Supportive Care in Cancer

https://doi.org/10.1007/s00520-018-4225-y

ORIGINAL ARTICLE

The timeliness of patients reporting the side effects of chemotherapy

Ian Olver 1,5 & Mariko Carey 2,3,4 & Allison Boyes 2,3,4 & Alix Hall 3,4 & Natasha Noble 2,3,4 & Jamie Bryant 2,3,4 &
Justin Walsh 2,3,4 & Rob Sanson-Fisher 2,3,4

Received: 24 November 2017 / Accepted: 25 April 2018

# Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract
Purpose To explore the actions cancer patients reported they would take in response to a range of common side effects of
chemotherapy and whether these were considered appropriate based on current guidelines and evidence; and to explore the
sociodemographic and cancer-related variables associated with patients selecting the appropriate action (immediate medical
attention or reporting) for two potentially life-threatening side effects: fever, and unusual bleeding and bruising.
Methods Four hundred thirty-six medical oncology and haematology patients receiving chemotherapy completed two
surveys to provide demographic, disease and treatment characteristics, and details on how they would respond if
they experienced a range of specified side effects of chemotherapy (for example, nausea and vomiting, fatigue, and
skin rash or nail changes). The proportion of patients reporting the appropriate action for each side effect was
calculated. Multiple logistic regressions examined the patient demographic and cancer characteristics associated with
selecting the appropriate action (seeking immediate medical attention) for two potentially life-threatening side effects
of chemotherapy: high fever of 38 °C or more, and unusual bleeding or bruising.
Results Two thirds of patients indicated that they would seek immediate medical attention for high fever (67%), but only 41%
would seek immediate attention for bleeding or bruising. Cancer type and time since diagnosis were significantly associated with
patients indicating that they would seek immediate medical attention for high fever; while time since diagnosis was the only
variable significantly associated with patients reporting that they would seek immediate medical attention for unusual bleeding or
bruising. For chronic side effects, like skin rash or nail changes, and tingling or numbness, which usually do not require urgent
reporting, only between 12 and 16% would report them immediately. A significant proportion of patients reported that they
would Bdo nothing^ about fatigue or tiredness (24%). By comparison, less than 10% patients reported that they would do nothing
for the other side effects investigated.
Conclusions Tools need to be created so that patients better understand the side effects after being treated with chemotherapy and
what action they should take.

Keywords Chemotherapy . Side effects . Timeliness . Self-reporting . Quality of life

Relevance Understanding how patients intend to report side effects

allows the development of educational tools to help patients better
understand the side effects of chemotherapy and the need to promptly
report potentially life-threatening side effects.

* Ian Olver 3
Priority Research Centre for Health Behaviour, University of
[email protected] Newcastle, Callaghan, NSW, Australia
4
Hunter Medical Research Institute, New Lambton Heights, NSW,
1
Sansom Institute for Health Research, University of South Australia, Australia
Adelaide, SA 5000, Australia 5
Sansom Institute for Health Research, University of South Australia,
2
Health Behaviour Research Collaborative, School of Medicine and P7-17 Playford Building, City East Campus, GPO Box 2471,
Public Health, Faculty of Health and Medicine, University of Adelaide, SA 5001, Australia
Newcastle, Callaghan, NSW, Australia
 Support Care Cancer

Introduction The aims of this study were:

Chemotherapy is an important component of cancer treatment (a) To report the actions patients perceived they would take
and has contributed to improvements in cancer survival rates. in response to a range of common side effects of
Chemotherapy is associated with a range of side effects, with chemotherapy;
nausea and vomiting, fatigue, anaemia, hair loss, and changes (b) To explore the sociodemographic and cancer-related var-
in taste and smell being among some of the most common iables associated with patients selecting the appropriate
experienced [1]. action (immediate medical attention or reporting) for the
It is important that patients report adverse effects from che- two potentially life-threatening side effects of fever, and
motherapy to their care provider. Some side effects may be able unusual bleeding and bruising.
to be prevented or reduced, such as nausea and vomiting
through the use of antiemetic prophylaxis [2]. Other side effects
such as anaemia may require an adjustment in dosage or inter-
val for subsequent cycles of treatment but are less urgent [2].
Some serious side effects such as fever, infection, and unusual Methods
bruising or bleeding should be reported immediately to the
health care team because of their association with prolonged Setting The study was conducted in three medical oncology
hospitalisation, reduced quality of life, and death [3–5]. clinics (located in Victoria, Tasmania and Western Australia)
The reasons for some patients not reporting side effects and three haematology clinics (located in Queensland, New
may relate to a lack of awareness or education [6]. Many of South Wales and Victoria) in Australia. Clinics were in met-
the side effects of chemotherapy are experienced at home, ropolitan areas and all were public hospitals. Ethics approvals
particularly as hospital stays become shorter [7] and most were granted by the University of Newcastle Ethics
recently patients are increasingly being treated as day patients. Committee (H-2010-1324), the Cancer Institute NSW
Therefore, patients and their families must be aware of what Population & Health Services Research Ethics Committee
side effects to expect, and how to manage them, including (ref. 2011/10/351), and the relevant hospital ethics
when to seek medical advice [8]. Yet many patients and their committees.
families may regard side effects as inevitable and not be aware
that adverse effects can be alleviated. Some patients believe Sample Eligible patients were those attending a participating
that Bgood^ patients do not complain, or they do not want to treatment clinic who had a confirmed diagnosis of cancer,
distract their doctor from administering treatment [9]. were aged 18 years or older, were English speaking, and were
In Australia, chemotherapy education for medical oncolo- able to provide informed consent. Only participants who had
gy and haematology patients is typically provided by chemo- received chemotherapy treatment for their cancer were includ-
therapy nurses and involves both written and verbal informa- ed in the final sample. Those attending the clinic for the first
tion. Chemotherapy education usually takes place prior to time, or who were too unwell to complete the survey, were
commencing chemotherapy if possible a day or two prior. excluded.
Recall of side effects also becomes an issue if the reporting
is delayed [10]. For example, in a study by Coolbrandt et al., Procedure A research assistant provided a consecutive sample
respondents reported fewer chemotherapy side effects and of eligible patients with written information about the study
fewer severe side effects when self-report was delayed until when patients presented for their outpatient oncology appoint-
the next hospital visit, compared to when symptoms were self- ment. Informed consent was sought from all participants. Two
reported on each of the 7 days immediately following chemo- surveys were conducted in an effort to minimise clinic disrup-
therapy administration [10]. There is also a discrepancy be- tion and reduce participant burden. Participants were asked to
tween patient and clinician reports of symptoms, with clini- complete a brief paper and pencil survey in the clinic while
cians often underestimating both the number and severity of they waited for their appointment. The survey included ques-
symptoms [11]. The use of structured side effect symptom tions on sociodemographic, disease and treatment characteris-
lists rather than open-ended spontaneous reporting by the pa- tics. Those who did not have time to complete the first survey
tient can help to improve reporting of side effects [12]. For in clinic were given the option of completing it at home and
example, one study found a tenfold difference in the number mailing the survey back to the researchers using a reply-paid
of symptoms reported when a structured questionnaire was envelope supplied. Participants were asked to complete a sec-
used, compared to spontaneous patient reporting. Another ond mail out survey approximately 1 month later. The second
study reported the rates of detecting adverse drug reactions survey contained questions on self-management actions for
varied between 16% for spontaneous reporting, 24% for gen- chemotherapy side effects. Non-responders were followed
eral enquiry, to 62% for specific questioning [9, 13]. up by letter 3 and 6 weeks later.
Support Care Cancer

Measures earlier appointment but they do not, in general, require urgent

reporting [16, 17].
Demographic variables Age, gender, education, indigenous
status, marital status, country of birth, home post code, living Statistical analysis
situation, employment status, private health insurance status
and concession card status, were obtained by patient self-re- Sociodemographic, disease and treatment characteristics, and
port. Concession cards are government issued cards which patient survey responses were summarized as frequencies and
allow access to lower cost health services and medicines. percentages.
Two multivariable logistic regression analyses were under-
Disease and treatment variables Cancer type and time since taken to assess patient demographic and disease characteris-
diagnosis were assessed via patient self-report. Studies show a tics associated with patients indicating that they would call or
high level of agreement between self-reported cancer charac- go to hospital immediately for the serious side effects high
teristics and medical records [14, 15]. fever; and unusual bleeding or bruising. For the logistic re-
gressions, a binary outcome variable was defined. Selecting
Knowledge of appropriate actions to take for chemotherapy Bcall or go to hospital immediately^ was coded as correct and
side effects Participants who had had chemotherapy were pre- all other responses were coded as incorrect.
sented with the following introduction: BChemotherapy often A hypothesis-driven approach was used for the selection of
has side-effects. The following questions ask about how you patient demographic and cancer characteristics included in the
would manage chemotherapy side effects at home. You may multivariable logistic regressions: sex, age, education, conces-
not have experienced all of these side-effects. If this is the case, sion card, cancer type (which was coded as breast, colorectal,
please still tell us what you think you would do if this happened non-Hodgkin lymphoma and other based on the largest sub-
to you.^ Respondents were asked BWhat action would you take groups) and time since diagnosis and treatment centre was
if, following chemotherapy you experienced….^and were pre- accounted for in the models through the clustered jackknife
sented with a list of common side effects of chemotherapy. method [18–20]. Adjusted odds ratios (ORs) with 95% CIs
Response options included call or go to the hospital immedi- and type 3 p values are presented. Associations with p < 0.05
ately; call or go to the hospital if it has not improved after a few were considered statistically significant. Listwise deletion was
hours; call or go to the hospital if it has not improved after a few used to remove observations with missing data so that only
days; make an earlier appointment with my cancer doctor; wait participants with complete data on all relevant variables were
until my next appointment with my cancer doctor, do nothing. included in the multivariable analyses.
Participants attending medical oncology centres also had an A retrospective analysis shows that for the fever outcome
additional response category Bmake an appointment with my (130 out of 399 reported incorrect response), for covariates
GP in the next day or two.^ with 50% prevalence the study has 80% power to detect OR
Appropriate responses based on American Cancer Society of < 0.55 or > 1.82 at 5% significance. For the bleeding out-
and Cancer Council Australia advice which informs patients come: (321 out of 484 reported the incorrect response), for
and relatives about the related toxicities from evidence-based covariates with 50% prevalence the study has 80% power to
guidelines were assigned to each side effect by the expert detect OR of <0.56 or > 1.78 at 5% significance.
medical oncologist on the team (IO) [16, 17]. The descriptions
of symptoms used and time frames mirror the language com-
monly used in patient educational materials disseminated by
these reputable cancer organizations. The potentially rapidly Results
life-threatening side effects which should be reported imme-
diately are a high fever, and bleeding or bruising. Symptoms Figure 1 provides an overview of the recruitment process of
that should be reported promptly (usually within a few hours), eligible consenting participants into this study. Of the 1138
if severe, included flu-like symptoms, sore mouth, or soreness eligible patients identified, 898 (79%) provided consent, of
in the vein, because they can become serious if they progress. which 436 completed both surveys and indicated that they
Other side effects like diarrhoea, nausea and vomiting, consti- had had chemotherapy treatment and were thus included in
pation and pain can be alleviated with treatment. In order to the analysis.
improve quality of life during treatment, such symptoms There were no significant differences between consenters and
should be reported if they persist over hours or days, depend- non-consenters with regard to sex p = 0.23) or age (p = 0.53).
ing on their severity. Other side effects such as a rash or nail Just over half of participants in this study were female, aged
changes, tingling or numbness, and fatigue, are more chronic between 55 and 74 years and had a vocational training, uni-
and due to a gradual onset of cumulative toxicity after therapy. versity or other level education. The most commonly reported
Worry about the severity of these may occasion making an cancer type was breast cancer, followed by colorectal cancer.
 Support Care Cancer

Fig. 1 Overview of patient 1,601 paents screened for 463 ineligible paentsa
recruitment. Patients could be
eligibility
classified as ineligible for First me aending the clinic (n=231)
multiple reasons, thus individual Non-English speaking (n=133)
categories may not add up to total
Too unwell (n=38)
number of ineligible patients
Unable to provided consent (n=4)
1,138 eligible paents Unable to complete survey
idenfied independently (n=15)
Other reasons (n=53)

898 (79%) eligible paents

provided consent

566 (63%) consenng

paents completed both
surveys

436 paents who completed

both surveys had
chemotherapy and were
included in this study

A detailed description of the sociodemographic, disease char- days, except for a sore mouth or throat, where only 38.7%
acteristics of participants is presented in Table 1. indicated they would report it within a few days or sooner.
For potentially treatable side effects where treatment can
improve quality of life, pain was most likely to be reported
Proportion of patients reporting they would take quickly, with 56% reporting immediately or within hours.
the recommended action in response Sixty percent of participants would report diarrhoea within
to chemotherapy side effects days or sooner, but only 37.2% with constipation would con-
tact the hospital within days, and 28% would seek advice from
Table 2 describes the actions that patients reported they would their GP within days.
take in response to the range of potential chemotherapy side For more chronic side effects like skin rash or nail changes,
effects. The two side effects that can rapidly become life which usually do not require urgent reporting, only 12%
threatening and should be reported immediately are high fe- would report them immediately. By comparison, less than
ver, and unusual bleeding and bruising. Most patients (67%) 10% patients reported they would do nothing for other side
indicated that they would call or go to the hospital immediate- effects, except for fatigue or tiredness where 24% would not
ly for high fever, but only 41% would report unusual bleeding take steps to report it.
or bruising immediately.
Other side effects that should be reported within a few Characteristics associated with correct action
hours if they are severe are flu-like symptoms, a sore mouth on the most serious side effects of high fever
or throat, and persisting soreness in the vein. A sore mouth or or unusual bleeding or bruising
throat can indicate mouth ulcers, which are a potential portal
for infection and therefore need symptomatic treatment. Results are presented from the multivariable logistic regres-
Persisting soreness in the vein can indicate extravasation sion analyses assessing the association between patient char-
(leakage of fluid into the tissues). Most of the respondents acteristics and patient’s indicating that they would call or go to
indicated they would report these side effects within a few hospital immediately for the potentially rapidly life-
Support Care Cancer

Table 1 Sociodemographic and disease characteristics of participants threatening side effects of high fever (Table 3) and unusual
Variable Subgroup Total (N = 436) bleeding or bruising (Table 4).
A total of 399 patients had complete data for the side effect
Sex Male 187 (43%) high fever and were thus included in this analysis. As shown in
Female 249 (57%) Table 3, cancer type and time since cancer diagnosis were the
Age < 55 133 (31%) only two characteristics found to be statistically significantly
55 to 74 245 (57%) associated with patients indicating that they would call or go to
≥ 75 54 (13%) hospital immediately in the case of experiencing a high fever.
Indigenous status Non-indigenous 423 (98%) Patients with non-Hodgkin lymphoma had more than four
Indigenous 7 (1.6%) times the odds (OR 4.01; 95% CI 1.2 to 13.6) of selecting the
Marital status Married or partner 276 (64%) recommended action than patients diagnosed with Bother^ can-
Single, divorced, separated 155 (36%) cers (p = 0.004). Patients diagnosed with cancer more than
or widowed 24 months ago had significantly lower odds of selecting the
Highest level of High school or below 206 (48%) recommended action (OR 0.51; 95% CI 0.3 to 0.9) compared
education
Vocational training, university 225 (52%) to patients who were diagnosed 12 or less months ago.
or other A total of 394 patients had complete data for the side effect
Country of birth Australia 306 (71%) unusual bleeding or bruising and were thus included in this
Others 127 (29%) analysis. As shown in Table 4, time since diagnosis was the
Insurance Yes 166 (38%) only characteristic found to be statistically significantly asso-
No 266 (62%) ciated with patients indicating that they would call or go to
Concession card Yes 253 (59%) hospital immediately for this side effect. Specifically, patients
No 177 (41%) who were diagnosed 24 months or more ago had statistically
Rurality City 364 (84%) significantly lower odds of selecting the recommended action
Regional or remote 68 (16%) (0.53; 95% CI 0.3 to 0.8) than patients who were diagnosed 12
Living arrangements Lives with spouse 297 (69%) or less months ago.
Lives alone 92 (21%)
Lives with other family 29 (6.7%)
members Discussion
Unrelated 12 (2.8%)
Other 2 (0.5%) It is critical that patients undergoing chemotherapy have a
Employment Home duties, unemployed, 265 (61%) thorough understanding of the potential toxicities of their
retired, disabled
Full or part time work 144 (33%)
treatment, and that they know what action to take if they
Other 25 (5.8%)
experience a side effect. Taking the appropriate action will
improve quality of life while receiving chemotherapy, and
Cancer type Breast 120 (28%)
may also improve survival [3–5]. This study shows that there
Colorectal 52 (12%)
is scope for increasing the proportion of patients reporting
Lung 27 (6%)
Prostate 14 (3%) potentially serious side effects in a timely manner.
Melanoma 7 (2%) Fever and unusual bleeding or bruising are two of the most
Non-Hodgkin lymphoma 46 (11%) potentially serious side effects of chemotherapy. These life-
Myeloma 27 (6%) threatening side effects should receive immediate medical at-
AML 18 (4%) tention. Neutropenic fever can lead to longer hospitalisation
CLL 15 (3%) times if antibiotic treatment is delayed, and between 2 and
MDS 7 (2%)
21% of patients will die if left untreated [21, 22]. Unusual
CML 6 (1%)
bleeding or bruising, often due to low platelet counts after
ALL 2 (0%)
chemotherapy, can increase patient risk of life-threatening
Haematology other 13 (3%)
than above
spontaneous haemorrhage, and also limit future chemotherapy
Multiple cancers 64 (15%) doses and frequency [23]. While in this study, it is encourag-
Missing 6 (1%) ing that both fever and unusual bleeding or bruising had the
Time since diagnosis 12 months or less 163 (38%) two highest proportions of patients reporting that they would
13–24 months 76 (18%) contact the hospital immediately; the number of participants
24+ months 195 (45%) who indicated that they would not take immediate action for
these side effects is concerning. Almost one third of partici-
pants indicated that they would not take immediate action for
 Support Care Cancer

Table 2 Number and percentage of respondents endorsing each response to experiencing a chemotherapy side effect (n = 436)

What action would you Call or go Call or go to Call or go to hospital Make an appointment Wait until my next Do nothing
take if, following chemotherapy, to hospital hospital if it has not if it has not improved with my GP in the appointment with
you experienced... immediately improved in a after a few days next day or two OR my cancer doctor
few hours make an earlier
appointment with my
cancer doctor

Nausea or vomiting 35 (8.5%) 141 (34%) 86 (21%) 69 (17%) 58 (14%) 21 (5.1%)

Diarrhoea 15 (3.7%) 101 (25%) 131 (32%) 74 (18%) 60 (15%) 26 (6.4%)
Fatigue or felt tired 5 (1.2%) 15 (3.6%) 45 (11%) 72 (18%) 176 (43%) 98 (24%)
Constipation 8 (2.0%) 21 (5.2%) 123 (30%) 113 (28%) 102 (25%) 40 (9.8%)
A sore mouth or throat 12 (3.0%) 31 (7.7%) 113 (28%) 108 (27%) 114 (28%) 26 (6.4%)
A high fever (38 °C or more) 276 (67%) 91 (22%) 19 (4.6%) 22 (5.3%) 3 (0.7%) 3 (0.7%)
A rash or other skin and 49 (12%) 45 (11%) 94 (23%) 81 (20%) 118 (29%) 19 (4.7%)
nail changes
Pain or burning 96 (24%) 128 (32%) 78 (19%) 54 (13%) 40 (9.9%) 9 (2.2%)
Tingling or numbness 64 (16%) 83 (20%) 72 (18%) 62 (15%) 110 (27%) 16 (3.9%)
Unusual bleeding or bruising 167 (41%) 112 (27%) 53 (13%) 52 (13%) 17 (4.2%) 7 (1.7%)
Flu-like symptoms such 119 (29%) 92 (22%) 84 (21%) 65 (16%) 41 (10%) 8 (2.0%)
as fever or cough
Soreness in my vein (where the 94 (24%) 105 (26%) 84 (21%) 39 (9.8% 58 (15%) 19 (4.8%)
chemotherapy was given)

fever, and less than half of participants indicated that they effects and potential adverse events related to their treatment.
would seek immediate help for unusual bleeding or bruising. Alternatively, it is possible that the findings indicate that pa-
These findings suggest that a sizeable proportion of che- tients are aware of what to report but do not intend to do so.
motherapy patients may not be well informed or do not recall The suggestion of not being informed or not recalling or un-
or do not understand the information given about serious side derstanding the information given is more likely and aligns
with research by Hershman and colleagues who performed a
Table 3 Sociodemographic, disease and treatment characteristics
associated with participants, selecting that they would Bcall or go to the Table 4 Sociodemographic, disease and treatment characteristics
hospital immediately^ in response to high fever associated with selecting that you could ‘call or go to the hospital
immediately’ in response to unusual bleeding or bruising
Variable Subgroup OR (95% CI) p
Variable Subgroup OR (95% CI) p
Sex Female 1.65 (0.8 to 3.3) 0.0691
Sex Female 1.33 (0.6 to 2.9) 0.3594
Male
Male .
Age 55 to 74 1.15 (0.3 to 3.8) 0.1499
Age 55 to 74 0.76 (0.4 to 1.4) 0.0566
≥ 75 0.67 (0.1 to 6.6)
≥ 75 0.68 (0.4 to 1.0)
< 55
< 55 .
Education Vocational training, 0.76 (0.3 to 1.8) 0.4219
university or other Education Vocational training, 0.76 (0.4 to 1.4) 0.2721
High school or below university or other
High school or below .
Cancer type Breast 0.96 (0.5 to 1.8) 0.0036
Cancer type Breast 1.26 (0.8 to 2.1) 0.6363
Colorectal 1.40 (0.5 to 4.1)
Colorectal 0.89 (0.3 to 2.8)
Non-Hodgkin 4.01 (1.2 to 13.6)
lymphoma Non-Hodgkin 1.30 (0.3 to 5.3)
Other lymphoma
Other .
Time since cancer 13–24 months 0.47 (0.1 to 2.4) < .0001
diagnosis Time since diagnosis 13–24 months 0.68 (0.2 to 2.3) 0.0001
24+ months 0.51 (0.3 to 0.9) 24+ months 0.53 (0.3 to 0.8)
12 m or less 12 m or less .
Concession card Yes 1.26 (0.6 to 2.6) 0.4077 Concession card Yes 1.30 (0.8 to 2.1) 0.1783
No No
Support Care Cancer

study to explore patients’ perceptions of physician-patient dis- A potential limitation of the current study is that it only
cussions of adverse events [24]. Common side effects, includ- included patients treated in urban hospitals. Further research
ing tiredness, nausea and vomiting, and loss of appetite, were with more diverse samples of participants including more of
discussed with patients prior to chemotherapy more than 80% lower socioeconomic status may be warranted. We acknowl-
of the time. However less common but more serious adverse edge the need to balance the desire to pursue rigorous detail in
events, such as fever with low white cell counts, were a patient reported survey with the practical need to have the
discussed less frequently. Also, whereas 76% of patients re- survey simple enough to be able to be completed by a large
ported having discussed neutropenia, only 68% reported un- sample of patients attending a cancer treatment clinic. In the
derstanding the information Bcompletely^ or Bvery well^. The current study, patients were not asked to consider varying
events most commonly discussed in Hershman’s study were levels of symptom severity, given that this would have added
the ones most often experienced. Other studies have found length and complexity to the survey. This is a further study
correlations between the expectation of subjective toxicities limitation, as the actions of patients are likely to be influenced
and the subsequent reporting of that toxicity [25]. The issue of by their perceptions of symptom severity. Similarly, no dis-
not retaining or understanding education sessions is tinction was made between symptoms of nausea and
underpinned by a literature review in the nursing literature vomiting, which may otherwise present different experiences
which shows that there are many ways to provide information for the patient and need for urgency of action. Finally, the
but retention is based on individual patients’ preference [26]. survey was developed specifically for this study, and while
In the Australian setting, although there is a paucity of studies based on a review of the literature and consultation with ex-
reporting the evaluation of nursing education about anticancer perts in the field, it is not a validated tool for the assessment of
chemotherapy, a small survey showed that 70% of the re- patient responses to potential side effects of chemotherapy
sponders Bagreed^ or Bstrongly agreed^ that the education treatment. Such a tool should be developed for future research
had been beneficial [27]. in this area.
Longer time since diagnosis was associated with lower Given the importance of timely and accurate reporting of
odds of selecting the recommended action to take in response side effects experienced following chemotherapy, it is critical
to unusual bleeding or bruising. This may reflect a familiarity that strategies are implemented to ensure patients know how
with toxicities by patients over time and greater confidence in to appropriately respond to serious side effects they may ex-
dealing with these. It is also possible that patient education for perience. Several interventions incorporating electronic
managing potentially serious side effects is emphasised more methods of patient self-report of side effects have been tested.
for newly diagnosed patients, but that such information may These mobile healthcare solutions allow patients to report, in
not be retained by patients over time. Health professionals real time, the prevalence and severity of the side effects they
may erroneously assume that patients who have had a previ- are experiencing with feedback of this information to
ous course of chemotherapy already have this knowledge. healthcare providers that are then able to initiate appropriate
Compared to participants with other diagnoses, those with management of the reported symptoms [30]. The findings of
non-Hodgkin lymphoma were more likely to indicate that they this study support the need for ongoing efforts to improve
would seek help immediately in response to a high fever. This patient responses for appropriately managing the side effects
finding may be the result of differences between education of chemotherapy. An in-depth qualitative study would offer a
provided to those with potentially curable cancers who are useful next step in determining why patients may or may not
often treated more intensively, and others. It may also reflect choose to act on specific symptoms.
differences in the types of information provided by
haematologists and oncologists.
No sociodemographic characteristics were associated with Acknowledgements Our thanks to the participating cancer treatment cen-
tres; Rochelle Smits, Alison Zucca, Heidi Turon and Hannah Small for
actions in response to the hypothetical chemotherapy side ef- research support; Sandra Dowley for data management; and Tiffany
fects. This contrasts with findings from previous studies. For Evans or statistical assistance. I acknowledge the support of Christopher
example, Hershman et al. found that black patients had less Oldmeadow, Senior Statistician Hunter Medical Research institute in
physician discussion than white patients, and previous re- reviewing the manuscript.
search shows a lower survival rate in black patients, but little
Funding information This research was supported by a National Health
is known about the link between adverse events and survival and Medical Research Council (NHMRC) Project Grant (ID 1010536), a
[24, 28]. In a study, recording adverse events to cancer therapy Strategic Research Partnership Grant (CSR 11-02) from Cancer Council
using patient journals, women spontaneously reported many NSW to the Newcastle Cancer Control Collaborative (New-3C), and
more side effects than men [29], although in the current study infrastructure funding from the Hunter Medical Research Institute
(HMRI). A/Prof Mariko Carey was supported by a NHMRC TRIP
women were not more likely to report serious side effects than Fellowship (APP1073031). Dr. Allison Boyes was supported by a
men. This may reflect differences in the study methods, for NHMRC Early Career Fellowship (APP1073317) and Cancer Institute
example, prospective reporting versus hypothetical reporting. NSW Early Career Fellowship (13/ECF/1-37).
 Support Care Cancer

Compliance with ethical standards reporting of toxicity symptoms during chemotherapy. J Clin Oncol
2:3552–3561
13. Olsen H, Klemetsrud T, Stokke HP, Tretli S, Westheim A (1999)
Conflicts of interest The authors declare that they have no conflict of
Adverse drug reactions in current antihypertensive therapy: a general
interest.
practice survey of 2586 patients in Norway. Blood Press 8:94–101
14. Bergman MM, Byers T, Freedman DS, Mokdad A (1998) Validity
Informed consent Ethical approval was granted by the Ethics of self-reported diagnoses leading to hospitalization: a comparison
Committees of each of the hospitals in this study, and all procedures of self-reports with hospital records in a prospective study of
performed involving human participants were in accordance with the American adults. Am J Epidemiol 147:969–977
ethical standards of the institutional research committee and with 15. Gupta V, Gu K, Chen Z, Lu W, Shu XO, Zheng Y (2011)
the1964 Declaration of Helsinki and its later amendments. Written in- Concordance of self-reported and medical chart information on
formed consent was obtained from all individual participants included cancer diagnosis and treatment. BMC Med Res Methodol 11:72.
in the study. https://doi.org/10.1186/1471-2288-11-72
16. American Cancer Society. Managing chemotherapy-related side ef-
fects (2017) https://www.cancer.org/treatment/treatments-and-side-
effects/physical-side-effects.html last accessed 27th Feb 2018
References 17. Cancer Council Australia. Understanding Chemotherapy. https://www.
cancercouncil.com.au/wp-content/uploads/2014/05/UC-pub-
1. Wagland R, Richardson A, Ewings S, Armes J, Lennan E, Hankins Chemotherapy-CAN708-lo-res-Aug-2016.pdf accessed 27th Feb 2018
M, Griffiths P (2016) Prevalence of cancer chemotherapy-related 18. Adams RJ, Appleton SL, Hill CL, Dodd M, Findlay C, Wilson DH
problems, their relation to health-related quality of life and associ- (2009) Risks associated with low functional health literacy in an
ated supportive care: a cross-sectional survey. Support Care Cancer Australian population. Me J Aust 191:530–534
24:4901–4911 19. Galdas PM, Cheater F, Marshall P (2005) Men and health seeking
2. Roila F, Molassiotis A, Herrstedt J, Aapro M, Gralla RJ, Bruera E, behaviour: literature review. J Adv Nurs 49:616–623
Clark-Snow RA, Dupuis LL, Einhorn LH, Feyer P, Hesketh PJ, 20. Jung HP, Baerveldt C, Olesen F, Grol R, Wensing M (2003) Patient
Jordan K, Olver I, Rapoport BL, Ruhlmann CH, Walsh D, Warr characteristics as predictors of primary health care preferences: a
D, van der Wetering M (2016) MASCC and ESMO guideline up- systematic literature analysis. Health Expect 6:160–181
date for the prevention of chemotherapy- and radiotherapy-induced 21. Perron T, Emara M, Ahmed S (2014) Time to antibiotics and out-
nausea and vomiting and of nausea and vomiting in advanced can- comes in cancer patients with febrile neutropenia. BMC Health
cer patients. Ann Oncol 27(Suppl 5):v119–v133 Serv Res 14:162. https://doi.org/10.1186/1472-6963-14-162
3. Llopis-Salvia P, Sarrio-Montes G, Garcia-Liopis P, Bargues-Ruiz A 22. Herbst C, Naumann F, Kruse EB, Monsef I, Bohlius J, Schulz H,
(2010) Chemotherapy dose intensity due to adverse drug reactions Engert A (2009) Prophylactic antibiotics or G_CSF for the preven-
in the oncology outpatient setting. J Oncol Pharm Pract 16:256–261 tion of infections and improvement of survival in cancer patients
4. Vandyk AD, Harrison MB, Macartney G, Ross-White A, Stacey D undergoing chemotherapy. Cochrane Database Syst Rev Jan 21(1):
(2012) Emergency department visits for symptoms experienced by CD007107. https://doi.org/10.1002/14651858.CD007107.pub2
oncology patients: a systematic review. Support Care Cancer 20: 23. Kuter DJ (2015) Managing thrombocytopenia associated with can-
1589–1599 cer chemotherapy. Oncology (Williston Park) 29:282–294
5. McKenzie H, Hayes L, White K, Cox K, Fethney J, Boughton M, 24. Hershman D, Calhoun E, Zapert K, Wade S, Malin J, Barron R
Dunn J (2011) Chemotherapy outpatients’ unplanned presentations (2008) Patients’ perceptions of physician-patient discussions and
to hospital: a retrospective study. Support Care Cancer 19:963–969 adverse events with cancer therapy. Arch Drug Info 1:70–78
6. Binkley JM, Harris SR, Levangie PK, Pearal M, Guglielmino J, Kraus 25. Whitford HS, Olver IN (2012) When expectations predict experi-
V, Rowden D (2012) Patient perspectives on breast cancer treatment ence: the influence of psychological factors on chemotherapy tox-
side effects and the prospective surveillance model for physical reha- icities. J Pain Symptom Manag 43:1036–1050
bilitation for women with breast cancer. Cancer 118:2207–2216 26. Valenti RB (2014) Chemotherapy education with patients with can-
7. Kearney N, Kidd L, Miller M, Sage M, Khorrami J, McGee M, cer: a literature review. Clin J Cancer Nurs 18:637–640
Cassidy J, Niven K, Gray P (2006) Utilising handheld computers to 27. Blancahrd G, Cox Y (2014) An evaluation of oncology patient
monitor and support patients receiving chemotherapy: results of a education: a quality improvement exercise. HNE Handover: For
UK-based feasibility study. Support Care Cancer 14:742–752 Nurses and Midwives [S.I>], v. 7. N. 1. Oct 2014 ISSN 2201-
8. Valenti RB (2014) Chemotherapy education for patients with can- 179X Available at http://journals.sfu.ca/hneh/index.php/hneh/
cer: a literature review. Clin J Oncol Nurs 18:637–640 article/view/296 (last accessed 18 Feb 2018)
9. Homsi J, Walsh D, Rivera N, Rybicki LA, Nelson KA, Legrand SB, 28. Smigal C, Jemal A, Ward E, Cokkinides V, Smith R, Howe HL,
Davis M, Naughton M, Gvozdjan D, Pham H (2006) Symptom Thun M (2006) Trends in breast cancer by race and ethnicity: up-
evaluation in palliative medicine: patient report vs systematic as- date 2006. CA Cancer J Clin 56:168–183
sessment. Support Care Cancer 14:444–453 29. Olver IN, Eliott JA, Long L, McKinnon M, Rumsby G (2012) The
10. Coolbrandt A, Van den Heede K, Vanhove E, De Bom A, Milisen impact of receiving treatment for cancer at a large metropolitan
K, Wildiers H (2011) Immediate versus delayed self-reporting of teaching hospital as recorded by patients using unstructured
symptoms and side effects during chemotherapy: does timing mat- journals. J Cancer Educ 27:625–630
ter? Eur J Oncol Nurs 15:130–136 30. Lamnguis-Eklöf A, Crafoord M-T, Christiansen M, Fjell M,
11. Olver IN, Matthews J, Bishop JF, Smith R (1994) The roles of Sundberg K (2017) Effects of an interactive mHealth innovation
patient and observer assessments in anti-emetic trials. Eur J for early detection of patient-reported symptom distress with focus
Cancer 1994(30A):1223–1227 on participatory care: randomized, controlled trials in patients with
12. Basch E, Artz D, Dulko D, Scher K, Sabbatini P, Hensley M, Mitra prostate and breast cancer. BMC Cancer 17:466. https://doi.org/10.
N, Speakman J, McCabe M, Schrag D (2005) Patient online self- 1186/s12885-017-3450-y