Objectives To assess the sustainability, clinical utility and acceptability to clinicians and parents of a tele-homecare programme for infants with major congenital heart disease (CHD), and to evaluate the impact on healthcare resource use.
Design Randomised control trial.
Setting UK tertiary congenital cardiac centre.
Participants 83 infants with major CHD.
Intervention Participants were randomised to one of three groups: video-conferencing support (n=35), telephone support (n=24) and a control group (n=24). Patients in the two intervention groups received regular, standardised remote consultations. Video-conferences (VCs) were facilitated by Integrated Systems Digital Network lines and replaced by home broadband connections later in the study.
Main outcome measures Healthcare resource use, utilisation including hospitalisation, clinicians' opinions on utility and quality of interventions, parental opinions on quality of interventions.
Results Clinicians were more confident making medical decisions following VCs compared with telephone consultations (p=0.01). Both VC and telephone support were very well received, but parents expressed significantly higher levels of satisfaction with VC support (p=0.001). Healthcare resource use was 37% lower in the video-conferencing group compared with both telephone support and control groups (p<0.001), as was the risk of hospitalisation (p=0.006). Direct health service costs were significantly lower in the video-conferencing group (p<0.05).
Conclusions A tele-medicine home support programme for families of infants with major CHD is feasible, sustainable and effective. Home support with video-conferencing is superior to telephone consultations. Parents are highly satisfied with tele-homecare. Tele-homecare significantly reduces health service utilisation and may reduce health service costs.
- Congenital heart disease
- home support
- paediatric cardiology
- paediatric surgery
- interventional cardiology
- non-coronary intervention
- device closure
Statistics from Altmetric.com
- Congenital heart disease
- home support
- paediatric cardiology
- paediatric surgery
- interventional cardiology
- non-coronary intervention
- device closure
Tele-home care has become established as a useful support mechanism for adults with specific chronic illnesses.1–3 In particular, there is evidence that home monitoring is associated with reduced hospital admissions and mortality in patients with congestive heart failure.2 ,4 Recent reports suggest that remote monitoring of implantable cardiac defibrillators is safe, time-saving and cost-saving.5 ,6
While the scale of paediatric chronic illness is much smaller than in adult and elderly populations, the past 20 years has seen a significant increase in the number of children with chronic complex illness.7 This is especially true of paediatric cardiology. Advances in care means that babies with even the most severe forms of congenital heart disease (CHD) now have the possibility of life beyond the neonatal period.7 ,8 However, patients with complex CHD, typified by single ventricle physiology are never ‘cured’, and the most vulnerable period can be between the first and second stages of surgical palliation.9
A pilot project conducted by this research team suggested that a tele-medicine home support programme was feasible and well received.10 The primary objectives of this study were to rigorously examine the sustainability, clinical utility and acceptability to clinicians and parents of a home support programme for families of babies with major CHD following discharge from hospital. Most importantly, this study aimed to assess the impact on healthcare resource use and the attendant costs.
This study was performed at the department of paediatric cardiology, Royal Belfast Hospital for Sick Children. Ethical approval was granted from the Queen's University Belfast Research Ethics' Committee. Informed consent was obtained from all participants. All paediatric cardiology admissions between August 2005 and October 2008 were considered for inclusion in the study. Patients were eligible if they had complex CHD or congestive heart failure, as well as infants with corrected CHD who had a difficult perioperative course. The essential inclusion criterion was if the attending paediatric cardiologist felt that the clinical condition of the patient would require significant support following discharge from hospital. The attending paediatric cardiologist was not in the study following patient selection. The only exclusion criterion was if there was no fixed address at which to install the equipment.
Study groups and randomisation process
There were three study groups: two in the intervention- video-conference (VC) and telephone group, and one control group. Participants were randomly allocated (computer generated random numbers) to one of the three study groups depending on the availability of VC equipment. If VC equipment was available, a 2:1:1 randomisation weighted towards the VC group was employed. If no codec was available, then participants were randomised on a 1:1 basis between the telephone and control groups. On nine occasions, VC equipment was not available. Patients were enrolled for a predicted period of 10 weeks. However, there was some flexibility permitted depending on the clinical status of the patient.
The intervention took the form of scheduled VC consultations. The video consultation consisted of an initial enquiry if there were any ongoing concerns. This was followed by systematic questioning relating to feeding, weight gain, respiratory status, general form and review of medications. The clinician then visually assessed the patient. Pulse oximetry was obtained from patients in whom variations in oxygen saturation were felt to indicate deviations in their cardiac status.
A summary of the clinical assessment was related to the parent along with the clinician's recommendations for management until the next assessment. Parents were encouraged to ask questions. Video consultations were arranged on a weekly or twice-weekly basis depending on the wishes of the parents or clinician. There was also the facility of ‘urgent’ consultations whereby parents could request a video consultation before the next one scheduled if they had any concern regarding the clinical condition of their baby. These were not ‘emergency’ consultations, that is, in situations where the parent was concerned that their child was acutely unwell. This condition was stipulated very clearly at the outset and repeated periodically throughout the study period. The study protocol did not stipulate specific parameters (eg, oxygen saturations/weight loss) to prompt urgent consultation as the cohort comprised a heterogeneous group of lesions.
Telephone consultation group
The telephone consultations followed a parallel schedule and format to the VCs. The only difference was that the clinician was unable to visually assess the patient.
Parents in this group were informed that they had been randomised to the control group and would still receive the same level of care from the paediatric cardiology team as if they were not involved in the study.
Commercially available tele-medicine equipment was utilised. A Tandberg 880TM codec (Lysaker, Norway) in the hospital and a Tandberg 1000TM ‘Classic’ (Lysaker, Norway) in their homes. The pulse oximeter distributed to selected patients was the Masimo SET (Neuchatel, Switzerland).
VCs were performed across standard tele-medicine links. Initially, Integrated Systems Digital Network lines (ISDN 6, 384 Kbps) were used as we had experience using this connection. During the last 12 months of the study following pilot testing, the tele-link was changed to an Asynchronous Digital Subscriber Loop (ADSL, 256 Kbps) which is a form of internet connection.
Qualitative data was collected prospectively from families and clinicians by means of structured questionnaires, devised by the research team, involving a Likert scale. Healthcare resource usable data was acquired from the relevant hospital and primary care databases.
Comparisons between the video-conferencing group and telephone group were analysed using an independent samples t test, or the corresponding non-parametric test (Mann–Whitney U test). Comparisons across all three study groups, including healthcare resource use, were analysed using χ2 test, analysis of variance, or the appropriate non-parametric test (Kruskal–Wallis test). Linear regression was used to compare the mean average number of NHS episodes per 10 weeks in the video, telephone and control groups after adjusting for age, sex, distance from hospital, physiology and surgical status. Logistic regression was used to compare the odds of being admitted to hospital at least once in the video, telephone and control groups (over the study period) after adjusting for age, sex, distance from hospital, physiology and surgical status. This analysis was performed using SPSS for Windows version 16.0. The perspective of the cost analysis was that of the payer (ie, the NHS). Healthcare resource use (hospital bed days, emergency department visits, general practitioner visits, paediatric outpatient visits, paediatric cardiology outpatient visits and specialist nurse contact (visits and telephone contact)) was collected prospectively during the randomised controlled trial (RCT) Resource use was valued in £ sterling using unit costs based on the financial year 2006/2007 and obtained from national sources and Unit Costs of Health and Social Care.11 ,12
Due to the tendency for cost data to be positively skewed, non-parametric bootstrapping was used to analyse the difference in healthcare costs between the study groups. This analysis was performed using STATA V.10 for Windows.
Between 31 August 2005 and 31 October 2008, 85 patients and families were eligible for inclusion in the study. Two families refused consent leaving 83 patients and families recruited to the study. Thirty-five patients were randomised to the VC group, 24 to the telephone group and 24 to the control group. Patient demographics are shown in table 1.
Over the 41-month study period, 370 video consultations were conducted involving 35 patients and families. Twelve urgent video consultations were requested (3.2%). Family involvement in the home support program lasted on average 12.1 weeks (SD 5.7 weeks). A successful audiovisual connection was made at the first attempt in 342 consultations (92.4%).
Technical problems were noted during 62 video consultations (16.8%). Technical problems resulted in an aborted video consultation on only five occasions (1.4%). The mean video consultation lasted 10.5 min (SD=2.0 min). Over the 41-month study period, 273 telephone consultations were conducted involving 24 families. Eight urgent telephone consultations were requested (3.0%). Family involvement in the home support program lasted on average 11.3 weeks (SD 3.5 weeks). The mean consultation time was 8.0 min.
Clinician evaluation of video consultations
The clinical utility of the 370 video consultations rated highly on a five-point Likert scale. The clinician strongly agreed that the parents' concerns could be appreciated (median=4.5/5, IQR=4.3–4.9), that the child could be confidently assessed (median=4.5/5, IQR=4.3–4.8), and that the parents' concerns could be adequately addressed across the tele-link (median=4.6/5, IQR=4.3–4.8). The clinician agreed that the video consultations were beneficial (median=4.3, IQR=4.1–4.5).
Comparison of video-conferencing support with telephone support
The VC group was involved in the home support program slightly longer compared with the telephone group, but this was not statistically significant (mean=12.1 vs 11.3 weeks, independent samples t test p=0.08). There was no significant difference in the number of consultations per family between the VC and telephone group (mean=10.6 vs 11.4, independent samples t test p=0.51). However, the video consultations were significantly longer than the telephone consultations (mean difference=2.5 min, 95% CIs 1.6 to 3.5 min, independent samples t test p=0.04) (table 2).
A similar proportion of video and telephone consultations included the parent expressing a specific concern (median per patient=63% vs 68%, χ2 test p=0.19). However, there was a significant difference in the proportion of patients advised to seek further healthcare assessment on one or more occasions (VC=40% vs telephone=96%, χ2 test p<0.01). In 97% of video consultations the clinician felt he was able to address the parental concerns compared with 64% of telephone consultations. This finding is supported by the clinician's opinion as to the benefit of visually assessing the patient. At the conclusion of 30% of video consultations, the clinician stated that the ability to visually assess the patient had influenced their management plan. Similarly, following 27% of telephone consultations, the clinician stated that the inability to visually assess the patient had affected their management plan.
Table 3 demonstrates that parents were extremely positive about the video consultations with a median four-point Likert scale=4 across all four questions. In particular, parents felt the tele-medicine equipment was easy to use. This is especially gratifying as one of the main concerns at the inception of this study was parents having difficulties with the tele-medicine equipment. Parents in the telephone group were also very positive about telephone support. However, direct comparison of video with telephone consultations demonstrates that parents felt that video consultations were superior in terms of facilitating communication and overall benefit (p=0.001).
Healthcare resource use
Overall, the rate of health service episodoes was significantly reduced in the VC group (mean=7.7 per 10 weeks) with approximately 40% fewer health service episodes than either telephone (mean=13.3 per 10 weeks) or control groups (mean=12.9 per 10 weeks) (table 4). Linear regression of this data demonstrates that patients in the VC group had significantly fewer health service episodes per week than the telephone support or control group patients, despite adjusting for all the variables described in table 1. From the linear regression, the adjusted mean in the VC group compared with the control group (95% CI) was −0.5 (−0.8, −0.3) and the telephone group compared with the control group was +0.1 (−0.2, +0.4). Most importantly, the probability of being admitted at least once to hospital was significantly lower in the VC group (6/35, 17%) compared with the telephone (11/24, 46%) and control (13/24, 54%) (p=0.004) groups. Logistic regression adjusting for the same variables demonstrates that the probability of hospital admission remains significantly lower in the VC group. The adjusted OR for the VC group compared with the control group was (95% CIs) 0.15 (0.04 to 0.57) p=0.006, and in the telephone group compared with the control group, it was 0.73 (0.18 to 2.89) p=0.65.
Cost of healthcare resource use
A mean cost per patient for health service resource use was calculated based on patient-level data. Table 5 summarises the cost to the NHS of healthcare resource use for the three study groups. Healthcare resource use costs are considerably lower in the video-conferencing group compared with either telephone group (mean difference/patient = £1563; 95% CI £502 to £2600) and control group (mean difference/patient = £2760; 95% CI £732 to £5478). The main cost driver was inpatient bed days. Although the mean cost/patient in the telehone group was lower than the control group, it was not statistically significant (mean = £1197; 95% CI £ −937 to £3920).
This study clearly demonstrates that a tele-medicine home support programme for infants with major CHD is feasible, sustainable and effective. Clinicians are more confident in making clinical judgements via a VC than a telephone consultation. Parents are very positive about home support provided by both video and telephone consultations, but VC support was considered superior in facilitating communication and the overall benefit conferred. The VC equipment is also easy to use. Home video-conferencing support was associated with a 37% reduction in health service episode per patient compared with both telephone support and control groups. In particular, the risk of hospital admission was significantly lower in the VC support group. Attendance at the family doctor and emergency department were also significantly lower in the video-conferencing group compared with the control group. Consequently, mean healthcare resource use costs per patient were significantly lower in the video-conferencing group compared with both telephone (£1560) and control groups (£1200). These savings need to be considered alongside the costs of implementing and running a home support programme, which will be presented in a subsequent separate analysis.
It is a strength of this study that 98% (83/85) of individuals agreed to participate, and that no one dropped out. However, the authors acknowledge that the relatively small sample size divided into three groups has limited the power of the study, and is responsible for some of the differences in demographic patterns across the study groups. Another weakness in the study design is that the clinician was not blinded to which group the patient was in.
There is good and increasing evidence that tele-homecare can be an effective method of healthcare delivery. A recent meta-analysis of all home-based tele-health research indicated a moderate positive effect on health outcomes (mean weighted effect size = 0.5, p<0.01). Subanalysis revealed particular benefits in patients with mental health illness and chronic adult heart disease, but no significant effect in diabetes mellitus.13 The potential for tele-homecare in the management of chronic obstructive pulmonary disease has also been highlighted, and trials are now in progress.14 ,15 Home monitoring is now established as an adjunct to the ongoing management of specific adult populations: for example, chronic heart failure, hypertension and implantable cardiac defibrillators/pacemakers.2 ,4–6 ,16 In 2000, a major US healthcare organisation (Kaiser Permanente) published a recommendation that home tele-health should be incorporated into the standard plan for subgroups of patients.17 It is recommended that tele-homecare be targeted at conditions that require close monitoring, clinical assessment and early intervention to avoid adverse events, such as hospitalisation and emergency visits.13 In addition, parental expectations have increased regarding the quality of life experienced by their child, and the level of specialist input that should be provided (Little Hearts Matter, personal communication, 2009). This is particularly relevant to paediatric cardiology, which is a highly specialised and centralised field of acute medicine. It seems logical that members of the paediatric cardiology team, who have been involved in the daily management of these patients, are ideally placed to monitor progress and deal with any problems as they arise.
There is very little evidence of home support/monitoring trials in a paediatric setting. A small number of studies are reported in the literature relating to home ventilation, neonates and diabetes, but on closer inspection their design and scope are particularly limited. However, Ghanayem et al devised a simple system of monitoring pulse oximetry and weight in infants with hypoplastic left heart syndrome between the first and second surgical stages of palliative reconstruction. This trial found a significant reduction in inter-stage mortality compared with historical controls.18 ,19 A pilot project of our own home support programme demonstrated that home video-conferencing via ISDN lines was technically feasible, that clinicians were satisfied with the image quality in order to adequately assess the patient, and that parental anxiety was reduced.10 ,20
We believe this home support programme is effective because it permits rapid access to expert opinion and identification of potentially serious issues. At the outset of this project, there were misgivings regarding the possibility of falsely reassuring parents, and the consequent risk that sick patients would be kept at home. This concern has reduced through our experience. There were no incidents of clinical risk noted throughout the study. It also facilitates resolution of more minor problems which cause considerable parental anxiety, and when presented to the non-specialist in primary and secondary care settings, may result in hospital admission. There are several important advantages to keeping these vulnerable patients at home: avoiding exposure to infections in crowded waiting areas and hospital wards, and minimising family disruption, particularly, associated with hospitalising a child.21 Parents feel confident that their baby has been clinically assessed by a health professional familiar with the anatomy, physiology and ongoing health issues particular to their baby. By the end of the tele-consultation, parents feel that their concerns have been understood and addressed satisfactorily. The facility to visually assess the patient, and thereby evaluate general well being, respiratory rate, work of breathing, hydration level and wound integrity, is crucial to the success of this home support programme. The facility for accurate pulse oximetry and heart rate measurement during the consultation are also valuable adjuncts in making a confident clinical assessment. The home support programme is now nurse-led, and has become integrated into the routine postdischarge care for all suitable patients at our unit.
In the future, Skype and other similar voice over internet protocol services may facilitate clinical-quality video-conferencing, but probably not yet. Another important consideration is that the video-conferencing hardware also impacts on the visual quality of the consultation. The camera built into modern laptop computers/tablets may not be of sufficient quality to permit satisfactory visual examination of the patient. Improvements in internet coverage, bandwidth and consistency (including mobile phone communication standards, eg, 4G), in conjunction with rapid advances in mobile computers are exciting developments which will most likely enhance tele-medicine home support programmes. However, it is vital that such innovations are evaluated in order to confirm that they deliver clinical-quality consultations.
This tele-homecare programme could be expanded to incorporate other patient populations, such as unstable angina, congestive heart failure, stroke rehabilitation, chronic respiratory disease and mental health disorders, apart from paediatric subspecialties.
A tele-medicine home support programme for families of infants with major CHD is feasible, reliable and effective. Home support, facilitated via video-conferencing, is superior compared with telephone consultations. Parents are highly satisfied with tele-homecare. Tele-homecare significantly reduces primary and secondary healthcare utilisation and, in particular, is associated with reduced risk of hospital admission with a consequent reduction in health service costs.
The authors would like to acknowledge the technical and financial support of Questmark Limited, and the statistical advice of Dr Chris Cardwell.
Funding This work was supported by: (1) The Paediatric Cardiology Charitable Funds of the Royal Belfast Hospital for Sick Children; was paid the research salary. (2) Questmark Limited video-conferencing company. This took the form of not charging for the use of the video-conferencing equipment and technical support. It also included paying the cost of phone line rental. None of the researchers have received any payments from this company.
Competing interests This study was partially funded by a videoconferencing company Questmark limited.
Ethics approval This study was conducted with the approval of the research ethics committee of Queen's University, Belfast.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement The data presented is complete; an additional cost analysis data was conducted. This data will be prepared for future publication.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.