A systematic review of economic analyses of telehealth services using real time video communication

Telehealth is the delivery of health care services at a distance, using information and communication technology (ICT). Telehealth became a separate field of study from the 1970's[1], and innovation increased from the 1990's, due to the development of new technologies such as cellular phones and the internet[2]. The field was originally known as telemedicine, but was later broadened to telehealth[3], although these terms continue to be used interchangeably. It is a subset of e-health, which encompasses all uses of ICT in health, including electronic records and decision support systems, however telehealth is particularly characterised by the geographical separation of patient and provider[4]. Tulu[5] categorises telehealth according to its purpose, such as clinical, educational or administrative; its healthcare discipline area; the environmental setting; the type of communication infrastructure used; and the delivery modality. Telehealth applications are very diverse, ranging from home care for chronic diseases, to remote primary care and subspecialist services such as paediatric cardiology. Telehealth can be delivered synchronously, also known as real-time, where the participants interact with each other simultaneously, and asynchronously, also known as store-and-forward, in which information such as X-rays or photographs are collected, transmitted, and then utilised at a later time.

Economic analysis is of central importance to telehealth because the main rationales for its introduction have been to decrease the cost of delivering health care, make more efficient use of the health workforce, and improve timely and equitable access to services. Expansive promises have been made about the potential to achieve these ends. For example Cusack[6]modelled cost savings of $4.3 billion a year if telehealth were implemented to facilitate consultations between healthcare providers in the USA, and this is without considering savings associated with the provision of care direct to the patient. However, the ability to deliver these results needs to be tested in the real world.

Three reviews published in 2000 covered the early economic analyses of telehealth[7‐9], and each reported that the studies were methodologically flawed; lacking appropriate outcome measures, consistency and clarity. Thus, no conclusions could be made about the cost-effectiveness of telehealth at that time. In 2001, Roine[10] included economic assessments in a general review of telemedicine, and noted that most were poor quality cost comparisons of short term pilot projects, however there was evidence for cost savings in teleradiology. Whitten's review in 2002[11] appraised the quality of telehealth articles containing cost data, also finding that most were simple cost comparisons, with no definitive comparison of telemedicine to traditionally organised care. They also noted the difficulty of generalising results of individual economic studies due to the variability of applications and the effect of unique local factors on each telehealth service. Jennet reviewed the broader socio-economic impacts of telehealth in 2003[12] and combining cost, cost-effectiveness and health system utilisation measures into one category, concluded that there was evidence of benefit of at least fair to good quality in paediatrics, geriatrics, home care, radiology, mental health and rehabilitation. Hailey's review[13] of the benefits of telemedicine in 2004 rated 25 health economic analyses, using Drummond's criteria[14], finding 13 studies met 5 or more criteria and were rated as fair to good. These indicated cost savings in radiology, geriatric care, and intensive care, and conflicting evidence in dermatology.

From 2004, telehealth reviews diverge into specific areas of clinical practice, some of which contained economic outcomes. Two reviews of telepsychiatry concluded that cost-effectiveness could not be demonstrated because the volume of consulting was too low[15, 16], while another found conflicting results of both increased and decreased costs[17]. Brebner[18] reviewed telehealth provision of accident and emergency support to primary care, and found seven studies with economic data, all indicating cost-effectiveness, however he concluded that the case was far from proven. A review of the use of telehealth in intensive care units found two clinical trials which showed cost savings[19]. Pare [20] summarised a number of reviews of home care for chronic disease, and reported that very few detailed economic analyses had been done, leading to no confirmation of economic viability. However, in regard to heart failure, Martinez concluded that home monitoring reduced costs of hospital admissions[21], and Seto reported initial costs, but substantial long term cost savings[22]. Recently, Bergmo[23] reviewed the quality of economic evaluations in telemedicine, and echoed the earlier findings of highly diverse evaluations, many of which did not adhere to standard economic evaluation techniques. Specifically, statistical, sensitivity, and marginal analyses, and information on the perspective of the studies were often lacking. Whereas this review pointed out methodological deficiencies, it did not aim to draw conclusions about the cost-effectiveness of telemedicine.

These reviews have covered a broad variety of telehealth applications, from synchronous videoconferencing, to remote monitoring, telephone follow-up, call centre advice lines, email and web-based systems. Each mode of telehealth has requirements for particular technology, staffing, services, and means of organisation, and consequently has different cost components, and can deliver particular types of outcomes. Not surprisingly, it has been difficult to compare results across studies and offer consistent guidance for cost-effective health services development. Therefore this review will focus on one mode of telehealth: real-time or synchronous video communication.

Real-time video can be regarded as the traditional form of telehealth, and despite new developments such as interactive software for chronic disease management, "smart houses" to measure resident's activity levels, and a range of monitoring devices such as scales, glucometers, and peak flow meters now able to send data remotely, real-time video remains in common use across a wide range of disciplines, particularly in mental health, primary care, specialist consulting, and multidisciplinary teamwork. It has particular requirements, and hence associated costs, for video screens, connectivity of sufficient bandwidth and reliability for real time communication, physical space at each location, and health providers' time to deliver the services. Research has been able to directly compare remote consultations with in-person consultations because the health providers' activities are similar to usual care. The intention of this review is to synthesise the economic analyses of this distinct form of telehealth, and determine whether conclusions can be made about its value.

The literature on telehealth is extensive, however there are a relatively small number of economic analyses that met the review criteria. Identification and review of evidence of equivalence for all economic studies that did not claim or cite equivalence was beyond the capacity of this study, and so a pragmatic approach was taken that excluded economic studies that either did not claim equivalence or provided no direct or indirect (cited) evidence of effectiveness. Cited evidence of equal health outcomes was reviewed to establish that the evidence was relevant to the setting of the economic analysis. This approach resulted in the exclusion of some well-conducted cost analyses[65‐68]. Identification of evidence around the effect of the services analysed in these studies would increase the pool of evidence on the cost-effectiveness of telehealth services.

Those studies in which the only health outcomes reported were patient satisfaction ratings were also excluded because telehealth studies report almost uniformly high patient satisfaction, which is likely to be a positive bias due to social desirability and acquiescence[69]. Finally, studies were excluded in which the health outcomes were not comparable, such as where participants receiving telehealth were asked to rate the success or accuracy of the teleconsultation, but this was not done for the usual care comparator.

Quality assessment using Chiou's criteria[24] gave similar results to those of Bergmo's recent review of economic analyses in telehealth[23]. Bergmo reported that studies were most lacking in providing information on the perspective of the study, statistical comparison, sensitivity analyses, and marginal analysis. The criterion least fulfilled in this review, allowing sufficient time to assess all important outcomes, was not used by Bergmo, who developed a new set of criteria based on issues mentioned in the telehealth literature. By comparison, Chiou's criteria were developed and validated by a study of health economists. Also in line with earlier reviews[11, 23], there was great variability in methodology and outcome measures. The majority of studies were cost-consequences analyses, utilising outcomes specific to each service, and therefore a quantitative meta-analysis could not be attempted.

Overall, nearly two thirds of the studies showed cost savings in utilising telehealth, according to the perspectives adopted by each. In particular, whenever the patient perspective was assessed, telehealth was found to be cost-saving, however when the health services perspective was considered alone, which was possible for 35 of the studies, the proportion reporting cost savings reduced to half. Most studies showed similar health outcomes, about a third showed improved outcomes, and only two indicated that telehealth was less effective. Surprisingly, there was no obvious trend for the more recent studies to indicate more cost savings, which was expected due to the reduction in technology costs over time. When seeking to ascertain the particular circumstances within which telehealth was most cost-effective, the organisational model of care was found to be the most relevant factor.

Reviewing 36 economic analyses of the delivery of health services by synchronous video communication indicates that this form of telehealth can offer value to health care, and it suggests that key factors associated with this are the settings and particular models of health service delivery. The health outcomes of the patients were either equal to or better than conventional care, with two minor exceptions that did not compromise quality of care. Therefore the decision as to whether or not to introduce a telehealth service can be made using cost-effectiveness criteria and consideration of the model of care.

It is concluded that synchronous video delivery is cost-effective for home care, and for on-call hospital specialists, and it can be cost-effective for regional and rural health care, depending upon the particular circumstances of the service. However, it is not cost-effective, from the health services perspective, for local delivery of service between hospital specialists and primary care, particularly due to additional health care staffing. Across settings, equipment and connectivity costs have been major factors in setting up telehealth services, but even as these costs reduce, this will not necessarily make telehealth more cost-effective, unless the other factors such as health workforce and facility space are also addressed.

Improvement in the quality of economic analyses is also needed to provide data for more accurate modelling of the effects of widespread introduction of telehealth into the health care system.