Background

The growing burden of the group of diseases that comprise cardiovascular disease (CVD) on national healthcare systems is well recognised amongst healthcare professionals internationally. Despite a trend towards decreasing mortality, CVD remains the major cause of death worldwide [1]. In Australia, CVD accounts for a greater proportion of deaths than any other disease group, and accounts for the highest proportion of direct health care expenditure [2]. Improved acute care, advanced intervention techniques, early diagnostic procedures, and worldwide demographic ageing have resulted in an increasing number of patients living with chronic CVD [3, 4]. As a result, chronic heart failure (HF) is becoming an increasing health issue [5]. Almost half the individuals admitted to hospital for HF are re-admitted within 6 months [6, 7]. These frequent hospital re-admissions contribute significantly to the high costs of HF management [6, 7]. Up to 70% of HF cases, are preceded by hypertension [5]. Implementation of evidence based hypertension management has the potential to reduce the prevalence of HF and, subsequently, reduce HF burden. Given the interrelationship and progressive nature of cardiac disease in primary care, there are advantages in considering the impact of management strategies across the disease silos.

The increasing burden associated with CVD necessitates the investigation of innovative models to provide evidence-based care to promote early identification of exacerbations and early intervention to minimise their severity. Among the available models is telemonitoring, which consists of remote monitoring of patients to enable clinicians to intervene when there is evidence of clinical deterioration in an attempt to avoid hospitalisation [8]. Telemonitoring moves patient care out of a clinical setting into the patient’s home by utilising telecommunications technologies such as the Internet, telephone, or videoconferencing to transmit physiological data and information about current symptoms from the patient to health care professionals [8]. Shifting the burden of care to the patient’s home also facilitates supported self-care, giving the patient enhanced autonomy and control of their health care [8]. Telemonitoring programs can be flexible, individually tailored, and have the potential to provide access to specialist care for a larger number of patients across a much greater geography when compared to usual care [8–10].

Currently the literature reports multiple systematic reviews investigating the effectiveness of telemonitoring in both HF and its antecedent, hypertension. Given the large amount of literature available, a systematic review of reviews was conducted to synthesise and critically evaluate this current Level 1 evidence around the use of telemonitoring in the primary care setting.

Methods

This review was conducted using the Preferred Reporting Items of Systematic Reviews Meta-Analyses (PRISMA) guidelines [11]. A systematic review design was selected to limit any bias in the selection and reporting of evidence.

Search strategy

A comprehensive and systematic search was undertaken using CINAHL, MEDLINE and the Cochrane Library electronic databases. The search strategy included the Medical Subject Headings (MeSH) terms; telehealth, telemedicine, heart failure, hypertension, cardiovascular disease and systematic review. As a relatively new form of technology, reviews were retrieved if they were published between 2000 and 2013. Secondary searching of the reference lists of retrieved papers and of the Internet via the Google scholar search engine was undertaken to identify any additional reviews that met the inclusion criteria.

Selection procedures

Papers were retrieved based on whether the title and abstract or, if required, the full manuscript met the inclusion criteria for this review. Papers identified through the search were assessed based on the inclusion and exclusion criteria by two independent reviewers (RP and EH). Where there was discrepancy, the reviewers discussed the issues and reached consensus.

Inclusion criteria

Systematic reviews were included if they met the following inclusion criteria. Firstly, the paper needed to report a systematic review or meta-analysis of original intervention studies. Reviews that did not meet the definition of a systematic review or meta-analysis, or were expert commentaries were not included [8]. Secondly, included reviews needed to explore the impact of telemonitoring on the health outcomes of adults with known CVD as distinct from those with other chronic diseases. An a priori decision was made to not include studies with paediatric samples. Thirdly, given the focus on telemonitoring as distinct from home monitoring, reviews were excluded if there was no evidence that home based measures were transmitted to a healthcare provider. Finally, due to resource constraints and ease of access, reviews that were published in any language other than English, or reviews that were unpublished were excluded from the analysis.

Data extraction

Two authors (RP and EH) developed a standardised form for extracting data and relevant information from the 13 included reviews. Whilst developed specifically for this review, the form was based on tools used in other systematic reviews and systematic reviews of reviews [12, 13]. The form contained 10 categories regarding the characteristics and results of the included reviews. All data were initially extracted by one author (RP), this extraction was then checked by a second author (EH) for accuracy. Where differences were identified, the authors discussed issues and reached a consensus decision.

Methodological quality

A two-stage process was undertaken to evaluate both the type of evidence contained in each included review, and the quality of the review process used. In the first stage of this process, the level of evidence was graded using the National Health & Medical Research Council (NHMRC) hierarchy (Table 1). Whilst systematic reviews of randomised controlled trials are considered the highest level of evidence on this hierarchy, other reviews are ranked only as high as their included studies [14].

Table 1 NHMRC hierarchy of evidence[14]
Full size table

In the second stage of this process, the quality of included systematic reviews and meta-analyses were assessed using the component and total scores from the Overview Quality Assessment Questionnaire (OQAQ). The OQAQ has been previously validated as a measure of quality in research reviews [15, 16]. Two reviewers independently scored each paper using OQAQ (RP and SM) and any disparities were discussed with a third reviewer (EH) until consensus was reached.

Results

Ninety-nine papers were identified by the search strategy and assessed against the review inclusion criteria. All reviewers agreed that 13 papers met the inclusion criteria for the review. Papers were excluded if they did not distinguish individuals with CVD from those in other disease groups, the intervention did not involve information transmission between the consumer and health provider or did not report a formal systematic review (references to these papers can be provided on request). The flowchart in Figure 1 outlines the process for selecting included papers.

Figure 1
figure 1

Process of review selection.

Full size image

Of the 13 included reviews, nine were graded as Level 1 on the NHMRC Hierarchy [8, 10, 17–23]. The remaining reviews included a combination of RCTs, observational, descriptive, case series and/or non-randomised controlled studies, therefore, were considered to be Level II to IV evidence [24–27]. Seven reviews [8, 10, 17, 18, 23, 26, 28] scored five or above on OQAQ, indicating minor or no methodological flaws [15] (Table 2). Reviews that failed to score above a five predominantly did so due to the poor assessment of validity in primary studies and/or the authors failing to use a validated quality scoring system. The quality features of included systematic reviews are presented in Table 2.

Table 2 OQAQ Scores for the methodological quality of included reviews
Full size table

As can be seen in Tables 3 and 4, 9 reviews investigated telemonitoring for heart failure [8, 10, 21–27], three focused on telemonitoring interventions for hypertension [17, 19, 20] and a single review looked at risk factor reduction in CVD [18]. The number of studies synthesised in the included reviews varied between 9 [20, 21] and 56 [25] studies. Whilst the number of included participants was not accurately reported in two reviews [24, 26], in the remaining reviews the number of participants ranged from 2,401 to 9,946 [8, 10, 17–22, 25, 27, 28].

Table 3 Summary table of included reviews - hypertension
Full size table
Table 4 Summary table of included reviews – heart failure
Full size table

Definitions of telemonitoring varied, with some reviews of hypertension management only citing studies that used telemonitoring to transmit physiological data [20, 22] and other reviews included studies of telemonitoring combined with additional support, such as counselling, and/or education [8, 10, 17–19, 21, 24–28]. In some reviews it was possible to compare outcomes between various types of interventions [8, 21, 25], whilst in others various interventions were combined in the analysis [23]. This variation is a reflection of the current state of the literature and the significant variety of operational definitions of telemonitoring between studies included in the reviews.

A diverse range of outcomes was measured across the reviews. Those reviews which investigated hypertension primarily used blood pressure and medication use as outcomes (Table 3), whilst those investigating HF used a combination of outcomes including, mortality, hospital admissions, quality of life, cost, acceptability (Table 4). All reviews reported benefits associated with telemonitoring, however the significance of the identified benefits differed between studies and reviews. No review reported negative effects of telemonitoring or harm to patients. All reviews noted significant variance between included studies with four reviews citing statistically significant heterogeneity [8, 18, 19]. It is essential that this be considered in the interpretation of these data.

Synthesis of reviews of hypertension & CVD management

Table 3 provides a summary of the four included reviews that focused on telemonitoring in hypertension and CVD management.

Blood pressure

The four reviews which investigated hypertension and CVD management all used blood pressure as an outcome measure [17–20]. The operational definition of blood pressure differed across reviews with all reviews reporting systolic blood pressure [17–20], three reporting diastolic pressure [17, 19, 20], one review reporting mean arterial pressure [17] and another reporting blood pressure normalisation [19]. All reviews reported significant reductions in blood pressure with the various interventions (Table 3). Given the concomitant changes in extraneous variables such as medication use and lifestyle risk factors that occurred around the interventions, it is difficult to ascertain which specific aspects of the intervention led to the improvements in blood pressure seen within these reviews and included studies.

Medication use

Three reviews which investigated hypertension and CVD management reported medication use as an outcome measure [17, 19, 20]. There was some variation in finding of medication use between reviews. Agarwal et al. [17] reported that 10 studies demonstrated reduced medication use (RR 2.02, 95% CI 1.32 to 3.11) and 12 studies reported no greater increase in medication (RR 0.94, 95% CI 0.75 to 1.19). In contrast, Omboni & Guarda [19] reported five studies that demonstrated increased use of antihypertensive medications (WMD +0.22, 95% CI: +0.02, +0.43). It should be noted that this increase in antihypertensive use may have been a positive finding if patients had been sub-optimally medicated and were now receiving best practice pharmacotherapy.

In their review, Verberk et al. [20] compared the outcomes of those who had their antihypertensive treatment modified and those who did not have modified treatment during the study in both the intervention and treatment groups. This comparison demonstrated that treatment modification was associated with significantly lower systolic BP compared to non-modification (5.1 mmHg ± 2.9 mmHg lower).

Synthesis of reviews of heart failure management

Table 4 provides a summary of the nine included reviews that focused on telemonitoring in HF management.

Hospital admission

Eight of the reviews of HF management used hospitalisation as an outcome measure [8, 10, 21–26]. All these reviews reported that telemonitoring was associated with reduced hospitalisations [8, 10, 21–27]. The reduction in hospitalisation was reportedly as high as 50% in one review [21]. However, three reviews reported included studies that did not demonstrate a statistically significant reduction in hospitalisation rates [10, 21, 22]. Indeed in their review, Giamouzis et al. [22] identified four included studies in which the telemonitoring group had more rehospitalisation’s than the usual care group, although these findings were either non-significant or the significance was not reported.

In terms of HF related hospitalisations, Inglis et al. [8] reported that both telemonitoring and structured telephone support reduced the number of patients having a HF related admission to hospital (structured telephone support RR 0.77, 95% CI 0.68 to 0.87, P < 0.0001 and telemonitoring RR 0.79, 95% CI 0.67 to 0.94, P = 0.008). Similarly, Chaudhry et al. [21] found that both video conferencing and regular nurse phone calls reduced 6-month HF related admissions to hospital compared to usual care. However, there were no between group differences between video conferencing and regular nurse support.

Length of stay

Length of stay (LOS) was reported as an outcome in five included reviews [8, 23–26]. However, in three of these reviews, fewer than six included studies reported LOS as an outcome [8, 24, 25]. In these three reviews all but one study reported reduced length of hospital stay following the intervention [25].

The other two reviews included 12 studies each that reported LOS as an outcome, although had conflicting findings. The recent review published by Klersy [23] concluded that length of stay was not different between remote patient monitoring (combined telemonitoring and structured telephone support) and usual care for either HF (95% CI 20.12–0.13, P = 0.88) or all-cause hospitalisation (95% CI 20.18–0.02, P = 0.83). However, the older review by Martínez et al. [26] reported that telemonitoring reduced length of hospitalisation.

Mortality

Six reviews of HF management used all-cause mortality as an outcome measure [8, 10, 21, 22, 24, 26]. As can be seen from Table 5, a range of interventions improved mortality. In their comparative reviews, both Clark et al. [10] and Inglis et al. [8] demonstrated that all-cause mortality was more significantly reduced in studies using telemonitoring rather than structured telephone support.

Table 5 Summary of all-cause mortality
Full size table

Quality of life

Six reviews of telemonitoring in HF reported quality of life (QOL) as an outcome measure [8, 10, 23–26]. All of these reviews described some studies that had demonstrated improved quality of life (Table 1). There was limited information provided about the measures used to evaluate QOL. Only Klersy et al. [23] reported quality of life in terms of quality adjusted life years (QALYS). This review calculated that remote patient monitoring was associated with a gain of 0.06 QALYs – 0.02 due to reduced mortality and 0.04 due to reduced hospitalisations [23].

Cost

Two reviews of telemonitoring in HF focused on cost as an outcome measure [23, 27]. Both of these reviews concluded that telemonitoring reduced costs compared to usual care. This reduction in costs was attributed to the reduction in hospital admissions. Seto [27] also reported a 3.5% saving in direct patient costs as a result of the reduced need for patients to travel to services. Additionally, Seto [27] identified that over half of patients were willing to make a financial contribution to allow them to use telemedicine.

Discussion

This systematic review of systematic reviews found that the use of telemonitoring for patients with hypertension and HF was associated with multiple benefits. The reviews of telemonitoring for hypertension demonstrated that various telemonitoring interventions were able to effect significant reductions in blood pressure. The reviews of telemonitoring for HF demonstrated a reduced risk of mortality [8, 10, 21, 22, 24, 26], fewer hospitalisations [8, 10, 21, 22, 24–27], reduced health care costs [23, 27] and improved quality of life compared to usual care [8, 10, 24–26]. It has been reported in two syntheses of the effectiveness of telemonitoring across disease groups that the evidence supporting the efficacy of telemonitoring is most favourable for HF and hypertension compared to other forms of chronic disease [29–31].

A key limitation in the included reviews is the heterogeneity of the interventions reported in included studies and reviews [8, 18, 19]. Telemonitoring interventions are frequently multi-dimensional, containing a range of elements including the transmission of physiological data, coaching, telephone support, video-consultations, nurse interventions and web based communications [30]. The rapid technological advancements that have been seen in the last decade may also impact on the ability to compare older and newer studies using different technology [30]. Whilst some reviews reported comparisons between the various types of telemonitoring interventions, in others it was more difficult to differentiate the outcomes from the various interventions. Further research needs to tease out the specific aspects of telemonitoring interventions that are essential to improving health outcomes.

A further limitation was the lack of reporting of or use of unvalidated outcome measures. Whilst some reviews reported the range of outcomes measures used in included studies, others provided limited data on which to evaluate how the various outcomes were measured.

A final limitation is that, as a systematic review of reviews, this paper only included studies that had been reported within systematic reviews. This means that some more recent research is not included in the analysis. Whilst this recent literature includes several papers which affirm the value of telemonitoring [32–35], it also includes some large trials of telemonitoring in HF and hypertension which report negative results [36–39]. These divergent findings highlight the need for future research to more carefully interrogate how and why telemonitoring interventions work to improve outcomes.

Whilst the specific elements of the interventions that underpinned the positive outcomes are not clear, it is postulated that the benefits of telemonitoring reflect a combination of improved implementation of and adherence to guideline therapies, early identification of complications, and a positive impact on patient psychology [8, 10]. Wotton [31] asserts that telemedicine is effective as it facilitates integration and case-management within chronic disease, both strategies well known to improve guideline adherence and early intervention. Further investigation into how telemedicine can best be integrated into existing clinical primary care to enhance case-management will assist in developing new models of care. Future research exploring the impact of telemonitoring on adherence to treatment guidelines and early identification of exacerbation, will increase our understanding of how telemonitoring interventions improve outcomes.

The benefits of telemonitoring can also be attributed to patient empowerment and its flow on effects [40]. Providing patients with the tools and education to monitor their own symptoms, vital signs, and fluctuating medication needs, empowers the patient to take an active role in their own healthcare. Improved self-management has long been recognised as a strategy to enhance outcomes in chronic disease. Given that patients largely find telemonitoring acceptable and easy to use [8, 10, 24, 26, 41–43], further research focusing on the impact of telemonitoring on self-management and the use of telemonitoring as a patient education tool would inform new models of care [44].

Conclusions

In summary, this systematic review of systematic reviews found that telemonitoring has the potential to reduce the burden associated with hypertension and HF in primary care. However, further well-designed research is required to facilitate our understanding of how this intervention improves various outcomes and to allow the essential components of a telemonitoring intervention to be identified.