Content uploaded by Balazs Krisztian Fule
Author content
All content in this area was uploaded by Balazs Krisztian Fule on Nov 03, 2014
Content may be subject to copyright.
Conclusions
•Remote monitoring of physiology in
children is feasible.
•Accuracy of SpO2 and heart rate
measurements were acceptable
•Accuracy of respiratory rate measurements
improved after software update, however
did not yet reach target level. Further
improvements are needed.
•Development of specific paediatric sensors
can help to improve patient comfort and
compliance.
Reference
1. Nangalia V, Critical Care 2010;14:233.
Acknowledgements
The sensors , gateways and the software
platform for data visualization and recording
was provided by Isansys Lifecare Ltd.
Data analysis and reporting was performed
by the authors independently.
No financial conflict of interest of the authors
declared.
Remote Monitoring of Physiologic Parameters is Feasible in Children
Background
Remote monitoring of physiologic
parameters has been established in adults;
however there is a lack of paediatric data.
We have conducted a pilot study to assess
feasibility of remote monitoring in children.
Aims
• To prove feasibility of remote monitoring in
children.
• To assess duration and accuracy of data
collection
• To compare data before and after an
interim software (SW) update
Methods
Written consent was gained and the study
was registered with the Institutional Review
Board.
Patients on a cardiology ward were
monitored with remote ECG (LifeTouch®,
Isansys) and SpO2 (Wristox®, Nonin)
monitoring devices. (Figure 1.)
The recorded physiologic data was
transmitted via Bluetooth connection to a
gateway located near to the bedside. The
data was further transmitted from the
gateways to the hospital network on Wi-Fi.
Periods of normal transmission were
identified retrospectively by a PICU
physician. Heart rate (HR), respiratory rate
(RR) and SpO2 from the remote sensors was
compared to routine intermittent
measurements (Phillips SureSigns) recorded
by bedside nurses, during periods of normal
transmission (Figure 2).
Participant feedback was collected.
We have calculated
• bias (routine monitoring - remote data)
• precision (SD of bias),
• accuracy (root mean square, ARMS) and
• percent error (PE)
ARMS<4% for SpO2 , PE<10 for HR and RR
was considered as acceptable agreement
between measurements. Results for HR and
RR were compared before and after a
software update for LifeTouch.
Figure 1. Remote sensors used in the study: ECG
(LifeTouch®, Isansys) and SpO2 (Wristox®, Nonin)
Fule BK, McTiernan C, Cameron T, Laker S, Jarvis N, Tucker K, Matam BR, Duncan HP
Paediatric Intensive Care Unit, Birmingham Children’s Hospital, Birmingham UK
Figure 3. Bias , precision and percent error for heart rate
before and after SW update. Red line shows limit of
agreement (10% PE)
Figure 4. Bias , precision and percent error for respiratory
rate before and after SW update. Red line shows limit of
agreement (10% PE)
Figure 5. Bias , precision and percent error for respiratory
rate before and after SW update. Red line shows limit of
agreement (ARMS 4).
Results
1818 hours of monitoring in 41 patients (age
between 5 months and 14 years) during 43
sessions were recorded. 22 sessions were
recorded following software update. 505
hours (28%) of data transmission was
considered as normal.
Agreement between standard and remote
monitoring:
•HR showed good agreement (PE 6.9 vs
3.7 pre-, and post SW update, Figure 3.)
•SpO2 was acceptable (ARMS=3.88 overall;
3.49 pre-, 4.41 post SW update Figure 4.).
•RR measurements improved following SW
update but remained inaccurate (PE 31 vs
22%. Figure 5.).
Participant feedback suggested discomfort
being the main reason for not wearing the
sensors. Remote ECG sensor was better
tolerated than pulse-oximetry.
Figure 2. A typical session of remote monitoring.
Continuous lines represent remote monitoring, markers
represent routine intermittent monitoring. Normal
transmission is indicated by a yellow line, patient not
wearing the ECG sensor by a purple line at the bottom of
HR chart.
