The use of wearable devices in chronic disease management to enhance adherence and improve telehealth outcomes: A systematic review and meta-analysis

Abstract

Introduction

Wearable device (WD) interventions are rapidly growing in chronic disease management; nevertheless, the effectiveness of these technologies to monitor telehealth outcomes has not been adequately discussed. This study aims to evaluate the effects of WDs in adherence and other health outcomes for people with chronic obstructive pulmonary disease (COPD), diabetes mellitus (DM), and cardiac disease (CD).

Methods

CINAHL, PsycINFO, CENTRAL, and EMBASE were searched for randomized controlled trials (RCTs) and non-RCTs from 1937 to February 2020. Studies comparing interventions with the use of WD were assessed for quality in RCTs and a meta-analysis was performed.

Results

Eleven studies were included in this review. All of the interventions involved WD use with educational support such as goal setting, virtual social support, e-health program, real-time feedback, written information, maintain diary, and text messaging. The meta-analysis showed no difference in adherence (p = .38). The DM group showed effects of more than a 2% reduction in weight when WDs were implemented for three months (risk ratio = 2.20; 95% confidence interval (CI) 1.38 to 3.50; p = .0009), as well as blood glucose (mean difference (MD) = –32.39; 95% CI = –48.07 to –16.72; p < .0001), haemoglobin A_1c (MD = –0.69; 95% CI = –1.28 to –0.10; p = .02), and physical exercise time in the CD group (MD = 9.53; 95% CI = 0.59 to 18.47; p = .04).

Discussion

WD with educational support may be particularly useful for people with DM and CD to enhance support beyond usual care. The results of this review showed insufficient evidence to support the use of WD for COPD to enhance telehealth outcomes for disease management.

Keywords

Wearable device chronic obstructive pulmonary disease diabetes cardiac disease systematic review meta-analysis

Introduction

The wearable medical device market is rapidly growing, with 27.9% annual growth predicted from 2020 to 2027, and home healthcare and remote monitoring is expected to influence demand for these devices.¹ A wearable device (WD) is defined as a sensory device that can be attached to clothing or worn as an accessory, allowing tracking of health information without hindrance.^2–5 WDs were essentially developed for fitness and health promotion use and to promote wellness, and they have the ability to accurately track health-related activities.⁶ These devices have the potential to capture information with real-time synchronization and monitoring, allowing healthcare providers to monitor and assess health conditions, communicate effectively, manage risk factors, and gain feedback – all of which help enable self-management.^7,8

Recent studies have reported that a WD could be used for continuous monitoring and early diagnosis for patients who have chronic diseases, including cardiac disease (CD),^9–11 diabetes mellitus (DM),¹¹ chronic obstructive pulmonary disease (COPD),¹¹ Parkinson’s disease,¹² cancer,¹³ and mental disorders.⁹ WDs could also be useful in the areas of gerontology¹⁴ and mobility support.¹⁵ To help monitor chronic conditions, they can also track the status of multiple health-related parameters such as step counting, energy expenditure, activity-time duration, sleep status, body temperature, oxygen saturation, and electrocardiogram.^9–16

A non-invasive WD offers an efficient and cost-effective solution that allows patients to monitor health-related parameters without expensive costs.¹⁷ Although previous studies have explored how this solution could benefit home-monitoring-based telehealth/telenursing, the effectiveness of using these technologies as well as users’ adherence and type of education for continuous disease management and telehealth outcomes have not been adequately discussed.

Telemonitoring health status with technology in chronic disease management

Chronic diseases are conditions characterized by a long duration and slow progression.¹⁸ They are currently increasing and represent 71% of all causes of death worldwide, killing 41 million people each year.¹⁹ Concerning older adults, age-related frailty and illness are commonly affected by environmental and lifestyle factors.²⁰ Therefore, the ability to monitor people’s daily status non-intrusively to find physical changes has merit.

Telemonitoring is an important aspect of telehealth that includes remotely collecting and sending people’s data for interpretation, and it can be used in ongoing healthcare.²¹ It involves self-monitoring and/or healthcare providers monitoring from a distance to improve health behaviour.²² Daily telemonitoring using non-WDs for people with chronic diseases such as COPD, DM, and CD has a positive effect on health outcomes^23–25 and in detecting early-stage changes.²⁶

In recent years, several types of wearable technologies have been introduced to support people with chronic conditions in home care settings, and several reports suggest that they have the potential to assist in predicting clinical outcomes.^{12,13,27–31} One study³² reported that data can be transmitted to healthcare providers to interpret and provide feedback in the management of chronic diseases. It was suggested that WDs are acceptable and feasible for people with chronic conditions. However, various studies have reported only the effectiveness of using WDs without intervention from healthcare providers. Further, the current evidence relating to the effectiveness of using WDs for chronic disease management, patient outcomes, and adherence is unclear. Moreover, while educational support plays an important role in chronic disease management,³³ the use of WDs and educational support for chronic disease management in telehome-care have not been discussed. Therefore, the current study focused on exploring WD-use interventions to monitor chronic conditions and evidence. This systematic review and meta-analysis evaluated the adherence to and effectiveness of the use of WDs and educational interventions to improve telehealth outcomes for people with chronic diseases like COPD, DM, and CD.

Methods

In this systematic review, we followed the steps in the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement.³⁴ The protocol was registered in the International Prospective Register for Systematic Reviews (PROSPERO) CRD 42018111632.

Published work search strategies and sources

We performed a comprehensive literature search of published work. Search dates for databases were not limited. The databases used in this study were the Cumulative Index of Nursing and Allied Health Literature Plus with Full Text database (including PsycInfo and MEDLINE) through the EBSCOHOST interface, PubMed, EMBASE, and the Cochrane Central Register of Controlled Trials. All databases, key words, and combinations used in the search are shown in Table 1. The search for published work was performed twice (in October 2018 and February 2020) to facilitate the inclusion of updated published work. Published conference proceedings that included abstracts were also reviewed. Gray literature such as unpublished articles or those that were not easily accessible via databases were not included. A manual search was performed to identify article references of research papers published in engineering journals.

Table 1.

Searched database and keyword combinations.

1. CINAHL Plus with Full Text, 1937 to 4 February 2020

S1 “wearable devices” OR “wearable technology” OR “apple watch*” OR fitbit OR garmin OR “smart watch*” OR “waist belt*” OR “fitness tracker*” OR MH Fitness Trackers

S2 TI randomized OR AB randomized OR MH treatment outcomes OR PT clinical trial

S3 (copd OR MH Pulmonary Disease, Chronic Obstructive OR chronic AND obstructive AND lung AND disease) OR (diabetes OR MH Diabetes Mellitus+ OR MH Diabetes Insipidus+) OR (MH Heart Failure+ OR “heart failure”)

S4 S1 AND S2 AND S3

2. PubMed, 1948 to 4 February 2020

#1 “wearable electronic devices”[MeSH Terms] OR (wearable[All Fields] AND electronic[All Fields] AND devices[All Fields]) OR (wearable[All Fields] AND technology[All Fields]) OR “apple watch”[All Fields] OR fitbit[All Fields] OR garmin[All Fields] OR “smart watch”[All Fields] OR “waist belt”[All Fields] OR “fitness trackers”[All Fields]

#2 randomized controlled trial[Publication Type] OR (randomized[Title/Abstract] AND controlled[Title/Abstract] AND trial[Title/Abstract])

#3 (“pulmonary disease, chronic obstructive”[MeSH Terms] OR (pulmonary[All Fields] AND disease[All Fields] AND chronic[All Fields] AND obstructive[All Fields])) OR (copd[All Fields]) OR (“diabetes mellitus”[MeSH Terms] OR “diabetes insipidus”[MeSH Terms] OR diabetes[All Fields]) OR (“heart diseases”[MeSH Terms] OR (heart[All Fields] AND diseases[All Fields]) OR (heart[All Fields] AND failure[All Fields]))

#4 #1 AND #2 AND #3

3. EMBASE, until 4 February 2020

#1 “wearable technology” OR “apple watch*” OR fitbit OR garmin OR “smart watch”/exp OR “smart watch*” OR “waist belt*” OR “wearable device*”

#2 randomized AND controlled AND trial

#3 (“chronic obstructive lung disease”/exp OR (chronic AND obstructive AND lung AND disease) OR copd) OR (diabetes OR “diabetes insipidus”/exp OR “diabetes mellitus”/exp) OR ((heart AND failure) OR “congestive cardiac failure” OR “congestive heart failure”/exp)

#4 #1 AND #2 AND #3 AND [embase]/lim

4. CENTRAL, until February 2020

#1 ((wearable electronic devices) OR (wearable technology) OR “apple watch” OR fitbit OR garmin OR “smart watch” OR “waist belt” OR “fitness trackers”):ti,ab,kw

#2 MeSH descriptor: [Wearable Electronic Devices] explode all trees

#3 #1 OR #2

#4 ((pulmonary disease chronic obstructive) OR copd):ti,ab,kw

#5 MeSH descriptor: [Pulmonary Disease, Chronic Obstructive] explode all trees

#6 MeSH descriptor: [Diabetes Mellitus] explode all trees

#7 MeSH descriptor: [Heart Diseases] explode all trees

#8 #4 OR #5 OR #6 OR #7

#9 #3 AND #8

The study selection and data extraction process

The suitability of each study was assessed in three steps. First, two authors independently reviewed the abstracts and screened and applied the selection criteria to identify suitable studies. Second, two authors independently reviewed the full text of the articles. Third, each article was analysed using the Cochrane public health group data extraction and assessment template. Suitability of selected papers for the review and meta-analysis was discussed by the whole team and disagreements were resolved by consensus.

Eligibility criteria included i) original and peer-reviewed studies written in English; ii) studies involving participants with COPD, DM, or CD who were living at home; iii) studies observing participants utilizing WDs; iv) studies related to any type of WD (the type of WD was not considered); v) studies that implemented WDs for more than two weeks with health interventions; and vi) studies that utilized a randomized controlled trial (RCT) and controlled clinical trial design. WDs in this study were defined as sensory devices that can be worn or attached to clothing and that can remotely track and monitor health information synchronously.

Exclusion criteria included i) qualitative, observational, and quasi-experimental studies; ii) pre-and post-studies without comparators; iii) prospective cohort studies; iv) studies including other chronic diseases; and v) studies involving daily monitoring without WDs or interventions of another telecommunication technology (videophones for face-to-face communication with participants at their residence or telephone support only).

Quality assessment

The risk of bias in the included studies was assessed using the Cochrane Handbook for Systematic Reviews of Intervention in six specific domains: sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and other bias.³⁵ The publication bias was assessed using funnel plots.³⁵

We included articles in the meta-analysis if they reported outcome measures for the following parameters: primary outcome – number of participants who maintained the WD for the intervention period (adherence); and secondary outcomes – number of admitted participants, weight control (more than 2% weight lost and body mass index (BMI) reduction), physical activity (walking steps/day and exercise duration), and biomarkers (fasting blood glucose (FBG) and haemoglobin A_1c (HbA_1c)) of participants with DM.

Statistical analysis

The meta-analysis was performed using Review Manager 5.3 software (The Nordic Cochrane Center, Copenhagen). Data were synthesized into forest plots, and a random-effect model was utilized to analyse intervention effects across all studies. The risk ratio (RR) was determined for outcome measures of dichotomous variables, and the mean difference (MD) or standard MD was calculated for continuous data. To confirm the reliability of the summary estimate, 95% confidence intervals (CIs) were calculated. The variables in each study were distilled from the data to facilitate the intention-to-treat analysis of the original group. We recorded missing data and drop-out rates for each RCT and reported the number of participants included in the meta-analysis as the reported number in the study. A sub-group analysis was performed for groups of participants diagnosed with DM, CD, or COPD. Statistical heterogeneity was analysed using the I² statistic: I² ≥ 50% indicates significant heterogeneity. Thus, a random-effect model was used – otherwise (i.e. I² ≤ 49%), a fixed-effect model was used.

Results

Flow of the study selection

We found 339 studies in the database search and two studies in the manual search. Of these, the full text of 192 studies were screened and 11 RCTs met the inclusion criteria for this review (Figure 1).^36–46 One RCT targeted participants with COPD,⁴⁴ seven with DM,^{36–38,41,43,45,46} and three with CD,^39,40,42 with a total of 655 participants (Table 2).

Figure 1.

PRISMA flow diagram for article selection.

Table 2.

Characteristics of the reviewed articles.

Author, year of publication, country of origin	Research design, number of arms	Duration of implementation (months)	Target disease	Wearable device, tracked information	Study purpose	Participant characteristics	Interventions
Bender et al., 2018³⁶USA	RCT, 2 arms	6 (3 months intervention + 3 months follow-up)	Type II DM	Fitbit Zip, wrist band, real-time steps	To assess the feasibility and potential efficacy for Fit&Trim intervention of a diabetes prevention program-based culturally adapted mobile phone-based weight loss lifestyle intervention including virtual social support (Facebook) for weight loss in Type II DM among Filipino Americans with overweight/obesity	n = 67, Filipino American adults who were overweight or obese (BMI for Asians > 23 kg/m²)Intervention group:n = 33; mean age 41.3 (SD 12.1) years; M/F 51.5/48.5%Control group:n = 34; mean age 42.1 (SD 12.2) years; M/F 44.1/55.9%	Intervention group:Phase 1 (baseline to 3 months), set individual tailored goals for weight, diet, and physical activity step and returned for three monthly office visits and completed the intervention at the 3-month visit. Received weekly posting of discussion topics on the study’s Facebook group site. Phase 2 (3 to 6 months), transitioned to a 3-month maintenance period with in-person weight checks and personal coaching, completing the study at the 6-month visitControl group (active waitlist control group):Phase 1 (baseline to 3 months), only asked to wear the Fitbit Zip on their own with no education or coaching and received education on Hepatitis A and B at the office visit. Phase 2 (3 to 6 months), received the Fit&Trim intervention
Bender et al., 2017³⁷USA	RCT, 2 arms	6 (3 months intervention + 3 months follow-up)	Type II DM	Fitbit Zip, wrist band, real-time steps	To assess feasibility and potential efficacy of a pilot, RCT of a culturally adapted mHealth weight loss lifestyle intervention (Pilipino Americans Go4Health) for overweight Filipino Americans with Type II DM	n = 45, Filipino adults; ≥18 years; based on the diabetes prevention program criteria and American Heart Association metabolic syndrome risks, overweight or obese (BMI for Asians > 23 kg/m²); physician diagnosis of Type II DM; own a smartphone, tablet, or laptop with Internet access; and English language proficientIntervention group:n = 22; mean age 57.4 (SD 9.8) years; M/F 37/63%Control group (waitlist):n = 23; mean age 57.7 (SD 10.0) years; M/F 40/60%	Intervention group:In phase 1 (baseline to 3 months), trained on using the Fitbit accelerometer to self-monitor real-time physical activity steps and associated app with diary to self-report daily food/calorie intake and weekly weight, joined the study’s private Facebook group for virtual social support, coaching, and weekly education topics posted by research staff. In phase 2 (at the 3-month office visit), transitioned to a 3-month follow-up and were removed from the private Facebook group. Follow-up office visits at 4 and 6 monthsControl group:In phase 1, received only the Fitbit accelerometer and training for daily wear and returned for 1- and 3-month office visits when they received hepatitis B and C education, respectively. In phase 2 at the 3-month office visit, transitioned to receive the Pilipino American Go4Health intervention and returned for 3 office visits at 4, 5, and 6 months
Bentley et al., 2016³⁸UK	RCT, 3 arms	3 months	Type II DM	AiperMotion 500, worn at the hip, physical activity, self-reported nutritional intake, energy burned and consumed	To initiate the process of evaluating a specific mHealth intervention, in order to develop more robust evidence on whether this intervention could be used to help people with Type II DM lose weight and control their HbA_1c levels	n = 27, overweight and obese members of the public in South Yorkshire; mean age 52.9 (SD 8.6) years; M/F 44.4/55.6%Intervention group 1:n = 9; mean age 50.8 (SD 8.1) years; M/F 77.8/22.2%Intervention group 2:n = 9; mean age 52.1 (SD 7.8) years; M/F 55.6/44.4%Control group:n = 9; mean age 55.7 (SD 9.9) years; M/F –/100%	Intervention group:Group 1: 90-min group training, appropriate behaviours to lose weight and control their HbA_1c, maximizing physical activity by using wearable mHealth device. The device records physical activity level and information about food consumption, and provides motivational feedback based on energy balance. They could individually choose their own weight-loss goal to increase motivation, and send emails to the research team weekly. Emailed replies delivered tailored positive encouragementGroup 2: same intervention except send emails to the research team weekly and emailed replies delivered tailored positive encouragement to themControl group:90-min group training, appropriate behaviours to lose weight and control their HbA_1c
Deka et al., 2019³⁹USA	RCT, 2 arms	1.5 (56 days)	NYHA class, I–III HF	Fitbit Charge HR, wrist band, real-time steps, heart rate, activity minutes	To investigate the effects of group social support by Internet-based synchronized face-to-face video and objective physical activity feedback on adherence to recommended exercise guidelines	N = 30, two cardiology practices in the Midwest, stable in the past 30 days and receiving standard pharmacological treatment; mean age 64.7 (SD 11.5) years, M/F 63.3/36.7%Intervention group:n = 15; mean age 61.7 (SD 11.3) years; M/F 66.7/33.3%Control group:n = 15; mean age 67.8 (SD 11.4) years; M/F 60/40%	Intervention group:Provide a handout on self-care in heart failure, an exercise routine, and a Fitbit Charge HR, and asked to wear it. They record their exercise sessions using both the Fitbit Charge HR and exercise diaries and they connected to Vidyo software, once a week, for 8 weeks, for a 45-minute face-to-face online group discussion/education sessionControl group:Provided a handout on self-care in heart failure, an exercise routine, and a Fitbit Charge HR and asked to wear it daily, and record their exercise sessions using both the Fitbit Charge HR and exercise diaries
Duscha et al., 2018⁴⁰USA	RCT, 2 arms	3 months	Cardiovascular disease, completing 36 on-site cardiac rehabilitation sessions	Fitbit Charge, wrist band, real-time steps, heart rate, activity minutes	To evaluate if a mHealth program could sustain the fitness and physical activity levels gained during cardiac rehabilitation in the pilot study	n = 25; age>18 years old; completing 36 on-site cardiac rehabilitation sessions, clinically stable, and ownership of a smartphoneIntervention group:n = 16; mean age 59.9 (SD 8.1) years; M/F 81.2/18.8%Control group:n = 9; mean age 66.5 (SD 7.2) years; M/F 66.7/33.3%	Intervention groupA 12-week mHealth program was implemented using physical activity trackers and an exercise prescription by daily step count, health coaching, and text messaging following cardiac rehabilitationControl group:Receive standard care as ordered by their treating physician and did not receive any specific lifestyle recommendations from study personnel
Grewal et al., 2015⁴¹USA–Qatar	RCT, 2 arms	1	Diabetic peripheral neuropathy	LegSys™, body-worn sensor belt, balance and balance training	To investigate the effect of sensor-based interactive balance training on postural stability and daily physical activity in older adults with diabetic peripheral neuropathy	n = 39 older adults with diabetic peripheral neuropathy from Arizona and Doha, mean age 63.7 (SD 8.2) years; BMI mean 30.6 (SD 6); M/F 46/54%Intervention group:n = 19; mean age 62.58 (SD 7.98) years; BMI mean 31.78 (SD 7.53); M/F 42.1/57.9%Control group:n = 20; mean age 64.90 (SD 8.5) years; BMI mean 29.58 (SD 4.24); M/F 50/50%	Intervention group:Received sensor-based interactive exercise training tailored for people with diabetes, twice a week for 4 weeks. The exercises focused on shifting weight and crossing virtual obstacles. Body-worn sensors were implemented to acquire kinematic data and provide real-time joint visual feedback during the trainingControl group:Not reported
Kidholm et al., 2016⁴²Denmark	RCT, 2 arms	3	Artery sclerosis, coronary artery bypass surgery, valve surgery, and heart failure	Fitbit Ultra, wrist band, number of steps	To develop and test an individualized cardiac telerehabilitation program designed to increase participation in rehabilitation, improve patient quality of life, reduce the number of admissions and the need for acute care	n = 141 cardiac patients diagnosed with artery sclerosis, coronary artery bypass surgery, valve surgery, and heart failureIntervention group:n = 72; mean age 62.46 (SD 12.31; range 31 − 88 years; IQR 54–72) years; M/F 77.8/22.2%Control group:n = 69; mean age 62.67 (SD 11.72; range 39 − 85 years; IQR 54 − 72) years; M/F 79.7/20.3%	Intervention group:Received training in the use of the different devices and the digital rehabilitation plan, and a doctor prescribed how often they needed to measure blood pressure, pulse, and weight, most often twice a week, whereas steps were measured every day. After 2 weeks in the program, each patient was visited by a project assistant.Control group:Usual care; followed a traditional rehabilitation program at the hospital or healthcare centre
Kooiman et al., 2018⁴³The Netherlands	RCT, 2 arms	2.8 (84 days)	Type II DM	Fitbit Zip, wrist band, number of steps	To determine the efficacy of an online self-tracking program on physical activity, glycated haemoglobin, and other health measures in patients with type II diabetes	n = 72, type II diabetes, 18 years of age or older, with HbA_1c of 7.5% (58 mmol/mol) or greater, access to the Internet, and the ability to use a computerIntervention group:n = 40; mean age 56.8 (SD 11.4) yearsControl group:n = 32; mean age 55.8 (SD 11.4) years	Intervention group:Received usual care plus a Fitbit Zip and access to the online self-tracking (eHealth) program aimed to optimize knowledge about a healthy lifestyle, increase awareness of individual physical activity, increase self-efficacy for exercise, and ultimately optimize and maintain a healthy lifestyleControl group:Received only usual care
Orme et al., 2018⁴⁴UK	RCT, 3 arms	0.5 (post discharge)	COPD	LUMO, waist belt, step count, posture, physical activity, and strength	To test feasibility and the acceptability of an education and self-monitoring intervention using wearable technology to reduce sedentary behaviour for individuals with COPD admitted to hospital for an acute exacerbation	n = 33, aged 40–85 years; acute exacerbation of CPOD as the reason for hospitalization; fewer than 4 exacerbations to admission in the previous year; mean age 71.0 (SD 20.0) years; M/F 30/70%Intervention group:1) WD with educational support (n = 10)2) WD with education plus sedentary time vibration feedback (n = 12)Control group:Usual care (n = 11)	Intervention group:The WD with education group received verbal and written information about reducing their time in sedentary behaviour, sitting face-to-face with a study researcher. The WD with education plus feedback group received the same education component along with real-time feedback on their sitting time, stand-ups, and steps at home through a waist-worn inclinometer linked to an app.Control group:Received only usual care
Polgreen et al., 2018⁴⁵USA	RCT, 3 arms	6 (180 days)	Type II DM	Fitbit, wrist band, number of steps, energy expenditure	To determine whether automatic text-message reminders would improve Fitbit adherence and/or increase physical-activity levels long-term; and whether regular text-message reminders plus goal setting would improve Fitbit adherence and/or increase physical-activity levels	n = 138, adults aged 19–75; obese (BMI > 30 kg/m²); fasting glucose of 100 mg/dl or higher in the last year; or had been diagnosed with type II diabetes (not taking insulin)Intervention group:1) Reminders; n = 44; mean age 47.4 (SD 15.1) years; M/F 22.7/77.3%2) Goal setting; n = 46; mean age 43.0 (SD 16.0) years; M/F 21.7/78.3%Control group:Fitbit only; n = 48; mean age 44.6 (SD 16.7) years; M/F 25.5/74.5%	Intervention group:1) Fitbit with reminders group; sent text-message reminders to wear their Fitbit2) Fitbit with both reminders and goal-setting group; sent a daily text message asking for a step goal. All subjects had three in-person visits (baseline, 3 and 6 months)Control group:Given Fitbit but not given any extra information or sent any text message. Both groups were given brochure about “healthy weight loss” from National Institute of Health
Shenoy et al., 2010⁴⁶India	RCT, 2 arms	2	Type II DM	Polar S410™, wrist band, heart rate, and number of steps	To analyse the effects of 8 weeks of aerobic walking using a heart rate monitor and pedometer for monitoring exercise intensity on glycaemic outcomes, fasting blood glucose, cardiovascular fitness, and well-being in type II diabetes patients	n = 40, diagnosed with type II diabetes; aged 40–70 years; not taking insulin; without physical limitations; were not enrolled in any other physical activity program previously or simultaneously and with the duration of diabetes between 1 and 10 yearsIntervention group:n = 20; mean age 53.15 (SD 4.4) years; M/F 75/25%; duration of diabetes 5.3 (SD 0.89) yearsControl group:n = 20; mean age 51.0 (SD 5.4) years; M/F 70/30%; duration of diabetes 5.0 (SD 1.4) years	Intervention group:The subjects performed aerobic walk using pedometer and heart rate monitor to achieve a target of 150-min/week moderate intensity of aerobic physical activity (50–70% of maximum heart rate), approximately 3000 steps. They were given a time schedule and were asked to walk on a regular basis of 5 days a week with pedometer and heart rate monitor and a diary was maintained for them to note heart rate and step count for each sessionControl group: They underwent no training but continued with medication as before and were not engaged in any kind of active exercise intervention during the entire study period

WD: wearable device; RCT: randomized controlled trial; DM: diabetes mellitus; COPD: chronic obstructive pulmonary disease; BMI: body mass index; Hb: haemoglobin; M: male; F: female; SD: standard deviation; IQR: interquartile range.

Characteristics of the reviewed studies

The studies were published in six different countries including the US (n = 5),^{36,37,39,40,45} the UK (n = 2),^38,44 the Netherlands (n = 1),⁴³ Denmark (n = 1),⁴² India (n=1),⁴⁶ and US–Qatar (n = 1).⁴¹ Eight RCTs^{36,37,39–43,46} set two arms, and three RCTs^38,44,45 set three arms for intervention and control groups. The duration of the interventions varied from two weeks to six months. The types of WDs included wristbands^{36,37,39,40,42,43,45,46} and belts worn around the waist,^38,41,44 and they tracked real-time steps,^{36–40,42–46} heart rate,^39,40,46 physical activity,^38–40,44 balance and balance training,⁴¹ and energy expenditure.^38,45 Participants’ mean ages varied from 41 years³⁶ in the DM group to 71 years⁴⁴ in the COPD group. The sample size varied from a minimum of 25 participants⁴⁰ to a maximum of 141,⁴² and nine studies had small samples with fewer than 100 participants^{36–41,43,44,46} (Table 2).

Types of intervention

All interventions involved the use of a WD plus an additional type of education intervention. In the DM group, educational interventions involved goal setting,³⁶ virtual social support,^36,37 motivational feedback using a social networking service or email,^36–38 an electronic health (e-Health) program,⁴³ real-time feedback,⁴¹ text message reminders,⁴⁵ and diary maintenance.⁴⁶ In the CD and COPD groups, the interventions involved online group discussion,³⁹ handouts or written information,⁴⁴ text messaging,⁴⁰ and office visits.⁴² Educational interventions with WD use were compared to those that used WDs without an educational intervention or those that provided usual care without an WD as controls (Table 2).

Risk of bias in the included studies

The summary of risk of bias for each study is shown in Table 3. All studies reported randomization; however, all studies showed an uncertain risk.^36–46 One study had a high risk of bias in allocation concealment³⁹ and blinding of participants and personnel,³⁶ and two studies had a selective reporting bias.^36,37 However, the funnel plot showed no publication bias.

Table 3.

Risk of bias assessment in each study.

	Random sequence generation	Allocation concealment	Blinding of participants and personnel	Blinding of outcome assessment	Incomplete outcome data	Selective reporting bias	Other bias
Bender et al., 2018³⁶	+	?	–	+	+	–	+
Bender et al., 2017³⁷	+	+	?	?	+	–	+
Bentley et al., 2016³⁸	+	+	?	+	?	+	?
Deka et al., 2019³⁹	+	–	?	?	+	+	+
Duscha et al., 2018⁴⁰	+	?	?	?	?	?	?
Grewal et al., 2015⁴¹	+	?	?	?	+	+	+
Kidholm et al., 2016⁴²	?	?	?	?	+	?	+
Kooiman et al., 2018⁴³	+	?	?	?	+	+	+
Orme et al., 2018⁴⁴	+	?	?	?	+	+	+
Polgreen et al., 2018⁴⁵	+	?	?	+	+	+	+
Shenoy et al., 2010⁴⁶	+	?	?	?	+	+	+

(+) = low risk of bias; (?) = uncertain risk of bias; (−) = high risk of bias.

Synthesis of results

The primary and four secondary outcomes were synthesized (Table 4). Other outcomes could not be integrated because they reported different outcome measures or did not provide standard deviation results. A meta-analysis was performed to compare use of WD with educational support versus the control group for each outcome.

Table 4.

Wearable device (WD) studies and reporting clinical outcomes.

Clinical outcomes	Number of studies	Total number of participants	Reference
Primary outcome
Adherence (maintain WD for the intervention period)	11	655	(36–46)
Secondary outcomes
Number of admitted participants	3	204	(39,42,44)
Weight control
Weight loss more than 2%	2	106	(36,37)
Body mass index	2	85	(37,46)
Physical activity
Walking steps/day	5	306	(37,40,41,4
Duration of physical exercise (minutes)	3	80	3,45)
Biomarkers			(39–41)
Fasting blood sugar	2	85	(37,46)
Haemoglobin A_1c	2	85	(37,46)

Adherence of intervention

Adherence was evaluated in all of the studies,^36–46 which were meta-analysed by disease sub-group. Three comparator arms of a COPD study⁴⁴ and a DM study⁴⁵ were integrated into two arms: WD with education intervention versus usual care group or only WD use without education. An adherence of waistbelt WD use was no different compared with the COPD group that did not use a WD (p = .80).⁴⁴ In the DM group, the group that used a wristband WD^{36–38,41,43,45,46} were more likely to display high adherence than the group provided with a body-worn sensor belt intervention⁴¹; however, there was no significant difference (p = .31).

In the CD group, all three studies^39,40,42 used a wristband WD; however, there was no significant difference in adherence between groups (p = .92). Overall, there was no significant difference between intervention and control groups in adherence (RR = 0.97; 95% CI = 0.92 to 1.03; p = .38), and no statistical heterogeneity was observed between sub-groups (I² = 0%; Figure 2).

Figure 2.

Forest plot of comparison for primary outcome (adherence of intervention): wearable device use plus educational support versus wearable device use only/usual care in the number of completions of intervention.

Number of admitted participants

Three RCTs^39,42,44 were evaluated for the number of hospitalized participants and integrated into the meta-analysis. One study with a waist belt WD was targeted for COPD,⁴⁴ and two studies with wristband WDs were targeted for CD.^39,42 No difference in WD use was observed in hospitalized participants (RR = 1.33; 95% CI = 0.89 to 1.99; p = .17; I² = 0%; Figure 3).

Figure 3.

Forest plot of comparison for secondary outcomes (number of admitted participants): wearable device use plus educational support versus wearable device use only/usual care in the number of admitted participants during intervention.

Weight control

Two RCTs that targeted DM^36,37 were evaluated for more than 2% weight loss and BMI reduction, and integrated into the meta-analysis. Both studies provided wristband WDs and set goals for weight, diet, physical activity, office visits, and weekly posting of discussion topics on a Facebook group site. Only the control group were asked to wear the WD. There was a significantly higher number of participants with a weight loss > 2% of baseline weight in the WD group that received a goal-setting education intervention (RR = 2.20; 95% CI =1.38 to 3.50; p = .0009; I² = 0%); however, there was no difference in BMI (MD = –1.89; 95% CI = –5.20 to 1.41; p = .26; I² = 79%; Figure 4).

Figure 4.

Forest plot of comparison for secondary outcomes (weight control: number with > 2% weight loss and BMI): use of wearable device plus educational support versus wearable device use without education/usual care for the weight control.

Physical activity

Five RCTs^{37,40,41,43,45} evaluated the number of steps/day and included a meta-analysis. Most studies compared wristband/waist belt WD use with education interventions like goal setting, interactive training, text messaging, and e-Health programs. There was no significant difference between groups in number of steps/day (MD = 333.48; 95% CI = –415.83 to 1082.79; p = .38; I² = 0%). Three studies – one for patients with DM⁴¹ and two for patients with CD^39,40 – were identified concerning physical exercise duration (minutes) and meta-analysed. The DM study provided a sensor belt WD with a training program, and a CD study provided a wristband WD and conducted face-to-face weekly online groups or provided health coaching and text messaging in an intervention group. A significant difference was observed in the exercise duration in the CD sub-groups (MD = 9.53; 95% CI = 0.59 to 18.47; p = .04; I² = 0%; Figure 5).

Figure 5.

Forest plot of comparison for secondary outcomes (physical activity: walking steps per day and duration of physical exercise): use of wearable device plus educational support versus wearable device use only/usual care for the physical activity.

Biomarkers

Two studies^37,46 that targeted DM were evaluated for FBG and HbA_1c and integrated into the meta-analysis. Both studies were implemented for two to three months, provided virtual social support and coaching to increase patients’ number of steps,³⁷ and prompted them to keep a diary.⁴⁶ There were significant differences in FBG (MD = –32.39; 95% CI = –48.07 to –16.72; p < .0001; I² = 68%) and HbA_1c (MD = –0.69; 95% CI = –1.28 to –0.10; p = .02; I² = 60%; Figure 6).

Figure 6.

Forest plot of comparison for secondary outcomes (biomarkers: fasting blood glucose and HbA_1c): use of wearable device plus educational support versus wearable device use only/usual care for the biomarkers.

Discussion

Characteristics and quality of RCTs

There were only 11 RCT studies available, and the studies were typically conducted in the US and Europe rather than in East Asian countries. The number of eligible participants was 655, which is limited. All the studies were published between 2010 and 2019. We identified a total of 69 outcomes from 11 articles in this review; however, only eight outcomes were integrated into the review because other outcomes used different measurement tools or did not provide standard deviations.

Summary of the intervention

All the reviewed studies compared the WD-use interventions and simultaneously provided educational support during said interventions. Of the two types of devices used in the studies, the wristbands seemed to result in less operational difficulty for the participants. The frequency of the WD-use intervention showed some difference; a wristband and a waist/hip belt WD was used continuously for the intervention period, and a lower-back body-worn sensor belt was used for 45-minute training sessions twice a week for four weeks. These interventions were compared to interventions with the use of WDs without an education group or only provided with usual care with no WD use.

Most RCTs set two arms for the intervention and control groups; however, the control groups comprised both no WD use and the use of WD without educational support. The educational support given to the intervention group showed some variety; thus, it was grouped into two types: “in-person education,” such as goal setting, providing handouts or written information, or visits to nurses’ offices by participants; and “electronic-style educational support,” such as virtual social support using a social network service, e-Health/mobile health (mHealth) programs, real-time feedback, and text message reminders. Therefore, the interventions in the reviewed article have the potential of clinical heterogeneity. To be safe, no adverse events were reported in the studies.

Summary of the participants

The mean age of the participants was the youngest in the DM group and the oldest in the COPD group, which showed a characteristic of the disease process. Specifically, 43% of the COPD group were readmitted to the hospital within three months before being included in the study, indicating that their prognoses may have been more unstable as compared to the other two disease groups. Furthermore, the difference in the needs for using WDs depending on the age of the participants was considered.

Summary of evidence

Selection, performance, and detection bias were considered in the studies and the quality of evidence in this review was low. This meta-analysis showed that, among the DM group, the use of WDs and goal-setting educational support over three months was more effective than only using WDs without education to reduce weight by more than 2%; and using WDs with some follow-up support implementation resulted in reduced FBG and HbA_1c as compared with the group receiving usual care and not using WDs. The same result was true in the CD group concerning physical exercise duration. The results may be particularly noteworthy concerning the intervention; however, the available evidence was quite limited. For the other outcomes, we found no intervention effects on any of the primary/secondary outcomes: adherence, number of admitted participants, BMI, and steps/day. The effective intervention period, type of educational support, and disease group could not be clearly discussed in this review owing to the small number of studies.

This review concludes that the use of WDs has the potential to improve weight loss, FBG, and HbA_1c for people with DM; moreover, their use may improve exercise duration in the CD group if they are provided with some educational support. As suggested by selection bias, the results may differ for different target age groups. The age of people with chronic disease was wide-ranging, and the need for self-care by using WDs was considered to be different. Moreover, in this meta-analysis, many interventions, such as class of technologies and education protocols, and many suspected comparisons make the results prone to having larger alpha errors owing to multiple comparisons. Nonetheless, the p-value for the outcomes of physical exercise duration was significant (p = .04).

Implication for home care and telemonitoring for older adults with chronic conditions

In the context of chronic disease management and the chronic care model to improve telehealth outcomes, it is indicated that not only self-management support but also design of the healthcare delivery system, decision support, clinical information system in the healthcare organization, and community resources are important.⁴⁷ Therefore, using new healthcare technology such as WDs may provide benefits by adding to a new telehealth delivery system to support people with chronic disease, reinforcing their disease management. Moreover, it is important that cost, accuracy, and user-friendliness are considered before adopting this new technology, specifically for older adults who may need to use it for the rest of their lives.

Owing to the increase of life expectancy, an aging population with chronic diseases has expanded the demand for telehealth and telemonitoring. The DM group were rather young and vigorous compared to the COPD and CD groups in this review, and the type of education was adapted to each target group’s overall severity of their conditions. There are several distinct exercise recommendations, such as i) reduced sedentary time in the COPD group; ii) increased walking, weight reduction, and physical exercise in the DM group; and iii) other-device use, online discussions, or rehabilitation education for the CD group. However, the results of this review provided little evidence of the effectiveness of the use of WDs for disease management. Future studies may change these conclusions. Although the available evidence is quite limited, use of WDs provides an opportunity to utilize new technology for real-time telemonitoring of chronic conditions. Therefore, educational support to improve older adults’ telehealth outcomes by using WDs should be considered.

Limitations

We only reviewed 11 RCT studies, which were split into three disease categories; therefore, there is a risk of selection bias. The small number of participants also suggests a reporting bias. Owing to multiple comparisons of clinical and educational interventions, some of the significant pooled effects may turn out to be non-significant after statistical adjustment for multiple comparisons. Thus, the effects may have been overestimated, and it should be interpreted with caution. We could not discuss which type of educational support would be appropriate to be provided simultaneously to WD users. Moreover, available evidence of use for chronic disease management was quite limited. However, the use of WDs was demonstrated to be a useful technology with no reported harmful results.

Conclusions

Eleven RCTs with 657 participants were integrated into a meta-analysis that analysed the effectiveness of WDs used with educational support to improve adherence to healthcare guidelines and telehealth outcomes. This meta-analysis showed that, in the DM group, the use of WDs in combination with educational support was more effective than the use of only WDs with no education to reduce weight by more than 2% when delivered with educational support. Further, the use of WDs was associated with reduced FBG and HbA_1c compared with regular care and no-use of a WD in the CD group. WD use with online support improved the physical exercise duration time when implemented for two to three months. However, currently, there is not enough evidence to support the use of WDs because they have not been sufficiently shown to affect telehealth outcomes for chronic disease management.

What is already known about the topic?

The WD market is rapidly growing, and it has the potential to follow information with real-time synchronization with healthcare providers and people with chronic conditions to monitor health conditions along with effective communication to manage risk factors and provide feedback for self-management.

Recently, several types of wearable technology were initiated to be adopted by people with chronic conditions and were suggested to have the potential to assist in predicting clinical outcomes.

What this study adds

This meta-analysis showed that the use of WDs with educational support rather than use of WDs without education had the potential to reduce weight and improve exercise time, FBG, and HbA_1c when implemented for two to three months and targeted to people with DM/CD; however, the available evidence was quite limited. Adherence and other outcomes showed no difference between wearing a device or not among people with chronic disease.

Footnotes

Acknowledgements

The authors would like to thank Ms Julie Hansen, research librarian, University of Queensland, and Ms Naoko Matsumoto, St Luke’s International University, Tokyo. This systematic review was registered to PROSPERO through the National Institute for Health Research (ID CRD42018111632). The author contributions were as follows: all authors made a direct contribution to this study; Kamei was responsible for the study concept and design, literature database search, screen, text review and data integration, and drafting the manuscript. Edirippulige carried out the database search of the literature with PI, text review and interpretation of data, critical revision of the manuscript. Kanamori was responsible for screen, text review, data extraction and integration, and critical revision of the manuscript. Yamamoto was responsible for screen, text review and data extraction of the literature, and critical revision of the manuscript. All authors approved the final version of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: this study was funded by the JSPS KAKENHI, Tokyo, Japan (Grant-in-Aid for Scientific Research (A) 19H01082).

ORCID iDs

Tomoko Kamei

Sisira Edirippulige

References

Grand View Research. Wearable medical device market size, share & trends analysis report by product type, by site, by application and segment forecasts, 2020 – 2027, https://www.grandviewresearch.com/industry-analysis/wearable-medical-devices-market (2020, accessed 25 March 2020).

Lunney

Cunningham

Eastin

MS.

Wearable fitness technology: a structural investigation into acceptance and perceived fitness outcomes. Comput Hum Behav 2016; 65: 114–120.

Mezghani

Exposito

Drira

, et al. A semantic big data platform for integrating heterogeneous wearable data in healthcare. J Med Syst 2015; 39: 185.

Raskovic

Martin

Jovanov

Medical monitoring applications for wearable computing.

Comput J 2004; 47(4): 495–504.

Anliker

Ward

Lukowicz

, et al. AMON: a wearable multiparameter medical monitoring and alert system. IEEE Trans Inf Technol Biomed 2004; 8(4): 415–427.

Feehan

L M

Geldman

Sayre

, et al. Accuracy of Fitbit devices: systematic review and narrative syntheses of quantitative data. J Med Internet Res 2018; 6(8): e10527.

Bonato

Advances in wearable technology and its medical applications. Conf Proc IEEE Eng Med Biol Soc 2010; 2021–2024.

Dobkin

Dorsch

The promise of mHealth: daily activity monitoring and outcome assessments by wearable sensors. Neurorehabil Neural Repair 2011; 25(9):788–798.

Baig

GholamHosseini

Moqeem

, et al. A systematic review of wearable patient monitoring systems – current challenges and opportunities for clinical adoption. J Med Syst 2017; 41: 115.

10.

Hickey

Freedson

PS.

Utility of consumer physical activity trackers as an intervention tool in cardiovascular disease prevention and treatment. Prog Cardiovasc Dis 2016; 58(6): 613–619.

11.

Ummels

Beekman

Theunissen

, et al. Counting steps in activities of daily living in people with a chronic disease using nine commercially available fitness trackers: cross-sectional validity study. JMIR Mhealth Uhealth 2018; 6(4): e70.

12.

Cancela

Pastorino

Tzallas

, et al. Wearability assessment of a wearable system for Parkinson’s disease remote monitoring based on a body area network of sensors. Sensors 2014; 14(9): 17235–17255.

13.

Burnham

Yaeger

, et al. Using wearable technology to predict health outcomes: a literature review. J Am Med Inform Assoc 2018; 25(9): 1221–1227.

14.

Armstrong

Najafi

Shahinpoor

Potential applications of smart multifunctional wearable materials to gerontology. Gerontology 2017; 63(3): 287–298.

15.

Costa

SEP

Rodrigues

JJPC

Silva

BMC

, et al. Integration of wearable solutions in AAL environments with mobility support. J Med Syst 2015; 39(12): 184.

16.

Kamisalic

Fister

Turkanovic

, et al. Sensors and functionalities of non-invasive wrist-wearable devices: a review. Sensors 2018; 18(6): 1714.

17.

Majumder

Mondal

Deen

MJ.

Wearable sensors for remote health monitoring. Sensors. 2017; 17: E130.

18.

World Health Organization. Noncommunicable diseases, https://www.who.int/news-room/fact-sheets/detail/noncommunicable-diseases (2018, accessed 31 March 2020).

19.

World Health Organization. Preventing noncommunicable disease, https://www.who.int/health-topics/noncommunicable-diseases#tab=tab_1 (2019, accessed 31 March 2020).

20.

Franceschi

Garagnani

Morsiani

, et al. The continuum of aging and age-related diseases: common mechanisms but different rates. Front Med 2018; 5(61): 1–23.

21.

NZ Telehealth Center. About telehealth; telemonitoring, https://www.telehealth.org.nz/what-is-telehealth/telemonitoring/ (2019, accessed 25 March 2020).

22.

McLaughlin

Why telemedicine and remote patient monitoring demand will skyrocket in 2019, https://www.redoxengine.com/blog/telemedicine-remote-patient-monitoring-demand-2019/ (2019, accessed 25 March 2020).

23.

Kamei

Yamamoto

Kajii

, et al. Systematic review and meta-analysis of studies involving telehome monitoring-based telenursing for patients with chronic obstructive pulmonary disease. JJNS 2013; 10(2): 180–192.

24.

Kim

Park

Lee

, et al. Comparative effectiveness of telemonitoring versus usual care for type 2 diabetes: a systematic review and meta-analysis. J Telemed Telecare 2018; 25(10): 587–601.

25.

Clarke

Shah

Sharma

Systematic review of studies on telemonitoring of patients with congestive heart failure: a meta-analysis. J Telemed Telecare 2011; 17(1): 7–14.

26.

Kamei

Yamamoto

Kanamori

, et al. Detection of early‐stage changes in people with chronic diseases: a telehome monitoring‐based telenursing feasibility study. Nursing & Health Sciences 2018; 3(20): 313–322.

27.

Steins

Dawes

Esser

, et al. Wearable accelerometry-based technology capable of assessing functional activities in neurological populations in community settings: a systematic review. J Neuroeng Rehabil 2014; 11: 36.

28.

Liaqat

de Lara

, et al. Feasibility of using a smartwatch to intensively monitor patients with chronic obstructive pulmonary disease: prospective cohort study. J Med Internet Res 2018; 6(6): e10046.

29.

Deka

Pozehl

Norman

, et al. Feasibility of using the Fitbit® Charge HR in validating self-reported exercise diaries in a community setting in patients with heart failure. Eur J Cardiovasc Nurs 2018; 17(7): 605–611.

30.

Petrofsky

McLellan

Prowse

, et al. The effect of body fat, aging, and diabetes on vertical and shear pressure in and under a waist belt and its effect on skin blood flow. Diabetes Technol Ther 2010; 12(2): 153–160.

31.

Weatherall

Paprocki

Meyer

, et al. Sleep racking and exercise in patients with type 2 diabetes mellitus (Step-D): pilot study to determine correlations between Fitbit data and patient-reported outcomes. JMIR mHealth and uHealth 2018; 6(6): e131.

32.

Fukuoka

Vittinghoff

Hooper

A weight loss intervention using a commercial mobile application in Latino Americans-Adelgaza trial. TBM 2018; 8: 714–723.

33.

Gorina

Limonero

Álvarez

Effectiveness of primary healthcare educational interventions undertaken by nurses to improve chronic disease management in patients with diabetes mellitus, hypertension and hypercholesterolemia: a systematic review. Int J Nurs Stud 2018; 86: 139–150.

34.

PRISMA. PRISMA statement, http://www.prisma-statement.org/PRISMAStatement/ (2015, accessed 25 March 2020).

35.

Higgins

JPT

Thomas

Cochrane handbook for systematic review of interventions, the Cochrane Collaboration, https://training.cochrane.org/handbook/current (2019, accessed 25 March 2020).

36.

Bender

Cooper

Flowers

, et al. Filipinos fit and trim–a feasible and efficacious DPP based intervention trial. Contemp Clin Trials Commun 2018; 12: 76–84.

37.

Bender

Cooper

Park

, et al

. A feasible and efficacious mobile-phone based Lifestyle intervention for Filipino Americans with type 2 diabetes: randomized controlled trial.

JMIR Diabetes 2017; 2(2): e30.

38.

Bentley

Otesile

Bacigalupo

, et al. Feasibility study of portable technology for weight loss and HbA1c control in type 2 diabetes. BMC Med Inform Decis Mak 2016; 6: 92.

39.

Deka

Pozehl

Williams

, et al. MOVE-HF: an internet-based pilot study to improve adherence to exercise in patients with heart failure. Eur J Cardiovasc Nur 2019; 18(2): 122–131.

40.

Duscha

Piner

Patel

, et al. Effects of a 12-week mHealth program on peak VO₂ and physical activity patterns after completing cardiac rehabilitation: a randomized controlled trial. Am Heart J 2018; 199: 105–114.

41.

Grewal

Schwenk

Lee-Eng

, et al. Sensor-based interactive balance training with visual joint movement feedback for improving postural stability in diabetics with peripheral neuropathy: a randomized controlled trial. Gerontology 2015; 61: 567–574.

42.

Kidholm

Rasmussen

Andreasen

, et al. Cost-utility analysis of a cardiac telerehabilitation program: the teledialog project. Telemedicine and e-Health 2016; 22(7): 553–563.

43.

Kooiman

TJM

de Groot

Hoogenberg

, et al. Self-tracking of physical activity in people with type 2 diabetes: a randomized controlled trial. Comput Inform Nurs 2018; 36(7): 340–349.

44.

Orme

Weedon

Saukko

, et al. Findings of the chronic obstructive pulmonary disease- sitting and exacerbations trial (COPD-SEAT) in reducing sedentary time using wearable and mobile technologies with educational support: randomized controlled feasibility trial. JMIR mHealth and uHealth 2018; 6(4): e84.

45.

Polgreen

Anthony

Carr

, et al. The effect of automated text messaging and goal setting on pedometer adherence and physical activity in patients with diabetes: a randomized controlled trial. PLoS ONE 2018; 13(5): e9579701.

46.

Shenoy

Guglani

Sandhu

JS.

Effectiveness of an aerobic walking program using heart rate monitor and pedometer on the parameters of diabetes control in Asian Indians with type 2 diabetes. Primary Care Diabetes 2010; 4(1): 41–45.

47.

Reynolds

Dennis

Hasan

, et al. A systematic review of chronic disease management interventions in primary care. BMC Fam Pract 2018; 19(11): 1–13. doi:10.1186/s12875-017-0692-3.