Professional Documents
Culture Documents
DOI 10.1007/s00520-004-0622-5
Phyllis M. Lau
Kay Stewart
Michael Dooley
ORIGINAL ARTICLE
Introduction
In the practice of oncology, adverse drug reactions
(ADRs) of cancer treatments have become almost synonymous with the treatments themselves. The low therapeutic index of chemotherapeutic agents and the predictable and common adverse events of cancer treatments
mean that these events are seen as an unavoidable com-
627
found that the incidence of ADRs is high and is dominated by reactions that are predictable and preventable [5,
10, 17, 24, 26]. From their review of Australian studies,
Roughead and colleagues found that between 32 and 69%
of drug-related admissions reported were classified as
definitely or possibly preventable [22].
The magnitude of ADRs endured by oncology patients
is significant. The 1992 Australian National Hospital
Morbidity Data (excluding NT) reported that 11% of
ADRs in Australian hospitals were associated with antineoplastic and immuno-suppressive drugs [19]. The 1995
Quality in Australian Health Care Study, which reviewed
medical records of over 14,000 admissions to 28 hospitals
in New South Wales and South Australia, also implicated
anti-neoplastics amongst the most common agents responsible for medication-related hospitalisations [27].
Advances in the development of supportive care in oncology, for example, 5HT3 antagonists for the control of
chemotherapy-induced nausea, have led to reductions in
drug-related toxicity. However, there is a need to quantify
the frequency and severity of ADRs experienced by oncology patients to enable the development and implementation of intervention strategies. Appropriate and adequate use of preventative measures to combat ADRs and
the optimal management of these reactions in cancer patients are important issues in the quality use of medicines.
The impact of ADRs on the well-being of cancer patients has been a much discussed topic in both public and
medical arenas. There is evidence to show that medical
perspectives do not necessarily coincide with those of
patients who are being treated[8, 9]. Patients are more
likely to report bothersome but not necessarily clinically
important symptoms [12, 16]. It is important that patients
perspectives and psychosocial issues be considered when
attempting to identify, prioritise and develop strategies to
combat the problem [15]. Discrepancies between patients
and clinicians perception of what constitutes a significant
adverse reaction is worthy of further investigation.
Strategies to reduce the burden of ADRs and to improve the quality use of medicines in health care are
clearly needed [14]. The British Audit Commissions
2001 report Spoonful of Sugar highlighted the important proactive role that pharmacists play in medicine
management [2]. The 2002 report by the Clinical Oncological Society of Australia, The Cancer Council Australia and the National Cancer Control Initiative similarly
highlighted the contribution of pharmacists to the quality
of cancer treatment and patient care [7]. An increased
understanding of the prevalence and nature of ADRs,
as well as an appreciation of patients perspectives on
quality of life, would enable us as a profession to identify and target specific areas for vigourous pharmaceutical
care.
The aim of this study was to assess the incidence,
predictability, preventability and severity of ADRs in hos-
628
Table 1 Naranjo algorithm or adverse drug reaction probability scale. The total score calculated from this table defines the category as:
Possibly (total score 14), Probably (total score 58), Definitely (total score >9)
Data collection
The study involved both ongoing patient interviews and review of
patient medical records.
Patient interview
Patients were interviewed on alternate days during their hospital
stay and asked questions about adverse symptoms as a result of
drug treatment. Symptoms included those present on admission,
regardless of whether they related to reasons for admission, and
those that occurred during hospitalisation. Patients were also asked
to rate on a seven-point scale how they thought each symptom
impacted on their well-being. Labels were used to define only the
extreme categories with 0 being no or hardly any impact and 6
being totally changed my life [13].
Review of patient medical records
Patient medical records were reviewed daily to identify possible
ADRs. All symptoms reported during interviews and in medical
records were scrutinised for possible relationship to drugs by extensive review of the patients medical records, which included past
and present medical history, past medication histories, results of
any investigative or assessment procedures, notes recorded by
members of the attending medical team, medication administration
charts, and prescription records. Patients health status, diseases and
disease progression were all noted during the assessment of each
symptom.
For each ADR, a complete search on the ADR and the drug(s)
involved was accessed through the on-line MICROMEDEX
Health-care Series [25]. Information found or not found was verified with other resources, namely product information (PI), Australian Adverse Drug Reactions bulletins prepared quarterly by the
Australian Adverse Drug Reaction Advisory Committee (ADRAC)
and published by the Therapeutic Goods Administration (TGA),
PREMEDLINE and MEDLINE databases [21]; American Hospital
Formulary Service Drug Information (AHFS DI) [1]; Australian
Prescription Products Guide (APPG) [4]; and Meylers Side Effects
of Drugs [11].
Assessment algorithm and criteria
ADRs identified either from patient interview or from patient
medical records review alone, were assessed for causality, predictability, preventability and severity.
Yes
No
Do not know
+1
+2
+1
0
1
0
0
0
0
+2
1
1
+1
+1
1
+2
+1
0
0
0
0
0
0
0
+1
+1
0
0
0
0
Incidence description
Predictability
1/10
1/100 and <1/10
1/1000 and <1/100
1/10,000 and <1/1000
<1/10,000
Very common
Common
Uncommon
Rare
Very rare
Predictable
Predictable
Not predictable
Not predictable
Not predictable
Causality criteria
The Naranjo Algorithm or Adverse Drug Reaction Probability
Scale (Table 1) was used to evaluate the causality relationship
between a likely ADR and a drug [18]. Only symptoms with a
Naranjo score of 1, that is, at least possibly related to one or more
drugs, were recorded as ADRs, which were classified into three
categories: possibly (Naranjos scores 14), probably (Naranjos
score 58) or definitely (Naranjos score >9) related to a drug. They
were then assessed for predictability and preventability.
Predictability criteria
The Council for International Organizations of Medical Sciences
guidelines for preparing core clinical-safety information on drugs
were adapted for assessment of predictability (Table 2) [6]. Product
Information and references including the on-line MICROMEDEX
Health-care Series, ADRAC On-Line, MEDLINE, AHFS drug information, APPG and Meylers Side Effects of Drugs were used to
determine the documented incidence rate of an ADR.
Preventability criteria
The criteria of Schumock and Thornton were modified to assess
and categorise ADRs into definitely, probably or not preventable
(Table 3) [23]. The factor toxic serum drug level was omitted
from the original criteria, and two questions (4 and 5, section B) on
the use of preventative measures were added.
629
Proportion %
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
24.4
16.7
16.1
11.9
9.5
8.3
3.6
4.8
0.6
1.8
0.6
0.6
0.6
0
0
0
0
0
0.6
Proportion %
Breast
Haematological
Gastrointestinal
Lung
Dermatological
Head/neck
Urology
Sarcoma
Gynaecological
Brain
Others
Unknown primaries
22.3
14.0
13.4
12.5
11.5
8.5
6.2
4.9
4.0
1.3
1.1
0.4
Results
Interview sample
Patients from 276 admissions were selected to be approached for interviews. Of these, 194 were available to
be informed about the study and 183 consented to participate. In total, 167 patients from 171 (93.4%) admissions were interviewed. The interview sample comprised
admissions of 87 women and 80 men whose ages ranged
from 20 to 92 years with an average of 59.014.8 years.
The proportions and types of cancer at the time of admission are illustrated in Table 4.
630
Preventability of ADRs
Over half (53.4%) of ADRs were classified as not preventable, 45% probably preventable and 2% definitely
preventable. For the ten most common ADRs, only three
cases were assessed as definitely preventabletwo cases
of drowsiness, which were caused by inappropriate opiate
doses, and one case of diarrhoea, which was caused by
simultaneous use of three laxatives. High proportions of
constipation, nausea vomiting, diarrhoea and mucositis
were assessed as probably preventable. Alopecia and
anorexia were always assessed as not preventable (Fig. 3).
Patients impact rating of the ten most common ADRs
Fig. 2 Predictability of the ten most common adverse drug reactions (ADRs)
Predictability of ADRs
Eighty-eight percent of the ADRs were assessed as predictable. The predictability of the ten most common
ADRs closely matched the frequency (Fig. 2).
Patients perception of how ADRs impacted their wellbeing ranged from no impact at all to totally changed
my life, with an average of 3.5 on a seven-point scale.
The proportions of patients who responded in each category are shown in Table 6. Not all patients were able to
rate their ADRs. This question was not pursued further in
interviews when patients expressed difficulty. Except for
myelosuppression and alopecia, high proportions of the
common ADRs were rated above mid-point on the scale.
Patients who responded to the question in relation to
631
Proportion %
7.0
12.7
11.0
15.9
17.2
16.8
19.3
Discussion
It is clear from the results that ADRs are common in
hospitalised oncology patients. Extrapolating the results
from the study, about three quarters of admissions to Peter
MacCallum Cancer Centre have at least one ADR either
on admission or during hospital stay, and over 40% have
three or more ADRs. Comparatively, figures from this
study are higher than those reported in the 1992 Australian National Hospital Morbidity Data because of the
difference in methods of event identification [19]. The
national database is a collection of confidential summary
records for admitted patients separated from public and
632
633
Conclusions
ADRs are common in hospitalised oncology patients,
predictable in many instances but definitely preventable
in only a few. There are, however, many occasions where
improved use of preventative measures has the potential
References
1. American Hospital Formulary Service
drug information (1984) Board of Directors of the American Society of
Hospital Pharmacists, Bethesda
2. Audit Commission (2001) A spoonful
of sugarmedicines management in
NHS hospitals. Audit Commission,
London
3. Australian adverse drug reactions bulletin (1998) Australian Adverse Drug
Reaction Advisory Committee
(ADRAC)
4. Australian prescription products guide
(2000) 29th edn. Australian Pharmaceutical Publishing
5. Bates D, Leape L, Petrycki S (1993)
Incidence and preventability of adverse
drug events in hospitalized adults.
J Gen Intern Med 8:289294
6. Council for International Organisation
for Medical Science (CIOMS) (1994)
WGI, Guidelines for preparing core
clinical-safety information on drugs.
Council for International Organisation
for Medical Sciences, Geneva
7. Clinical Oncological Society of Australia (2002) Optimising cancer care in
Australia. The Cancer Council Australia
and the National Cancer Council Initiative, Melbourne, pp 1122
8. Coates AS, Abraham SB, Kaye T et al
(1983) On the receiving endpatient
perception of the side-effects of cancer
chemotherapy. Eur J Cancer Clin Oncol
19:203208
9. Dewland P, Dewland J (1999) At the
coalface, but on the receiving end.
J Med Ethics 25:541546