting.10–11 A study conducted in a
expenditure and potentially additional hospital with highly computerised infor-
medication errors.3–5 mation systems found that high rates of
A previous study found that almost half ADRs occurred, including where previous
of all patients, within the hospital setting, similar ADRs were recorded, despite the
self-report some type of drug “allergy” with development and implementation of elec-
the majority being an ADR to penicillin or tronic systems, and found that this was due
an opioid.6 As patients may report expected to factors that are still open to human error,
acceptable adverse effects as “allergies”, such as drug selection and drug dosing.12
there is a disconnect between what a patient The aim of this study was to identify
considers an “allergy” and an allergy that discrepancies in the recording of ADR
restricts prescribing choices. This may information between electronic medication
cause an acceptable first-line treatment to charts and paper-based medication charts
be withheld—one that would not actually to highlight if resources are required to
have caused true harm to the patient, and minimise this potential problem; further
an inferior alternative to be utilised.7 The to identify the number and type of medica-
nature and severity of an ADR experienced tions implicated and to determine whether
will influence the decision of prescribing these varied between the ward type (general
appropriateness on future occasions, surgical versus general medical).
however this decision can only be accurate
when based on a reasonable minimum level Method
of information. It is no longer acceptable
practice to just record the name of the
Study context
The study hospital is a tertiary New
agent implicated, rather the advantages of
Zealand hospital with approximately 400
documenting the drug thought to be impli-
beds. A number of electronic and paper-
cated, type of reaction, severity and date
based systems are utilised within this
of occurrence have been recognised.1,4 The
hospital to manage patient information
importance of documenting this infor-
and medical notes. The central electronic
mation has been highlighted in publications
integrated clinical management system
from the American Medical Association8
(CMS), gathers information from other
and locally in the Health Safety and Quality
electronic systems and acts as the primary
Commission documents9 which reinforce this
electronic tool for medical staff to monitor
process by physicians as part of medicines
patient progress and retrieve medical
reconciliation at points of transfer of care,
notes. Patient information contained in
given the significant implications for patient
the CMS may include history of admissions
safety as well as appropriate patient therapy.
and discharges, laboratory tests, X-rays and
The increasing use of technology, and ADR documentation.
in particular electronic records, has the
Prescribing of medications at this
potential to assist in capturing more
hospital is split between traditional paper
detailed ADR information. This may range
drug charts and an electronic prescribing
from simple prompts (in stand-alone
software system (EPSS); the rollout of the
systems) to retained patient profiles—where
EPSS is being conducted in a systematic
information captured can pre-populate
manner with the addition of individual
electronic prescribing records, including a
wards in turn. All patients have a file in the
real-time allergy alert when a potentially

93
NZMJ 28 October 2016, Vol 129 No 1444
ISSN 1175-8716 © NZMA
Te Hauora mō ngā Iwi Katoa
www.nzma.org.nz/journal
ARTICLE

CMS, irrespective of which ward they are documented, the patient was excluded. This
admitted to and the prescribing system used. process was continued until 50 patients from
There are several ways that ADRs may be each specialty area (surgical or medical)
recorded; 1) hand written onto the paper- were included in the study.
based medication chart; 2) entered into the Demographic information such as gender,
EPSS; 3) entered into the CMS system as part age and the number of documented ADRs for
of patient notes; or 4) CMS can “import” ADR each patient was recorded. Where possible,
information from the EPSS—however, the the implicated drug or drug class was
synchronicity of information between these recorded, and where documented, the mani-
two electronic systems is not always complete. festation of the ADR. Additionally, the date of
ADRs may be recorded from the patient the ADR was recorded (if documented).
self-reporting these at time of admission, or Identification of discrepancies occurred
from clinicians accessing previous records when agents and associated reactions were
(electronic and paper based). not the same in each information source.
Data collection These discrepancies occurred because one
Following ethical approval, medication information source may have stated the
charts from patients on the two general agent, whereas the other source may have
surgical wards (that use paper-based had the agent and the level of reaction (or
prescribing charts) were viewed and medi- even the date). Discrepancies like these
cation charts from patients on the two meant that completeness of information
general medical wards (that use electronic viewed depended on the information
prescribing) were viewed. The ADR infor- source used by the clinician. For patients
mation from each patient chart was recorded who had discrepancies, the information
and the ADR information for each patient source with the highest quantity of infor-
was subsequently reviewed on the CMS. mation was determined (by comparison of
the information) and identification of the
If any of the information sources had one
source that provided the largest number
or more ADRs documented, the patient was
of reported ADRs. ADRs reported were not
included in the study, but if no information
independently investigated and verified for
was documented in any of these sources
accuracy during the study.
or ‘no known drug allergies’ or ‘NKDA’ was

Figure 1: Age distribution of patients with ADRs from the medical () and surgical wards ().

94
NZMJ 28 October 2016, Vol 129 No 1444
ISSN 1175-8716 © NZMA
Te Hauora mō ngā Iwi Katoa
www.nzma.org.nz/journal
ARTICLE

Figure 2: Number of medication classes implicated in ADRs from patients in the medical () and surgi-
cal wards ().

Results other analgesics, as well isolated reports


from varied agents such as allopurinol, alen-
In both wards 114 patients were reviewed dronate, oxybutynin, metformin, digoxin,
to find 50 patients with at least one docu- quinine and more.
mented ADR. However, upon revision of the
In the medical wards, (EPSS) 44/49 (90%)
data, one patient from the medical ward had
patients had discrepancies identified
been duplicated, bringing the number in this
between their chart and CMS regarding their
group to 49.
ADR history. On the surgical wards, (paper
The gender split between the wards was chart) 49/50 (98%) patients had identified
very similar; the medical wards were 51% discrepancies. Of the 44 patients’ documents
females and the surgical wards were 52% in the medical wards, the electronic chart
females. The age distribution is shown in held the most information for 33 (75%)
Figure 1, the difference in the age distribu- patients, CMS stated a more complete ADR
tions most likely reflects the ward specialty. history for 10 (23%) patients; and for one
The number of documented ADRs to a patient who had two different drug reac-
specific drug class or substance is shown in tions, one appeared on the electronic chart
Figure 2. These appear similar between the and one on CMS. In the surgical wards, the
two types of wards where the majority of paper drug chart held the most information
patients have only a few (one, two or three) 31/49 times (63%), CMS had the most infor-
documented ADRs. Of note is that one outlier mation 15/49 times (31%) and the content
had 11 documented ADRs to specific agents. differed but the volume was the same in
Where the specific ADR was documented, 3/49 (6%) patient profiles.
the agent responsible was tabulated (see
Table 1) as well as the reaction being Discussion
recorded (see Table 2). This study found that there are high
Antibiotics represented the largest numbers of ADRs reported and documented
contributor with n=85 of the 197 agents within the hospital setting, which is positive
(43%), followed by penicillins n=51 (26%) in that it indicates that prescribers (or
and opioids n=15 (8%). The ‘other medicines’ other healthcare professionals) recognise
included predominantly single reports from the importance of this information and its
many classes such as statins, beta-blockers, documentation. This incidence is higher
calcium channel blockers, antiemetics and than reported in other studies—only 228

95
NZMJ 28 October 2016, Vol 129 No 1444
ISSN 1175-8716 © NZMA
Te Hauora mō ngā Iwi Katoa
www.nzma.org.nz/journal
ARTICLE

Table 1: Agent documented in recorded adverse drug reactions.

Medical ward electronic Surgical ward paper-based

Penicillins 25 26

Cephalosporins 7 1

Sulfur-containing antibiotics 5 1

Other antibiotics 9 11

Opioids 9 6

NSAIDs 7 9

ACE inhibitors 2 2

Diuretics 2 1

Other medicines 26 30

Iodine and contrast media 3 0

Non-drug allergy* 7 8

Total 102 95
*Includes foods, sticking plasters.

Table 2: Adverse drug reaction detail recorded.

Medical ward electronic Surgical ward paper-based

Rash and/or urticaria 19 21

Nausea and/or vomiting 10 6

Anaphylaxis 6 7

Misc GI upset* 3 8

Swelling or angioedema 6

Headache 5

GI bleed* 4

Hypertension 4

Fever 4

Flushing 2 2

Shortness of breath 3

Cough 3

‘Unwell’ 3

Diarrhoea 3

Palpitations/tachycardia 3

Asthma/bronchospasm 2

Tight chest 2

Other 14 14

Total 86 68
*GI = gastrointestinal
*Misc = miscellaneous

96
NZMJ 28 October 2016, Vol 129 No 1444
ISSN 1175-8716 © NZMA
Te Hauora mō ngā Iwi Katoa
www.nzma.org.nz/journal
ARTICLE

patients needed to be screened in order to be improved by better importing of one


obtain a total of 100 participants, which is database into another. For example, all
44% of patients having a documented ADR. documented ADRs on CMS should automat-
Previous studies have found that ADRs are ically populate the electronic medication
present in 10–20% of hospitalised patients.13 chart. In this way the patient management
The medications implicated in these ADRs system is a permanent historical record of
were similar between the two wards and previously reported and witnessed ADRs.
are broadly representative of the medi- The option of a ‘drop-down’ or reminder
cations known to be implicated in drug box to check for ADRs when compiling the
allergies due to their potential for allergic electronic medication chart should assist
reaction and/or common usage eg opioids in keeping this information current and
and β-lactams. The percentages reported has the added advantage of date-stamping
differ slightly from other published liter- when the event was entered, as well the
ature;6 this study found lower reports of healthcare professional’s name and their
ADRs to opioids (7.6% vs 19%), comparable designation (eg pharmacist) is retained with
reports to penicillin (26% vs 27%) and the entered ADR. Additionally, the fact that
higher reports of ADRs to antibiotics overall the prescriber has to physically transcribe
(43% vs 37%). The lower rate to opioids the ADR information held in the patient
may be due to practitioners recognising management system onto the paper chart
that mild type A ADRs, eg nausea, as well as is fraught with the potential for ADRs to be
mild itch, due to histamine release, are not overlooked or mis-transcribed.
a contraindication to rechallenge with these It is important to acknowledge that
agents. There were differences noted in the no system is perfect and that electronic
types of reactions documented between the entering of data is not always fool-proof,
wards and this may be due to differences in and with the implementation of electronic
the specialty demographics. For example, recording an erroneous ADR may be perpet-
the medical wards had patients who were uated. Physicians have identified that the
significantly older, as older people are less fragmented nature of accessing electroni-
likely candidates for surgery; the impact of cally stored information causes details to be
this may warrant further investigation. It missed. Also, communication and coor-
was positive to see that the ‘level’ of reaction dination processes can lead to situations
had been documented for many patients, where information entered into the system
showing that healthcare professionals is not accessed by other practitioners.14 This
recognised the implication of differential fragmentation of information was seen
reactions. For example, the demarcation when one information source reported
of an anaphylactic reaction from that of a one drug reaction and a different source
simple rash can assist in deciding whether a reported another, which did occur with
patient can be rechallenged with caution or several patients.
if there is a true contraindication.
Lastly, these electronic systems are only
There are some limitations to this study: tools and healthcare professionals must
firstly it is important to note that in this use their professional clinical judgement
study there was no threshold limit on when deciding on a treatment, particularly
severity of ADR or confirmation of ADR, if the patient has had a previous reaction.
secondly the two wards represent different Until electronic systems become better
specialties as well as different systems aligned with greater sharing of infor-
which makes it difficult to compare directly mation, prescribers cannot simply use the
and lastly the numbers of patients limit the patient’s medicine chart as the information
ability to perform further statistical eval- repository for noting ADRs and prescribing
uation. However, this exploratory piece of medications. This highlights the ineffi-
work has highlighted the lack of adequacy ciencies in having to check multiple patient
of recording of ADR information and that information sources and the potential for
discrepancies can occur. patients to inadvertently receive medi-
The differences between the elec- cations that they have previously had an
tronic charts and the electronic patient adverse reaction to.
management systems should be able to

97
NZMJ 28 October 2016, Vol 129 No 1444
ISSN 1175-8716 © NZMA
Te Hauora mō ngā Iwi Katoa
www.nzma.org.nz/journal
ARTICLE

Conclusion a patient management system, but this


is only of use if this information can be
Almost half of the patients in this study accessed at the time of prescribing. Discrep-
reported ADRs. However, for this infor- ancies between ADR information stored
mation to translate into the decision to in different systems exist, but information
modify prescribing behaviour, full infor- sharing between electronic and non-elec-
mation regarding the nature and severity of tronic sources could be prioritised in order
the reaction is needed. The potential exists to allow full and complete information to
for more detailed information regarding be collected, stored and utilised; and reduce
the nature of the ADR to be recorded in the current inadequacies.

Competing interests:
Nil.
Acknowledgements:
The authors wish to thank the staff at the study hospital for their assistance with data access
and collection.
Author information:
Wilson Shen, School of Pharmacy, University of Otago, Dunedin; Bernice Wong, School of
Pharmacy, University of Otago, Dunedin; Jessica Yi Ping Chin, School of Pharmacy, University
of Otago, Dunedin; Michael Lee, School of Pharmacy, University of Otago, Dunedin; Carolyn
Coulter, Pharmacy Department, Dunedin Hospital, Dunedin; Rhiannon Braund, Associate Pro-
fessor, School of Pharmacy, University of Otago, Dunedin.
Corresponding author:
Rhiannon Braund, Associate Professor, School of Pharmacy, University of Otago, Dunedin.
rhiannon.braund@otago.ac.nz
URL:
http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2016/vol-129-no-1444-
28-october-2016/7051

REFERENCES:
1. Radford A, Undre S, 6. Khalil H, Leversha A, cians’ experiences using
Alkhamesi NA, et al. Khalil, V. Drug allergy commercial e-prescribing
Recording of drug documentation—time systems. Health Affairs.
allergies: are we doing for a change? Int J Clin 2007; 26:393–404.
enough? J Eval Clin Pharm. 2011; 33:610–613. 11. Kaushal R, Kern LM,
Pract. 2007; 13:130–137.
7. Turner RD. (2006) Are Barron Y, et al. Electronic
2. Edwards IR, Aronson JK. we aware of hospital prescribing improves
Adverse drug reactions: patients’ drug allergies? medication safety in
definitions, diagnosis, J Clin Pharm Ther. community-based office
and management. Lancet.
2006; 31:649–650. practices. J Gen Intern
2000; 356:1255–1259.
8. The American Medical Med. 2010; 25:530–536.
3. Sastre J, Manso L,
Association. The physi- 12. Nebeker JR, Hoffman JM,
Sanchez-Garcia S, et al.
cians role in medicines Weir CR, et al. High rates
Medical and economic
reconciliation—issues, of adverse drug events
impact of misdiagnosis
of drug hypersensitivity strategies and safety in a highly computerized
in hospitalized patients. principles. https://bcpsqc. hospital. Arch Int Med.
J Allergy Clin Immunol. ca//documents/2012/09/ 2005; 165:1111–1116.
2012; 129:566–567. AMA-The-physi- 13. Gomes ER, Demoly P.
cian%E2%80%99s-role- Epidemiology of hyper-
4. Picard M, Begin P, Bouch-
ard H, et al. Treatment in-Medication-Recon- sensitivity drug reactions.
of patients with a history ciliation.pdf. [accessed Curr Opin Allergy Clin
of penicillin allergy in 20 Sept 2014]. Immunol. 2005; 5:309–316.
a large tertiary-care 9. Health Safety and Quality 14. Ash JS, Berg M, Coiera
academic hospital. J Commission New Zealand. E. Some unintended
Allergy Clin Immunol Medicines Reconciliation. consequences of infor-
Pract. 2013; 1:252–257. http://www.hqsc.govt.nz/ mation technology in
5. Jones TA, Como JA. our-programmes/medi- health care: the nature of
Assessment of medication cation-safety/projects/ patient care information
errors that involved drug medicine-reconciliation/. system-related errors. J
allergies at a university [accessed 20 Sept 2014]. Am Med Inform Assoc.
hospital. Pharmacother-
10. Grossman JM, Gerland 2004; 11:104–112.
apy. 2003; 237:855–860.
A, Reed MC, et al. Physi-

98
NZMJ 28 October 2016, Vol 129 No 1444
ISSN 1175-8716 © NZMA
Te Hauora mō ngā Iwi Katoa
www.nzma.org.nz/journal

Potrebbero piacerti anche