Wearables for Home Post-Operative Monitoring: Proof of Concept
We hypothesise that wearables could transform perioperative
care, including through facilitating early-discharge, a key tenant
of Enhanced Recovery After Surgery (ERAS), by enabling remote
patient monitoring of physiological & functional variables.
Utilising a commercially available fitness tracker (Charge 4,
Fitbit ®) and Application Program Interface (API) it was possible
to import measured data into R for further statistical
Both physiological (heart rate (HR)) and functional (sleep
duration, step count) variables were accessible. Whilst the
device includes a pulse oximeter, this is unavailable via the API
Owing to COVID-19 this methodology was trialled on a single
volunteer, over a 24-hour period, and serves only as a proof of
Other methods of outlier detection could include reference
ranges (e.g. NEWS2 –visualised in Figure 1) or comparison with
peers (e.g. 95th centiles). Following identification this may
prompt increased follow-up, e.g. telephone call or ‘push-
notification’ to patient outcome reporting app.
Furthermore, captured data contributes to a core & extended
perioperative outcome set (Myles et al., 2016) for further high-
Remote patient monitoring, via wearable fitness trackers, is
achievable and has potential to provide meaningful benefit, by
facilitating early discharge and detection of complications, as
part of an ERAS pathway. Following proof of concept we hope to
trial this tool clinically.
Table 1: Hypothesised Parameter Abnormalities
HR Steps Sleep SpO2*
Pain ↑ ↓ ↓ -
Infection ↑-/↓ --/↓
↑- - ↓
*Not available via API at present
Anaesthetic Registrar, Surrey & Sussex Healthcare NHS Trust
References & Code
R code for data import & analysis freely available at: https://ctomlinson.net/tag/EBPOM2020/
Royal College of Physicians. National Early Warning Score (NEWS) 2: Standardising the assessment of acute-illness severity in the NHS. Updated report of a working party. London: RCP, 2017.
P. S. Myles, M. P. W. Grocott, O. Boney, S. R. Moonesinghe, on behalf of the COMPAC-StEP Group. Standardizing end points in perioperative trials: towards a core and extended outcome set, BJA: British Journal of Anaesthesia, Volume 116, Issue 5, May 2016, Pages 586–589.
From the recorded parameters we may expect to see patterns
of abnormalities consistent with post-operative complications,
for example pain illustrated by a tachycardia and reduction in
activity and sleep (Table 1). This could be greatly enhanced with
SpO2, via an API update.
Figure 1: Heart Rate, Step Count & Time Slept
Figure 1 displays Heart Rate (red), Steps (blue) 15-minutely and
sleep (grey shading) over 24 hours. HR reference ranges from
the National Early Warning Score 2 are represented by yellow,
orange & red horizontal ribbons, corresponding to NEWS scores
of 1, 2 & 3, respectively (RCP, 2017).
Once can easily ‘eyeball’ a resting HR of ~53, a peak of ~145
during exercise and that HR was generally within normal limits
on the NEWS2 score. They undertook two periods of exercise,
lasting ~90 & ~60 mins, respectively, and slept for ~7.5 hours,
with little disturbance.