REVIEW ARTICLE OPEN Clinical outcomes of digital sensor alerting systems in remote monitoring: a systematic review and meta-analysis Fahad M. Iqbal 1,2 ✉ , Kyle Lam 1,2 , Meera Joshi 1,2 , Sadia Khan 3 , Hutan Ashrafian 1,2 and Ara Darzi 1,2 Advances in digital technologies have allowed remote monitoring and digital alerting systems to gain popularity. Despite this, limited evidence exists to substantiate claims that digital alerting can improve clinical outcomes. The aim of this study was to appraise the evidence on the clinical outcomes of digital alerting systems in remote monitoring through a systematic review and meta-analysis. A systematic literature search, with no language restrictions, was performed to identify studies evaluating healthcare outcomes of digital sensor alerting systems used in remote monitoring across all (medical and surgical) cohorts. The primary outcome was hospitalisation; secondary outcomes included hospital length of stay (LOS), mortality, emergency department and outpatient visits. Standard, pooled hazard ratio and proportion of means meta-analyses were performed. A total of 33 studies met the eligibility criteria; of which, 23 allowed for a meta-analysis. A 9.6% mean decrease in hospitalisation favouring digital alerting systems from a pooled random effects analysis was noted. However, pooled weighted mean differences and hazard ratios did not reproduce this finding. Digital alerting reduced hospital LOS by a mean difference of 1.043 days. A 3% mean decrease in all-cause mortality from digital alerting systems was noted. There was no benefit of digital alerting with respect to emergency department or outpatient visits. Digital alerts can considerably reduce hospitalisation and length of stay for certain cohorts in remote monitoring. Further research is required to confirm these findings and trial different alerting protocols to understand optimal alerting to guide future widespread implementation. npj Digital Medicine (2021)4:7 ; https://doi.org/10.1038/s41746-020-00378-0 INTRODUCTION With our ever-ageing population, a result of significant improve- ments in healthcare delivery, medicine, personal & environmental hygiene, a greater burden is placed on our primary and secondary care healthcare facilities 1 . The rising costs of healthcare delivery require novel strategies to improve healthcare service provision 2 , particularly one that proves to be cost-effective and is widely accepted by citizens. Telemedicine, a concept since the 1970s, has evolved to be synonymous with terms such as digital health, e-health, m-health, wireless health, and, remote monitoring, among others. Indeed, over 100 unique definitions have been uncovered for ‘telemedi- cine’, a variation, which is likely to be attributed to the progression of these technologies 3,4 . Remote monitoring allows people to continue living at home rather than in expensive hospital facilities through the use of non-invasive digital technologies (such as wearable sensors) to collect health data, support health provider assessment and clinical decision making 5 . Several randomised trials have demonstrated the potential for remote monitoring in reducing in-hospital visits, time required for patient follow-up, and hospital costs in individuals fitted with cardiovascular implantable electronic devices 6–8 . Vital signs including, heart rate (HR), respiratory rate (RR), blood pressure (BP), temperature, and oxygen saturations, are consid- ered a basic component of clinical care and an aide in detecting clinical deterioration; changes in these parameters may occur several hours prior to an adverse event 9,10 . With wearable sensors being light-weight, small, and discrete they can be powerful diagnostic tools for continuously monitoring important physiolo- gical signs and offer a non-invasive, unobtrusive opportunity for sensor alerting systems to remotely monitor patients, driving the potential to improve timeliness of care and health-related outcomes 11 . Feedback loops and alerting mechanisms allow for appropriate action following recognition of clinical deterioration. Current alerting mechanisms for remote monitoring include alert trans- mission to a mobile device; automated emails generated to a healthcare professional; video consultation; interactive voice responses; or web-based consultations 12 . The feedback loops can be relayed to nurses, pharmacists, physicians, counsellors, and physicians but also to patients 13 . Earlier recognition of deteriora- tion, through alerting mechanisms, has potential to improve clinical outcomes, such as hospitalisation, length of stay, mortality, and subsequent hospital visits, through earlier detection but has been inadequately studied. A recent systematic review reported outcomes for remote monitoring undertaken in individuals in the community with chronic diseases (e.g., hypertension, obesity, and heart failure), but many of the included studies were of low quality and under- powered; the meta-analyses were on obesity related intervention outcomes (body mass index, weight, waist circumference, body fat percentage, systolic blood pressure, and diastolic blood pressure), consisting of few studies 13 . Additionally, the evaluation of feedback loops and alerting mechanisms following recognition of abnormal parameters was not the main focus of this study, a pivotal phase where intervention could influence clinical out- comes. With the search performed in 2006, and the rapid evolution of such a field, an updated systematic review aimed at digital alerting mechanisms is warranted, with the inclusion of wider medical and surgical cohorts for generalisability. The aim of this systematic review is to identify studies evaluating digital 1 Division of Surgery, Imperial College London, St. Mary’s Hospital, London W2 1NY, UK. 2 Institute of Global Health Innovation, Imperial College London Faculty Building, South Kensington Campus, Kensington, London SW7 2AZ, UK. 3 Division of Cardiology, West Middlesex University Hospital, London TW7 6AF, UK. ✉ email: [email protected] www.nature.com/npjdigitalmed Seoul National University Bundang Hospital 1234567890():,;
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
REVIEW ARTICLE OPEN
Clinical outcomes of digital sensor alerting systems in remotemonitoring: a systematic review and meta-analysisFahad M. Iqbal 1,2✉, Kyle Lam 1,2, Meera Joshi1,2, Sadia Khan3, Hutan Ashrafian 1,2 and Ara Darzi 1,2
Advances in digital technologies have allowed remote monitoring and digital alerting systems to gain popularity. Despite this,limited evidence exists to substantiate claims that digital alerting can improve clinical outcomes. The aim of this study was toappraise the evidence on the clinical outcomes of digital alerting systems in remote monitoring through a systematic review andmeta-analysis. A systematic literature search, with no language restrictions, was performed to identify studies evaluating healthcareoutcomes of digital sensor alerting systems used in remote monitoring across all (medical and surgical) cohorts. The primaryoutcome was hospitalisation; secondary outcomes included hospital length of stay (LOS), mortality, emergency department andoutpatient visits. Standard, pooled hazard ratio and proportion of means meta-analyses were performed. A total of 33 studies metthe eligibility criteria; of which, 23 allowed for a meta-analysis. A 9.6% mean decrease in hospitalisation favouring digital alertingsystems from a pooled random effects analysis was noted. However, pooled weighted mean differences and hazard ratios did notreproduce this finding. Digital alerting reduced hospital LOS by a mean difference of 1.043 days. A 3% mean decrease in all-causemortality from digital alerting systems was noted. There was no benefit of digital alerting with respect to emergency department oroutpatient visits. Digital alerts can considerably reduce hospitalisation and length of stay for certain cohorts in remote monitoring.Further research is required to confirm these findings and trial different alerting protocols to understand optimal alerting to guidefuture widespread implementation.
npj Digital Medicine (2021) 4:7 ; https://doi.org/10.1038/s41746-020-00378-0
INTRODUCTIONWith our ever-ageing population, a result of significant improve-ments in healthcare delivery, medicine, personal & environmentalhygiene, a greater burden is placed on our primary and secondarycare healthcare facilities1. The rising costs of healthcare deliveryrequire novel strategies to improve healthcare service provision2,particularly one that proves to be cost-effective and is widelyaccepted by citizens.Telemedicine, a concept since the 1970s, has evolved to be
synonymous with terms such as digital health, e-health, m-health,wireless health, and, remote monitoring, among others. Indeed,over 100 unique definitions have been uncovered for ‘telemedi-cine’, a variation, which is likely to be attributed to the progressionof these technologies3,4. Remote monitoring allows people tocontinue living at home rather than in expensive hospital facilitiesthrough the use of non-invasive digital technologies (such aswearable sensors) to collect health data, support health providerassessment and clinical decision making5. Several randomisedtrials have demonstrated the potential for remote monitoring inreducing in-hospital visits, time required for patient follow-up, andhospital costs in individuals fitted with cardiovascular implantableelectronic devices6–8.Vital signs including, heart rate (HR), respiratory rate (RR), blood
pressure (BP), temperature, and oxygen saturations, are consid-ered a basic component of clinical care and an aide in detectingclinical deterioration; changes in these parameters may occurseveral hours prior to an adverse event9,10. With wearable sensorsbeing light-weight, small, and discrete they can be powerfuldiagnostic tools for continuously monitoring important physiolo-gical signs and offer a non-invasive, unobtrusive opportunity forsensor alerting systems to remotely monitor patients, driving the
potential to improve timeliness of care and health-relatedoutcomes11.Feedback loops and alerting mechanisms allow for appropriate
action following recognition of clinical deterioration. Currentalerting mechanisms for remote monitoring include alert trans-mission to a mobile device; automated emails generated to ahealthcare professional; video consultation; interactive voiceresponses; or web-based consultations12. The feedback loopscan be relayed to nurses, pharmacists, physicians, counsellors, andphysicians but also to patients13. Earlier recognition of deteriora-tion, through alerting mechanisms, has potential to improveclinical outcomes, such as hospitalisation, length of stay, mortality,and subsequent hospital visits, through earlier detection but hasbeen inadequately studied.A recent systematic review reported outcomes for remote
monitoring undertaken in individuals in the community withchronic diseases (e.g., hypertension, obesity, and heart failure), butmany of the included studies were of low quality and under-powered; the meta-analyses were on obesity related interventionoutcomes (body mass index, weight, waist circumference, body fatpercentage, systolic blood pressure, and diastolic blood pressure),consisting of few studies13. Additionally, the evaluation offeedback loops and alerting mechanisms following recognitionof abnormal parameters was not the main focus of this study, apivotal phase where intervention could influence clinical out-comes. With the search performed in 2006, and the rapidevolution of such a field, an updated systematic review aimed atdigital alerting mechanisms is warranted, with the inclusion ofwider medical and surgical cohorts for generalisability. The aim ofthis systematic review is to identify studies evaluating digital
1Division of Surgery, Imperial College London, St. Mary’s Hospital, London W2 1NY, UK. 2Institute of Global Health Innovation, Imperial College London Faculty Building, SouthKensington Campus, Kensington, London SW7 2AZ, UK. 3Division of Cardiology, West Middlesex University Hospital, London TW7 6AF, UK. ✉email: [email protected]
www.nature.com/npjdigitalmed
Seoul National University Bundang Hospital
1234567890():,;
alerting systems used in remote monitoring and describe theassociated clinical outcomes.
RESULTSStudy characteristicsA total of 2417 citations were retrieved through literaturesearches. An additional two articles were found from bibliographycross-referencing. Full-text review was performed for 128 articleswith 33 meeting the inclusion criteria for analysis, of which, 21were randomised controlled trials with the remaining prospectiveor retrospective studies. Of the 33 included studies, 23 allowed formeta-analysis. The characteristics of included studies is shown inTable 1. A PRISMA flow diagram can be seen in Fig. 1.
Hospitalisation and inpatient admissionsSix studies demonstrated a mean decrease in hospitalisation/inpatient admissions of 9.6% (95% CI 4.9–14.3%, I2= 96.4%, Fig. 2)favouring digital alerting systems from a pooled random effectsanalysis. However, pooled WMD reported no change in hospita-lisation from six studies (WMD 0.061; 95% CI −0.197–0.318, I2=78%)14–19. Pooled HRs for all-cause hospitalisation similarlydemonstrated no significant difference (HR 0.916; 95% CI0.781–1.074, I2= 0%)20,21.Six additional studies, reporting on cardiovascular related
hospitalisation, revealed no significant relationship with digitalalerting (mean decrease 10.1%; 95% CI −24.9–4.7%, I2= 95.6%and pooled HRs 0.907; 95% CI 0.757–1.088, I2= 2.4%)20,22–25
MortalityA total of 16 papers were included; pooled random effects analysisdemonstrated a 3% mean decrease in all-cause mortality fromdigital alerting systems (95% CI 2–3%, Fig. 3) from 12 studies;there was high heterogeneity with this analysis (I2= 94.4%).However, pooled HRs of five studies reported no change in all-cause mortality (HR 0.89; 95% CI 0.79–1.01, I2= 30.3%)20,21,25–27.A sub-group cardiovascular cohort pooled random effect
analysis failed to demonstrate a relationship between cardiovas-cular mortality and digital alerting (mean decrease 0.9%, 95% CI−0.6–2.4%, I2= 25.7%)20,24.
Length of stayTen studies were included; digital alerting reduced hospital LOSby a mean difference of 1.043 days (95% CI 0.028–2.058 days, p <0.001, I2= 95.5%)14–18,24,28–31. Three studies reported on LOS inchronic obstructive pulmonary disease (COPD) cases found nobenefit of digital alerting (mean difference 0.919 days; 95% CI−1.878–3.717 days, p= 0.213, I2= 35.3%)17,30,31.
Emergency department visitsEight studies were included; pooled random effects analysis of EDvisits demonstrated no statistical benefit of digital alerting (meandifference 0.025; 95% CI −0.032–0.082, I2= 51.8%)14,16–19,22,28,32.
Outpatient and office visitsFive studies were included; pooled random effects analysisdemonstrated no benefit of digital alerting (mean difference0.223 days; 95% CI −0.412–0.858, I2= 95.7%)14,17,18,28,32. Sub-group data from Ringbaek et al. (respiratory and non-respiratory)and Lewis et al. (primary care chest and non-chest related visits)were combined for this analysis.Similarly, no statistically significant mean decrease in outpatient
visits was noted from three additional studies27,33,34.
Sub-group analysis of a respiratory cohort demonstrated amean difference of 1.346 days (95% CI 0.102–2.598, I2=93.8%)17,28.
Risk of bias assessmentThe assessment of risk of bias for included randomised trials ispresented in Fig. 4.Allocation was random across all 20 studies with 15
adequately stating the method used for generating randomsequence17–21,28,30,31,33–39. Vianello et al.31 utilised a dedicatedalgorithm to check for imbalances for baseline variables withclear randomisation sequence methods detailed. However,concealment measures were not mentioned, resulting in ajudgement of ‘some concerns’ for risk of bias for randomisa-tion. Three additional studies were given the same judgementdue to lack of concealment descriptions15,19,35. Ringbaeket al.17 clearly described their method for randomisation butinformation on concealment was not given and baselinedemographic differences were noted between groups; as such,randomisation was judged to be at high risk of bias. Similarly,randomisation for Scherr et al.24 was deemed to be at high riskof bias.Sink et al.39 blinded participants with digital alerts not
forwarded to healthcare providers in the control arm. This, aresult of their automated telephone intervention collecting self-reported symptom data rather than continuous physiologicalparameter recording through wearable sensors or smart devices,as utilised by the other trials, made participant blinding possible. Alow risk of bias was, therefore, judged.The risk of attrition bias was deemed low across all included
studies with missing numbers clearly reported and deemed to nothave impacted the overall results. There was mostly a completefollow-up of all participants.Insufficient information was provided to assess whether other
important risk of biases exists in four studies so were judged assome concerns17,20,23,31. Basch et al.35 clustered groups intocomputer experienced and computer in-experienced but numbersacross various arms were unequal for selected outcome measures.Therefore, a judgement of high risk of bias was given. Comparably,Scherr et al.24 performed multiple analyses with both intention-to-treat and per-protocol. Only the latter revealed significant resultsfavouring their telemonitoring system.Overall, seven studies were deemed to be at low risk, ten
studies had some concerns, and the remaining were judged ashigh risk of bias.
Alerting mechanisms and response to alertsTable 2 summaries the alerting mechanisms utilised within thestudies. Mechanisms include text messaging, email notifications,alerts on telemonitoring hubs/web-based platforms, as well as,trialling audible alerts to study participants rather than healthcareprofessionals.
DISCUSSIONThis meta-analysis provides evidence that digital alerting mechan-isms used for remote monitoring are associated with reductions inhospitalisation and inpatient admissions. All pooled studies wereprospective with the majority being randomised trials. However,most studies included were low in quality (Table 1) and only twostudies had follow-up periods beyond 12 months20,21. Theincluded studies were particularly heterogenous meaning thatthe results should be interpreted cautiously but may suggest thatdigital alerting in remote monitoring could be beneficial across avariety of patient cohorts. Pooled mean differences, however, didnot reproduce this finding. The included studies consisted oflonger follow-up periods14–19. One possible explanation could be
F.M. Iqbal et al.
2
npj Digital Medicine (2021) 7 Seoul National University Bundang Hospital
1234567890():,;
Table1.
Characteristicsofincluded
studieswithqualityscore
(Jad
ad&New
castle-O
ttaw
aScale).
Author
Year
Title
Journal
Design
NFo
llow-up
Score
Baker
etal.26
2013
Effectsofcare
man
agem
entan
dtelehealth:A
longitudinal
analysisusingmed
icaredata
JAm
GeriatSoc
Retrospective
3534
2years
High
Basch
etal.35
2016
Symptom
monitoringwithpatient-reported
outcomes
during
routinecancertreatm
ent:Arandomised
controlledtrial
JClin
Oncol
RCT
766
6monthsforqualityoflife;
12monthsformortality
Lowa
Bekelman
etal.36
2015
Prim
aryResultsofthePa
tien
t-Cen
tred
Disease
Man
agem
ent
(PCDM)forHeart
Failu
reStudy:
ARan
domised
Clin
ical
Trial
JAMAIntern
Med
RCT
384
12months
Lowa
Biddisset
al.55
2009
Pred
ictingnee
dforinterven
tionin
individualswithco
ngestive
heart
failu
reusingahome-based
telecare
system
JTelemed
Telecare
Prospective
4518
(5)months(average,
SD)
Moderate
Bohm
etal.20
2016
Fluid
statustelemed
icinealerts
forheartfailu
re:a
randomised
controlledtrial
EurHeartJ
RCT
1002
18months
Lowa
Calvo
etal.37
2014
Ahometelehealthprogrammeforpatients
withsevere
COPD
:Th
ePR
OMET
Estudy
RespirMed
Cluster
RCT
597months
Lowa
Chen
etal.14
2013
Clin
ical
outcomean
dco
st-effective
nessofasynch
ronous
telehealthserviceforseniors
andnonseniors
with
cardiovasculardiseases:quasi-e
xperim
entalstudy
JMed
Internet
Res
Prospective
141
6monthsbefore
andafter
Moderate
Del
Hoyo
etal.33
2018
AWeb
-Based
Teleman
agem
entSystem
forIm
provingDisease
Activityan
dQualityofLife
inPa
tien
tsWithComplex
Inflam
matory
BowelDisease:Pilo
tRan
domised
ControlledTrial
JMed
Internet
Res
RCT
6324
wee
ksLo
wa
Den
iset
al.56
2019
Prospective
studyofaweb
-med
iatedman
agem
entoffebrile
neu
tropen
iarelatedto
chem
otherap
y(Bioco
nnect)
Supp
ortCa
reCa
ncer
Prospective
413wee
ksModerate
Godleskiet
al.45
2012
Hometelemen
talhealthim
plemen
tationan
doutcomes
using
electronic
messaging
JTelemed
Telecare
Prospective
766monthsbefore
andafter
Moderate
Heidbuch
elet
al.15
2015
EuroEco(EuropeanHealthEconomicTrialo
nHomeMonitoring
inICD
Patien
ts):Aprovider
perspective
infive
European
countriesonco
stsan
dnet
finan
cial
impactoffollo
w-upwith
orwithoutremote
monitoring
EurHeartJ
RCT
303
24(±2)
months
Lowa
Kotooka
etal.21
2018
Thefirstmulticen
ter,randomised
,controlledtrialofhome
telemonitoringforJapan
esepatients
withheart
failu
re:h
ome
telemonitoringstudyforpatients
withheart
failu
re(HOMES
-HF)
HeartVessels
RCT
181
15(0–31
)months
(mean,ran
ge)
Lowa
Leeet
al.38
2019
Telemed
icine-Based
Rem
ote
HomeMonitoringAfter
Liver
Tran
splantation:R
esultsofaRan
domised
Prospective
Trial
Ann
Surg
RCT
100
90days
Lowa
Lewiset
al.28
2010
Does
hometelemonitoringafterpulm
onaryrehab
ilitation
reduce
healthcare
use
inoptimised
COPD
apilo
trandomised
trial
COPD
RCT
4026
wee
kstelemonitoring+
26wee
kswithout(total
52wee
ks)
Lowa
Licskaiet
al.46
2013
Developmen
tan
dpilo
ttestingofamobile
healthsolutionfor
asthmaself-man
agem
ent:asthmaactionplansm
artphone
applicationpilo
tstudy
CanRespirJ
Prospective
223monthsbefore
andafter
Moderate
Luthje
etal.22
2015
Arandomised
studyofremote
monitoringan
dfluid
monitoringfortheman
agem
entofpatients
withim
planted
cardiacarrhythmia
dev
ices
Europa
ceRCT
176
15months
Lowa
Martin-Lesen
de
etal.16
2017
Telemonitoringin-homeco
mplexch
ronic
patients
from
primarycare
inroutineclinical
practice:
Impactonhealthcare
resources
use
EurJGen
Pract
Prospective
2812
monthsbefore
andafter
Moderate
McElroyet
al.29
2016
Use
ofdigital
healthkits
toreduce
read
missionaftercardiac
surgery
JSurg
Res
Prospective
443
30days
Moderate
F.M. Iqbal et al.
3
Seoul National University Bundang Hospital npj Digital Medicine (2021) 7
Table
1continue
d
Author
Year
Title
Journal
Design
NFo
llow-up
Score
Mousa
etal.34
2019
ResultsofTelehealthElectronic
MonitoringforPo
stDisch
arge
Complicationsan
dSu
rgical
Site
Infectionsfollo
wingArterial
Revascu
larizationwithGroin
Incision
Ann
Vasc
Surg
RCT
3030
days
Lowa
Oeffet
al.47
2005
[Monitoringmultiple
cardiovascularparam
etersusing
telemed
icinein
patients
withch
ronic
heart
failu
re]
Herzschrittm
acherther
Elektrop
hysiol
Prospective
2412
monthsbefore
andafter
Moderate
Pedoneet
al.23
2015
Efficacy
ofaPh
ysician-Led
Multiparam
etricTelemonitoring
System
inVe
ryOld
AdultswithHeart
Failu
reJAm
GeriatSoc
RCT
906months
Lowa
Pinnock
etal.30
2013
Effectiven
essoftelemonitoringintegratedinto
existingclinical
services
onhospital
admissionforexacerbationofch
ronic
obstructivepulm
onarydisease:R
esearcher
blin
d,m
ulticen
tre,
randomised
controlledtrial
BMJ
RCT
256
12months
Lowa
Pinto
etal.48
2010
Hometelemonitoringofnon-in
vasive
ventilationdecreases
healthcare
utilisationin
aprospective
controlledtrialof
patients
witham
yotrophic
lateralsclerosis
JNeurolN
eurosurg
Psychiatry
RCT
393years
Lowa
Ringbaeket
al.17
2015
Effect
oftelehealthcare
onexacerbationsan
dhospital
admissionsin
patients
withch
ronic
obstructivepulm
onary
disease:a
randomised
clinical
trial
IntJCh
ronObstruct
Pulm
onDis
RCT
281
6months
Lowa
Santiniet
al.49
2009
Rem
ote
monitoringofpatients
withbiven
triculardefi
brillators
throughtheCareLinksystem
improve
sclinicalman
agem
entof
arrhythmiasan
dheart
failu
reep
isodes
JInterv
Card
Electr
Prospective
6711
(6–20
)months
(med
ian,ran
ge)
Moderate
Scherret
al.24
2009
Effect
ofhome-based
telemonitoringusingmobile
phone
tech
nologyontheoutcomeofheart
failu
repatients
afteran
episodeofacute
decompen
sation:ran
domised
controlledtrial
JMed
Internet
Res
RCT
108
6months
Lowa
Seto
etal.18
2012
Mobile
phone-based
telemonitoringforheart
failu
reman
agem
ent:Arandomised
controlledtrial
JMed
Internet
Res
RCT
100
6months
Lowa
Sinket
al.39
2018
Effectiven
essofanovel,au
tomated
telephoneinterven
tionon
timeto
hospitalisationin
patients
withCOPD
:Arandomised
controlledtrial
JTelemed
Telecare
RCT
168
8months
Higha
Smee
tset
al.25
2017
Bioim
ped
ance
Alertsfrom
CardiovascularIm
plantable
Electronic
Dev
ices:O
bservational
StudyofDiagnostic
Relev
ance
andClin
ical
Outcomes
JMed
Internet
Res
Prospective
282
34months(m
ean)
High
Steven
tonet
al.32
2012
Effect
oftelehealthonuse
ofseco
ndarycare
andmortality:
Findingsfrom
theWhole
System
Dem
onstratorcluster
randomised
trial
BMJ
Cluster
RCT
3154
12months
Lowa
Steven
tonet
al.27
2016
Effect
oftelehealthonhospital
utilisationan
dmortalityin
routineclinical
practice:
Amatch
edco
ntrolcohortstudyin
anearlyad
optersite
BMJOpen
Retrospective
1432
10.4
months(ave
rage)
High
Vianello
etal.31
2016
Hometelemonitoringforpatients
withacute
exacerbationof
chronic
obstructivepulm
onarydisease:a
randomised
controlledtrial
BMCPu
lmMed
RCT
334
12months
Lowa
Yountet
al.19
2014
Arandomised
trialofwee
klysymptom
telemonitoringin
advancedlungcancer
JPa
inSymptom
Man
age
RCT
253
12wee
ksLo
wa
RCTrandomised
controlledtrial.
a Jad
adscale.
F.M. Iqbal et al.
4
npj Digital Medicine (2021) 7 Seoul National University Bundang Hospital
that difference in cohorts analysed, with the latter containingmore individuals suffering from chronic medical conditions (e.g.,COPD, heart failure) compared to the former, which encompassedacute surgical cohorts with shorter follow-up periods.A study in 2016 reported that avoidable hospitalisation
increased by a factor of 1.35 for each additional chronic conditionand 1.55 for each additional body system affected40,41. Clearly, achronic disease cohort is particularly susceptible to recurrenthospitalisations and, while digitisation may play role in changinghealthcare delivery, hospital departmental factors (e.g., seniority of
clinician reviewing, busyness of department, community servicedelivery) and external factors (e.g., patient education andactivation, behavioural insights towards digitisation, social supportavailable) are likely to significantly contribute and may impactwidespread deployment of novel digital technologies42.Hospital length of stay was found to be reduced with digital
alerting. This is likely a result of earlier recognition of deteriorationresulting in prompt clinical review and treatment administration; arecent systematic review concluded that digital alerts similarlyreduced hospital length of stay in sepsis by 1.3 days43. This review
Fig. 2 Forest plot hospitalisation. Forest plot of studies reporting hospitalisation and inpatient admissions.
Fig. 1 PRISMA flow diagram. Search and study selection process for this review.
F.M. Iqbal et al.
5
Seoul National University Bundang Hospital npj Digital Medicine (2021) 7
adds further support to the literature demonstrating the benefit ofdigital alerting in remote settings across medical and surgicalcohorts.A small reduction in all-cause mortality from digital alerting
systems was noted. A relationship not reproduced from pooledhazard ratios, which may be explained by the difference of studyqualities included in the analyses. Only three studies includedwere high quality; Of which, significant weighting was given to a2013 study by Baker et al.25–27 utilising the Health Buddytelemonitoring platform, which has since become obsolete. Earlyiterations of digital alerting and telemonitoring platforms maysuffer significant pitfalls, preventing successful use, a possibleexplanation for the described relationships.Visits to the emergency departments demonstrated no benefit
of digital alerting mechanisms from pooled mean differences.Earlier recognition of deterioration should prevent presentation toemergency departments and inpatient hospitalisations with non-urgent reviews scheduled for outpatient visits. Despite this, therewas no change in overall outpatient or clinic visits. However,respiratory sub-group data did demonstrate a reduction inoutpatient visits though the analysis was a culmination of onlytwo studies. Further randomised trials for specific medical cohortsand conditions may address the benefit of digital alerting inaffecting outpatient visits. Additionally, research capturing sched-uled and unscheduled presentations to hospital, includingemergency department visits, outpatient visits, and hospitalisa-tions would be vital in addressing whether workloads can bealtered across these departments.Despite the significance of the outcomes assessed, our analysis
had limitations based on the variety of methodologies used andoverall study quality, with the majority scoring low. One of thechallenges of this review was the relatively broad study into theeffectiveness of digital alerting on clinical outcomes. While thisallowed us to examine the similarities across various alertingmechanisms, it created significant heterogeneity. The justificationof which was to determine effectiveness of alerting toolspragmatically across various cohorts, determining their overallefficacy as a tool to assist clinical decision making. Nevertheless,
this limitation, largely a result of the paucity of high-qualityliterature, is to be acknowledged. The paucity in high quality,robust, literature limits the conclusions drawn in our review. Theincluded non-randomised trials, due to their observational nature,are prone to selection biases, particularly pre-post implementationdesigns, which can be theoretically confounded by longitudinalchanges in healthcare provision. Moreover, integrated feedbackloops and responses to alerts are likely to feed into the Hawthorneeffect44, an additional source of bias. Nonetheless, a great numberof variables allowing for comprehensive characterisation of thedigital alerting literature has been conducted which, to theauthors’ knowledge, has not been undertaken previously.Further research to answer several important questions is
required. First, the optimal frequency of alerting; a range ofremote monitoring schedules were utilised for data collection,including continuous15, daily16–18,21–25,27,29,30,34,38,39,45–49, onlyduring office working hours (Monday–Friday)28,31,32, and weekly19.Indeed, given the diverse methodology in the literature, responsetime variation would be expected with potential for missing earlysigns of acute deterioration. Studies with less intense monitoringschedules may be suited for a cohort of individuals less prone toacute deterioration, regardless, a ‘window of opportunity’ presentsitself for missing clinical deterioration in less frequent schedules.Second, which team members to be alerted and what nature ofalert to be utilised. Alerts were frequently generated when pre-established thresholds, often tailorable, were breeched or forconcerning responses to symptom questionnaires resulting inweb-platform-based notifications, email alerts, telephone calls,texts, or pagers sent to members of a healthcare team (Table 2). Incontrast, Santini et al.49 used audible alarms to alert patients whenthresholds were breeched, empowering individuals to contacttheir responsible physician for further assessment. It is unlikelythat one type of alert will be suitable for all individuals but furtherwork identifying the most rapidly acknowledged and actionablealerts is required, including the exploration of alerts sent toindividuals alongside healthcare professionals.In conclusion, this review provides evidence that digital alerts
used in remote monitoring can reduce hospital length of stay,
Fig. 3 Forest plot mortality. Forest plot of studies reporting all-cause mortality.
F.M. Iqbal et al.
6
npj Digital Medicine (2021) 7 Seoul National University Bundang Hospital
mortality, and may reduce hospitalisations. Digital technologiescontinue to innovate and have the capacity to change currenthealthcare provision, particularly in the current COVID era. There isneed for large, robust, multicentre, randomised trials studyingdigital alerting mechanisms in a varied cohort of individuals. Trialsshould seek to cycle different alerting protocols to understandoptimal alerting to guide future widespread implementation notonly within secondary and tertiary care settings but, importantly,in primary care, as implementation of new technologies withinhome settings has potential to truly revolutionise healthcaredelivery.
METHODSSearch strategy and databasesThis systematic review was conducted in accordance to thePreferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines50. The review was registered at theInternational Prospective Register of Systematic Reviews (PROS-PERO ID: CRD42020171457).
A systematic search was performed using electronic databasesthrough Ovid in Medline, EMBASE, Global health, health manage-ment information consortium (HMIC), and PsychINFO databaseswithout language restriction. The appropriate MeSH terms andfree text all field search was performed and combined withappropriate Boolean operators for “home”, “monitoring”, “remotesensing”, “self-monitor*”, “self-track*”, “remote monitor*”, “homemonitor*”, “biosensing techniques”, “wireless technology”, “tele-medicine”, “monitoring, physiologic”, “monitoring, ambulatory”,“home care services”, “ehealth”, “mhealth”, “telehealth”, “digital”,“mobile”, “social networking”, “internet”, “smartphone”, “cellphone”, “wearable electronic devices”, “internet”, “electronicalert*”, “alert*”, “messag*”, “text messaging”, “inform”, “commu-nicat*”, “communication”, “patient-reported outcome measures”,“outcome and process assessment”, “outcome”, “treatment out-come”, “outcome assessment”, “fatal outcome”, “adverse outcomepathways”, “patient outcome assessment”, “morbidity”, “mortality”,“length of stay”, “patient admission”, “readmission”. Furtherstudies not captured by the search were identified throughbibliometric cross-referencing.
Fig. 4 Risk of bias. Graphical display of the risk of bias results.
F.M. Iqbal et al.
7
Seoul National University Bundang Hospital npj Digital Medicine (2021) 7
Table2.
Studych
aracteristicsofalertingmechan
ismsan
dresponses.
Study
Cohort
Dataco
llected
Digital
alertingmechan
ism
Response
toalerts
Control
Baker
etal.26
HF;
COPD
;DM
Vital
signs;symptom
questionnaire;
men
talhealthquestionnaire;
HealthBuddyelectronic
dev
icewithfour
buttonsto
collect
dataan
dupload
edto
aweb
-portal
whichrisk
stratifies
responses.
Careman
ager
review
:specifics
not
men
tioned
.Retrospective
lymatch
ed
Basch
etal.35
Onco
logy
Self-reported
symptoms
Self-reportingthroughweb
-based
interface(STA
R).E-mailalerts
triggered
when
asymptom
worsen
edby>2points
orreached
anab
solute
grade>3.
Nurses
perform
edinterven
tions:(1)
telephoneco
unselling,(2)
med
ication
chan
ges,(3)
Emergen
cy/hospital
referral)
Usual
clinic
visits
withcliniciansto
discu
sssymptoms.
Bekelman
etal.36
HF
BP;
HR;weight;self-reported
symptoms;mood
Daily
telemonitoringusinghome-based
equipmen
t.Th
etelemonitoringsystem
assigned
arisk
toeach
response
onthe
system
.
Med
ium-riskindicators
werereview
edbynurses
forfurther
action.A
llhigh-risk
indicators
wereactedonbyco
ntacting
thepatientforassessmen
t.
Usual
care
Biddisset
al.55
HF
BP;
HR;weight;qualityoflife
questionnaire;symptom
questionnaire
Biometrics
entereddaily
into
the
‘Doc@
home’healthmonitor.Th
edatawere
tran
smittedat
nightthroughtelephone.
Alertsgen
erated
ifpre-estab
lished
thresholdscrossed
.
Monitoringpractitionersco
ntacted
patientforfurther
assessmen
t.-
Bohm
etal.20
HF
Intrathoracicfluid
statusmonitoring
OptiVolfluid
index
alert,ch
anges
inthoracicim
ped
ance
resultingfrom
accu
mulationofintrathoracicfluid
gen
erated
atext
messagealertto
responsible
physician.
Datawerereview
edremotely,a
ndthe
patientco
ntacted
within
2workingdays
byphoneto
evaluatean
dtake
appropriatemeasures
Usual
care
withouttelemonitoring.
Calvo
etal.37
COPD
Oxygen
saturation;HR;B
P;spirometry;p
eakexpiratory
flow
Daily
monitoringofbiometrics
tran
sferred
throughTele-M
odem
™to
clinical
monitoringteam
.Aredalertwas
gen
erated
ifpre-estab
lished
thresholds
werebreeched
inMPM
™callcentre
system
.
Anurseco
ntacted
thepatientto
verify
thealert.Fo
llowingthis,thealertwas
escalatedto
aPn
eumologist.Actions
include:
(1)telephonead
vice,(2)
home
visits,(3)
emergen
cydep
artm
entvisits
Usual
care
Chen
etal.14
Coronaryheart
disease;
HF;
arrythmia;an
gina;
synco
pe;
DM
BP;
HR;EC
G;oxygen
saturations;
bloodgluco
seReal-tim
etran
smissionofbiometrics
tohealthreco
rdcloudsunder
synch
ronous
surveillance
bytheTelehealthCen
tre.
Alertingmechan
ism
notspecified
.
Nursecase
man
agersco
ntacted
the
patientwhen
abnorm
aldatatran
smitted
withad
vice
ascertained
from
acardiologist.
Pre-im
plemen
tation
Del
Hoyo
etal.33
Inflam
matory
bowel
disease
Weight;vitalsigns;qualityoflife
NOMHADweb
-based
homeplatform
used.
Electronicco
mmunicationco
uld
take
place
betwee
nhealthcare
provider
andusers.
Individualised
alerts
weregen
erated
for
abnorm
alvalues.
After
receivingan
alert,thespecialised
med
ical
staff,reco
mmen
ded
action
plans:(1)med
icationad
justmen
t,(2)
telephonecalls,(3)
in-personvisits
Usual
care
inacco
rdan
cewithlocal
andnational
guidelines.
Den
iset
al.56
Onco
logy
Temperature;symptom
questionnaire
Bioco
nnectweb
applicationallowingdaily
biometrictran
smission.Ifalgorithmic
thresholdstriggered
,automatic
notificationsweresentto
thephysician
Med
ical
team
calledthepatientfor
assessmen
t.Actionsinclude:
(1)quick
planned
hospitalisation(byp
assED
),(2)
stay
athomean
dbloodtest
taken,(3)
antibioticad
ministration
-
Godleskiet
al.45
Men
talhealth
Symptom
andbeh
aviour
questionnaire;substan
ceab
use
questionnaire
HealthBuddyelectronic
messagingdev
ice
usedto
answ
erquestionsdailybypressing
largebuttonsonfrontofdev
ice.
Nurse
practitioner
review
edtran
smitteddataan
dco
ntacted
thepatientbytelephonefor
concerningresponses.
Actionsincluded
:(1)telephone
assessmen
t,(2)med
icationad
justmen
t,(3)inpatientvisit,(4)em
ergen
cydep
artm
entvisit
Pre-im
plemen
tation
Heidbuch
elet
al.15
CIED
CIED
metrics
Continuous,au
tomatic
remote
monitoring
withfreq
uen
cyofdataan
alysisan
dthe
response
toalerts
leftto
theinvestigator’s
discretion.
Alertsresultingin:(1)
hospital
admissions,(2)internal
discu
ssions,(3)
phonecalls,(4)
visits
tophysician,(5)
web
-rev
iew
Usual
care
(inoffice
regularvisits)
F.M. Iqbal et al.
8
npj Digital Medicine (2021) 7 Seoul National University Bundang Hospital
Table
2continue
d
Study
Cohort
Dataco
llected
Digital
alertingmechan
ism
Response
toalerts
Control
Kotooka
etal.21
HF
Weight;BP;
HR;b
odyco
mposition
Karad
aKarte™
telemonitoringsystem
that
tran
smitteddatadaily
tothecentral
web
server
viatheinternet.Ifpre-estab
lished
param
eter
thresholdsexceed
ed,
monitoringnurses
would
notify
the
physician
Physicianactionsincluded
:(1)
telephone
guidan
ce,(2)
med
icationch
anges,(3)
warningthreshold
adjustmen
t,(4)
hospital
admission
Usual
care
(inacco
rdan
cewiththe
2010
Japan
eseCircu
lationSo
ciety
Guidelines)
Leeet
al.38
Tran
splant(live
r)Temperature;B
P;bloodgluco
se;
weight;symptom
questionnaire;
med
icationuse.
Tabletwithbluetooth
dev
ices
tran
smitted
datadaily
tocentral
web
server
viathe
internet.Differen
talertingalgorithms
trialled.
Alertsresponded
bythenursecare
coordinatoran
descalatedto
care
provider.T
reatmen
torclinic
visit
initiatedifap
propriate.
Usual
care:logvitalsignsdaily
for
90days.Instructionsprovided
for
deterioration
Lewiset
al.28
COPD
Temperature;o
xygen
saturations;
HR;sym
ptom
questionnaire
Telemonitoringhub(Doco
bo™)
tran
smittingbiometrics
toaweb
-based
system
(doc@
HOME).Analertinge-mail
was
sentto
theco
mmunityteam
ifpre-
established
thresholdswereexceed
ed.
Thech
ronic
disease
man
agem
entteam
calledpatients
onreceiptofthisalerting
e-mailforfurther
assessmen
tduring
workinghours
(Mondays-Fridays,9a.
m.–5p.m
.)
Usual
care
Licskaiet
al.46
Asthma
Symptom
questionnaire;peak
expiratory
flow;m
edicationuse.
Theserver
analysed
biophysical
inputs
daily.E
-mailalerts
weresentformoderate
andhigh-riskdays;an
dasthmaco
ntrol
assessmen
tdisplayedas
green
,yello
wor
redzonewiththeco
rrespondingasthma
man
agem
entad
vice.
Asthmaco
ntrolassessmen
tdisplayedas
green
yello
worredzonean
dgave
appropriateasthmaman
agem
ent
advice.
Pre-im
plemen
tation
Luthje
etal.22
HFwithCIED
Bioim
ped
ance
measuremen
tsfrom
CIED
OptiVolfluid
index
alert,im
ped
ance
value
takendaily
andco
mpared
witharoving
reference
value-built
into
theCIED.
Phoneassessmen
twithalertingpatient
was
conducted
.Ifsignsofclinical
decompen
sation,admitto
hospital,ifno
signsofdecompen
sation,a
djust
diuretic
med
ication.
Usual
care
Martin-Lesen
de
etal.16
HF;ch
roniclungdisease
BP;
oxygen
saturations;HR;R
R;
weight;symptom
questionnaire
Daily
self-monitoringofparam
eterssent
usingsm
artphones
toaspecificWeb
-platform
.When
pre-estab
lished
threshold
values
werecrossed
,red
oryello
walerts
weretriggered
.
Notspecified
Pre-im
plemen
tation
McElroyet
al.29
Cardiacsurgery
Oxygen
saturation;H
R;BP;
weight;
symptom
questionnaire;ambulation
data;
adheren
ceto
med
ication
Abnorm
albiometrics,c
oncerningsurvey
responses,misseddigital
check-ins
registeredthroughadigital
healthkit
triggered
anau
tomated
notificationto
the
healthcare
team
.
Actionsinclude:
(1)video
chat/phone
call,
(2)med
icationad
justmen
t,(3)
education,(4)referral
tonurse
practitioner/doctor/em
ergen
cydep
artm
ent.
Disch
argeed
ucationbooklet;
med
icationed
ucationcards;
interactivevitalsignsan
dweightlog;
phonecallwithin
48hofdisch
arge
andev
ery4–
5daysfor30
days.
Mousa
etal.34
Peripheral
arterial
disease
(withgroin
incision)
Temperature;w
eight;BP;
oxygen
saturation;sym
ptom
questionnaire;
surgical
site
pictures
Sensormetrics
wereupload
edto
tablets
withtheEn
form
®ap
plication,syn
cingto
aweb
-portal.Alertsweregen
erated
for
values
that
exceed
edpre-estab
lished
thresholds.
Experiencednurses
contacted
patients
byphoneorusedtheap
p-in
tegrated
messagingforassessmen
tfollo
wing
concerningalerts.
Usual
care
Oeffet
al.47
HF
Weight;BP;
HR/rhythm;R
R;ox
ygen
saturations;symptom
questionnaire
Daily
telemonitoringtran
smissionof
biometrics.Alertsweregen
erated
when
individualised
limitswereexceed
ed.
Actionsinclude:
(1)discu
ssionwith
doctor;(2)med
icationad
justmen
t;(3)
planned
hospital
admission
Pre-im
plemen
tation
Pedoneet
al.23
HF
BP;
oxygen
saturations;weight;HR
Geriatriciansev
aluated
thedatadaily
once
tran
smittedthroughthetelemonitoring
kit.Alertsweregen
erated
ifdataexceed
edan
individualised
prespecified
rangean
dweredisplayedonthemonitoringsystem
.
Actionstaken:(1)
sched
uledoffice
appointm
ents,(2)
acute
care
wardreview
Usual
care
Pinnock
etal.30
COPD
Oxygen
saturation;d
aily
symptom
questionnaire
(dyspnoea,sputum
purulence/volume,
cough,
whee
ze,fev
er)
Algorithms,based
onthesymptom
score,
alertedtheclinical
monitoringteam
throughsecu
reinternet
connection,u
sing
atouch
screen
telemonitoringkit(Lothian),
ifdaily
read
ingshad
notbee
nsubmitted
daily
oracertainscore
obtained
.
Actioninclude:
(1)initiatingpatient
contact.(2)
homevisit,(3)co
mmen
cing
rescuetreatm
ent,(4)im
med
iate
admission.
Usual
care
withouttelemonitoring
F.M. Iqbal et al.
9
Seoul National University Bundang Hospital npj Digital Medicine (2021) 7
Table
2continue
d
Study
Cohort
Dataco
llected
Digital
alertingmechan
ism
Response
toalerts
Control
Pinto
etal.48
Amyo
trophic
lateral
sclerosiswith
respiratory
failu
reonNIV
NIV
data(IPAP,expiratory
positive
air
pressure;inspiratory/exp
iratory
ratio;b
acku
prate;v
entilation
sensitivities;risetime
Datatran
smissionwithamodem
through
TCP/IP
protoco
loccurred
.Alldatathat
wereSD
±1ofthemeanvalues
of
unpublished
pilo
tdatagen
erated
alerts.
Amessagewas
sentto
thephysician
whoco
uld
decideonpossible
setting
chan
ges,sch
edule
anoffice
visitor
phonecall,
orco
nduct
areal-tim
eco
mmunication.
Man
agem
entofNIV
settingswere
perform
edthroughregularvisits
Ringbaek
etal.17
COPD
Spirometer;o
xygen
saturations;
weight;self-reportingsymptoms
(dyspnoea,sputum
colour/vo
lume/
purulence)
Dataweretran
smitteddailyto
acallcentre
throughtelemonitoringeq
uipmen
t:categorisedan
dprioritisedwithalerts
gen
erated
ifvalues
werealarming.
Contact
initiatedbytherespiratory
nurse
duringworkingdays(M
onday–Friday,
9a.m.to3p.m
.).
Usual
care
Santiniet
al.49
HF;
arrythmias
Patien
tactivity;H
Ran
dvariab
ility;
intrathoracicim
ped
ance
Daily
tran
smissionthroughCareLinkwith
anau
dible
alarm
toalertthepatientwhen
aprogrammab
lethreshold
iscrossed
.
Ifthepatientwas
alertedorfeltworse,to
contact
theresponsible
physicianwho
requestad
ditional
dev
icetran
smissions,
unsched
uledvisits
orem
ergen
cyroom
admissions.
-
Scherret
al.24
HF
BP;
HR;w
eight;med
icationuse
Datatran
smittedusingamobile
telemonitoringkit(Zope)
daily.V
alues
outsideindividually
adjustab
leborders
resulted
inan
email/text
alert.
Physiciansco
ntacted
thepatientdirectly
viathemobile
phoneto
confirm
the
param
etersan
dad
just
med
ication.
Usual
care
withouttelemonitoring
Seto
etal.18
HF
Weight;BP;
ECG;symptom
questionnaires
Daily
tran
smissionofbiometrics
toa
mobile
phone,
then
tran
sferredto
adata
repository.Ifpre-estab
lished
thresholds
crossed
,emailalerts
sentto
acardiologist.
Dep
enden
toncardiologist.Actions
includeretakingmeasuremen
ts,
chan
gingmed
ication,attending
emergen
cydep
artm
entorcalling91
1.
Usual
care:v
isitingclinic
betwee
nonce
every2wee
ksto
once
every
3–6months.
Sinket
al.39
COPD
Self-reported
symptoms
Daily
automated
messages/callsdaily
from
acentral
server
toco
mmunicatedisease-
specificbiometricdataonExpCOPD
.The
designed
messagealgorithmsuse
Bayesian
branch
inglogic
togen
eratealerts
totext,
email,pag
er,o
rphone.
Follo
wingan
alert,themed
ical
residen
tco
ntacted
thepatientforassessmen
tan
d/orinitiatedap
propriate
interven
tion.
Receive
dthesamedaily
automated
messagewithoutalerts.
Smee
tset
al.25
HFwithCIED
Bioim
ped
ance
measuremen
tsfrom
CIED
Daily
alerttran
smissionsgen
erated
when
pre-defi
ned
alarm
thresholdswerecrossed
.OptiVolan
dCorVuealgorithmsfor
bioim
ped
ance
alerts
gen
eration.
Phoneco
ntact
initiatedbyanurse.
Subsequen
tprotoco
lised
actionwas
takenin
consultationwithaHF
specialist.
CIED
withoutbioim
ped
ance
alerts
gen
erated
.
Stev
enton
etal.32
COPD
;HF
Oxygen
saturations;bloodgluco
se;
weight;symptom
questionnaires
Readingstakenat
thesametimeeach
day
forupto
5daysper
wee
k,symptom
questionsan
ded
ucational
messages.
Monitoringcentres
(withspecialist
nurses
andmatrons),usedprotoco
lised
responses.
Usual
care
Stev
enton
etal.27
COPD
;HF;
DM
Weight;oxygen
saturation;BP;
temperature;b
loodgluco
se;peak-
flow;coag
ulation;1-lead
ECG
Readingstakenan
dau
tomatically
tran
smittedto
atriagecentrethrough
‘mym
edic’telemonitoringhub.
Ifsetthresholdswereexceed
ed,p
atients
wereco
ntacted
;escalationto
aphysician
forfurther
planwas
initiated.
Usual
care
Vianello
etal.31
COPD
HR;o
xygen
saturation
Alternateday
reco
rdingofobservations
throughatelemonitoringkit.Alerts
gen
erated
when
individualised
pre-
established
thresholdscrossed
.
Apulm
onaryspecialistcalledthepatient
forassessmen
tduringnorm
alworking
hours
(Monday–Friday,0
800–
1600
).Actionsinclude:
1.Modify
med
ication,2
.Homevisitbydistrictnurse,
3.Se
tupan
office
appointm
ent,4.
Escalate
avisitto
theEm
ergen
cyDep
artm
ent.
Usual
care
withouttelemonitoring
Yountet
al.19
Advancedlungcancer
Symptom
questionnaire
Wee
klycalls
placedusingtelephonebased
interactivevo
iceresponse
system
for
symptom
monitoring,responsesen
tered
usingthetelephonekeyp
ad.
Responsesmee
tingapre-defi
ned
threshold
forasymptom
gen
erated
ane-
mailtothesite
nurse.Pa
tien
tsco
ntacted
forassessmen
t.
Symptomsmonitoredwee
klybutno
automated
delivery
COPD
chronicobstructivepulm
onarydisease,H
Fheartfailu
re,D
Mdiabetes
mellitus,CIED
cardiacim
plantable
electronicdev
ice,NIV
non-in
vasive
ventilation,H
Rheartrate,BPbloodpressure,RRrespiratory
rate,
STARSymptom
Trackingan
dRep
orting.
F.M. Iqbal et al.
10
npj Digital Medicine (2021) 7 Seoul National University Bundang Hospital
All identified studies were uploaded to Covidence, a Cochranesupported systematic review package tool51. Initial screening wasconducted by one investigator and verified by a second todetermine if the eligibility criteria were met. Discrepancies werediscussed and resolved by consensus. Studies meeting theinclusion criteria underwent full-text screening; supplementalreferences were scrutinised for additional relevant articles.
Study selection criteria and outcome measuresStudies published containing the primary and secondary outcomeslisted below were included. No language restrictions were placed.Included study participants were adults (aged 18 years or over)discharged home with a digital alerting system (i.e., wearable sensor,non-invasive wireless technology, telemedicine, or remote monitor-ing). The last search was performed in October 2019.Abstracts, conference articles, opinion pieces, editorials, case studies,
reviews, and meta-analyses were excluded from the final review.Studies with inadequate published data relating to the primary andsecondary outcome measures were additionally excluded.
Data extractionThe primary outcome measure was hospitalisation and inpatientvisits. Secondary outcome measures include mortality, hospital lengthof stay (LOS), emergency department visits, and outpatient visits.All included study characteristics and outcome measures were
extracted by one investigator and verified by a second. All full-textreports of studies identified as potentially eligible after title andabstract review were obtained for further review.
Quality assessment (risk of bias)Methodological quality of randomised trials (RCTs) was assessed withthe Jadad Scale52. The scores range from 0–5; scores <3 wereconsidered low quality and scores ≥3 were considered high quality52.The risk of bias Cochrane tool was used to assess internal validity; thisassesses: (i) randomisation sequence allocation; (ii) allocationconcealment; (iii) blinding; (iv) completeness of outcome data; and(v) selective outcome reporting, classifying studies into low, high orunclear risk of bias53. Non-randomised trials were assessed using theNewcastle-Ottawa scale54. It comprises three variables: (i) patientselection; (ii) comparability of study groups; and (iii) assessment ofoutcomes. Scores range from 0–9, scores ≤3 were considered lowquality, between 4–6 moderate quality, and ≥7 high quality. Qualityassessment was assessed by one reviewer and validated by a second.
Data analysisA standard, hazard ratio, and proportion of means meta-analyseswere performed using Stata (v15.1. StataCorp LCC, TX). Effect sizeswere transformed into a common metric (e.g., days for time). Apercentage change for outcomes between control and interven-tion arms were calculated where possible. Hospitalisation andinpatient admissions were grouped into one variable.Continuous variables were compared through weighted mean
differences (WMD) with 95% CI. Where only the median wasreported, it was substituted for mean. Where range was reported,it was converted to standard deviation through division of four. Asassumption of normal distribution was made for this to occur.Forest plots were generated for all included studies.Data were pooled using a random effects model and
heterogeneity was assessed with the I2 statistic. We considereda value <30% as low heterogeneity, between 30 and 60%moderate, and over 60% as high.
DATA AVAILABILITYThe datasets generated during and/or analysed during the current study are availablefrom the corresponding author on reasonable request.
Received: 8 September 2020; Accepted: 1 December 2020;
REFERENCES1. Majumder, S., Mondal, T. & Deen, M. J. Wearable sensors for remote health
monitoring. Sensors (Basel, Switz.) 17, 130 (2017).2. Prince, M. J. et al. The burden of disease in older people and implications for
health policy and practice. Lancet 385, 549–562 (2015).3. Sood, S. et al. What is telemedicine? A collection of 104 peer-reviewed per-
spectives and theoretical underpinnings. Telemed. e-Health 13, 573–590(2007).
4. Vegesna, A., Tran, M., Angelaccio, M. & Arcona, S. Remote patient monitoring vianon-invasive digital technologies: a systematic review. Telemed. J. E Health 23,3–17 (2017).
5. Martirosyan, M., Caliskan, K., Theuns, D. A. M. J. & Szili-Torok, T. Remote mon-itoring of heart failure: benefits for therapeutic decision making. Expert Rev.Cardiovasc. Ther. 15, 503–515 (2017).
6. Crossley, G. H., Boyle, A., Vitense, H., Chang, Y. & Mead, R. H. The CONNECT(Clinical Evaluation of Remote Notification to Reduce Time to ClinicalDecision) trial: the value of wireless remote monitoring with automaticclinician alerts. J. Am. Coll. Cardiol. 57, 1181–1189 (2011).
7. Slotwiner, D. et al. HRS Expert Consensus Statement on remote interrogation andmonitoring for cardiovascular implantable electronic devices. Heart Rhythm 12,e69–e100 (2015).
8. Varma, N., Michalski, J., Stambler, B. & Pavri, B. B. Superiority of automatic remotemonitoring compared with in-person evaluation for scheduled ICD follow-up inthe TRUST trial-testing execution of the recommendations. Eur. Heart J. 35,1345–1352 (2014).
9. DeVita, M. A. et al. “Identifying the hospitalised patient in crisis”–a consensusconference on the afferent limb of rapid response systems. Resuscitation 81,375–382 (2010).
10. Smith, G. B. In-hospital cardiac arrest: is it time for an in-hospital ‘chain of pre-vention’? Resuscitation 81, 1209–1211 (2010).
11. Agoulmine, N., Deen, M. J., Lee, J.-S. & Meyyappan, M. U-health smart home. IEEENanotechnol. Mag. 5, 6–11 (2011).
12. Kitsiou, S., Paré, G. & Jaana, M. Effects of home telemonitoring interventions onpatients with chronic heart failure: an overview of systematic reviews. J. Med.Internet Res. 17, e63–e63 (2015).
13. Noah, B. et al. Impact of remote patient monitoring on clinical outcomes: anupdated meta-analysis of randomized controlled trials. NPJ Digital Med. 1, 20172(2018).
14. Chen, Y. H. et al. Clinical outcome and cost-effectiveness of a synchronous tel-ehealth service for seniors and nonseniors with cardiovascular diseases: quasi-experimental study. J. Med. Internet Res. 15, e87 (2013).
15. Heidbuchel, H. et al. EuroEco (European Health Economic Trial on Home Mon-itoring in ICD Patients): a provider perspective in five European countries on costsand net financial impact of follow-up with or without remote monitoring. Eur.Heart J. 36, 158–169 (2015).
16. Martin-Lesende, I. et al. Telemonitoring in-home complex chronic patients fromprimary care in routine clinical practice: Impact on healthcare resources use. Eur.J. Gen. Pract. 23, 135–142 (2017).
17. Ringbaek, T. et al. Effect of tele health care on exacerbations and hospitaladmissions in patients with chronic obstructive pulmonary disease: a randomizedclinical trial. Int. J. Copd. 10, 1801–1808 (2015).
18. Seto, E. et al. Mobile phone-based telemonitoring for heart failure man-agement: A randomized controlled trial. J. Med. Internet Res. 14, 190–203(2012).
19. Yount, S. E. et al. A randomized trial of weekly symptom telemonitoring inadvanced lung cancer. J. Pain. Symptom Manag. 47, 973–989 (2014).
20. Bohm, M. et al. Fluid status telemedicine alerts for heart failure: a randomizedcontrolled trial. Eur. Heart J. 37, 3154–3163 (2016).
21. Kotooka, N. et al. The first multicenter, randomized, controlled trial of hometelemonitoring for Japanese patients with heart failure: home telemonitoringstudy for patients with heart failure (HOMES-HF). Heart Vessels 33, 866–876(2018).
F.M. Iqbal et al.
11
Seoul National University Bundang Hospital npj Digital Medicine (2021) 7
22. Luthje, L. et al. A randomized study of remote monitoring and fluid monitoringfor the management of patients with implanted cardiac arrhythmia devices.Europace 17, 1276–1281 (2015).
23. Pedone, C., Rossi, F. F., Cecere, A., Costanzo, L. & Antonelli Incalzi, R. Efficacy of aphysician-led multiparametric telemonitoring system in very old adults withheart failure. J. Am. Geriatrics Soc. 63, 1175–1180 (2015).
24. Scherr, D. et al. Effect of home-based telemonitoring using mobile phone tech-nology on the outcome of heart failure patients after an episode of acutedecompensation: randomized controlled trial. J. Med. Internet Res. 11, e34 (2009).
25. Smeets, C. J. et al. Bioimpedance alerts from cardiovascular implantable elec-tronic devices: observational study of diagnostic relevance and clinical outcomes.J. Med. Internet Res. 19, e393 (2017).
26. Baker, L. C. et al. Effects of care management and telehealth: A longitudinalanalysis using medicare data. J. Am. Geriatrics Soc. 61, 1560–1567 (2013).
27. Steventon, A., Ariti, C., Fisher, E. & Bardsley, M. Effect of telehealth on hospitalutilisation and mortality in routine clinical practice: a matched control cohortstudy in an early adopter site. BMJ Open 6, 5–7 (2016).
28. Lewis, K. E. et al. Does home telemonitoring after pulmonary rehabilitationreduce healthcare use in optimized COPD a pilot randomized trial. COPD: J.Chronic Obstr. Pulm. Dis. 7, 44–50 (2010).
29. McElroy, I. et al. Use of digital health kits to reduce readmission after cardiacsurgery. J. Surg. Res. 204, 1–7 (2016).
30. Pinnock, H. et al. Effectiveness of telemonitoring integrated into existing clinicalservices on hospital admission for exacerbation of chronic obstructive pulmonarydisease: researcher blind, multicentre, randomised controlled trial. BMJ (Online)347, 10–16 (2013).
31. Vianello, A. et al. Home telemonitoring for patients with acute exacerbation ofchronic obstructive pulmonary disease: a randomized controlled trial. BMC Polm.16, 9–15 (2016).
32. Steventon, A. et al. Effect of telehealth on use of secondary care and mortality:findings from the Whole System Demonstrator cluster randomised trial. BMJ(Online) 344 (2012).
33. Del Hoyo, J. et al. A web-based telemanagement system for improving diseaseactivity and quality of life in patients with complex inflammatory bowel disease:pilot randomized controlled trial. J. Med. Internet Res. 20, e11602 (2018).
34. Mousa, A. Y. et al. Results of telehealth electronic monitoring for post dischargecomplications and surgical site infections following arterial revascularization withgroin incision. Ann. Vasc. Surg. 57, 160–169 (2019).
35. Basch, E. et al. Symptom monitoring with patient-reported outcomes duringroutine cancer treatment: a randomized controlled trial. J. Clin. Oncol. 34,557–565 (2016).
36. Bekelman, D. B. et al. Primary results of the patient-centered disease manage-ment (PCDM) for heart failure study: a randomized clinical trial. JAMA Intern. Med.175, 725–732 (2015).
37. Calvo, G. S. et al. A home telehealth program for patients with severe COPD: ThePROMETE study. Respiratory Med. 108, 453–462 (2014).
38. Lee, T. C. et al. Telemedicine based remote home monitoring after liver trans-plantation: results of a randomized prospective trial. Ann. Surg. 270, 564–572(2019).
39. Sink, E. et al. Effectiveness of a novel, automated telephone intervention on timeto hospitalisation in patients with COPD: A randomised controlled trial. J. TelemedTelecare. 26, 132–139 (2018).
40. Ida, S., Kaneko, R., Imataka, K. & Murata, K. Relationship between frailty andmortality, hospitalization, and cardiovascular diseases in diabetes: a systematicreview and meta-analysis. Cardiovasc. Diabetol. 18, 81 (2019).
41. Dantas, I., Santana, R., Sarmento, J. & Aguiar, P. The impact of multiple chronicdiseases on hospitalizations for ambulatory care sensitive conditions. BMC HealthServ. Res. 16, 348 (2016).
42. Pope, I., Burn, H., Ismail, S. A., Harris, T. & McCoy, D. A qualitative study exploringthe factors influencing admission to hospital from the emergency department.BMJ Open 7, e011543 (2017).
43. Joshi, M. et al. Digital alerting and outcomes in patients with sepsis: systematicreview and meta-analysis. J. Med. Internet Res. 21, e15166 (2019).
44. McCambridge, J., Witton, J. & Elbourne, D. R. Systematic review of the Hawthorneeffect: new concepts are needed to study research participation effects. J. Clin.Epidemiol. 67, 267–277 (2014).
45. Godleski, L., Cervone, D., Vogel, D. & Rooney, M. Home telemental healthimplementation and outcomes using electronic messaging. J. Telemed. Telecare18, 17–19 (2012).
46. Licskai, C., Sands, T. W. & Ferrone, M. Development and pilot testing of a mobilehealth solution for asthma self-management: asthma action plan smartphoneapplication pilot study. Can. Respir. J. 20, 301–306 (2013).
47. Oeff, M., Kotsch, P., Gosswald, A. & Wolf, U. [Monitoring multiple cardiovascularparamaters using telemedicine in patients with chronic heart failure]. Herzs-chrittmach 16, 150–158 (2005).
48. Pinto, A. et al. Home telemonitoring of non-invasive ventilation decreaseshealthcare utilisation in a prospective controlled trial of patients withamyotrophic lateral sclerosis. J. Neurol. Neurosurg. Psychiatry 81, 1238–1242(2010).
49. Santini, M. et al. Remote monitoring of patients with biventricular defibrillatorsthrough the CareLink system improves clinical management of arrhythmias andheart failure episodes. J. Inter. Card. Electrophysiol. 24, 53–61 (2009).
50. Moher, D., Liberati, A., Tetzlaff, J. & Altman, D. G. Preferred reporting items forsystematic reviews and meta-analyses: The PRISMA statement. Int. J. Surg. 8,336–341 (2010).
51. Innovation, V. H. Covidence Systematic Review Software (2017).52. Jadad, A. R. et al. Assessing the quality of reports of randomized clinical trials: is
blinding necessary? Controlled Clin. Trials 17, 1–12 (1996).53. Sterne, J. A. et al. RoB 2: a revised tool for assessing risk of bias in randomised
trials. BMJ 366, l4898 (2019).54. Stang, A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of
the quality of nonrandomized studies in meta-analyses. Eur. J. Epidemiol. 25,603–605 (2010).
55. Biddiss, E., Brownsell, S. & Hawley, M. S. Predicting need for intervention inindividuals with congestive heart failure using a home-based telecare system. J.Telemed. Telecare 15, 226–231 (2009).
56. Denis, F. et al. Prospective study of a web-mediated management of febrileneutropenia related to chemotherapy (Bioconnect). Support Care Cancer 27,2189–2194 (2019).
ACKNOWLEDGEMENTSInfrastructure support for this research was provided by the NIHR Imperial BiomedicalResearch Centre (BRC) and the NIHR Imperial Patient Safety Translational ResearchCentre (PSTRC).
AUTHOR CONTRIBUTIONSF.M.I. drafted the manuscript. F.M.I. and K.L. independently screened and reviewed allincluded articles and graded the quality of included studies. H.A. performed themeta-analysis. K.L., H.A., M.J., S.K., and A.D. all contributed to significant amendmentsto the final manuscript.
COMPETING INTERESTSThe authors declare no competing interests.
ADDITIONAL INFORMATIONCorrespondence and requests for materials should be addressed to F.M.I.
Reprints and permission information is available at http://www.nature.com/reprints
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claimsin published maps and institutional affiliations.
Open Access This article is licensed under a Creative CommonsAttribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in anymedium or format, as long as you giveappropriate credit to the original author(s) and the source, provide a link to the CreativeCommons license, and indicate if changes were made. The images or other third partymaterial in this article are included in the article’s Creative Commons license, unlessindicated otherwise in a credit line to the material. If material is not included in thearticle’s Creative Commons license and your intended use is not permitted by statutoryregulation or exceeds the permitted use, you will need to obtain permission directlyfrom the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
© The Author(s) 2021
F.M. Iqbal et al.
12
npj Digital Medicine (2021) 7 Seoul National University Bundang Hospital
Related Documents
