Telemedicine for diabetes and heart failure: an evidence review

THIS fourth article in a series on the digital health revolution discusses the evidence from systematic reviews of randomised controlled trials relating to remote monitoring and telemedicine interventions for people with diabetes and heart failure.

The majority of randomised controlled trials in this area focus primarily on the health outcomes arising from telemedicine interventions; however, there are also important questions beyond the scope of this article relating to whether telemedicine secures equity of access to health care and reduces costs.

Definitions

The World Health Organization has defined telemedicine broadly as:

“The delivery of health care services, where distance is a critical factor, by all health care professionals using information and communication technologies for the exchange of valid information for diagnosis, treatment and prevention of disease and injuries, research and evaluation, and for the continuing education of health care providers, all in the interests of advancing the health of individuals and their communities”.

In light of this broad definition, the World Health Organization has tended to use the terms “telemedicine” and “telehealth” interchangeably and synonymously. However, many studies distinguish between these terms on the basis of the level of intervention by the doctor or other health provider.

This article uses the terms “remote monitoring”, “telemedicine” and “telehealth” consistently with how they have appeared in the systematic reviews discussed. At other times, the term “telemedicine” is used to capture a broad range of interventions.

Cochrane review of telemedicine

A 2015 Cochrane review by Flodgren and colleagues assessed the effectiveness, acceptability and costs of telemedicine interventions in patients with any clinical condition. The review included 93 randomised controlled trials of telemedicine interventions used in direct patient care (ie, when the patient is remote from the clinician and communication is interactive) versus usual care (ie, face to face care or telephone consultation). Twenty-one trials (23%) included patients with diabetes and 25 trials (27%) were in heart failure.

In patients with diabetes, high certainty evidence indicated that telemedicine interventions such as remote monitoring and/or education for self‐management using videoconferencing improve the control of blood glucose versus usual care at median 9 months follow-up (glycated haemoglobin [HbA_1c], mean difference [MD] -0.31; 95% CI, -0.37 to -0.24; 16 trials; 2768 participants, I² = 42%; high certainty evidence). Moderate certainty evidence indicated that telemedicine interventions probably reduce low-density lipoprotein cholesterol (MD, -12.45 mg/dL; 95% CI, -14.23 to -10.68; 4 trials; 1692 participants, I² = 0%; moderate certainty evidence) and probably reduces blood pressure (MD, -4.33 mmHg systolic blood pressure; 95% CI, -5.30 to -3.35; 4 trials; 1770 participants; I² = 17%; moderate certainty evidence) at median 9 months follow-up compared with usual care.

In patients with heart failure, telemedicine interventions reviewed led to little or no difference in all-cause mortality at median 6 months follow-up versus usual care (risk ratio, 0.89; 95% CI, 0.76–1.03; 16 trials; 5239 participants; I² = 44%; high certainty evidence). Admissions to hospital ranged from a decrease of 64% to an increase of 60% at median 8 months follow-up (11 trials; 4529 participants; I² = 67%; moderate certainty evidence); however, telemedicine probably slightly improves quality of life at median 3 months follow-up compared with usual care (MD, -4.39, 95% CI, -7.94 to -0.83; 5 trials; 482 participants, I² = 0%; moderate certainty evidence).

The quality of care, cost and acceptability of telemedicine to patients and clinicians was unclear due to limited data reporting these outcomes. An update of the review will be available later in 2019.

Systematic reviews of telehealth interventions in diabetes

At least one overview (Lee 2018) and 15 systematic reviews of telehealth interventions in diabetes have been published in the past 5 years (So 2018, Rush 2018, Faruque 2017, Heitkemper 2017, Lee 2017, Su 2016, Ming 2016, Su 2016, Isaacs 2016, Rasekaba 2015, Mushcab 2015, Huang 2015, Greenwood 2014, Zhai 2014, Cotter 2014, Suksomboon 2014).

The 2018 overview by Lee and colleagues contains four systematic reviews (including 29 unique randomised controlled trials) of remote patient monitoring applications versus usual care in adults with type 2 diabetes. They focused on telehealth interventions in which patients transmitted (electronically or verbally) self-monitored blood glucose results to a health care professional or specialist team at an offsite monitoring centre for evaluation and feedback. Two systematic reviews in the overview were rated as moderate quality using the Assessment of Multiple Systematic Reviews (AMSTAR) tool, and two reviews were rated as low quality. All four reviews found telehealth interventions improved glycaemic control in people with type 2 diabetes compared with usual care. Additional meta-analyses demonstrated that telehealth applications slightly improve HbA_1c compared with usual care (MD, -0.55; 95% CI, -0.73 to -0.36; 25 trials; 3370 participants; I² 82%) although substantial heterogeneity across trials was found. The greatest effect was observed in telephone-delivered interventions, followed by internet blood glucose monitoring system interventions, and automatic data transmission using a mobile phone or telehealth unit. The heterogeneity in effects could not be explained due to minimal description about interventions and participant characteristics in trials in the systematic reviews (and primary trials were not retrieved by authors of the overview).

Overview and systematic reviews of telehealth interventions in heart failure

At least one overview (Bashi 2017) and 11 systematic reviews of telehealth interventions in heart failure have been published in the past 5 years (Lin 2016, Conway 2014, Grustam 2014, Gorst 2014, Kotb 2015, Lin 2017, Cajita 2016, Nakamura 2014, Brons 2018, Yun 2018, Jones 2016).

The 2017 overview of systematic reviews of remote monitoring interventions in heart failure by Bashi and colleagues identified 19 reviews, 14 of which assessed the effects of telephone-based monitoring, one on video monitoring, two on mobile phone monitoring, one on personal digital assistant devices, and four assessed home telehealth interventions. Of these, one review was high quality, 12 moderate quality and six were low quality, as assessed by the authors of the overview using AMSTAR. Based on these reviews, the authors reported that telemonitoring and home telehealth appear effective in reducing heart failure rehospitalisation and mortality, but other interventions (eg, mobile phone-based monitoring and videoconferencing) require further investigation. Limitations of the overview were that outcome data were retrieved from the systematic reviews (not primary trials), and individual trials may have been included in more than one systematic review and therefore been accounted for more than once, and systematic reviews not reported in English were not included in the overview.

Implications for research and health care delivery in Australia

The volume of randomised trials and systematic reviews being conducted in this area is extensive. Evidence from systematic reviews and overviews indicate that telemedicine can improve blood glucose control in people with diabetes and provide similar health outcomes in the management of heart failure as to face to face or telephone delivery of care.

However, telemedicine is an emerging field with new technologies and strategies evolving rapidly, such that, by the time a trial is completed, the intervention may be out of date. The evidence base for the cost-effectiveness and acceptability of telemedicine interventions is also limited. This suggests the need for other trial designs, including pragmatic or adaptive randomised designs and observational studies.

For rural and remote patients, an acceptable outcome might be that telemedicine has a “similar” outcome (eg, using a non-inferiority design) to conventional consultation. Even a slightly inferior outcome might be acceptable as, in many cases, this might be better than the outcomes secured with no interaction whatsoever. Indeed, there are various telemedicine programs already being run in Australia for rural and remote patients, including programs for cardiology patients at Monash Heart and for the management of diabetes at the Baker Heart and Diabetes Institute.

Those clinicians looking to implement a telemedicine component into their practice may benefit from the Cochrane telemedicine review findings, which will be updated later this year, and the Royal Australian College of General Practitioners telehealth guidelines.

Dr Denise O’Connor is a National Health and Medical Research Council (NHMRC) Translating Research into Practice Fellow and Senior Research Fellow in the Monash Department of Clinical Epidemiology, Cabrini Institute, at Monash University, and editor with Cochrane Effective Practice and Organisation of Care Group. Her research interests include research translation and models of health care service delivery.

Professor Le0nard C. Gray is the Director of the Centre for Health Services Research and Professor in Geriatric Medicine at the University of Queensland. He leads the NHMRC Centre for Research Excellence in Telehealth. His research interests include models of care, assessment and clinical decision support systems, aged care systems, telehealth and health informatics. 

Bianca Phillips is a Victorian academic lawyer conducting medical law research. She completed her Master of Laws at the University of Melbourne with her thesis on telemedicine, and is currently completing a doctoral thesis on the law making of the digital health revolution. She has authored articles on the digitisation of medicine in both legal and medical publications. She can be found on Twitter @biancarphillips.

The statements or opinions expressed in this article reflect the views of the authors and do not represent the official policy of the AMA, the MJA or InSight+ unless that is so stated.