Home-Based e-Health Platform for

original report

Multidimensional Telemonitoring of
Symptoms, Body Weight, Sleep, and
Circadian Activity: Relevance for
Chronomodulated Administration of
Irinotecan, Fluorouracil-Leucovorin, and
Oxaliplatin at Home—Results From a Pilot
Study
Purpose To assess the impact of chronomodulated irinotecan fluorouracil-leucovorin and oxalipla-
abstract

Pasquale Innominato
Sandra Komarzynski
Abdoulaye Karaboué
Ayhan Ulusakarya
Mohamed Bouchahda
Mazen Haydar
Rachel Bossevot-
Desmaris
Magali Mocquery
Virginie Plessis
Francis Lévi
Author affiliations and
support information (if
applicable) appear at the
end of this article.
Corresponding author:
Francis Lévi, MD, PhD,
Cancer Chronothera-
py Team, Division of
Biomedical Sciences and
Cancer Research Centre,
Office B-166, Warwick
Medical School, Warwick
University, Gibbet Hill
Rd, Coventry CV4 7AL,
United Kingdom; e-mail:
F.Levi@warwick.ac.uk.
tin (chronoIFLO4) delivered at home on the daily life of patients with cancer in real time using a
home-based e-Health multifunction and multiuser platform. This involved multidimensional tele-
monitoring of circadian rest-activity rhythm (CircAct), sleep, patient-reported outcome measures,
and body weight changes (BWCs).
Patients and Methods Patients received chronoIFLO4 fortnightly at home. Patients completed the
19-item MD Anderson Symptom Inventory on an interactive electronic screen, weighed them-
selves on a dedicated scale, and continuously wore a wrist accelerometer for CircAct and sleep
monitoring. Daily data were securely teletransmitted to a specific server accessible by the hospi-
tal team. The clinically relevant CircAct parameter dichotomy index I < O and sleep efficiency (SE)
were calculated. The dynamic patterns over time of patient-reported outcome measures, BWC, I <
O, and SE informed the oncology team on tolerance in real time.
Results The platform was installed in the home of 11 patients (48 to 72 years of age; 45% men;
27% with performance status = 0), who were instructed on its use on site. They received 26 cy-
cles and provided 5,891 data points of 8,736 expected (67.4%). The most severe MD Anderson
Symptom Inventory scores were: interference with work (mean: 5.1 of 10) or general activity
(4.9), fatigue (4.9), distress (4.2), and appetite loss (3.6). Mean BWC was −0.9%, and mean SE
remained > 82%. CircAct disruption (I < O ≤ 97.5%) was observed in four (15%) cycles before
chronoIFLO4 start and in five (19%) cycles at day 14.
Conclusion The patient-centered multidimensional telemonitoring solution implemented here was
well accepted by patients receiving multidrug chemotherapy at home. Moreover, it demonstrated
that chronoIFLO4 was a safe therapeutic option. Such integrated technology allows the design of
innovative management approaches, ultimately improving patients’ experience with chemothera-
py, wellbeing, and outcomes.
Clin Cancer Inform. © 2018 by American Society of Clinical Oncology

© 2018 by American Society of Clinical Oncology ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics 1

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
 INTRODUCTION
The routine collection of patient-reported out-
come measures (PROMs) in an ambulatory set-
ting has been repeatedly shown to be feasible in
patients with cancer receiving chemotherapy.1-5
Moreover, using information and communica-
tions technology in health,6-11 such patient mon-
itoring provides additional data that can drive
decisions during clinical practice.1,12-14
The majority of the studies conducted so far
systematically collected PROMs on a weekly
basis or less frequently.15,16 Objective evalua-
tions of physiologic and/or behavioral functions
have been the exception.17-21 Recent progress
in wearable sensor technology, however, allows
routine tracking of physiologic parameters in
the patient’s own environment.22,23 Thus, along-
side PROMs, continuous, objective measures
of relevant physiologic parameters with bio-
sensors could provide additional information
and help increase safety in domestic cancer
care.4,17-19 Such patient-centered solutions ought
to integrate objective and subjective parameter
records, data teletransmission, and real-time
analysis of time series, through a dedicated plat-
form. End-users are both the patients and the
health care providers.22,24,25 Although interactive
telemedicine is keeping its promise of similar or
improved outcomes in comparison with conven-
tional health care,26 these complex multidimen-
sional telemonitoring solutions have remained
thus far limited to small pilot investigations,
allegedly paving the path to larger studies with
further refined technologies.
The European project inCASA (Integrated Net-
work for Completely Assisted Senior Citizen's
Autonomy) in the Seventh Framework Program
of the European Union aimed at developing cit-
izen-centric technologies and a service network
to improve the health condition and daily life of
patients with a chronic disease, thus minimiz-
ing hospitalizations. The clinical relevance of the
technology development was assessed through
pilot studies in different diseases; our pilot in
cancer demonstrated the feasibility of such data-
dense systematic home telemonitoring.4
Here, we show the relevance of this platform
for enabling the remote monitoring of patients
receiving a notoriously toxic multidrug regimen
at home. We report the dynamics of objective
measures and subjective PROMs throughout
the 14-day duration of chemotherapy courses,
2ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics
Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083
Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
combining chronomodulated irinotecan, fluoro-
uracil-leucovorin and oxaliplatin (chronoIFLO4)
at home. Indeed, this combination of irinotecan,
fluorouracil-leucovorin and oxaliplatin (FOLFIRI-
NOX/FOLFOXIRI) has been reported to result
in severe toxicities for approximately 60% of
patients.27,28 Nonetheless, this triplet is among
the most active regimens as first-line treatment
of advanced or metastatic pancreatic or colorec-
tal adenocarcinomas.29,30 We assessed tolera-
bility with classic criteria and further evaluated
the effects of such triplet chronomodulated
chemotherapy on patients’ physiology, behav-
ior, and symptoms. A secondary end point was
the determination of the temporal relationships
between continuous objective measures and
repeated selected PROMs, corresponding to
the most frequent and difficult-to-treat systemic
symptoms.31-33
PATIENTS AND METHODS
e-Health Platform
The inCASA platform4 was installed at the patient
home and connected to the Internet network
by a biomedical engineer. Each patient was
instructed on how to use the platform and the
related equipment and was given a form with
contact details for technical or health-related
issues.
The platform contained the following equipment
(Fig 1): a touch-screen computer (ASUS Eee-
top ET1611; ASUSTEK, Taipei, Taiwan) with the
SARA software (Telefonica Investigacion y Desar-
rollo SA, Granada, Spain), a body weight scale
(UC321-PBT; A&D Medical, San Jose, CA) con-
nected via Bluetooth to the SARA application,
and a modified watch-sized wrist accelerometer
(actigraph; Micro MotionLogger; Ambulatory
Monitoring, Ardsley, NY), whose recorded accel-
erations per minute over the 24 hours were trans-
mitted daily by the patient to the computer using
an infrared USB dongle. The SARA application
included an electronic version of the MD Ander-
son Symptom Inventory (MDASI) questionnaire34
for self-assessment of 13 frequent core symp-
toms and six items assessing interference with
activities of daily living, to be completed once a
day. The MDASI questionnaire was selected for
its validity and for the recall period for symptom
severity of 1 day, making it fit for its use on a
daily basis. All recorded data (actigraphy, body
weight, and MDASI) were transmitted daily via
 from 10:15 p.m. until 9:45 a.m.on the following
chronolFLO4
100 CPT
day, with peak delivery at 4:00 a.m. This pro-
OHP file, used in previous studies,36,37 was performed
90 FU using a four-reservoir, two-channel programma-
Drug Administration Rate (mg/m2/h)

LV
80 ble-in-time infusion pump (Melodie; Domocare,
Montmirail, France). The treatment lasted a
70
total of 80 hours (from 2:00 a.m. on day 1 until
60 10:00 a.m. on day 4), and it was repeated every
2 weeks.
50

40
Data Monitoring
30
Visual displays of the rest-activity patterns and
20
quantification of estimated sleep parameters
10 were available on a secured central server,
accessible by the oncology nursing and medi-
0
0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 cal staff. A first prospective alert algorithm was
Time of Day (clock hour) designed and implemented, on the basis of pre-
set threshold values for body weight loss and
MDASI cluster scores increment. The graphical
Fig 1. Administration the Internet. Patients were asked to use the plat- displays and warning alerts were checked daily
profile of the chrono- by the designated hospital cancer nurse special-
form for 30 days.4
modulated irinotecan
ists through a dedicated dashboard. In cases of
fluorouracil-leucovorin
and oxaliplatin (chronoI- lack of data transmission for > 24 hours, alert
FLO4) schedule, showing
Patients about high symptom severity, quick body weight
irinotecan (CPT, red), loss, or apparent deterioration of the rest-activ-
The patients were recruited at the chronother-
oxaliplatin (OHP, gold),
apy clinics in the medical oncology department ity rhythm, the oncology nurse would phone the
fluorouracil (FU, dark
blue), and folinic acid of Paul Brousse Hospital (Villejuif, France), patient and organize any appropriate interven-
(LV, light blue). within the inCASA cancer pilot.4 For the current tion required. The intervention spanned from
report, patients with advanced or metastatic col- telephone reassurance, a home visit by a techni-
orectal or pancreatic cancer with an indication cian or a nurse, a referral to the patient’s general
to be treated with triplet chemotherapy were practitioner or oncologist, to an emergency visit
offered to take part in the study. Availability of at the outpatient clinics or a hospital admission.
an Internet connection at home and signature
of an informed consent were mandatory to be
Data Analysis
eligible. The inCASA study was approved by the
local institutional review board and conducted MDASI scores were analyzed individually as they
according to the Declaration of Helsinki.35 Each were rated by the patients. Each item ranged
patient could be monitored for more than one from 0 (ie, symptom not present) to 10 (ie, as
course of fortnightly chronochemotherapy. bad as imaginable).34 Hence, higher values indi-
cated more severe symptoms. Daily body weight
change (BWC) was calculated with reference to
ChronoIFLO4
the first recorded value. Rest-activity recordings
In the therapeutic regimen chronoIFLO4, each were analyzed using the Action 4 and AW2 Soft-
drug was administered according to a time-stip- ware (Ambulatory Monitoring, Ardsley, NY) to
ulated semisinusoidal profile (Fig 1). Irinotecan compute sleep parameters on each individual
was delivered on day 1, from 2:00 a.m. until night, as well as the dichotomy index I < O.38 I
8:00 a.m., with peak delivery at 5:00 a.m. This < O was computed as the percentage of activity
was followed by a 3-day alternation of 12-hour epochs when in bed whose values were lower
infusions of oxaliplatin (25 mg/m2/d), from than the median level of activity when out of bed.
10:15 a.m. until 9:45 p.m., with peak delivery at Here, I < O was calculated over 72 hours, with
4:00 p.m., and fluorouracil (900 mg/m2/d) and 3-day moving windows, throughout the 14 days
folinic acid (400 mg/m2/d), both administered of duration of each chronotherapy course.4 This

ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics 3

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
 parameter had demonstrated its clinical rele-
vance in several prior studies.39-43 The parame-
ters computed to assess sleep quality, duration,
and timing were, respectively, sleep efficiency
(SE), total sleep time (TST), and sleep mid-
point.41,44-46 The temporal patterns of each of the
19 MDASI items, BWC, sleep parameters, and
I < O were mapped along the 14 days of each
chronoIFLO4 cycle.
In addition, we used a multilevel generalized
linear model47 to identify those objective param-
eters that could independently predict for
selected PROMs. We chose systemic symptoms
and interference items on the basis of prior evi-
dence.48,49 Fatigue, insomnia, anorexia, interfer-
ence with activity, work, relations with others,
and enjoyment of life were selected as depen-
dent MDASI variables with a logit ordinal link,
whereas the four actigraphy parameters (I < O,
SE, TST), sleep midpoint, and BWC were jointly
used as independent variables, and the day of
the course (1 to 14) was added as a random-ef-
fect factor.
Finally, we computed time-series analyses47 to
evaluate the correlations between the temporal
patterns of the selected PROMs and those of
the actigraphy parameters, either at the same
time (no lag), or with 1-day lag, to determine the
predictive value of objective measures on the
next day’s symptom intensity. Thus, we used
an autoregressive integrated moving average
(ARIMA) model, with MDASI items as depen-
dent variables and each actigraphy parameter,
without lag or with a 1-day lag separately, as
independent variable. For validation purposes,
we also computed a generalized autoregressive
conditional heteroscedasticity model, with the
same specifications as the ARIMA model.
All analyses were performed using PASW
v24 (SPSS, IBM, Chicago, IL) and Stata v14
(StataCorp, College Station, TX) software pack-
ages. The threshold for statistical significance
was set at P < .05.
RESULTS
Study Population
Eleven (five men and six women) of 31 patients
participating in the inCASA cancer pilot received
chronoIFLO4 for advanced or metastatic colorec-
tal (n = 5) or pancreatic (n = 6) cancer. They were
age 48 to 72 years (median, 60 years) and had
4ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics
Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083
Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
a WHO Performance Status of 0 (n = 3) or 1 (n
= 8). Eight patients had liver metastases, seven
had received prior chemotherapy, and eight had
undergone surgery of primary tumor (Appendix
Table A1). Twenty-six courses of chronoIFLO4,
were administered during the study, including
25 being delivered at home (96.2%). Individual
patients received one to four courses of chronoI-
FLO4 while in the study. The theoretical cumula-
tive number of study days was 364.
Compliance and Satisfaction
Patients provided complete daily data for a total
of 182 of 364 patient-days, resulting in an over-
all full compliance of 50%. Individual patient
compliance ranged from 7% to 100% (median,
55%). Compliance at each specific day of the
cycle ranged from 39% to 92% (median, 73%).
Parameter-specific compliance was 70% for
actigraphy, 68% for body weight, and 66%
for MDASI. Thus, overall, 5,891 out of 8,736
expected data points were available (67.4%).
Informal and unstructured discussion during
outpatient clinic visits identified patient absence
from home because of planned or emergency
admissions or out-of-home trips and regular
technical failures (poor internet connection) as
the most common causes of missing data, sim-
ilar to the whole inCASA pilot population.4 How-
ever, to minimize intrusiveness into the patient’s
privacy, structured investigations for individual
causes of missed data were not conducted.
Patients were interviewed informally by the nurs-
ing staff at the outset of the study period and
expressed overall satisfaction regarding their
participation in this pilot experience. The three
surveyed hospital nurses who regularly used the
hospital dashboard expressed their satisfaction
with the system and reported that they believed
it improved the follow-up of the health condition
of patients monitored remotely.4
Dose Intensities
Triplet chronomodulated chemotherapy was
administered in all the 26 courses. Median
delivered doses were 160 mg/m2 (range, 120
to 180 mg/m2) for irinotecan, 2.4 g/m2 (1.8 to
2.7 g/m2) for fluorouracil, and 75 mg/m2 (45
to 75 mg/m2) for oxaliplatin. The full planned
protocol dose was administered in 11 courses
(42.3%) for irinotecan, four courses (15.4%)
Table 1. Incidence and Severity of Main Toxicities per Course
Toxicity 1 2
Anemia 5 (19.2)
Neutropenia 4 (15.4)
Thrombocytopenia 0 0
Nausea-vomiting 4 (15.4)
Diarrhea 5 (19.2)
Mucositis 2 (7.7)
Hand-foot syndrome 3 (11.5)
Sensory neuropathy 11 (42.3)
Fatigue 9 (34.6)
Anorexia 8 (30.8)
NOTE. Data are presented as No. of courses (%). Grade determined according to the National Cancer Institute Common Toxicity Criteria for Adverse Events v3 scale.
for fluorouracil, 22 courses (84.6%) for oxal-
iplatin, and all courses for leucovorin. No reduc-
tion below 60% of the protocol dose was ever
performed for any drug. Altogether, doses were
reduced for prior acute toxicity, as per routine
clinical practice.
Toxicity
No grade 5 toxicity occurred. No grade 3 to 4
toxicity was encountered, except for two uncom-
plicated neutropenic events (Table 1). All the
other toxic events were mild or moderate, with
grade 2 toxicity occurring in < 20% of patients.
Multidimensional Data Pattern
Overall, chronoIFLO4 induced minimal alter-
ations on the rest-activity circadian pattern, as
illustrated by the display of the mean rest-activ-
ity rhythm throughout the 14-day span of the 26
courses (Fig 2). Overall, activity was prominent
at daytime and rest at nighttime throughout the
whole chronochemotherapy course, although a
slight dampening was apparent during the sec-
ond week. Indeed, the dynamics of wrist actig-
raphy–derived parameters, including circadian
dichotomy index I < O, total sleep duration,
sleep efficiency, and midsleep clock hour (not
shown) revealed mild and transient disturbances
along the 14-day span (Fig 3). Thus, mean I < O
values ranged from 96.3% to 98.5%, with indi-
vidual patterns showing transitory spans with I
< O values indicating circadian disruption (ie, ≤
97.5%39,40), most often followed by recovery in
the last days of the cycle (Fig 3A). Average sleep
Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083
Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
Grade
3 4 Total Grades 1-4
1 (3.8) 0 0 6 (23.1)
1 (3.8) 1 (3.8) 1 (3.8) 7 (26.9)
0 0 0
1 (3.8) 0 0 5 (19.2)
5 (19.2) 0 0 10 (38.5)
3 (11.5) 0 0 5 (19.2)
2 (7.7) 0 0 5 (19.2)
2 (7.7) 0 0 13 (50.0)
5 (19.2) 0 0 14 (53.8)
2 (7.7) 0 0 10 (38.5)
efficiency remained between 81.9% and 90.8%
along the 14-day course (Fig 3B). Nonetheless,
sleep efficiency was < 70% in approximately one
tenth of the nights. Moreover, individual patterns
indicated that poor sleep efficiency was overall
consistently present in specific cycles, with few
instances of single-night low values (Fig 3B). The
average total sleep time ranged from 8.6 hours
to 9.7 hours, with a few individual instances of
extremely short or long sleep duration (Fig 3C).
The average timing of midsleep had a fairly nar-
row range (1:02 a.m. to 5:30 a.m.) across the
nights assessed, with an overall median value at
3:31 a.m. (data not shown).
On average, the patients lost 0.3% to 2.4% of
their initial weight (mean, 0.9%), with transient
grade 1 weight loss (ie, ≥ 5%) occurring in four
instances (1.6% of data) for three individual
patients (Fig 2). The three symptoms that were
rated as the most severe ones by the patients
included fatigue, whose mean severity score
was 4.9 (range, 3.4 to 6.1); distress, with mean
severity score of 4.2 (3.5 to 5.1); and appetite
loss, with mean severity score of 3.6 (range,
2.4 to 4.8). Patients rated that their symptoms
interfered at a higher level with their work (mean,
5.1), their general activity (mean, 4.9), and their
enjoyment of life (mean, 4.3; Fig 2). The mean
day-to-day patterns of the symptoms previously
shown as associated with circadian disruption48
are depicted in Fig 3D. Furthermore, all average
MDASI scores ranked below the midrange value
of 5.5. Only approximately 10% of individual
daily item ratings were ≥ 7, with consistently >
20% of ratings being 0 across the fortnight of the
cure (Fig 3E).
ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics 5
 Patient’s home Hospital
chronoIFLO4
Secure remote Day 250
1 2 3 4 5 6 7 8 9 10 11 12 13 14
service provider

Accelerations/min
Continuous wrist actigraphy 125

0
Value
0
1
2
3

MDASI Items
4
5
6
7
Once daily 8
e-MDASI 9
10

5.0

Body Weight Change

2.5

0.0

(%)
Once daily scale –2.5

–5.0

–7.5
Day 1 2 3 4 5 6 7 8 9 10 11 12 13 14
chronoIFLO4

Fig 2. InCASA (Integrated Network for Completely Assisted Senior Citizen's Autonomy) platform outline, including the sensors at the patient’s home
(left) and the output at the hospital interface, with the average plots of the actual data derived over the 14-day span of the course (right). Chemotherapy
(chronomodulated irinotecan fluorouracil-leucovorin and oxaliplatin [chronoIFLO4]) was administered between day 1 and day 4. Top panel: wrist acti-
graph, providing continuous recordings, displayed as average accelerations per minute. Middle panel: touch-screen desktop, with an electronic version
of the MD Anderson Symptom Inventory (MDASI) questionnaire, completed daily, providing 19 items, whose average is displayed as a heat map. It
also collected data from the actigraph via infrared connection and data from the scale via Bluetooth connection and was linked via the Internet to the
bespoken secure server. Bottom panel: electronic scale, used once daily, providing average (blue line) and individual (gray dots) body weight change
(yellow-shaded box represents a ≥ 5% loss, corresponding to grade 1 toxicity, according to the National Cancer Institute Common Toxicity Criteria for
Adverse Events).

Time-Series Analyses those of all actigraphy parameters. Moreover,

nightly increase in TST was correlated with daily
The multilevel generalized linear model identified
decrease in insomnia and anorexia severity, yet
I < O as an independent predictive factor for all
it was also correlated with daily increase in inter-
the selected PROMs, indicating that an increase
ference with work and activity.
in this objective parameter was associated with
a higher chance of less-severe symptoms (Table Finally, the ARIMA model was used to identify
2). Fatigue was independently predicted by I < the time courses of actigraphy parameters pre-
O and BWC, whereas insomnia was by all actig- dicting for the patterns of PROMs on the follow-
raphy parameters. Although the odds ratios for ing day. Thus, 1-day lagged I < O time series
SE and TST were < 1 for insomnia and anorexia, forecasted fatigue, insomnia, interference with
they were > 1 for interference with activity and activity and with work, invariably with a decrease
work, indicating a discordant impact on severity (ie, worsening) of I < O anticipating an increase
according to the PROM item (Table 2). in symptom severity (Table 3). Insomnia pat-
The ARIMA model revealed that the time course terns were inversely preceded by a deterioration
of I < O was negatively correlated with that of of all actigraphy parameters, whereas interfer-
all PROMs except interference with enjoyment of ence with relations was predicted by none of
life (Table 3). The temporal orderings of fatigue them. As for the synchronous model, the 1-day
were associated only with those of I < O, whereas lagged one revealed a negative coefficient for
the patterns of insomnia were associated with TST forecasting insomnia and anorexia, and a

6ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
 A chronolFLO4 D
Fatigue Interference with activity
Day 1 2 3 4 5 6 7 8 9 10 11 12 13 14
10 Disturbed sleep Sleepiness

Average MDASI Score

100
9 Appetite loss
8
95
7
I< 0 (%)

6
90
5
4
85
3
2
80
1
0 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14

B 100 Day
chronolFLO4
90
80
70 E
SE (%)

60
MDASI score 0 1-2 3-6 7-10
50
100

Percentage of Instances
40
90
30
80
20
70
60
C 15 50
13 40
TST (hours)

30
11
20
9 10
7 0
5 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Day 1 2 3 4 5 6 7 8 9 10 11 12 13 14 chronolFLO4 Day

chronolFLO4

Fig 3. Average patterns of

(A-C) three actigraphy-de-
rived parameters and (D)
selected MD Anderson
Symptom Inventory (MDASI)
items, as well as (E) distri-
bution of the MDASI item
scores along the 14-day
chemotherapy course. For
actigraphy parameters,
mean values per day (±
SEM) are represented. For
(A) the circadian dichotomy
index (I < O; light blue) and
(B) sleep efficiency (SE;
orange), individual patterns
are also represented (thin
gray lines). (C) For total
sleep time (TST; dark blue),
individual values are also
represented (gray dots).
(E) All 19 patient-reported
outcome measures scores
(each ranging from 0 to
10) were categorized into
four groups, and the daily
respective proportions are
represented. chronoIFLO4,
chronomodulated irinotec-
an, fluorouracil-leucovorin,
and oxaliplatin.
positive one predicting interference with activity
and work. Altogether, the absolute values of the
coefficients were similar between the synchro-
nous and the preceding actigraphy parameters
time series and had invariably the same direc-
tion, even though in a few instances a discor-
dance was observed. Using another forecasting
method yielded analogous results, with modest
differences in coefficients’ values between tests
(data not shown).
DISCUSSION
The inCASA e-Health platform enabled quanti-
tative and real-time measurements of multidi-
mensional parameters for ≥ 30 days that could
be useful for triggering proactive supportive
interventions and monitoring their efficacy. The
unique data set in this study, including 5,891
data points during 364 patient-days, helped pre-
cisely document the safety of chronomodulated
triplet chemotherapy delivered at home. The
results of the integrated telemonitoring approach
supported previous evidence of a satisfactory
tolerance for chronoIFLO.50 Such dense, remote,
and real-time monitoring of patient physiology,
behavior, and PROMs allowed a day-to-day mul-
tidimensional and accurate evaluation of the
impact of such an active yet potentially damaging
regimen on the daily life of the patient. We could
therefore observe a modest body weight loss after
chronoIFLO4. Furthermore, objectively assessed
sleep duration, quality, and timing varied only to
a narrow extent during and after chronoIFLO4.
Complaint of poor sleep during treatment was
associated with poor overall survival in patients
with colorectal cancer.51 Thus, the prevention of
iatrogenic sleep disruption represents a clinically
relevant end point for both well-being and sur-
vival. Better sleep quality should result from the
patient remaining at home to receive this most
effective yet aggressive combination chemother-
apy, rather than being admitted for it.52
In addition, this home-based solution provided
evidence for a temporal relationship between the
time course of the circadian rest-activity rhythm
and that of patient-rated fatigue, sleep prob-
lems, and interference with activity or work the

ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics 7

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
Table 2. Multilevel Generalized Liner Models for Objective Measures (actigraphy parameters and body weight change) as Independent Variables
and Patient-Reported Outcomes (MDASI items) as Dependent Variables
Objective Parameter
I<O SE TST SMP BWC
MDASI Item OR P OR P OR P OR P OR P
Fatigue 0.83 < .001† 0.98 .17 0.94 .51 0.74 .13 0.80 .013
Insomnia 0.81 < .001 0.96 .008 0.99 .018 0.94 .028† 0.91 .29
Anorexia 0.85 .001 0.94 < .001† 0.96 <.001 0.81 .28 1.07 .44
Interference with activity 0.82 .001 1.04 .008 1.01 .022 0.86 .45 1.08 .39
Interference with work 0.77 < .001 1.05 .002 1.01 .05 0.80 .26 1.04 .67
Interference with relations 0.90 .04 0.98 .27 0.92 .39 1.03 .87 0.85 .06
Interference with enjoyment 0.87 .018 1.06 < .001 1.21 .07 0.89 .57 1.03 .74
Abbreviations: BWC, body weight change; I < O, circadian dichotomy index; OR, odds ratio; SE, sleep efficiency; SMP, sleep midpoint; TST, total sleep time.

following day. These findings, albeit obtained in
a limited patient sample, bear the potential for
the development of timely proactive and per-
sonalized interventions. These can be attained
through coupling predictive algorithms of clini-
cal deterioration or improvement with partici-
patory and dynamic bespoken coaching for the
patient. Thus, such a model of care incorporat-
ing self-management could allegedly improve
patients’ well-being, contributing to the assim-
ilation of cancer to a chronic illness.53 How-
ever, dedicated time-series forecasting analysis
combining multiparametric data are required,
together with the testing of an array of potential
behavioral interventions, all integrated in patient
and health care professional platforms.4,18,21,54-56
This approach has the potential to provide ben-
efit not only to patients with cancer but also
patients with other chronic conditions, as pro-
posed in primary care and in particular employ-
ment conditions.18,57-59
Moreover, the combination of dose-dense contin-
uous monitoring of physiologic biorhythms and
repeated assessments of PROMs allows time-se-
ries correlative and predictive analyses, which
can help better understand the physio-patho-
genesis of patients’ complaints. We showed
here an example providing further evidence
for an association between circadian rhythms
and sleep function and systemic symptoms, in
agreement with existing literature.32,48,49,60-63
We acknowledge that a limitation of our study
was the approximately 30% rate of missing
data. This figure was, notwithstanding, similar to
that recently reported for electronic PROMs in
large studies and better than that observed with
daily collection.2,64 Most missing data here were
related either to technical issues or to the fixed
nature of the platform involving a touch-screen
desktop.4 Such limitations have now been
addressed in a second-generation mobile plat-
form (eg, a wireless tablet) involving improved
technologies, biosensors, and dedicated patient
education and services. The current system had
already been evaluated in terms of usability and
patients’ perception with the Whole Systems
Demonstrator Service User Technology Accept-
ability Questionnaire65 in a subset of end users,
whose results displayed an overall satisfaction
rate of 84%.4
A second-generation platform, on the basis of
the blueprint of the solution here tested, ought to
incorporate the monitoring and temporal analysis
of additional biorhythms, including, for instance,
skin temperature and heart rate variability,66-69
which can further increase the accuracy of the
predictive model and of appropriate response
through broadening the spectrum of physiologic
functions under remote surveillance. It should
also include a measure of the complex construct
of health-related quality of life alongside symp-
tom assessment.70
Altogether, the groundwork for patient-centered,
home-based cancer care involves the design of
effective and well-tolerated ambulatory chemo-
therapy regimens, alongside optimized integra-
tive telemonitoring. In this study, the dynamics
of multiple assessments concurred to estimate a
fairly limited impact of chronoIFLO4 on patients’
physiology and daily life, with usual satisfactory
recovery between courses.

8ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
Table 3. Autoregressive Integrated Moving Average Model for PROMs as Dependent Time Series and Actigraphy Parameters Temporal Courses as
Independent Variables, Used Either With No Lag or With 1-Day Lag

Actigraphy No Lag 1-Day Lag

PROM (MDASI) Parameter Coefficient P Coefficient P
Fatigue I<O −0.28 < .001 −0.20 .001
SE −0.03 .07 −0.02 .18
TST −0.08 .45 0.01 .97
SMP −0.43 .050 −0.60 .018
Insomnia I<O −0.29 < .001 −0.30 < .001
SE −0.07 < .001 −0.07 .001
TST −0.20 .039 −0.21 .042
SMP −0.54 .004 −0.55 .01
Anorexia I<O −0.21 < .001 −0.11 .050
SE −0.09 < .001 −0.08 < .001
TST −0.48 < .001 −0.46 < .001
SMP −0.32 .18 −0.15 .55
Interference with activity I<O −0.22 < .001 −0.16 .001
SE 0.02 .19 0.03 .07
TST 0.25 .023 0.28 .011
SMP −0.42 .06 −0.49 .023
Interference with work I<O −0.25 < .001 −0.20 < .001
SE 0.02 .20 0.02 .24
TST 0.23 .031 0.25 .028
SMP −0.52 .015 −0.41 .041
Interference with relations I<O −0.14 .015 −0.12 .06
SE −0.04 .06 −0.04 .07
TST −0.17 .16 −0.09 .40
SMP −0.12 .60 −0.01 .96
Interference with enjoyment I<O −0.05 .29 −0.03 .60
SE 0.05 .007 0.04 .034
TST 0.14 .17 0.14 .23
SMP −0.37 .10 −0.34 .11
Abbreviations: I < O, circadian dichotomy index; MDASI, MD Anderson Symptom Inventory; PROM, patient-reported outcome measure; SE, sleep efficiency; SMP sleep
midpoint; TST, total sleep time.

In conclusion, this pilot study confirmed the
feasibility and the clinical and scientific rele-
vance of a dense, multidimensional near–real-
time telemonitoring, even in patients receiving
an aggressive anticancer regimen. In addition,
chronoIFLO4 displayed an adequate safety pro-
file, despite being administered at home. Alto-
gether, these findings support the development
of integrated patient-centered solutions, accu-
rately telemonitoring relevant physiology and
capturing subjective assessments, thus provid-
ing innovative insight to improve patients’ experi-
ence with chemotherapy, safety, and outcomes.
DOI: https://doi.org/10.1200/CCI.17.00125
Published online on ascopubs.org/journal/cci on February
22, 2018.

AUTHOR CONTRIBUTIONS Provision of study material or patients: Pasquale Innominato,

Conception and design: Pasquale Innominato, Sandra
Komarzynski, Abdoulaye Karaboué, Francis Lévi
Financial support: Pasquale Innominato, Francis Lévi
Administrative support: Pasquale Innominato, Mohamed
Bouchahda, Francis Lévi
Abdoulaye Karaboué, Ayhan Ulusakarya, Mazen Haydar,
Magali Mocquery, Francis Lévi
Collection and assembly of data: Pasquale Innominato,
Sandra Komarzynski, Abdoulaye Karaboué, Ayhan Ulusa-
karya, Mazen Haydar, Rachel Bossevot-Desmaris, Magali

ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics 9

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
 Mocquery, Virginie Plessis, Francis Lévi Mohamed Bouchahda
Data analysis and interpretation: Pasquale Innominato, San- No relationship to disclose
dra Komarzynski, Abdoulaye Karaboué, Ayhan Ulusakarya,
Mohamed Bouchahda, Francis Lévi Mazen Haydar
Manuscript writing: All authors No relationship to disclose
Final approval of manuscript: All authors
Accountable for all aspects of the work: All authors Rachel Bossevot-Desmaris
No relationship to disclose
AUTHORS' DISCLOSURES OF
POTENTIAL CONFLICTS OF INTEREST Magali Mocquery
No relationship to disclose
The following represents disclosure information provided
by authors of this manuscript. All relationships are
Virginie Plessis
considered compensated. Relationships are self-held
No relationship to disclose
unless noted. I = Immediate Family Member, Inst = My
Institution. Relationships may not relate to the subject
Francis Lévi
matter of this manuscript. For more information about
Research Funding: Philips Respironics (Inst)
ASCO's conflict of interest policy, please refer to www.
Patents, Royalties, Other Intellectual Property: Institut
asco.org/rwc or ascopubs.org/jco/site/ifc.
National del la Santé et de lal Recherche Médicale patent
Pasquale Innominato on circadian dichotomy index automatic computation
Research Funding: Philips Healthcare (Inst) (Inst)
Travel, Accommodations, Expenses: Merck Serono
Sandra Komarzynski
No relationship to disclose ACKNOWLEDGMENT
Abdoulaye Karaboué We thank all the patients who participated in this study,
No relationship to disclose and A. Arbaud, J. Fursse, and J. Rovira-Simon for their
contribution to the development of the Integrated Network
Ayhan Ulusakarya for Completely Assisted Senior Citizen's Autonomy (inCA-
No relationship to disclose SA) platform.

Affiliations
Pasquale Innominato, North Wales Cancer Centre, Betsi Cadwaladr University Health Board, Denbighshire; Pasquale
Innominato, Sandra Komarzynski, and Francis Lévi, Warwick Medical School, Coventry, United Kingdom; Pasquale
Innominato, Sandra Komarzynski, Ayhan Ulusakarya, Mohamed Bouchahda, and Francis Lévi, Institut National de la Santé et
de la Recherche Médicale, Unit 935; Ayhan Ulusakarya, Mohamed Bouchahda, Mazen Haydar, Rachel Bossevot-Desmaris,
Magali Mocquery, Virginie Plessis, and Francis Lévi, Assistance Publique–Hôpitaux de Paris, Paul Brousse Hospital,
Villejuif; and Abdoulaye Karaboué, AK-SCIENCE, Research and Therapeutic Innovation, Vitry-sur-Seine, France.

Support
Supported by the European Commission through the Information and Communication Technologies Policy Support
Programme project inCASA (Integrated Network for Completely Assisted Senior Citizen's Autonomy) Contract No. CIP
250505, Framework Programme for Research and Technological Development (FP) 7 and the Coordinating Action on
Systems Medicine Contract No. 305033 CaSyM, FP7 SP1 HEALTH; the Conseil Régional d’Ile de France and Banque
Publique d’Investissement of France, through the Fonds Unique Interministériel 12 Contract No. PiCADo; and the
Institut de Recherche en Santé Publique Contract No. CLOCK-DOM1.

Prior Presentation
Presented in part at the European Society of Medical Oncology annual meeting, Madrid, Spain, September 8-12, 2017.

REFERENCES
1. Basch E, Deal AM, Kris MG, et al: Symptom monitoring with patient-reported outcomes during
routine cancer treatment: A randomized controlled trial. J Clin Oncol 34:557-565, 2016
2. Wysham NG, Wolf SP, Samsa G, et al: Integration of electronic patient-reported outcomes into
routine cancer care: An analysis of factors affecting data completeness. JCO Clin Cancer Inform
doi:10.1200/CCI.16.00043
3. Basch E, Dueck AC, Rogak LJ, et al: Feasibility assessment of patient reporting of symptomatic
adverse events in multicenter cancer clinical trials. JAMA Oncol 3:1043-1050, 2017
4. Innominato PF, Komarzynski S, Mohammad-Djafari A, et al: Clinical relevance of the first
domomedicine platform securing multidrug chronotherapy delivery in metastatic cancer patients
at home: The inCASA European Project. J Med Internet Res 18:e305, 2016

10ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
5. Basch E, Deal AM, Dueck AC, et al: Overall survival results of a trial assessing patient-reported
outcomes for symptom monitoring during routine cancer treatment. JAMA 318:197-198, 2017
6. Abernethy AP, Herndon JE II, Wheeler JL, et al: Improving health care efficiency and quality using
tablet personal computers to collect research-quality, patient-reported data. Health Serv Res
43:1975-1991, 2008
7. Odeh B, Kayyali R, Nabhani-Gebara S, et al: Optimizing cancer care through mobile health.
Support Care Cancer 23:2183-2188, 2015
8. Weaver A, Young AM, Rowntree J, et al: Application of mobile phone technology for managing
chemotherapy-associated side-effects. Ann Oncol 18:1887-1892, 2007
9. Pietanza MC, Basch EM, Lash A, et al: Harnessing technology to improve clinical trials: Study
of real-time informatics to collect data, toxicities, image response assessments, and patient-
reported outcomes in a phase II clinical trial. J Clin Oncol 31:2004-2009, 2013
10. Nardi EA, Lentz LK, Winckworth-Prejsnar K, et al: Emerging issues and opportunities in health
information technology. J Natl Compr Canc Netw 14:1226-1233, 2016
11. Girgis A, Durcinoska I, Levesque JV, et al: eHealth system for collecting and utilizing patient
reported outcome measures for personalized treatment and care (PROMPT-Care) among cancer
patients: Mixed methods approach to evaluate feasibility and acceptability. J Med Internet Res
19:e330, 2017
12. Strasser F, Blum D, von Moos R, et al: The effect of real-time electronic monitoring of patient-
reported symptoms and clinical syndromes in outpatient workflow of medical oncologists:
E-MOSAIC, a multicenter cluster-randomized phase III study (SAKK 95/06). Ann Oncol 27:324-
332, 2016
13. Basch E, Abernethy AP, Mullins CD, et al: Recommendations for incorporating patient-reported
outcomes into clinical comparative effectiveness research in adult oncology. J Clin Oncol
30:4249-4255, 2012
14. Kofoed S, Breen S, Gough K, et al: Benefits of remote real-time side-effect monitoring systems for
patients receiving cancer treatment. Oncol Rev 6:e7, 2012
15. Howell D, Molloy S, Wilkinson K, et al: Patient-reported outcomes in routine cancer clinical
practice: A scoping review of use, impact on health outcomes, and implementation factors. Ann
Oncol 26:1846-1858, 2015
16. Di Maio M, Basch E, Bryce J, et al: Patient-reported outcomes in the evaluation of toxicity of
anticancer treatments. Nat Rev Clin Oncol 13:319-325, 2016
17. Cheng C, Stokes TH, Wang MD: caREMOTE: The design of a cancer reporting and monitoring
telemedicine system for domestic care. Conf Proc IEEE Eng Med Biol Soc 2011:3168-3171,
2011
18. Gibbons C, Bailador Del Pozo G, Andrés J, et al: Data-as-a-service platform for delivering healthy
lifestyle and preventive medicine: Concept and structure of the DAPHNE project. JMIR Res
Protoc 5:e222, 2016
19. Pigini L, Bovi G, Panzarino C, et al: Pilot test of a new personal health system integrating
environmental and wearable sensors for telemonitoring and care of elderly people at home
(SMARTA Project). Gerontology 63:281-286, 2017
20. Jim HS, Small B, Faul LA, et al: Fatigue, depression, sleep, and activity during chemotherapy:
Daily and intraday variation and relationships among symptom changes. Ann Behav Med 42:321-
333, 2011
21. Low CA, Dey AK, Ferreira D, et al: Estimation of symptom severity during chemotherapy from
passively sensed data: Exploratory study. J Med Internet Res 19:e420, 2017
22. Chung AE, Jensen RE, Basch EM: Leveraging emerging technologies and the “internet of things”
to improve the quality of cancer care. J Oncol Pract 12:863-866, 2016

ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics 11

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
 23. Fisch MJ, Chung AE, Accordino MK: Using technology to improve cancer care: Social media,
wearables, and electronic health records. Am Soc Clin Oncol Educ Book 35:200-208, 2016
24. Abernethy AP, Etheredge LM, Ganz PA, et al: Rapid-learning system for cancer care. J Clin Oncol
28:4268-4274, 2010
25. Wildevuur SE, Simonse LW: Information and communication technology-enabled person-
centered care for the “big five” chronic conditions: Scoping review. J Med Internet Res 17:e77,
2015
26. Flodgren G, Rachas A, Farmer AJ, et al: Interactive telemedicine: Effects on professional practice
and health care outcomes. Cochrane Database Syst Rev 9:CD002098, 2015
27. Suker M, Beumer BR, Sadot E, et al: FOLFIRINOX for locally advanced pancreatic cancer: A
systematic review and patient-level meta-analysis. Lancet Oncol 17:801-810, 2016
28. Ficorella C, Bruera G, Cannita K, et al: Triplet chemotherapy in patients with metastatic colorectal
cancer: Toward the best way to safely administer a highly active regimen in clinical practice. Clin
Colorectal Cancer 11:229-237, 2012
29. Conroy T, Desseigne F, Ychou M, et al: FOLFIRINOX versus gemcitabine for metastatic pancreatic
cancer. N Engl J Med 364:1817-1825, 2011
30. Loupakis F, Cremolini C, Masi G, et al: Initial therapy with FOLFOXIRI and bevacizumab for
metastatic colorectal cancer. N Engl J Med 371:1609-1618, 2014
31. Dantzer R, Meagher MW, Cleeland CS: Translational approaches to treatment-induced symptoms
in cancer patients. Nat Rev Clin Oncol 9:414-426, 2012
32. Miller AH, Ancoli-Israel S, Bower JE, et al: Neuroendocrine-immune mechanisms of behavioral
comorbidities in patients with cancer. J Clin Oncol 26:971-982, 2008
33. Teunissen SC, Wesker W, Kruitwagen C, et al: Symptom prevalence in patients with incurable
cancer: A systematic review. J Pain Symptom Manage 34:94-104, 2007
34. Cleeland CS, Mendoza TR, Wang XS, et al: Assessing symptom distress in cancer patients: The
M.D. Anderson Symptom Inventory. Cancer 89:1634-1646, 2000
35. World Medical Association: World Medical Association Declaration of Helsinki: Ethical principles
for medical research involving human subjects. JAMA 310:2191-2194, 2013
36. Garufi C, Torsello A, Tumolo S, et al: Cetuximab plus chronomodulated irinotecan, 5-fluorouracil,
leucovorin and oxaliplatin as neoadjuvant chemotherapy in colorectal liver metastases: POCHER
trial. Br J Cancer 103:1542-1547, 2010
37. Lévi F, Karaboué A, Gorden L, et al: Cetuximab and circadian chronomodulated chemotherapy
as salvage treatment for metastatic colorectal cancer (mCRC): Safety, efficacy and improved
secondary surgical resectability. Cancer Chemother Pharmacol 67:339-348, 2011
38. Minors D, Akerstedt T, Atkinson G, et al: The difference between activity when in bed and out of
bed. I. Healthy subjects and selected patients. Chronobiol Int 13:27-34, 1996
39. Innominato PF, Giacchetti S, Bjarnason GA, et al: Prediction of overall survival through circadian
rest-activity monitoring during chemotherapy for metastatic colorectal cancer. Int J Cancer
131:2684-2692, 2012
40. Lévi F, Dugué PA, Innominato P, et al: Wrist actimetry circadian rhythm as a robust predictor of
colorectal cancer patients survival. Chronobiol Int 31:891-900, 2014
41. Natale V, Innominato PF, Boreggiani M, et al: The difference between in bed and out of bed
activity as a behavioral marker of cancer patients: A comparative actigraphic study. Chronobiol
Int 32:925-933, 2015
42. Chang WP, Lin CC: Correlation between rest-activity rhythm and survival in cancer patients
experiencing pain. Chronobiol Int 31:926-934, 2014
43. Sultan A, Choudhary V, Parganiha A: Worsening of rest-activity circadian rhythm and quality of
life in female breast cancer patients along progression of chemotherapy cycles. Chronobiol Int
34:609-623, 2017

12ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
44. Ancoli-Israel S, Cole R, Alessi C, et al: The role of actigraphy in the study of sleep and circadian
rhythms. Sleep 26:342-392, 2003
45. Palesh O, Aldridge-Gerry A, Zeitzer JM, et al: Actigraphy-measured sleep disruption as a predictor
of survival among women with advanced breast cancer. Sleep (Basel) 37:837-842, 2014
46. Palesh O, Haitz K, Lévi F, et al: Relationship between subjective and actigraphy-measured sleep
in 237 patients with metastatic colorectal cancer. Qual Life Res 26:2783-2791, 2017
47. Moran JL, Solomon PJ: Statistics in review. Part 2: Generalised linear models, time-to-event and
time-series analysis, evidence synthesis and clinical trials. Crit Care Resusc 9:187-197, 2007
48. Innominato PF, Mormont MC, Rich TA, et al: Circadian disruption, fatigue, and anorexia clustering
in advanced cancer patients: Implications for innovative therapeutic approaches. Integr Cancer
Ther 8:361-370, 2009
49. Mormont MC, Waterhouse J: Contribution of the rest-activity circadian rhythm to quality of life in
cancer patients. Chronobiol Int 19:313-323, 2002
50. Gholam D, Giacchetti S, Brézault-Bonnet C, et al: Chronomodulated irinotecan, oxaliplatin, and
leucovorin-modulated 5-fluorouracil as ambulatory salvage therapy in patients with irinotecan-
and oxaliplatin-resistant metastatic colorectal cancer. Oncologist 11:1072-1080, 2006
51. Innominato PF, Spiegel D, Ulusakarya A, et al: Subjective sleep and overall survival in
chemotherapy-naïve patients with metastatic colorectal cancer. Sleep Med 16:391-398, 2015
52. Dobing S, Frolova N, McAlister F, et al: Sleep quality and factors influencing self-reported
sleep duration and quality in the general internal medicine inpatient population. PLoS One
11:e0156735, 2016
53. McCorkle R, Ercolano E, Lazenby M, et al: Self-management: Enabling and empowering patients
living with cancer as a chronic illness. CA Cancer J Clin 61:50-62, 2011
54. Hrushesky WJ, Grutsch J, Wood P, et al: Circadian clock manipulation for cancer prevention and
control and the relief of cancer symptoms. Integr Cancer Ther 8:387-397, 2009
55. Bruggeman-Everts FZ, Wolvers MDJ, van de Schoot R, et al: Effectiveness of two web-based
interventions for chronic cancer-related fatigue compared to an active control condition: Results
of the “Fitter na kanker” randomized controlled trial. J Med Internet Res 19:e336, 2017
56. Kearney N, McCann L, Norrie J, et al: Evaluation of a mobile phone-based, advanced symptom
management system (ASyMS) in the management of chemotherapy-related toxicity. Support
Care Cancer 17:437-444, 2009
57. van Drongelen A, Boot CR, Hlobil H, et al: Process evaluation of a tailored mobile health
intervention aiming to reduce fatigue in airline pilots. BMC Public Health 16:894, 2016
58. Zan S, Agboola S, Moore SA, et al: Patient engagement with a mobile web-based telemonitoring
system for heart failure self-management: A pilot study. JMIR Mhealth Uhealth 3:e33, 2015
59. Murnane EL, Cosley D, Chang P, et al: Self-monitoring practices, attitudes, and needs of
individuals with bipolar disorder: Implications for the design of technologies to manage mental
health. J Am Med Inform Assoc 23:477-484, 2016
60. Ancoli-Israel S, Liu L, Rissling M, et al: Sleep, fatigue, depression, and circadian activity rhythms
in women with breast cancer before and after treatment: A 1-year longitudinal study. Support
Care Cancer 22:2535-2545, 2014
61. Fernandes R, Stone P, Andrews P, et al: Comparison between fatigue, sleep disturbance, and
circadian rhythm in cancer inpatients and healthy volunteers: Evaluation of diagnostic criteria for
cancer-related fatigue. J Pain Symptom Manage 32:245-254, 2006
62. Sephton S, Spiegel D: Circadian disruption in cancer: A neuroendocrine-immune pathway from
stress to disease? Brain Behav Immun 17:321-328, 2003
63. Yennurajalingam S, Tayjasanant S, Balachandran D, et al: Association between daytime activity,
fatigue, sleep, anxiety, depression, and symptom burden in advanced cancer patients: A
preliminary report. J Palliat Med 19:849-856, 2016

ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics 13

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
 64. Min YH, Lee JW, Shin YW, et al: Daily collection of self-reporting sleep disturbance data via a
smartphone app in breast cancer patients receiving chemotherapy: A feasibility study. J Med
Internet Res 16:e135, 2014
65. Hirani SP, Rixon L, Beynon M, et al: Quantifying beliefs regarding telehealth: Development of the
Whole Systems Demonstrator Service User Technology Acceptability Questionnaire. J Telemed
Telecare 23:460-469, 2017
66. Ortiz-Tudela E, Innominato PF, Rol MA, et al: Relevance of internal time and circadian robustness
for cancer patients. BMC Cancer 16:285, 2016
67. Martinez-Nicolas A, Ortiz-Tudela E, Rol MA, et al: Uncovering different masking factors on wrist
skin temperature rhythm in free-living subjects. PLoS One 8:e61142, 2013
68. Giese-Davis J, Wilhelm FH, Tamagawa R, et al: Higher vagal activity as related to survival in
patients with advanced breast cancer: An analysis of autonomic dysregulation. Psychosom Med
77:346-355, 2015
69. Melillo P, Orrico A, Scala P, et al: Cloud-based smart health monitoring system for automatic
cardiovascular and fall risk assessment in hypertensive patients. J Med Syst 39:109, 2015
70. Higginson IJ, Carr AJ: Measuring quality of life: Using quality of life measures in the clinical
setting. BMJ 322:1297-1300, 2001

14ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.
Appendix

Table A1. Patient Sociodemographic and Clinical Characteristics

Characteristic Patients
Age, years, median (range) 60 (48-72)
Sex
Male 5
Female 6
BMI
18.5-25 8
< 18.5 or > 25 3
WHO performance status
0 3
1 8
Primary tumor site
Colon-rectum 5
Pancreas 6
Metastatic sites
None 1
Liver 8
Lung 1
Peritoneum 1
Comorbidities
None 8
Diabetes 3
Prior surgery for
No prior surgery 3
Primary tumor only 5
Primary tumor and metastases 3
Prior chemotherapy
Adjuvant and metastatic 6
Adjuvant only 4
Metastatic only 1
No. of prior chemotherapy protocols for metastatic disease
0 4
1 4
≥2 3
NOTE. Data presented as No. unless otherwise noted.

ascopubs.org/journal/cci JCO™ Clinical Cancer Informatics 15

Downloaded from ascopubs.org by 91.160.130.83 on December 4, 2021 from 091.160.130.083

Copyright © 2021 American Society of Clinical Oncology. All rights reserved.