Sage Journals: Discover world-class research

Abstract

Objectives

This review aimed to present the clinical and health-care outcomes for patients with congenital heart disease (CHD) who use home monitoring technologies.

Methods

Five databases were systematically searched from inception to November 2020 for quantitative studies in this area. Data were extracted using a pre-formatted data-collection table which included information on participants, interventions, outcome measures and results. Risk of bias was determined using the Cochrane Risk of Bias 2 tool for randomised controlled trials (RCTs), the Newcastle–Ottawa Quality Assessment Scale for cohort studies and the Institute of Health Economics quality appraisal checklist for case-series studies.

Data synthesis: Twenty-two studies were included in this systematic review, which included four RCTs, 12 cohort studies and six case-series studies. Seventeen studies reported on mortality rates, with 59% reporting that home monitoring programmes were associated with either a significant reduction or trend for lower mortality and 12% reporting that mortality trended higher. Fourteen studies reported on unplanned readmissions/health-care resource use, with 29% of studies reporting that this outcome was significantly decreased or trended lower with home monitoring and 21% reported an increase. Impact on treatment was reported in 15 studies, with 67% of studies finding that either treatment was undertaken significantly earlier or significantly more interventions were undertaken in the home monitoring groups.

Conclusion

The use of home monitoring programmes may be beneficial in reducing mortality, enabling earlier and more timely detection and treatment of CHD complication. However, currently, this evidence is limited due to weakness in study designs.

Keywords

eHealth tele-cardiology telecare telehealth remote monitoring systematic review congenital heart disease home care clinical outcomes

Introduction

Congenital heart disease (CHD)¹ is an umbrella term for a broad range of birth defects that affect the function of the heart. Approximately 1% of babies are born with a heart of circulatory condition, which is usually a form of congenital heart disease. The mean prevalence of CHD globally was reported to be 8.224 (SD = 7.817–8.641) per 1000 for the years 1970–2017.² It was estimated that in 2017, there were 11,998,283 people living with CHD globally.³

Many patients with CHD will require lifelong follow-up in order to monitor for signs of deterioration. In light of this, novel strategies for managing these patients may be useful in terms of reducing the burden of repeated health-care attendances, but also by empowering patients to be proactive in the management of their condition.

Telehealth may be a useful tool in the management of cardiac patients. According to the World Health Organization, telehealth is ‘the use of telecommunications and virtual technology to deliver health care outside of traditional health-care facilities’.⁴ A systematic review of patients with heart failure reported that telehealth reduced all-cause mortality by 20% (95% confidence interval (CI) 0.68–0.94) and reduced heart failure hospitalisation by 29% (95% CI 0.60–0.83) compared to usual care. It would therefore be of interest to determine any health benefits afforded to patients with CHD through the use of home monitoring,⁵ whereby patients would monitor various physiological parameters at home, with the data obtained being reviewed by health-care staff. There has been one paper published to date⁶ which has reviewed the use of telehealth for monitoring this patient population. However, this paper was not a systematic review and did not undertake quality appraisal of the included papers. Additionally, this review was published in 2018 and, as such, only included papers published up to 2017, with only the PubMed database searched. This current review will therefore provide a rigorous examination of the literature in this area.

The aim of this systematic review was to provide an overview of the literature on home monitoring technologies used by patients with CHD. The objectives were (a) to present clinical and health-care outcomes for patients who use home monitoring technologies, and (b) to identify areas for research on the use of home monitoring technologies by patients with CHD.

Methods

Search strategy

A computerised search of the following databases from inception to November 2020 was performed: Medline, Embase, Cochrane, CINAHL and Scopus.

Search terms comprising Medical Subject Headings (MeSH), database-specific subject headings and key words were employed: EXP TELEMEDICINE OR telehealth* OR tele-health* OR telemedicine* OR tele-medicine* OR telemonitor* OR tele-monitor* OR telemanagement* OR tele-management* OR teleconsult* OR tele-consult OR telecare* OR tele-care* OR telepharmacy* OR tele-pharmacy* OR telenurs* OR tele-nurs* OR ‘remote monitor*’ OR ‘remote consult*’ OR ‘remote care*’ OR ‘mobile care*’ OR ehealth* or e-health* OR mhealth* OR m-health* OR ‘home monitor*’ OR ‘self monitor*’ OR ‘self manage*’ OR ‘home manage*’ AND EXP HEART DEFECTS, CONGENITAL OR congenital heart disease* OR congenital heart malformation* OR adult congenital heart disease* OR ACHD* OR grown up congenital heart disease* OR GUCH OR (congenital adj3 heart) OR (congenital adj3 cardiac) OR (cardiac adj3 malformation).

Grey literature was sought from the British Library (ETHoS), opengrey.eu, greylit.org, BMJ Best Practice, US Food and Drug Administration (FDA), National Institutes of Health, National Institute for Health and Care Excellence (NICE), Kings Fund, Nuffield Trust and Google Scholar. Hand searching of reference lists of articles was also performed.

The lead review author (R.C.) screened all records identified for inclusion/exclusion criteria. All identified publications were read as either abstracts or full texts. Papers were included if they were full text investigating home monitoring interventions actively used by the patient (or their carer) and published in the English language. Papers were excluded if they were abstracts only, citations, letters, case studies, investigating interventions used for the diagnosis of CHD, technical studies of telehealth equipment and home monitoring that was passively done without any action by the patient. Following the initial screening process, a full text evaluation of 25 articles was performed to determine correlation with the inclusion criteria. A meeting was held between the four members (R.C., C.M.H., S.Mc.F. and J.C.) of the review team to discuss findings, until inclusion was agreed by consensus.

The lead author (R.C.) extracted the necessary data from each included study into a pre-formatted data-collection table.

Quality assessment and risk of bias

Quality assessment and risk of bias was determined for all studies in this review. A meeting was held with all members (R.C., C.M.H., S.Mc.F. and J.C.) of the review team to determine the quality scores and risk of bias for all studies. Agreement on quality scores and risk of bias was gained by consensus. The quality appraisal tools that were used were as follows: the Cochrane Risk of Bias 2 tool⁷ for randomised controlled trials (RCTs), the Newcastle–Ottawa Quality Assessment Scale⁸ for cohort studies and the Institute of Health Economics (IHE) quality appraisal checklist⁹ for case-series studies.

Results

A total of 2025 articles were identified by various literature searches. Twenty-two articles were deemed appropriate for inclusion in this systematic review (Figure 1).

Figure 1.

PRISMA flow diagram depicting the study selection process.

Study characteristics

The main study characteristics are presented in Table 1. These studies were published between 2001 and 2020. The main types of study were RCT, cohort and case series. In total, they included 2809 participants, with the number in individual studies ranging between 14 and 494 participants.

Table 1.

Summary of study characteristics.

Authors	Year	Intervention type; comparator	Study population	Methodology	Number of participants	Outcomes	Risk of bias/quality score
Bingler et al.⁹	2018	SpO2 and weight HM using a PC-based App; compared to three-ring binder documentation	SV patients between S1P and S2P	Randomised crossover design. Discharged using the three-ring binder then randomised to use App either one or two months after discharge. Daily weight and twice-daily SpO2 monitored and documented.	31	Mortality: 0%Unplanned readmissions/health-care resource use: overall hospital length of stay trended lower (p = 0.07)	RoB: Some concerns
Castellanos et al.¹⁸	2016	SpO2 and weight HM; compared to retrospective review of SV patients receiving standard care during three years prior to initiation of the HM programme	SV patients following discharge after the Norwood procedure	Retrospective, single-institution review. SpO2 three times per day and daily weight measurements. Weekly telephone calls from nurse specialist.	126	Mortality: 3.8% (p = 0.001) Unplanned readmissions/health-care resource use: Increased by 35% (p < 0.05) Impact on treatment: 78% increase in percutaneous interventions (p < 0.01)	NOS: 8High quality
Christensen et al.²⁴	2004	INR home management; no comparator	Paediatric patients with CHD on long-term oral anticoagulation therapy	Case series. INR measured once a week. During the training phase, venous INR compared to CoaguChek INR. Calibration every fourth week using a control solution, and every six months the CoaguChek was mailed to the hospital for a check-up.	22	Mortality: 9%	IHE: 10Moderate quality
Christensen et al.²³	2001	INR home management; no comparator	Paediatric patients with CHD on warfarin or phenoprocoman	Case series. INR measured once a week. Calibration every fourth week using a control solution, and every six months the CoaguChek was mailed to the hospital for a check-up.	14	Mortality: 7%	IHE: 11Moderate quality
Cooper et al.¹²	2020	Heart rate, SpO2, weight and caloric intake HM; compared to usual care	Infants with complex CHD (Risk Adjustment for Congenital Heart Surgery (RACHS-1) category ≥2)	RCT. Parent–infant dyads randomly assigned to receive HM intervention or usual care.	219	Unplanned readmissions/health-care resource use: More infants in the intervention group were readmitted (66% vs. 57%) trended lowerMore infants in the intervention group had more than one hospital readmission (15% vs. 7%; p < 0.001)	RoB: Some concerns
Dobrolet et al.¹⁵	2011	SpO2 and weight HM; compared to historical controls	Newborns after S1P or placement of BT/central shunt	Prospective study. Patients followed from discharge to S2P. Daily oxygen saturation, daily weight. Scheduled telephone calls with nurse. Retrospective review of historical controls.	163	Mortality: 2.5% trended lowerImpact on treatment: Mean age at detection, admission and treatment for residual lesions was younger with HM (p < 0.05)	NOS: 6Moderate quality
Gardner et al.³¹	2019	SpO2, weight and feeding HM; compared to historical controls	CHD patients with shunt-dependent SV defects discharged after neonatal palliation	Daily HM of oxygen and weight, feeding log, weekly phone calls with nurse practitioner and regular reviews with clinicians	334	Mortality: 5.4% trended lowerUnplanned readmissions/health-care resource use: percentage of participants with at least one readmission: HM 63% vs. controls 46% (p < 0.001) Impact on treatment: S2P earlier with HM (p < 0.001)	NOS: 8High quality
Ghanayem et al.²⁷	2004	SpO2 and weight HM; compared to historical controls	Patients between S1P and S2P	Cohort study using historical controls. SpO2 and weight HM. Contact with physicians if breach in surveillance criteria e.g. SpO2 < 70%	89	Mortality: 0% (p = 0.02) Unplanned readmissions/health-care resource use: 58% breached criteria and sought medical attention; no data for comparisonImpact on treatment: S2P earlier with HM (p = 0.001)	NOS: 6Moderate quality
Ghanayem et al.¹⁴	2003	SpO2 and weight HM; compared to historical controls	SV patients during the interstage period	Cohort study using historical controls. SpO2 and weight HM. Contact with physicians if breach in surveillance criteria e.g. SpO2 < 70%	87	Mortality: 0% (p = 0.039) Impact on treatment: S2P earlier with HM (p = 0.016)	NOS: 7High quality
Hansen et al.¹³	2012	SpO2 and weight HM; compared to historical controls and those not discharged home after S1P	Infants with HLHS after S1P	Observational cohort study with three cohorts. Group A: daily SpO2 and weight HM with scheduled telephone calls with nurse practitioner; group B: historical; group C: infants who were not discharged after S1P.	162	Mortality: 2.2% (p = 0.042) Unplanned readmissions/health-care resource use: 31% breached criteria and were referred for observationImpact on treatment: 51% of patients who breached criteria underwent diagnostic catheterisation	NOS: 4Moderate quality
Harahsheh et al.¹⁹	2016	SpO2, weight and nutritional intake HM; compared to historical controls	Patients with SV who survived to discharge following Norwood operation	Cohort study. SpO2, weight and nutritional intake monitored and data submitted at least three times per week. Assessment by HM team if any red flags noted.	98	Mortality: 5.4% trended higherImpact on treatment: More major post-Norwood procedures (p = 0.02)	NOS: 9High quality
Kauw et al.²²	2019	ECG, BP and weight HM using an App; no comparator	ACHD patients with palpitations within the last three years or New York Heart Association heart failure classification ≥II and possession of a mobile device	Prospective observational study. Seven days of daily monitoring used for baseline. Then ECG once per month, with BP/weight measured twice per week.	109	Impact on treatment: recurrence of arrhythmia was detected and treated in 25% of patients with history of arrhythmia; 38% of patients with hypertension at baseline had treatment adjusted	IHE: 14High quality
Koole et al.²¹	2019	ECG, BP and weight HM using an App; no comparator	Patients with symptomatic CHD	Prospective observational studyECG, BP and weight done twice a week. Extra measurements in the event of symptoms. ECGs were assessed daily along with BP and weight parameters.	55	Unplanned readmissions/health-care resource use: 1.7% readmissionsImpact on treatment: medication changes made in 10.9% of patients based on HM	IHE: 13Moderate quality
McCrossan et al.¹⁰	2007	Videoconferencing home support with SpO2 HM in seven patients; compared to phone home support and standard care	Paediatric patients with new diagnosis of major CHD	RCT. Group A: weekly videoconferences with clinicians with the option for extra consultations if required; group B weekly telephone consultations with clinicians; group C: standard care.	66	Unplanned readmissions/health-care resource use: contact was advised less with videoconferencing (4%) compared to telephone support (22%)	RoB: Some concerns
McCrossan et al.¹¹	2012	Videoconferencing home support with SpO2 monitoring if clinically indicated; compared to phone support and standard care	Infants with complex CHD or congestive heart failure, as well as infants with corrected CHD who had a difficult preoperative course	RCT. Group A: scheduled weekly or twice weekly videoconferencing with SpO2 used in some patients (depending on clinical requirements) – VCs were scheduled on a weekly or twice-weekly basis; group B: telephone consultations weekly or twice weekly; group C: standard care.	83	Unplanned readmissions/health-care resource use: lower probability of being admitted at least once (p = 0.004)	RoB: Some concerns
Öhman et al.¹⁷	2013	SpO2 HM; compared to historical controls	SV patients with shunt as only source of pulmonary blood flow	Observational study. Daily SpO2 monitoring and recording of results in logbook. Telephone contact with hospital if oxygen saturations remained < 70% for >15 minutes	77	Mortality: 15% trended higherImpact on treatment: 38% of shunt complications were detected by HM and required intervention	NOS: 5Moderate quality
Oster et al.¹⁶	2015	SpO2 and/or weight HM; comparison of daily, weekly or no monitoring	Interstage patients between S1P and S2P	Retrospective cohort study using prospectively collected data from the Paediatric Cardiology Quality Improvement Collaborative from 2008 to 2012	494	Mortality: 8.1% trended lowerUnplanned readmissions/health-care resource use: no significant difference in unscheduled readmissions between groupsImpact on treatment: S2P earlier with SpO2 HM (p = 0.002) and with weight HM (p = 0.004)	NOS: 9High quality
Petit et al.³⁰	2011	SpO2 and weight HM; compared to historical controls	SV patients during the interstage period	Retrospective study of patients with SV. HM of SpO2 and daily weight with at least weekly contact with nurse practitioners. SpO2 < 75% or failure to gain weight on at least three successive days prompted OP visit or admission	225	Mortality: 8.3% trended lowerUnplanned readmissions/health-care resource use: unplanned readmissions and ED visits did not differ between groupsImpact on treatment: S2P earlier with HM (p < 0.001)	NOS: 7High quality
Rudd et al.²⁵	2014	SpO2and weight HM; no comparator	Patients discharged after S1P	Longitudinal observation study. Daily SpO2 and weight. Clinical team notified if there was a breach in surveillance criteria e.g. SpO2 < 75% or >90%	157	Mortality: 2% Unplanned readmissions/health-care resource use: 59% breached criteria; 57% required at least one admission Impact on treatment: early S2P was performed in 36 patients who were readmitted	IHE: 15High quality
Shirali et al.²⁸	2016	HR and SpO2 HM with PC based App; compared to historical controls	Babies during the interstage period	Retrospective observational study. SpO2, weight and any red flags documented in the App. Instant alerts sent for some red flags e.g. SpO2 < 70% or >93%	83	Mortality: 0% (p = 0.025) Unplanned readmissions/health-care resource use: unplanned readmissions, readmission days and emergency room visits trended lower	NOS: 8High quality
Siehr et al.²⁶	2014	HR, SpO2, weight and nutritional intake HM; no comparator	Infants with HLHS or other SV lesions with significant aortic obstruction requiring a Norwood procedure	Retrospective, single-institution review. Daily measurement and recording of HR, oxygen saturation, weight and enteral intake. Phone HM team if breach in criteria e.g. SpO2 < 75%	46	Mortality: 0% Unplanned readmissions/health-care resource use: 41% readmissionsImpact on treatment: 37% of patients required interventions triggered by HM	IHE: 13Moderate quality
Vergales et al.²⁰	2020	SpO2, weight, nutritional intake and output HM programme with digital platform; compared to historical non-digital HM programme	Infants who required a neonatal operation as part of staged SV palliation	Retrospective two-centre, observational study. Daily measurement of SpO2, weight, nutritional intake and output with data input to digital platform and reviewed by clinical team	69	Mortality: 0% trended lowerImpact on treatment: S2P earlier with digital HM 4.6 months vs 5.4 months (p < 0.001)	NOS: 7High quality

HM: home monitoring; SV: single ventricle; S1P: stage 1 palliation; S2P: stage 2 palliation; RoB: Cochrane Risk of Bias 2; NOS: Newcastle–Ottawa Scale; INR: international normalised ratio; CHD: congenital heart disease; IHE: Institute of Health Economics quality appraisal checklist; RCT: randomised controlled trial: HLHS: hypoplastic left heart syndrome; ECG: electrocardiogram; BP: blood pressure.

Risk of bias

The Cochrane Risk of Bias 2 tool was used for all RCTs (N = 4; Figure 2). All four studies^9–12 were assessed as having some concerns, with all indicating a moderate risk of bias. The most common deficit in methodological quality was in the domains for randomisation and selection of the reported result. No study reported whether allocation sequence was concealed until participants were enrolled and assigned to the intervention. No study reported whether the data were analysed in accordance with a pre-specified analysis plan or whether the outcome assessors were blinded to the intervention.

Figure 2.

Risk of bias in studies.

Newcastle–Ottawa Quality Assessment Scale

The Newcastle–Ottawa Quality Assessment Scale was used for the quality appraisal of all cohort studies (N = 12; Figure 3). The most common shortfalls in methodological quality related to the domains of comparability and outcome. Five studies scored zero in the domain of comparability.^13–17 These studies did not provide information on whether they controlled for factors that could affect the outcome measured. Only four studies¹⁶,^18–20 reported that the outcomes measured were controlled for multiple variables. One study¹³ was rated as low quality for the outcome domain. This study did not provide details of how the outcome was assessed or whether all subjects were accounted for.

Figure 3.

Newcastle–Ottawa Quality Assessment Scale.

IHE quality appraisal checklist

The IHE quality appraisal checklist for case-series studies was used to assess the quality of the remaining studies which used a case-series design methodology (N = 6; Figure 4). The most common deficiencies in methodological quality were for the domains of competing interests and sources of support, statistical analysis and study design. Two studies²¹,²² provided details on competing interests and sources of support, with the remaining four studies^23–26 not providing any information in relation to this. The domain for statistical analysis also identified three studies^23–25 which were judged to have a low level of quality based on the IHE quality appraisal checklist.

Figure 4.

Institute of Health Economics quality appraisal checklist.

Overall, the risk of bias results indicate that the RCTs were all rated as having some concerns; cohort studies were rated as being of moderate to high quality, with 64% bring rated as high quality; and case series were all rated as moderate to high quality, with 67% of these studies receiving a rating of high quality. Whilst many of the case studies and cohort studies were judged to be high quality, it must be borne in mind that the studies themselves were less robust by virtue of their design methodology compared to the randomised cohort studies.

Synthesis of results

Thematic analysis of outcome measures was undertaken by the lead author (R.C.) who systematically examined the pre-formatted data-extraction table to identify outcome measures recorded by each of the papers. Themes were discussed and agreed with the co-authors. There were three common outcome measures: mortality, unplanned readmissions/health-care resource use and impact on treatment.

Mortality

There were 17 studies that included mortality as an outcome measure (Table 2). Five studies found that home monitoring programmes were associated with a significant improvement in mortality.¹³,¹⁴,¹⁸,²⁷,²⁸ Mortality was reduced by between 10.2% and 17%. The study with the smallest reduction in mortality reported 12.4% mortality for historical controls compared to 2.2% for the study group.²⁹ The largest reduction in mortality reported 17% mortality for historical controls compared to 0% in the study group.²⁸

Table 2.

Summary of mortality rates.

Authors	Type of monitoring	Mortality
Castellanos et al., 2016¹⁸	SpO2, weight; scheduled phone calls with nurse	3.8% (p = 0.001)
Ghanayem et al., 2003¹⁴	SpO2, weight; contact with HM team if breach in criteria	0% (p = 0.039)
Ghanayem et al., 2004²⁷	SpO2, weight; contact with HM team if breach in criteria	0% (p = 0.02)
Hansen et al., 2012¹³	SpO2, weight; scheduled phone calls with nurse	2.2% (p = 0.042)
Shirali et al., 2016²⁸	HR, SpO2; instant alerts to HM team	0% (p = 0.025)
Dobrolet et al., 2011¹⁵	SpO2, weight; scheduled phone calls with HM team	2.5% Trended lower
Petit et al., 2011³⁰	SpO2, weight; scheduled phone calls with nurse	8.3% Trended lower
Gardner et al., 2019³¹	SpO2, weight; scheduled phone calls with nurse	5.4% Trended lower
Oster et al., 2015¹⁶	SpO2 and/or weight	8.1% Trended lower
Vergales et al., 2020²⁰	SpO2, weight, nutritional intake and output; digital platform with review by clinical team	0% Trended lower
Bingler et al., 2018⁹	SpO2, weight	0% in both groups
Christensen et al., 2001²³	INR	7% No control group
Christensen et al., 2004²⁴	INR	9% No control group
Siehr et al., 2014²⁶	HR, SpO2, weight, nutritional intake; contact with HM team if breach in criteria	0% No control group
Rudd et al., 2014²⁵	SpO2, weight; contact with clinical team if breach in criteria	2% No control group
Harahsheh et al., 2016¹⁹	SpO2, weight, nutritional intake; assessment by HM team if red flags	5.4% Trended higher (p = 0.071)
Öhman et al., 2019¹⁷	SpO2; contact with hospital if breaches in criteria	15% Trended higher

Five studies found that mortality rates in the home monitoring group trended lower, with mortality reduced by between 3.8% and 7.6%.¹⁵,¹⁶,²⁰,³⁰,³¹ The study with the biggest reduction in mortality reported a mortality rate for historical controls at 13% compared to 5.4% in the home monitoring group (p = 0.2).³¹ The study with the smallest reduction in mortality reported that mortality was reduced from 12.1% to 8.3% (p = 0.924) with home monitoring.³⁰ While there was a trend for lower mortality in these studies, the results were not significant. One study reported 0% mortality in both groups.⁹

There were four studies that reported mortality rates. However, there was no comparator.^23–26 Mortality rates in these studies ranged from 0% to 9%.

Two studies found that mortality trended higher for the home monitoring groups.¹⁷,¹⁹ One study reported that mortality in the home monitoring group was 15% versus 10% in the historical controls.¹⁷ In the other study, mortality in the home monitoring group was 5.4% versus 2.4% in the historical controls (p = 0.71).¹⁹

Unplanned readmissions/health-care resource use

Unplanned readmissions/health-care resource use was reported in 14 studies (Table 3). Four studies^23–26 reported data for the home monitoring cohort, but there were no data for comparison. Three studies reported data on breaches in surveillance criteria (e.g. a SpO2 < 70%).¹³,²⁵,²⁷ Between 31% and 59% of patients breached surveillance criteria, with up to 57% of patients requiring observation and being admitted to hospital. Readmissions were reported in two studies.²¹,²⁶ One study²⁶ reported that 41% of patients had 27 readmissions, and another study³² reported that there were two (3.6%) presentations to ED and 1 (1.8%) hospitalisation.

Table 3.

Summary of unplanned readmissions/health-care resource use.

Authors	Type of monitoring	Unplanned readmissions/health-care resource use
Castellanos et al., 2016¹⁸	SpO2, weight; scheduled phone calls with nurse	Increased by 35% (p < 0.05)
Ghanayem et al., 2004²⁷	SpO2, weight; contact with HM team if breach in criteria	58% breached criteria and sought medical attention; no data for comparison
Hansen et al., 2012¹³	SpO2, weight; scheduled phone calls with nurse	31% breached criteria and were referred for observation
Rudd et al., 2014²⁵	SpO2, weight; contact with clinical team if breach in criteria	59% breached criteria; 57% required admission at least once; no control group
Siehr et al., 2014²⁶	HR, SpO2, weight, nutritional intake; contact with HM team if breach in criteria	41% readmissions; no control group
Koole et al., 2019²¹	ECG, BP, weight	1.8% readmissions; no control group
Gardner et al., 2019³¹	SpO2, weight; scheduled phone calls with nurse	63% of HM participants required at least one readmission compared to 46% for controls (p < 0.001)
Cooper et al., 2020¹²	HR, SpO2, weight and caloric intake home monitoring	15% of infants in HM group had more than one hospital readmission compared to 7% in control group (p < 0.001); 66% of infants in HM group were readmitted compared to 57% in the control group
Shirali et al., 2016²⁸	HR, SpO2; instant alerts to HM team	Unplanned readmissions, readmission days and ER visits trended lower
Petit et al., 2011³⁰	SpO2, weight; scheduled phone calls with nurse	Unplanned readmissions and ED visits did not differ between groups
Oster et al., 2015¹⁶	SpO2 and/or weight	No significant difference in unscheduled readmissions between groups
McCrossan et al., 2007¹⁰	Videoconferencing with SpO2 monitoring in some patients	Contact was advised less with videoconferencing (4%) compared to telephone support (22%)
McCrossan et al., 2012¹¹	Videoconferencing with SpO2 monitoring in some patients	Lower probability of being admitted at least once (p = 0.004)
Bingler et al., 2018⁹	SpO2, weight	Overall hospital length of stay trended lower (p = 0.07)

Table 4.

Summary of impact on treatment.

Authors	Type of monitoring	Impact on treatment
Gardner et al., 2019³¹	SpO2, weight, feeding log; scheduled phone calls with nurse	S2P 22 days earlier with HM (p < 0.001)
Ghanayem et al., 2003¹⁴	SpO2, weight; telephone support if required	S2P earlier with HM: 4.3 ± 1.6 months versus 5.6 ± 2.1 months (p = 0.016)
Ghanayem et al., 2004²⁷	SpO2, weight; telephone support if required	S2P earlier with HM: 4.1 ± 1.5 months versus 5.6 ± 2.1 months (p = 0.001)
Petit et al., 2011³⁰	SpO2, weight; scheduled phone calls with nurse	S2P earlier with HM: 150 ± 52 days versus 120 ± 114 days (p < 0.001)
Oster et al., 2015¹⁶	SpO2 and/or weight	S2P 26 days earlier with SpO2 HM (p = 0.002); S2P 21 days earlier with weight HM (p = 0.004)
Vergales et al., 2020²⁰	SpO2, weight, nutritional intake and output; digital platform with review by clinical team	S2P earlier with HM with digital platform: 4.6 months versus 5.4 months (p < 0.001)
Rudd et al., 2014²⁵	SpO2, weight; contact with clinical team if breach in criteria	S2P was carried out early in 36 patients who were readmitted due to home monitoring events
Castellanos et al., 2016¹⁸	SpO2, weight; scheduled phone calls with nurse	78% increase in percutaneous interventions (p < 0.01)
Harahsheh et al., 2016¹⁹	SpO2, weight, nutritional intake; contact with HM team if breach in criteria	More major post-Norwood procedures (p = 0.02)
Hansen et al., 2012¹³	SpO2, weight; scheduled phone calls with nurse	51% of patients who breached criteria underwent diagnostic catheterisation; no data to compare
Siehr et al., 2014²⁶	HR, SpO2, weight, nutritional intake; contact with HM team if breach in criteria	37% of patients required interventions triggered by HM; no control group
Öhman et al., 2013¹⁷	SpO2; contact with hospital if breach in criteria	38% of shunt complications were detected by HM and required intervention
Dobrolet et al., 2011¹⁵	SpO2, weight; scheduled phone calls with HM team	Mean age at detection, admission and treatment for residual lesions was younger (2.7 ± 0.9 months) versus historical controls (6.6 ± 0.6 months; p < 0.05)
Kauw et al., 2019²²	ECG, BP, weight	Recurrence of arrhythmia detected and treated in 25% of patients with history of arrhythmia; 38% of patients with hypertension at baseline had treatment adjusted
Koole et al., 2019²¹	ECG, BP, weight	Medication changes made in 10.9% of patients based on HM

The remaining nine studies included data for a control group. There was a significant increase in unplanned readmissions,¹²,¹⁸,³¹ with rates up to 35% higher.¹⁸ However, the duration of the readmissions was shorter for the home monitoring group compared to controls, with a median of three days (interquartile range (IQR) 2–7 days) versus four days (IQR 2–10 days; p = 0.002).³¹ Conversely unplanned readmissions, readmission days and emergency room visits trended lower with the use of the home monitoring with an App, but this did not reach statistical significance.²⁸

Two studies found no difference in rates of unplanned readmissions/health-care resource use.¹⁶,³⁰ There was no significant difference in unscheduled readmissions (any cause) between weekly versus daily monitoring, or between no monitoring versus daily monitoring.¹⁶

Contact with health service professionals was advised 18% less often with videoconferencing compared to phone support (p < 0.01).¹⁰ The probability of being admitted at least once to hospital was 37% lower with videoconferencing support compared to standard care (p = 0.004).¹¹ Home monitoring using the app was associated with significantly shorter unplanned length of stay in the intensive care unit (1 (IQR 1–2) vs. 6 (IQR 1–16); p = 0.03).⁹

Impact on treatment

The impact on treatment was reported in 15 studies (Table 4). Seven studies reported that age at stage 2 palliation (S2P) was younger in the home monitoring group,¹⁴,¹⁵,²⁰,²⁶,²⁷,³⁰,³¹ and age at S2P was 22 days earlier with home monitoring (p < 0.002).³¹ S2P in the home monitoring groups was earlier (p = 0.016 and p = 0.001).¹⁴,²⁷ One study found that the age of S2P for the home monitoring group was 150 ± 52 days versus 120 ± 114 days for the historical controls (p < 0.001).³⁰ A study comparing weight and/or SpO2 monitoring against no monitoring reported that S2P was 26 days earlier for those with SpO2 monitoring versus no SpO2 monitoring (p = 0.002) and 21 days earlier for those with weight monitoring compared to no weight monitoring (p = 0.004).¹⁶ It was also reported that S2P was carried out early in 36 patients (23% of participants) who were readmitted due to home monitoring events.²⁵

Five studies reported on interventions and treatments that were undertaken due to breaches in home monitoring criteria.¹³,^17–19,²⁶ Two studies included historical controls and found that there were more interventions/procedures undertaken in the home monitoring group,¹⁸,¹⁹ with a 78% increase in percutaneous interventions (p < 0.01)¹⁸ and more major post-Norwood procedures (p = 0.02)¹⁹ in the home monitoring groups. Three studies did not have a comparator and reported that there were between 17% and 57% of patients who breached surveillance criteria and required intervention or diagnostic catheterisation.¹³,¹⁷,²⁶

One study found that the mean age at detection, admission and treatment of residual lesions after shunt placements was younger in the home monitoring group (p < 0.005).¹⁵ It was reported in one study that in 25% of patients with a previous diagnosis of arrhythmia, a recurrence of the arrhythmia was confirmed and treatment initiated.²² In the final study, it was reported that medication changes were made in 11% of patients based on home monitoring.²¹

Discussion

The period between S1P and S2P is known to be a precarious time for CHD patients, with haemodynamic instability.³² Just over half (56%) of the studies found a significant reduction or a trend for lower mortality with home monitoring.

Interstage mortality (i.e. mortality that occurs in patients between S1P and S2P) is reported in the literature to be between 11% and 19%.^33–35 Reported mortality rates compare favourably to the literature, with 10 demonstrating lower mortality.^13–16,¹⁸,²⁰,²⁷,²⁸,³⁰,³¹ A 15% mortality was noted by one study,¹⁷ while the remaining studies were <11% when home monitoring was used. This would imply that the use of home monitoring may have beneficial effects, which lead to a lower mortality in this patient population.

No minor or major clinical complications were reported for patients who were home monitoring their international normalised ration (INR) using a prothrombin time (PT)-INR point-of-care device,²³,²⁴ which suggests that mortality was unrelated to the INR levels.

Studies that reported outcomes for mortality included RCT (n = 1),⁹ cohort studies (n = 12)^{13–20,27,28,30,31} and case series (n = 4).^23–26 Whilst 59% of studies reported either a significant reduction or a trend for lower mortality, these were all cohort studies, with 9/10 studies using historical controls. Furthermore, 3/10 studies were rated as moderate quality. The two studies that reported a trend for higher mortality also used historical controls which is problematic and introduces bias into the research. Medical and surgical treatments may have improved between the two time periods, data may be recorded differently and so on. Therefore, the evidence is weak for this outcome.

Between 1.8% and 63% of patients had unplanned readmissions. In the context of the literature available, these results compare favourably. A study in 2016 reported that 65.5% of patients had at least one unplanned readmission in the interstage period.³⁶ Similarly, a more recent study in 2018 found that 75% patients had unplanned readmissions.³⁷ In 62.5% of the studies that reported on unplanned readmissions, rates were lower than reported in the literature for interstage patients.³⁶,³⁷ An important finding from this review is that no studies reported unplanned readmissions using home monitoring significantly more than expected from the wider literature.

Between 31% and 61% of patients required observation and/or evaluation as a result of data from home monitoring. Two studies¹⁰,¹¹ found that contact with health service professionals was advised less often. This suggests that home monitoring is a valuable resource for detecting clinical deterioration in this patient population.

The increases in unplanned readmissions and health-care resource use may be explained by the fact that subtle deteriorations in clinical status could be picked up more frequently due to the home monitoring protocols. These increases should be viewed alongside mortality rates. The two studies that found a significant increase in unplanned readmissions also reported either a significant reduction or a trend for lower mortality rates with the use of home monitoring.

Studies that reported data for this outcome included RCTs (n = 4),^9–12 cohort studies (n = 7)^{13,16,18,27,28,30,31} and case series (n = 3).²¹,²⁵,²⁶ Three of the four studies which reported lower unplanned readmissions/health-care resource use were RCTs, which were all judged to have some concerns regarding bias. The other study was a cohort study using historical controls and was rated as high quality. Two of the three studies that reported an increase in relation to this outcome were cohort studies, using historical controls, with one study being an RCT. The remaining studies reported either no differences between groups or had no controls. Based on the design methodologies and quality appraisal scores, there is a stronger evidence that home monitoring reduces unplanned readmissions/health-care resource use. However, the evidence remains limited for this outcome.

Of the 15 studies^13–19,²⁰^–²²,^25–27,³⁰,³¹ that reported on the impact that home monitoring had on treatment, 47% found that S2P was performed at a younger age in the home monitoring groups.¹³,¹⁴,²⁰,²⁵,²⁷,³¹ The timing of S2P varies among centres and is a clinically led decision. Literature on the optimal timing of S2P reports that timing differs depending on whether patients are low, intermediate or high risk. A study in 2018 reported S2P being performed after three months of age was associated with maximal 2 year survival in low/intermediate risk infants.³⁸ Additionally, another study reported that median age at S2P for the study cohort was 155 days (IQR 109–214 days).³⁹ The earlier timing of S2P may be a reflection of earlier detection of clinical deterioration or clinicians being providing with more data on the patient, enabling them to make a more informed decision on the timing of S2P.

In studies without a control group, the intervention rates were reported to be between 37% and 38% of patients. The rates of re-interventions in these studies compare favourably to the literature, with a retrospective analysis of 1157 interstage patients reporting that 50% of patients required reintervention.⁴⁰

Studies for this outcome included cohort studies (n = 11)^{13–20,27,30,31} and case series (n = 4),²¹^–²²,^25–26 with all but one being of high quality. There were 9/15 (60%) studies that found that either treatment was undertaken significantly earlier with home monitoring or significantly more interventions were undertaken in the home monitoring groups.¹⁴,¹⁶,^18–20,²⁵,²⁷,³⁰,³¹ Eight of these were cohort studies using historical controls.¹⁴,¹⁶,^18–20,²⁷,³⁰,³¹ Of the remaining studies with either no data for comparison or no control group, only one was rated as high quality, and this was a case series design, with the rest rated as moderate quality. However, whilst most studies reported an either earlier treatment or increased interventions, given that these were cohort studies which primarily used historical controls, the methodological shortfalls in the studies mean that the evidence is limited for this outcome.

Strengths and limitations of the review

This review restricted the search to English language articles and excluded passive forms of home monitoring, which may have caused some relevant research to be excluded. Additionally, this review only considered clinical and health-care outcomes, and as such, qualitative outcomes such as anxiety levels, patient perceptions and so on have not been considered.

There have been no systematic reviews published on this topic to date, and this review systematically examined the full scope of the literature on home monitoring by patients with CHD. As such, this review provides a comprehensive review and interpretation of the data available.

Areas for further research

This systematic review included patients with CHD at any age who were undergoing home monitoring of their CHD. Only 2/20 (9.1%) of these studies were undertaken in adults with CHD. Improvements in survival rates of patients with CHD have led to increasing numbers of children and adults living with complex CHD. Approximately 90% of babies born with cardiovascular abnormalities are expected to reach adulthood,⁴¹ and as surgical techniques continue to improve, this could increase further in the future. In light of this, further research should focus on home monitoring interventions for adults with CHD, as home monitoring of this patient population has the potential to relieve pressure in the health service.

Conclusion

This systematic review examining the clinical and health-care outcomes for CHD patients when using home monitoring programmes identified that home monitoring may be beneficial in reducing mortality, enabling earlier and more timely detection and treatment of CHD complications.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Rachel Crawford

References

NICOR, https://www.nicor.org.uk/national-cardiac-audit-programme/congenital-heart-disease-in-children-and-adults-congenital-audit/ (accessed December 2020).

Liu

Chen

Zühlke

, et al. Global birth prevalence of congenital heart defects 1970–2017: updated systematic review and meta-analysis of 260 studies. Int J Epidemiol 2019; 48: 455–463.

GBD 2017 Congenital Heart Disease Collaborators. Global, regional, and national burden of congenital heart disease, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet Child Adolesc Health 2020; 4: 185–200.

World Health Organization. Telehealth, https://www.who.int/sustainable-development/health-sector/strategies/telehealth/en/ (accessed May 2020).

Inglis

Clark

Dierckx

, et al. Structured telephone support or non-invasive telemonitoring for patients with heart failure. Cochrane Database Syst Rev 2015; CD007228.

Kauw

Koole

Van Dorth

, et al. eHealth in patients with congenital heart disease: a review. Expert Rev Cardiovasc Ther 2018; 16: 627–634.

Sterne

JAC

Savović

Page

, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ 2019; 366: l4898.

Wells

Shea

O’Connell

, et al. The Newcastle–Ottawa Scale (NOS) for assessing the quality if nonrandomized studies in meta-analyses, www.ohri.ca/programs/clinical_epidemiology/oxford.asp (2012, accessed July 2020).

Bingler

Erickson

Reid

, et al. Interstage outcomes in infants with single ventricle heart disease comparing home monitoring technology to three-ring binder documentation: a randomized crossover study. World J Pediatr Congenit Heart Surg 2018; 9: 305–314.

10.

McCrossan

Morgan

Grant

, et al. Assisting the transition from hospital to home for children with major congenital heart disease by telemedicine: a feasibility study and initial results. Med Inform Internet Med 2007; 32: 297–304.

11.

McCrossan

Morgan

Grant

, et al. A randomised trial of a remote home support programme for infants with major congenital heart disease. Heart 2012; 98: 1523–1528.

12.

Cooper

Marino

Fleck

, et al. Telehealth home monitoring and postcardiac surgery for congenital heart disease. Pediatrics 2020; 146: e20200531.

13.

Hansen

Furck

Petko

, et al. Use of surveillance criteria reduces interstage mortality after the Norwood operation for hypoplastic left heart syndrome. Eur J Cardiothorac Surg 2012; 41: 1013–101.

14.

Ghanayem

Hoffman

Mussatto

, et al. Home surveillance program prevents interstage mortality after the Norwood procedure. J Thorac Cardiovasc Surg 2003; 126: 1367–1377.

15.

Dobrolet

Nieves

Welch

, et al. New approach to interstage care for palliated high-risk patients with congenital heart disease. J Thorac Cardiovasc Surg 2011; 142: 855–860.

16.

Oster

Ehrlich

King

, et al. Association of interstage home monitoring with mortality, readmissions, and weight gain: a multicentre study from the National Pediatric Cardiology Quality Improvement Collaborative. Circulation 2015; 132: 502–508.

17.

Öhman

Strömvall-Larsson

Nilsson

, et al. Pulse oximetry home monitoring in infants with single-ventricle physiology and a surgical shunt as the only source of pulmonary blood flow. Cardiol Young 2013; 23: 75–81.

18.

Castellanos

Herrington

Adler

, et al. Home monitoring program reduces mortality in high risk sociodemographic single-ventricle patients. Pediatr Cardiol 2016; 37: 1575–1580.

19.

Harahsheh

Hom

Clauss

, et al. The impact of a designated cardiology team involving telemedicine home monitoring on the care of children with single-ventricle physiology after Norwood palliation. Pediat Cardiol 2016; 37: 899–912.

20.

Vergales

Peregoy

Zalewski

, et al. Use of a digital monitoring platform to improve outcomes in infants with a single ventricle. World J Pediatr Congenit Heart Surg 2020; 11: 753–759.

21.

Koole

Kauw

Winter

, et al. First real-world experience with mobile health telemonitoring in adult patients with congenital heart disease. Neth Heart J 2019; 27: 30–37.

22.

Kauw

Koole

Winter

, et al. Advantages of mobile health in the management of adult patients with congenital heart disease. Int J Med Inform 2019; 132: 104011.

23.

Christensen

Attermann

Hjortdal

, et al. Self-management of oral anticoagulation in children with congenital heart disease. Cardiol Young 2001; 11: 269–276.

24.

Christensen

Andersen

Maegaard

, et al. Oral anticoagulation therapy in children: successfully controlled by self-management. Heart Surg Forum 2004; 7: e321–e325.

25.

Rudd

Frommelt

Tweddell

, et al. Improving interstage survival after Norwood operation: outcomes from 10 years of home monitoring. J Thorac Cardiovasc Surg 2014; 148: 1540–1547.

26.

Siehr

Norris

Bushnell

, et al. Home monitoring program reduces interstage mortality after the modified Norwood procedure. J Thorac Cardiovasc Surg 2014; 147: 718–723.

27.

Ghanayem

Cava

Jaquiss

, et al. Home monitoring of infants after stage one palliation for hypoplastic left heart syndrome. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu 2004; 7: 32–38.

28.

Shirali

Erickson

Apperson

, et al. Harnessing teams and technology to improve outcomes in infants with single ventricle. Circ Cardiovasc Qual Outcomes 2016; 9: 303–311.

29.

Institute of Health Economics. Quality appraisal of case series studies checklist, www.ihe.ca/research-programs/rmd/cssqac/cssqac-about (2014, accessed July 2020).

30.

Petit

Fraser

Mattamal

, et al. The impact of a dedicated single-ventricle home-monitoring program on interstage somatic growth, interstage attrition, and 1-year survival. J Thorac Cardiovasc Surg 2011; 142: 1358–1366.

31.

Gardner

Mercer-Rosa

Faerber

, et al. Association of a home monitoring program with interstage and stage 2 outcomes. J Am Heart Assoc 2019; 8: e010783.

32.

Cross

Harahsheh

McCarter

, et al. Identified mortality risk factors with presentation, initial hospitalisation, and interstage period for the Norwood operation in a multi-centre registry: a report from the National Pediatric Cardiology – Quality Improvement Collaborative. Cardiol Young 2014; 24: 253–262.

33.

Ghanayem

Allen

Tabbutt

, et al. Interstage mortality after the Norwood procedure: results of the multicenter Single Ventricle Reconstruction trial. J Thorac Cardiovasc Surg 2012; 144: 896–906.

34.

Vojtovič

Tláskal

Gebauer

, et al. Long-term results of children operated for hypoplastic left heart syndrome in Children’s Heart Centre. Cor Vasa 2014; 56: e449–e455.

35.

Cua

Thiagarajan

Taeed

, et al. Improved interstage mortality with a modified Norwood procedure: a meta-analysis. Ann Thorac Surg 2005; 80: 44–49.

36.

Hanke

Joy

Riddle

, et al. Risk factors for unanticipated readmissions during the interstage: a report from the National Pediatric Cardiology Quality Improvement Collaborative. Semin Thorac Cardiovasc Surg 2016; 28: 803–814.

37.

Simsic

Phelps

Kirchner

, et al. Interstage outcomes in single ventricle patients undergoing hybrid stage 1 palliation. Congenit Heart Dis 2018; 13: 757–763.

38.

Meza

Hickey

McCrindle

, et al. The optimal timing of stage-2-palliation after the Norwood pperation. Ann Thorac Surg 2018; 105: 193–199.

39.

Hill

Rudd

Ghanayem

, et al. Center variability in timing of stage 2 palliation and association with interstage mortality: a report from the National Pediatric Cardiology Quality Improvement Collaborative. Pediatr Cardiol 2016; 37: 1516–1524.

40.

Buelow

Rudd

Tanem

, et al. Reintervention following stage 1 palliation: a report from the NPC-QIC registry. Congenit Heart Dis 2018; 13: 919–926.

41.

Moons

Bovijn

Budts

, et al. Temporal trends in survival to adulthood among patients born with congenital heart disease from 1970 to 1992 in Belgium. Circulation 2010; 122: 2264–2272.

A systematic review examining the clinical and health-care outcomes for congenital heart disease patients using home monitoring programmes

Abstract

Objectives

Methods

Conclusion

Keywords

Introduction

Methods

Search strategy

Quality assessment and risk of bias

Results

Study characteristics

Risk of bias

Newcastle–Ottawa Quality Assessment Scale

IHE quality appraisal checklist

Synthesis of results

Mortality

Unplanned readmissions/health-care resource use

Impact on treatment

Discussion

Strengths and limitations of the review

Areas for further research

Conclusion

Footnotes

Declaration of conflicting interests

Funding

ORCID iD

References