Impact of Telemedicine on the Clinical Outcomes among Elderly Patients: A Systematic Review

INTRODUCTION

Telehealth or telemedicine represents a method of providing health care services by using the currently available information communication technology (ICT) which allows patients to access health care services from a distance. Such technology allows patients to receive health care without needing to visit a doctor’s clinic, hospital, or any other health care facility unless the treatment of the patient’s condition necessitates the physical presence of a health care professional [1, 2]. The World Health Organisation (WHO) defines telemedicine as the provision of health care services through long-distance interaction between all health care professionals and the patient to discuss health matters pertaining to the patient’s diagnosis, treatment, and disease management or for research and evaluation by utilizing ICT to improve the health outcomes of individual patients and the community as a whole [3].

In 2017, the United Nations (UN) estimated the population of older people aged 60 or above, to be at 962 million which comprised 13% of the total world population [3]. The proportion of the older population is expected to increase from 1.4 billion in 2030 to 2.1 billion in 2050. The rise in the number of individuals in this age group raises many concerns particularly in the field of health care due to the prevalence of comorbidities and chronic conditions in older adults. The prevalence of multi-morbidity and polypharmacy rises with increasing age, particularly amongst those 60 years and older [4]. Thus, the treatment and healthcare of patients from the older age group have become increasingly challenging and demand a large proportion of resources from the healthcare system. Furthermore, the cost and expenditure for the treatment of chronic conditions which are more prevalent amongst older people will be expected to further increase along with the growth of this population [5].

Therefore, telemedicine is a promising avenue that could help reduce the cost of healthcare and the burden on the healthcare system from this population while improving their health benefits. The use of telemedicine has been shown to have promising benefits to patients using already available healthcare systems [6-16]. Thus, the introduction of telemedicine into the healthcare of older adults has the potential to serve as a new platform to educate older patients on health and disease management while being meticulously monitored by healthcare professionals. This would be able to improve self-care amongst older patients which would reduce the reliance of patients on healthcare professionals for minor health issues. The increased possibility of early detection of chronic diseases would promote earlier treatment and improve clinical outcomes for older patients.

Previous research [17-19] provides evidence that the implementation of telemedicine in health care stands to significantly reduce costs and unnecessary expenditures. The utilization of telemedicine is seen to improve patient compliance and lower unneeded use of resources, which in turn helps to cut down expenses on health care. However, the impact of telemedicine on clinical and self-reported outcomes of older patients has not been fully reviewed systematically.

Objective

The objective of this systematic literature review is to evaluate the impact of telemedicine and other forms of remote monitoring technology on the clinical and self-reported outcomes of older patients. In doing so, the effectiveness and feasibility of the use of telemedicine and remote monitoring technology can be assessed to determine the value of this technology in the field of healthcare.

MATERIALS AND METHODS

Search Strategy

A systematic search was conducted across six databases namely, PubMed, EMBASE (Ovid), EMCARE (Ovid), PsycINFO, and EBSCOhost by two independent reviewers using key phrases related to the research question as search terms. The terms ‘Geriatrics’, ‘Well-being’, and ‘Technology’ were used together with the Boolean phrase ‘AND’, along with alternative phrases that are synonymous with the key terms mentioned. Further description of the search strategy and the list of alternative phrases used.  The same phrases were used as search terms for every database. The terms were searched in all fields without having any restrictions in the year of publication, field of study, or any other search criteria.

Inclusion and Exclusion Criteria

The studies were included in the review if there were: 1) addressing the use of technology and the health outcome of older patients (60 years and older), 2) Measuring the effects of telemedicine or remote monitoring technology on any of the clinical or patient’s self-reported outcomes; and 3) randomized controlled trials and quasi-experimental studies. Non-English language papers were excluded. We used the taxonomy of classification of telemedicine as described in a previous study [20].

Study Selection

Articles were screened with the study titles followed by abstract and full-text screening.

Data Extraction Process

Data were extracted independently on the study characteristics, the study population, interventions, and outcomes using Google spreadsheets. Quantitative results from randomized controlled trials in the form of continuous and binary data were collected and sorted in separate sheets.

Quality Assessment

The quality of the studies that were selected to be included in the study was assessed and examined by one independent reviewer using the Cochrane Risk of Bias Tool [21]. Each criterion for each study was rated ‘low risk’ if there was a low chance of any bias or a low possibility that the existence of any bias would significantly affect the outcome of the study, ‘high risk’ if there was a high chance of any bias or a high possibility that the existence of any bias would significantly affect the outcome of the study, and ‘unclear risk’ if the existence of any bias in the study could not be determined due to particular reasons. The quality assessment was cross-checked by another independent reviewer by reviewing the assessment of three studies. There were no major issues or disputes brought up during the cross-check and the initial assessment of the quality of the included studies was retained.

Outcomes of Interest

The clinical outcomes included changes in blood pressure (measured in mmHg), low-density lipoprotein (LDL) blood levels (measured in mmol/L), body weight (measured in kg), and HbA1c levels (measured in %). The self-reported outcomes consisted of depression and health-related quality of life (HRQoL). Depression was assessed using the Geriatric Depression Scale (GDS) and Patient Health Questionnaire 9 (PHQ-9), whereas HRQoL was assessed by using the 36-Item Short Form Survey (SF-36). Only three components of the SF-36 were included in this systematic review, namely measures of bodily pain, general health, and social functioning.

Analysis

Results obtained involving clinical outcomes which included changes in blood pressure (systolic and diastolic blood pressure), LDL blood levels, HbA1c blood levels, and body weight along with self-reported patient outcomes which included depression and quality of life ratings were assessed through meta-analysis. The Review Manager 5.3 program was used to conduct the meta-analyses for each of the individual outcomes of interest by using a random effects model.

RESULTS AND DISCUSSION

A total of 147,333 unique titles were screened, followed by abstract screening and full-text screening (n=39). A total of seven studies were included (Figure 1).

Description of Included Studies

All the included studies (n=7) were published between year 2009 to 2017 [22-28]. There was one quasi-experimental study among the included articles Choi et al. (2014) [22], the other six studies were randomized controlled trials [23-28]. Four of the included studies were conducted in the United States [24, 25, 27, 28] while the other three studies were each conducted in Korea [22, 23, 26].

Clinical outcomes that included changes in blood pressure (systolic and diastolic blood pressure), LDL blood levels, HbA1c levels, and body weight were evaluated in five studies [22, 23, 26-28] whereas, self-reported patient outcomes including depression and quality of life were evaluated in five studies [22-26].

The comorbid conditions of the study population were specified in three of the included studies whereby one study evaluated chronically ill and depressed older patients primarily [25], the second study investigated heart disease patients and COPD patients [24], and the third study examined heart disease patients and diabetic patients [26]. However, patients in all three of the studies were reported to have multiple comorbidities besides the primary chronic illnesses being studied [24-26]. A detailed description of all the included studies in this review has been provided in Table 1.

Description of Interventions

The interventions of the included studies were provided by a range of primary healthcare professionals, such as nurses, general practitioners, psychologists, and neurologists. All seven of the included studies had teleconsultation, videoconferencing, or teleconferencing as part of the intervention for the telemedicine group of patients. A remote monitoring system to observe the symptoms of certain diseases or regular physiological measurements as part of the intervention was included in all included studies. Education about the patients’ chronic illness or general health was only provided as part of the intervention two studies [24, 25]. One study [27] included goal setting and action planning by a health coach or primary healthcare provider as part of the intervention for the telemedicine group of patients. The details of interventions given to the telemedicine group and usual care group are provided in Table 1.

Quality of Included Studies

Overall, the results and outcomes from the included studies possess high reliability as more than 50% of the studies reported a low risk of bias in each criterion assessed.  Three of the included studies were likely to have a high risk of bias in terms of random sequence generation [23, 28], while one study had unclear random sequence generation [22]. The random sequence generation represented the major source of any possible bias among all the components that were assessed, with only four studies (57.1%) reporting a low risk of bias [24-27], two studies reporting a high risk of bias [23, 28] and one study having an unclear risk of bias [22]. All seven studies included in this systematic review reported a low risk of bias for blinding of outcome assessment, selective reporting, and any other source of bias [22-28]. The general summary of the quality assessment and risk of bias of each study are presented in Figure 2, respectively.

Clinical Outcomes

Blood Pressure

Blood pressure was measured as systolic blood pressure and diastolic blood pressure in four studies [22, 23, 26, 28]. The mean change of systolic blood pressure from baseline to follow-up for the control groups and intervention groups in each study ranged from -4.12 mmHg to +6.9 mmHg and -18.82 mmHg to -5.43 mmHg respectively (Figure 3). The results of the meta-analysis showed a statistically significant reduction of systolic blood pressure -3.83 (95% CI -6.04, -1.61) as compared to usual care (Table 2). Whereas no significant reduction was observed for diastolic blood pressure (-0.92; 95%CI -2.14, 0.30) (Figure 3).

MD = mean difference; CI = confidence interval; GDS = Geriatric Depression Scale; PHQ = Patient Health Questionnaire 9; HRQoL = Health Related Quality of Life.

Blood LDL Levels

Two studies evaluated blood LDL levels as a clinical outcome [26, 28]. The control groups and intervention groups from both studies had a mean change in blood LDL levels ranging from -0.36 mmol/L to -0.39 mmol/L and -0.34 mmol/L to -0.4 mmol/L respectively (Table 2). A statistically insignificant reduction in LDL was observed when compared to usual care (-0.01; 95% CI -0.09, 0.07) (Figure 3).

HbA1c

Two studies included HbA1c blood levels as an outcome of interest [26, 28]. There was a greater improvement in the mean changes of HbA1c in the intervention groups of both studies in comparison to the control groups. The mean changes from baseline to follow-up for the control groups in both studies were -0.06% and +0.18% while -0.31% and +0.04% were the mean changes for the intervention groups. Overall statistically significant reduction in HbA1c level was observed for telemedicine intervention (-0.2; 95% CI -0.35, -0.04) (Figure 3).

Body Weight

The body weight of patients was measured as a clinical outcome in two studies [26, 27]. Telemedicine interventions had an insignificant effect on the patient’s body weight (0.15; 95% CI: -0.74, 1.05) (Figure 3).

Self-reported Patient Outcomes

Depression

A total of four studies evaluated the depressive symptoms and depression status of older patients in relation to telemedicine usage in the delivery of healthcare [22-25]. Several tools were used to measure the depression ratings of patients across all four studies such as the Centre for Epidemiologic Studies Depression (CES-D), Geriatric Depression Scale (GDS), Hamilton Depression Scale (HAM-D), and the Patient Health Questionnaire-9 (PHQ-9), whereby some studies used more than one tool [22-25]. Since two different measurement tools were used by two studies each, a meta-analysis was conducted for both measurement tools in assessing the depressive symptoms of patients. The meta-analysis produced statistically significant results for the PHQ-9 measurement tool [24, 25] but not for the GDS measurement tool [22, 23] (Figure 4).