An integrative review of chronic illness mHealth self-care interventions: Mapping technology features to patient outcomes

Abstract

Mobile health (mHealth)—hand-held technologies to address health priorities—has significant potential to answer the growing need for patient chronic illness self-care interventions. Previous reviews examined mHealth effect on patient outcomes. None have a detailed examination and mapping of specific technology features to targeted health outcomes. Examine recent chronic illness mHealth self-care interventions; map the study descriptors, mHealth technology features, and study outcomes. (1) Information extracted from PubMed, CINAHL, and Web of Science databases for clinical outcomes studies published 2010–January 2020; and (2) realist synthesis techniques for within and across case analysis. From 652 records, 32 studies were examined. Median study duration was 19.5 weeks. Median sample size was 62 participants. About 47% of interventions used solely patient input versus digital input; 50% sent tailored messages versus generic messages; 22% augmented the intervention with human interaction. Studies with positive clinical outcomes had higher use of digital input. Software descriptions were lacking. Most studies built interventions: only two incorporated target audience participation in development. We recommend researchers provide sufficient system description detail. Future research includes: data input characteristics; impact of augmentation with human interaction on outcomes; and development decisions.

Keywords

cellular phone smart phone clinical trial patient self-care home care and e-health patients with chronic illness or special needs

Introduction

Chronic disease, with its high prevalence, increased mortality, and associated health costs, is a public health burden internationally.¹ For example, in the United States, the chronic illness cost burden was estimated at 78% of total health care spending.² People living with chronic conditions have economic challenges including medication related costs and reduced ability to work.³ A technology category with significant potential to address this need is mobile health (mHealth)—application of mobile (e.g. hand-held) technologies to address health priorities.⁴ Accordingly, mHealth studies have been published, among which are behavioral interventions intended to improve patient chronic illness management.^5–7 This management encompasses tasks that persons need to take on so as to live well with chronic condition(s)⁸ such as following a specific diet, daily monitoring for physiologic changes (e.g. blood glucose), and response to those changes.

Lately, reviews have examined mHealth chronic illness self-care interventions’ impact on clinical outcomes,^6,7 whereas previous reviews were limited to reports on mHealth feasibility, acceptance, and usability^5,9 and impact on behavior change^9,10 for specific chronic conditions. The two recent articles focused on clinical outcomes are relatively small systematic reviews (in regard to number of studies) of mHealth randomized control trials (RCT) published between 2005 and 2016. Lee et al. included multiple common chronic diseases. Ten of 12 studies reported statistically significant improvement in health and clinical outcomes. Identified mHealth functionality characteristics were data input devices and automated text reminders.⁶ Whitehead et al. focused on apps for four chronic conditions (i.e. diabetes types 1 and 2, cardiovascular diseases, chronic lung diseases). Six of nine studies reported statistically significant improvement in clinical outcomes. Identified mHealth functionality were apps which tended to include data input, data transmission, and sending text messages; and receiving automated text reminders.⁷

However, mHealth interventions targeting self-care of chronic conditions have not had a detailed examination and mapping of the specific mHealth technology features (e.g. data input, messaging) to the health outcomes they target. This mapping would enable investigation of whether specific technology characteristics result in better patient outcomes.

Objectives

This review updates and expands previous reviews to examine mHealth features, viewed through the scope of self-care. Our objective was to: (1) identify empiric (i.e. observational or experimental) mHealth interventions studies explicitly designed for self-care of community-living adults with one or more chronic conditions which assessed the impact on patient outcomes, and among the retained studies (2) categorize and map the study descriptors, mHealth technology features, and study outcomes. We targeted studies sufficiently developed to include clinical outcomes, in contrast to evaluation limited to process outcomes such as feasibility or usability.

Methods

The authors were a health informatician with expertise in home health care and mHealth (PS), a nurse expert in geriatric patient self-care (HB), and a nurse with expertise in chronic conditions and software development for clinician chronic condition management (ES).

Study design

An integrative review methodology was used to synthesize the research results using a narrative analysis¹¹ for literature published from 2010 to January 20, 2020. The methodology was to: (1) identify applicable studies; (2) select for review studies that met criteria; (3) organize data from the retained studies; and (4) summarize and report the results. This paper is one of two papers addressing the topic area mHealth in chronic illness management as part of a larger examination of the mHealth human/technology interface. One paper addresses the human aspect or self-care behavior issues in mHealth studies.¹² This second paper is designed to address (1) the technology aspect or mHealth intervention technology features themselves (i.e. patient or digital device input, general or tailored patient feedback, software features) and (2) their relation to health outcomes and mHealth usage and satisfaction outcomes. Dissemination via this two-pronged approach was designed to support the advancement of the science.

Identify studies

Two authors (PS, ES) developed a search strategy with the assistance of a research librarian (see Acknowledgments). The authors identified a model article,⁵ which was a scoping review of studies focused on the design, development, and evaluation of self-care mHealth for older adults with chronic conditions living at home. From this article the authors gleaned keywords which were used in a PubMed search. This search returned mostly literature reviews, the reference lists of which were hand searched for studies.

Referring to the model article, the authors consulted the research librarian to develop a definitive list of key keywords (i.e. MeSH terms and text words which would also search in the MeSH field) and subject headings in order to capture a comprehensive list of potential sources. They identified and combined keywords to address four research question components: (1) mobile or electronic devices, (2) technology-based health care delivery, (3) chronic condition, and (4) an adult population. The research librarian structured and ran a search query using Boolean operators for each database of peer-reviewed scientific journals searched: PubMed, CINAHL, Web of Science. These databases were available at the librarian’s institution and chosen for their salience to finding mHealth intervention studies. Delimiters in the query were that the articles were published in English, after 2010. Literature reviews related to patient-facing (e.g. for use by patients) mHealth for chronic conditions were also reviewed to retrieve studies to be considered for inclusion.

Selection for review studies that meet criteria

The two authors (PS, ES) established inclusion criteria for article retention: observational or experimental research studies explicitly focused on self-care and with clinical outcomes, which included mHealth technologies designed for use by community-residing adults living with at least one physiological chronic condition. The mHealth technology could be developed by the researchers, and/or publicly available.

They performed the article selection from each database sequentially, as follows. They independently identified articles for review by searching titles and abstracts and resolving conflicts through consensus. Exclusion criteria were: (1) solely mental health or behavioral health conditions; (2) children (younger than 18 years); (3) study designs which lacked clinical outcomes such as feasibility, pilot studies, protocols, software description; (4) mHealth designed solely for provider use; (5) technologies out of scope: telehealth, social networking, telephonic and telecounseling, telerehabilitation; (6) non-chronic conditions such as rehabilitation (e.g. stroke; cancer), risk factors (e.g. smoking), or health promotion; and (7) absence of clinical outcome.

Articles that met inclusion and exclusion criteria through abstract review were independently reviewed in full by two reviewers (PS, ES). The reviewers held regular virtual meetings to discuss their decisions regarding inclusion/exclusion criteria of articles. A result of these meetings was the revision of both inclusion and exclusion criteria as the search progressed so as to better address the research question. The adjustments were to include only physiological conditions (e.g. exclude behavioral conditions), include studies with adults of any age (e.g. not limited to older adults); include studies of mHealth sufficiently developed so as to be studies in an observational or experimental trial with measurable clinical outcomes (e.g. exclude studies limited to user acceptance); and exclude studies which did not electronically generate messages to patients (instead only people such as clinicians or coaches sent messages).

Organize, summarize, and report the review results

The reviewers (PS, HB, ES) developed a data charting form, implemented as an Excel spreadsheet, to record selected data from the retained articles. The spreadsheet was revised as the review progressed.¹² The following information from articles identified through the review process described above was entered into the spreadsheet:

Publication information (Author, title, journal, date)

Condition(s)

Sample size

Study duration

Research setting country

Participant age range

mHealth user interface characteristics (e.g. patient input, electronic device input, personal data used, patient feedback—general or tailored messages)

Human interaction to augment technology

Software characteristics: algorithm; gamification

Outcome type (clinical, satisfaction, etc.)

Outcome measures

Main study findings

The two reviewers (PS, ES) abstracted data from the retained articles using this standardized form by charting half the articles and cross-checking the other author’s charted articles. They discussed charting differences until disagreements were resolved. A third author, HB, reviewed the resulting form for completeness and confirmation of technology elements.

The reviewers (PS, ES) synthesized the data using realist synthesis techniques^13,14: Analyzing data elements (i.e. technology and patient outcomes) across studies to determine the nature and relationships among and between these data elements to deduce higher order abstractions. This included examining interface characteristics, software characteristics, and human augmentation in light of study outcomes; and investigating patterns across studies. Study outcomes were classified into three types: clinical; performance of care processes (i.e. adherence, health care, or lifestyle changes measures); and usefulness, usage, and satisfaction. Studies were categorized as having positive, mixed, or null (no/null) outcomes based primarily on the reported clinical outcomes. Studies categorized as positive had statistical significance in the primary outcome measure. Studies that did not report clinical outcomes were classified by their reported clinical performance of care processes.

Results

The query used in the three databases, PubMed, CINAHL, and Web of Science, is shown in Figure 1. The search of articles published between 2010 and 2020, which also included those in relevant literature reviews, returned 652 unique studies, of which 32 studies were retained as shown in Figure 2.

Figure 1.

Search query used against the three databases, PubMed, CINAHL, and Web of Science.

Figure 2.

PRISMA flow study diagram.

Study descriptors

The location of most studies was North America (n = 17 studies (53%)) as shown in Table 1. The remaining studies were conducted in Asia (n = 6 (19%)), Europe and Scandinavia (n = 4 (13%)), Australia (n = 3 (9%)), and the Indian sub-continent (n = 2 (6%)). The study design of the majority of studies was randomized control trial (n = 20 (63%)). However, a few of these studies omitted the power analysis,^15–17 or appeared to be under-powered (for all¹⁸ or some¹⁹ measures) suggesting possible threats to the study findings validity. Additional study designs were single-armed trials (n = 9 (28%)), a (3%) three-armed trial,²⁰ a two-armed case-control trial,²¹ and an observational study.²² A few single-armed trials omitted the power analysis,²³ or appeared to be under-powered for all^24,25 or some²⁶ measures. The median of the studies’ duration was 19.5 weeks with a range of 4^27,28 to 52²⁹ weeks. The median of the studies’ sample size was 62 participants (range 6,²⁷ 710³⁰). Two studies focused on multiple chronic conditions,^31,32 one of which was in a population over 65 years old.³³ The remainder of the studies focused on 1 of 12 conditions, with diabetes (n = 10 studies (31%)) and a group of cardiovascular conditions (n = 8 (25%)) being the most frequent. The median age of the studies’ participants was 52 years of age, with a range of 29²¹ to 72³² years of age.

Table 1.

Descriptors of retained studies: location, study design, sample size, participant age, study duration, and chronic condition.

Article authors, location	Study design	Sample size	Mean age (standard deviation)	Study duration (weeks)	Condition
Adams et al.,⁴⁰ USA	RCT	64	35.1 (12.5)	26	Hypertension
Agarwal et al.,¹⁵ Canada	RCT	223	Intervention: 51.5 (10.6)	26	Diabetes
Arora et al.,⁴⁶ USA	RCT	128	50.7 (10.2)	26	Diabetes
Chandler et al.,¹⁷ USA	RCT	54	46.5 (9.9)	39	Hypertension
Chhabra et al.,⁴³ India	RCT	93	41.2 (14.1)^a	13	Chronic low back pain
Chow et al.,³⁰ Australia	RCT	710	58 (9.2)	26	HF
Cook et al.,²⁴ USA	Pre-post	60	50.1 (17)	18	Asthma
Fukuoka et al.,³⁷ USA	RCT	61	55.2 (9.0)	21	Obesity
Goh et al.,⁴⁷ Singapore	Pre-post	84	48.2 (8.5)	9	Diabetes
Isetta et al.,³⁸ Spain	Observational	60	56 (10)	6	Sleep apnea
Kim et al.,¹⁹ Korea	RCT	151	58.4 (8.9)^a	26	Diabetes
Kirwan et al.,³⁶ Australia	RCT	53	35.2 (10.43)	26	Diabetes
Kleinman et al.,³⁴ India	RCT	91	48.4 (9.2)	26	Diabetes
Lee et al.,⁶ Korea	Case-control	36	28.8	6	Obesity
Lim et al.,⁴¹ Korea	RCT	144	60	26	Diabetes
Liu et al.,¹⁶ Taiwan	RCT	89	52	26	Asthma
Mallow et al.,³¹ USA	Intervention	30	52 (10.0)	13	MCC
Martin et al.,²⁸ USA	RCT	48	41.4	4	Cardiovascular
Mira et al.,³² Spain	RCT	99	71.9 (7.1)	13	MCC, polypharmacy
Nundy et al.,²⁷ USA	Pre-post	6	50	4	HF
Ong et al.,⁴⁵ Canada	Pre-post	47	59.4 (14)	26	Kidney disease
Plow and Golding²⁰ USA	RCT	46	57.8 (9.48)	7	Musculoskeletal or neurological conditions
Quinn et al.,²⁹ USA	RCT	163	52.8	52	Diabetes
Ryan et al.,²⁵ Canada	Pilot	31	40.0 (13.9)	21	Diabetes
Selter et al.,³⁹ USA	Intervention single arm	93	46 (16)	13	Lower back pain
Seto et al.,²⁶ Canada	Pre-post	100	53.7 (13.7)^a	26	Heart failure
Sieber et al.,²² Germany	Uncontrolled, multicenter, observational	51	54.1 (12.6)	13	Diabetes
Skrepnik et al.,⁴⁴ USA	RCT	211	62.6 (9.4)	13	Osteoarthritis
Torbjørnsen et al.,⁴² Norway	RCT	151	57 (12)	18	Diabetes
Toro-Ramos et al.,²³ USA	Experimental	50	47.7 (10.3)	26	Prehypertension or hypertension
Varnfield et al.,¹⁸ Australia	RCT	120	55	30	Myocardial
Waki et al.,³⁵ Japan	RCT	54	57	13	Diabetes

Indicates median and/or standard deviation calculated by authors using published study data.

mHealth characteristics

mHealth interventions mentioned in the retained studies had characteristics which included the user interface, whether the intervention was stand-alone, software features, and development approach. Two user interface characteristics and whether the mHealth intervention was augmented with human interaction are shown in Table 2.

Table 2.

Retained studies’ characteristics of user interface and human augmentation, with study outcome categorization.

Article authors	Characteristics of the mHealth user interface			Stand alone or coach/provider input	Study outcome characterization
Article authors	Patient input	Digital device input (EHR, device)	Patient feedback (generic, tailored, unspecified)	Stand alone or coach/provider input	Study outcome characterization
Adams et al.⁴⁰	None specified	Continuous heart rate readings derived from the phone’s video camera via reflective photoplethysmography	Tailored: heart rate; motivational messages	Stand alone	Positive
Agarwal et al.¹⁵	Baseline health, blood glucose, exercise activity, food intake	None specified	Tailored: educational, motivational messages	Stand alone	Null
Arora et al.⁴⁶	None specified	None specified	Generic text messages	Stand alone	Null
Chandler et al.¹⁷	None specified	Digital device input (blood pressure, pill dispenser)	Generic: input reminders; Tailored: feedback of BP, heart rate levels; motivational/social reinforcement	Stand alone	Positive
Chhabra et al.⁴³	Surveys	Sensors collected daily activity data	Tailored: variation in activity level	Stand alone	Mixed
Chow et al.³⁰	None specified	None specified	Tailored: support and motivation	Stand alone	Positive
Cook et al.²⁴	Condition and self-knowledge self assessment	None specified	Generic: input reminders; Tailored: level of control	Stand alone	Positive
Fukuoka et al.³⁷	Weight, activity, caloric intake	None specified	Generic: input reminders, educational; Tailored: goals	Stand alone	Mixed
Goh et al.⁴⁷	Answers to daily symptom questions	Daily weight and blood pressure readings and weekly single-lead ECGs	Not described	Stand alone	Insufficiently described
Isetta et al.³⁸	Treatment adherence yes/no questions	None specified	Generic: input reminders; feedback on treatment adherence	Stand alone	Null
Kim et al.¹⁹	None specified	Bluetooth glucometer, activity tracker	Tailored: insulin level	Stand alone	Mixed
Kirwan et al.³⁶	Blood glucose levels, insulin dosages, other medications, diet, physical activities	None specified	Unspecified tailored weekly message	Stand alone	Mixed
Kleinman et al.³⁴	Blood glucose	None specified	Generic: task reminders; out-of-range follow-up questions	Coach to respond to patient questions, alerts	Mixed
Lee et al.²¹	Food consumption, activity	None specified	Tailored: nutrition/calories; avatar body type	Stand alone	Positive
Lim et al.⁴¹	None specified	Automatic upload of blood glucose data	Generic: reminders; Tailored: evaluation messages	Stand alone	Positive
Liu et al.¹⁶	Electronic diary: daily asthma symptom score, use of relievers, peak expiratory flow rate (PEFR), and PEFR variability	None specified	Tailored: assessment of asthma status, corresponding management advice	Stand alone	Positive
Mallow et al.³¹	Self-monitored readings	Bluetooth scale, glucometer, blood pressure cuff	Tailored: previous readings, reminders for using the self-monitoring devices and taking medications	Stand alone	Mixed
Martin et al.²⁸	Activity data	Wearable accelerometer, Bluetooth-enabled with compatible smartphones	Tailored: reinforcement and booster messages based on real-time activity	Stand alone	Positive
Mira et al.³²	None specified	Automated pillbox	Tailored: medication adherence; Unspecified: reminders, medication images	Stand alone	Null
Nundy et al.²⁷	None specified	None specified	Generic: education; Tailored: reminders	Stand alone	Null
Ong et al.⁴⁵	Medication management, symptoms, laboratory test results	BP	Tailored: condition assessment	No coach; clinician monitoring	Mixed
Plow and Golding²⁰	Physical activity, nutritional behaviors, progress to goals, symptoms	None specified	Display	Stand alone	Null
Quinn et al.²⁹	Blood glucose values, carbohydrate intake, medications, other diabetes management information	None specified	Tailored: educational, behavioral, and motivational messaging specific to the entered data	Coach	Mixed
Ryan et al.²⁵	Blood glucose, proposed carbohydrate intakes, planned activities	None specified	Tailored emails: appropriate insulin doses	Provider	Positive
Selter et al.³⁹	Pain, activity level, medication/coping mechanisms	None specified	Generic: education; Tailored: passive activity-level measurement	Coach	Null
Seto et al.²⁶	Symptoms	Weight, blood pressure, ECG	Tailored: message/alert based on the physiological and symptom information	Stand alone	Mixed
Sieber et al.²²	Daily diabetes routine	None specified	Data display	Stand alone	Positive
Skrepnik et al.⁴⁴	Pain, mood data	Wearable activity monitor	Tailored: step count, calories burned, sleep; Unspecified: motivational messages	Stand alone	Positive
Torbjørnsen et al.⁴²	Blood glucose, food habits, physical activity, personal goals	Bluetooth glucometer	Tailored: progress to goal	Coach	Null
Toro-Ramos et al.²³	Blood pressure, weight, meals, physical activity	None specified	Dashboard data display	Coach	Positive
Varnfield et al.¹⁸	Health diary, blood pressure, weight	Activity	Generic: motivational and educational materials	Coach	Positive
Waki et al.³⁵	Blood glucose, blood pressure, body weight, pedometer counts; voice/text messages about meals and exercise; photos of meals	None specified	Tailored: advice on lifestyle modification, matched to the patient’s input about food and exercise	Stand alone	Mixed

Aspects of the user interface were data input and patient messaging. Data input was characterized as patient input or via a digital device. About 15 studies (47%) described solely patient input, for example: blood glucose readings,^{22,25,29,34–36} food/caloric intake,^{15,20–23,25,29,35–38} exercise activity,^{15,19–23,25,35–39} body weight,^23,35,38 and symptom monitoring.^16,20,24 One of these studies enabled voice, text, and photo input.³⁵ Five studies^{17,19,32,40,41} (16%) described only digital input such as wireless transmission of blood pressure,^17,31 glycemic level,^19,41 weight, activity,¹⁹ or heart rate⁴⁰ readings, or medication administration.³² Nine studies (28%) described both patient and digital input.^{18,26–28,31,42–45} Examples of digital input included devices,^17,19,31,32 and sensors in^18,40 or not in cell phones.^43,44 Three studies (9%) did not provide a description of the data input.^27,30,46

Patient messaging was characterized as a generic message or display, a personalized or tailored message or display, or as unspecified as to either characteristic. About 16 studies (50%) described tailored messages solely. Examples include customized messages based on evidence or clinical targets^{15,16,19,21,26,29}; and social reinforcement and motivation based on recently input adherence levels,^31,40 and personal characteristics related to the condition (e.g. smoking).³⁰ Four studies (13%) described generic messages solely, such as reminders about the intervention.^34,38 Five studies described use of both tailored and generic messages.^{17,24,27,37,39} One study (3%) described tailored and also unspecified messages.⁴⁴ One study sent tailored emails, not mHealth messages.²⁵ Four studies did not mention messages: Three studies reported displays of user data,^20,22,23 one of which also sent tailored messages.³² One study did not describe messages.⁴⁷

Human interaction that augmented the mHealth intervention was mentioned in seven studies (22%). Clinicians, or trained coaches or mentors provided the interactions which included monitoring,^23,29,39 assessment,²⁵ support,^23,39,42 counseling,^18,23 and responding to patient questions and alerts.^18,34,39,42 One study limited coaching to responding to alerts and questions.³⁴

Description of the mHealth software was infrequent and lacked detail. Although most studies generated tailored messages, algorithms (e.g. a rules engine which generated tailored messages based on the patient data) were mentioned or described in nine studies (28%)^{16,17,19,27,29,30,35,41} including machine learning.⁴³ One study also mentioned gamification (i.e. rewards points).⁴³ One study used natural language processing of patient-entered text.³⁵

Researchers in five studies incorporated publicly available apps.^{15,20,22,23,36} One app included a commercial curriculum and human coach intervention platform.²³

Among studies that mentioned the mHealth development process, most content (logic and messages) was developed by the research team. Two studies described target audience participation.^17,32

Study outcomes

Categorization of study outcomes are shown in Table 3. One study⁴⁷ did not sufficiently describe any of the three outcome types which the study authors ascribed to attrition. Across outcome categories, more than half the studies were RCTs.

Table 3.

Characterization of retained studies’ outcomes: clinical, performance of care processes, usefulness/usage/satisfaction.

Article authors	Clinical outcomes	Performance of care processes: adherence/health care/lifestyle	Usefulness/usage/satisfaction of mHealth
Statistically significant positive difference in clinical outcomes
Adams et al.⁴⁰	Significant decrease in systolic BP maintained at 3 and 6 months; significant effect of dosage and adherence on BP	Reduced adherence by (higher) intervention dose (length of meditation); mixed adherence results	Usefulness: 94% found the app easy to use, and 96% reported it was easy to learn how to use the app
			Satisfaction: 89% of participants reported high satisfaction
Chandler et al.¹⁷	Significant decrease in systolic BP at 1, 3, 6, and 9 months; Significant decrease in diastolic BP at 3, 6, and 9 months	Greater increases in medication adherence at 1, 3, 6, and 9 months	A smaller sample of the intervention group rated overall satisfaction and usability as high
Chandler et al.¹⁷		Good protocol adherence to BP self-monitoring
Chow et al.³⁰	Intervention group had significant reductions in LDL-C, systolic blood pressure, BMI	Significant increases in physical activity and decrease in smoking	Usage: text messages were useful (91%), easy to understand (97%), and appropriate in frequency (86%)
Cook et al.²⁴	Improvement in Asthma Control Test scores and increased forced expiratory volume	Significant decrease in the number of courses of systemic steroids per patient	Usage: 6 months after the study began, 72% patients continued to use the app
Cook et al.²⁴			Satisfaction: patients found the app easy to use (93%), personalized (79%), and helpful in managing their asthma (74%)
Lee et al.⁶	Significant decreases in fat mass, weight, and BMI	Improved adherence to diet	Usefulness: majority of participants found app useful for obtaining information and managing diet process
Lee et al.⁶	Significant decreases in fat mass, weight, and BMI	Improved adherence to diet	Satisfaction: 58% agreed that the system was easy to use and the contents were interesting
Lim et al.⁴¹	Significant decrease in mean HbA1c levels in both intervention groups compared to control group; more significant decrease in app group; significantly less hypoglycemic episodes in intervention groups	Improved self-management by close and consistent supervision, prompt follow up, recommendations, consistent reminder messages	Satisfaction: 96.1% completed the study:
Liu et al.¹⁶	Significant improvements in pulmonary function; improved QoL; fewer exacerbations, less unscheduled visits	Intensive monitoring and management, increase in daily dose of corticosteriods	Usage: 72% adherent at 6 months
Martin et al.²⁸	Significant increased daily steps (in text receiving group)	Not described	Satisfaction: high satisfaction with activity tracker and text messages
Ryan et al.²⁵	Significant decrease in HbA1c	Not described	Usefulness: patient found the app simple to use (score of 8/10), and agreed with the bolus calculator suggested dose (8/10)
Ryan et al.²⁵	Significant decrease in HbA1c	Not described	Satisfaction: 66% still used app after study’s end
Sieber et al.²²	Significant decrease in HbA1c	Not described	Not described
Skrepnik et al.⁴⁴	Significant reduction in pain during walking	Increased number of steps per day, no difference in PAM score	Usefulness: satisfaction: 65.4% of patients and 67.3% of physicians would be likely or very likely to use/recommend device
Toro-Ramos et al.²³	Significant improvements in weight, and BP	Increased motivation	Usage: 80% completed the program
Varnfield et al.¹⁸	Significant difference in weight reduction, improved emotional state, and QoL at 6 weeks	No difference in 6-min walk test	Usage: higher uptake (80% vs 62%), adherence (94% vs 68%), and completion (80% vs 47%) rates
Varnfield et al.¹⁸		No difference in 6-min walk test	Satisfaction: >85% of the participants found the step counter to be motivational in reaching CR goals
Mixed clinical outcomes
Chhabra et al.⁴³	No significant difference in pain; both groups recorded a decline in disability, greater in app group; significant improvement in symptoms and general mobility	Not described	Usage: no participants discontinued intervention
Fukuoka et al.³⁷	Reduction in BP and weight loss; no significant effect on fasting lipid or glucose levels	Significant, increase in steps per day, reduction in hip circumference, intake of saturated fat and sugar-sweetened beverages in intervention group	Usefulness: usage declined over the 20-week trial but were comparable to or better than rates in similar trials
Kim et al.¹⁹	Significant reductions in HbA1c level, percentage of body fat and fasting plasma glucose in intervention group; no difference in severe hypo/hyperglycemic events, blood pressure or lipid profile between groups	No improvement of diabetes self-care activities	Satisfaction: participants were more satisfied with overall health after intervention than at baseline; app was well tolerated
Kirwan et al.³⁶	Significant decrease in HbA1c in the intervention group (no significant change in the control), no significant change in QoL	No significant change over time in self-efficacy, self-care activities	Usage: patients logged 24,720 diabetes parameters in total: 54.0% of the logs related to blood glucose levels, 33.0% to insulin, 12.0% to diet, and 1.0% to exercise
Kleinman et al.³⁴	Greater HbA1c reduction in the intervention versus control group; no difference in blood glucose levels, BP, or BMI	Improved medication adherence, blood glucose testing and communication with doctors	Usefulness: 75% of participants actively used the app at week 24
Kleinman et al.³⁴			Satisfaction: high satisfaction on all aspects of the app
Mallow et al.³¹	Significant decreases in glucose, BP, BMI, no significant reduction in weight	Not described	Not described
Ong et al.⁴⁵	Significant reduction in home BP readings, no differences in the proportions of patients in the target ranges for potassium, phosphate, or hemoglobin between baseline and exit	Not described	Usefulness: monthly adherence rates over 80%; no drop off of interest over time
Ong et al.⁴⁵		Not described	Satisfaction: all but two participants wished to continue using the app after the study; clinicians expressed satisfaction
Quinn et al.²⁹	Significant declines in HbA1c; differences were not observed for patient-reported diabetes distress, depression, diabetes symptoms, or blood pressure and lipid levels	Not described	Not described
Seto et al.²⁶	Improved QoL in app group; No significant differences in Brain natriuretic peptide (BNP) levels, left ventricular ejection fraction, or hospitalization	Improved post study self-care maintenance and management	Usage: required measurements were completed on average 5–6 days per week; by final week 89% were taking measurements at least three times per week
Waki et al.³⁵	Significant decline in HbA1c and fasting blood sugar, no significant difference in BMI, LDL, BP	No change in medication adherence and diabetes self-management	Usage: morning measurements stayed over 70%, but bedtime measurements, except for pedometer count, declined to around 50%
Waki et al.³⁵			Satisfaction: scores of 70%–100% on the Usability survey
No difference/null clinical outcomes
Agarwal et al.¹⁵	No reduction in HbA1c levels	No effect on self-efficacy, QoL, or health care utilization behaviors	Usefulness: overall, low app utilization, with a significant mean decrease over time
Agarwal et al.¹⁵	No reduction in HbA1c levels		Satisfaction: user ratings were completed by about half of participants—53% gave 4/5 stars, 39% gave 3/5 stars
Arora et al.⁴⁶	Non-significant decrease in Hba1c levels at 6 months (except for sub-group of Spanish speakers)	Increase in self-reported medication adherence, ER usage, self-efficacy, performance of self-care tasks	Usefulness: 94% of participants enjoyed the program; the majority believed program was a good way to learn about diabetes, enjoyed the program, and understood all the messages
Arora et al.⁴⁶			Satisfaction: very high satisfaction with program; 100% would recommend use to other
Isetta et al.³⁸	Not described	Regular users of APPnea (i.e. for >66% of the study) had significantly higher CPAP usage compared with less regular users (5.6_1.4 vs 4.3_1.3 h/night; p¼.008). Increased participation, improved compliance	Usefulness: 63% of participants used the app for >66% of the study period
Isetta et al.³⁸	Not described		Satisfaction: high satisfaction levels for the majority of users
Mira et al.³²	No significant difference in HbA1c, BP, and perceived health status	Increased treatment adherence. medication adherence	Usefulness: 59% reported that the app improved medication use, 30% reported partial satisfaction, 12% reported no satisfaction.
Mira et al.³²		Increased treatment adherence. medication adherence	Satisfaction: mean score of 80%
Nundy et al.²⁷	Not described	Significant improvement in self-management in maintenance and improvement in management; no sig change in self-care confidence	Usefulness: although all participants reported comfort in texting prior to enrollment, we observed low participant response rates and requests for additional training in texting
Nundy et al.²⁷	Not described		Satisfaction: although most participants reported high levels of satisfaction with the system, a few found the system difficult to use and not helpful in improving self-management
Plow and Golding²⁰	Non-significant differences between groups in physical function	Improved self-efficacy and self-regulation	Usefulness: improvements seen in both app and paper groups compared to control group
Selter et al.³⁹	Not described	Patient engagement	Usefulness: 38% completed the program; 76% found the notifications helped remember to exercise; 81% found system easy to use
Selter et al.³⁹	Not described	Patient engagement	Satisfaction: 62% rated overall experience good or excellent
Torbjørnsen et al.⁴²	No difference in HbA1c between groups after 4 months, no changes in QOL	Improvement in self-management in app group and app + counseling group compared to control group; app group exhibited improved skill and technique acquisition	Not described
Insufficiently described clinical outcomes
Goh et al.⁴⁷	Not described	Not described	Not described

About 13 studies (41% of all studies) reported statistically significant positive outcomes, for example in weight,^18,21,23,30 BP,^17,30,40 HbA1c,^22,25,41 LDL,³⁰ asthma control,²⁴ pain,⁴⁴ and pulmonary function.¹⁶ Ten studies reported performance of care processes: eight reported improved performance (e.g. Intensive monitoring and management, increase in corticosteroid daily dose¹⁶). One study reported no effect on care processes,¹⁸ and one study reported mixed results.⁴⁰ Of the 12 studies which reported user experience, usage, usefulness, and/or satisfaction was described as high.

Ten studies (31%) reported mixed study outcomes. For example, Fukuoka et al.³⁷ reported reduction in BP and weight loss, yet no significant effect on fasting lipid or glucose levels.³⁷ Among the six studies that reported performance of care processes, assessments were mixed. Three studies indicated no effect^19,35,36 (e.g. no significant change over time in self-efficacy, self-care activities³⁶) and three studies reported positive effects^26,34,37 (e.g. improved medication adherence, blood glucose testing, and communication with doctors³⁴). Seven of the eight studies that reported comparative usefulness/usage/satisfaction reported generally affirmative effects.^{19,25,26,34,35,43,45}

Eight studies (25%) reported null study outcomes.^{15,20,27,32,38,39,42,46} Five studies reported null clinical outcomes.^{15,20,32,42,46} For example, Torbjørnsen et al.⁴² reported no difference in HbA1c between groups after 4 months, and no changes in QOL. Three studies did not report clinical outcomes.^16,37,39 Six of the eight studies that reported performance of care processes described affirmative effects^{20,27,32,38,42,46} (e.g. increase in self-reported medication adherence and self-efficacy, performance of self-care tasks⁴⁶). Of the eight studies that reported usage and user satisfaction, two had affirmative effects.^38,46

Integration of study descriptors, mHealth technology features by study outcomes

In each study outcome category (positive, mixed, or null), the most often occurring clinical outcomes were blood glucose (14 studies) and blood pressure (11 studies) followed by weight (5 studies) reflecting the large number of diabetes and cardiovascular disease studies. Across outcome categories, studies reported clinical care processes such as performance of care or usefulness/usage/satisfaction as generally affirmative. Careful analysis of study features and the technology by outcome reveals several interesting findings.

Of the 13 studies with positive study outcomes the study descriptor, median study duration, was 21 weeks, slightly above the median for all retained studies (19.5). Study sample size was 60 participants, slightly below the median (62) of retained studies. Median study participant age was 50 years, slightly less than the median for retained studies.⁵¹ This analysis suggests that slightly longer but smaller sample size studies with more rigorous design in well-studied chronic illnesses in younger patients were more likely to report positive clinical outcomes.

Regarding mHealth technology features, 6 of the 13 studies used digital input: either alone or with patient input (three studies each). About 10 studies used tailored messages. Two studies used display output (and patient input). Of the three studies that used coaching, two included assessment and counseling for weight and blood pressure outcomes. Two studies used publicly available apps. One study described user participation in the mHealth development process. This synthesis suggests that digital input and tailored messages are the more successful user interfaces for supporting chronic illness management sufficiently to change clinical outcomes.

Of the 10 studies with mixed study outcomes, median study duration was 26 weeks, 6.5 weeks longer than the median for all retained studies. Study sample size of 57 participants was less than the median (62) of all studies. Median participant age for the 10 studies was 53 years, slightly above the median of retained studies. Of the studies that reported performance of care, the measure was mixed. This examination indicates studies with a design mostly similar to that of studies with positive study outcomes, with the exception of having a longer duration, were more likely to report mixed study outcomes.

Concerning mHealth features, nine studies used patient input: either alone (five studies) or with digital input (four studies). Nine studies used tailored messages. No study used display output. Two studies included coaching consisting of monitoring or replying to patients; neither included counseling. One study used a publicly available app. One study mentioned gamification with incentives for outcomes which included pain. The analysis suggests that studies that used patient input and tailored messages were more likely to reported mixed clinical outcomes.

Of the eight studies with null reported difference in study outcomes, median study duration of 13 weeks was much shorter than the median duration of all retained studies. Study sample size was 96 participants, much higher than the median (62) of retained studies. Study participant median age was 53.5, slightly above the median age of all retained studies. This exploration suggests studies of shorter duration with larger sample sizes with more rigorous design in a range of chronic illnesses in younger patients were more likely to report null outcomes.

As for mHealth technology, among the six studies which described data input, five used patient input. Five studies used tailored messages and four studies used generic messages (two studies used both). One study used a display output. Two studies included coaching for monitoring and support, and did not include counseling. Three studies used a publicly available app: one study outcome was physical function; another study did not report a clinical outcome. One study described user participation in development. This investigation suggests that studies that used patient input were more likely to report null outcomes.

Discussion

We conducted an integrative review of clinical studies of mHealth self-care applications for patients living with chronic conditions to examine the study descriptors, mHealth technology features, and study outcomes. We identified similarities and variations among the 32 retained studies, and offer suggestions as shown in Table 4. The resulting information enables researchers to identify gaps in knowledge and future avenues of inquiry as shown in Table 5.

Table 4.

Study results and related suggestions.

Study finding	Suggestion
Despite prevalence of multiple chronic conditions, all but one study focused on a single chronic condition	Development and testing of mHealth applications designed for multiple chronic conditions
Most studies investigated quantifiable clinical measures of conditions	Investigation of measures less easily quantified, for example, stress reduction
Studies lacked specifics of data input, message content, and tailored message algorithms	Provide system detail as per STAtement on the Reporting of Evaluation studies in Health Informatics (STARE-HI)⁵⁰
Differences in mHealth data input characteristics among studies	In data input design decisions, consider recruitment bias, participation burden, and the digital divide
Two studies incorporated target audience participation in development	In study design decisions, consider participatory development which has the potential to improve intervention effectiveness

Table 5.

Recommendations for future research.

Recommendation focus	Topic
Study design	Study duration sufficient for short-term and long-term clinical outcomes and patient engagement to emerge
Study design	How long patients are to remain engaged in mHealth app use
Study design	Appropriate sample size, for determining effect size
Study design	Use of participatory development to elicit and address the target audiences’ needs, using mHealth where applicable
Research question	Impact of digital input on bias related to participant recruitment, burden, adherence, engagement, and retention
Research question	Impact of human interaction on patient engagement and clinical outcomes
Research question	Equipoise for human interaction and high tech
Research question	Factors that promote user engagement in serious games

Study descriptors and mHealth characteristics similarities

All but one study focused on a single chronic condition in an adult population younger than 65 years. However, multiple chronic conditions is the most common chronic condition⁵¹ internationally.^48,49 Development and testing of mHealth self-care applications designed for multiple chronic conditions is warranted.

Among the dozen conditions targeted in the retained studies, the most common had quantifiable clinical measures such as blood glucose, blood pressure, weight, food intake, and exercise activity. Notable was an under examination of measures less easily quantified, such as stress reduction, although a few studies assessed symptom monitoring.

Description of mHealth characteristics was variable. Several studies lacked specifics of data input or messages. While most studies sent tailored messages based on data input, only some studies mentioned the software logic to determine which message to send when, and few studies described the logic. We recommend researchers provide a sufficient level of detail of the system description to permit the reader to understand how the system works, as per the STAtement on the Reporting of Evaluation studies in Health Informatics (STARE-HI).⁵⁰

Study descriptors and mHealth characteristics related to study outcomes

Among studies that reported measures, performance of care, and usefulness/usage/satisfaction outcomes were generally affirmative. These outcomes were not indicative of study outcome categorization.

Studies with positive or mixed outcomes tended to have a 20–26 week duration, an approximately 60 participant sample size, and use digital input. Whereas studies with null outcomes tended to have a shorter duration, larger sample size, and rely on patient input.

The study design differences, coupled with the wide range in study duration and sample size, highlight opportunities for future research. Questions include what is a sufficient duration for short-term and long-term clinical outcomes and patient engagement to emerge; and for how long do patients remained engaged in mHealth app use? For example, hemoglobin A1c, weight, and low-density lipoproteins change slowly: Measurement within months will not show large differences. Another question is an appropriate sample size, especially in regards to determining effect size.

Differences in mHealth data input characteristics among studies point to areas for further research. Dependence on patient input may introduce recruitment bias among older people uncomfortable with the technology. Furthermore, the participation burden of patient data input²⁴ coupled with the new hurdle of unfamiliar technology use may impact participant adherence, engagement, and retention. Digital input use, such as sensors or voice-activated data input⁵¹ may address these challenges: another avenue of future research. However, use of digital input technologies which require capabilities that may be unavailable to vulnerable populations²⁷ or people living in under-resourced communities⁴⁶ may introduce recruitment bias.

Some studies augmented technology with human interaction such as coaching (“high-touch”). Studies with positive outcomes that included coaching offered counseling, a feature not included in studies with mixed or null study outcomes. Questions of interest are whether high-touch leads to better patient engagement¹⁵ and clinical outcomes; and what is the high-touch/high tech equipoise? Prevalent use of tailored messages and infrequent use of coaching among studies, regardless of clinical outcome categorization, suggests these questions warrant further research.

mHealth development characteristics included build-or-buy approaches and participatory development. While most studies built mHealth interventions, only two incorporated target audience participation in development, with differing study outcomes. Participatory development has the potential to improve intervention effectiveness.⁵² Its infrequent use among the studies raises the question whether researchers are implementing in mHealth existing approaches while expecting better outcomes, rather than eliciting and addressing the target audiences’ needs and using mHealth where applicable.

The few studies that used the buy-approach, incorporating commercially available apps, had diverse study outcomes. This finding may suggest a lack of app effectiveness, indicative of the lack of formal assessment, regulatory review,⁵³ and testing outside of a small, relatively homogeneous population.¹⁵

An mHealth design feature mentioned in one study with mixed outcomes was gamification, a method that may address the challenge of initiating and maintaining patient engagement. DeSmet et al.’s⁵⁴ meta-analysis of serious games for healthy lifestyle promotion found small positive effects on healthy lifestyles, their determinants, and clinical outcomes. However, most games in the meta-analysis were designed for children and young adults, a younger population compared to those living with chronic conditions. Edwards et al.’s⁵³ systematic review of publicly available games for healthy behavior change found no relation between game content and user ratings. This finding indicates that the question of what promotes user engagement remains unanswered among gaming scientists, and is an opportunity for future research.

Practical implications

Those who use or intend to use mHealth self-care applications designed for patients living with multiple chronic conditions should keep in mind that these applications could benefit from further development and testing. Researchers should be aware that these applications often are inadequately described in the studies, that appropriate sample size for effect size estimation remains undetermined, and recruitment bias may not be accounted for in completed studies.

Limitations

Characteristic of literature reviews, this review does not claim to be comprehensive nor to fully implement PRISMA guidelines given the more targeted scope of the paper. Instead, it summarizes the research on mHealth for patient chronic illness self-management based on the search terms used, the databases included, and the review time period. That we used a medical librarian with extensive literature search experience and transparently reported the data acquisition process provides evidence for the dataset’s limitations and scope. Also, we acknowledge that positive results publication bias may have limited the studies available in the databases searched. We did not exclude studies that omitted power analysis or appeared underpowered, nor change authors’ study design classification to reflect sample size. As this was an integrative review, we did not report quality assessment which is required only in systematic reviews and meta-analyses.

Conclusions

We investigated recent chronic condition mHealth interventions experimental studies with a focus on technology features related to clinical outcomes. We offer suggestions about study duration and sample size, and data input and messaging design decisions. We provide these suggestions to advance mHealth science and to future technology developers and researchers for their consideration.

Footnotes

Acknowledgements

The research librarian, Janice Masud-Paul, for assistance with development of the search strategy and retrieval of articles from databases.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Paulina S Sockolow

References

Riley

Cowan

. Noncommunicable diseases country profiles 2014. World Health Organization, Geneva, 2014.

Bodenheimer

Chen

Bennett

. Confronting the growing burden of chronic disease: can the U.S. Health Care Workforce do the job? Health Aff (Millwood) 2009; 28(1): 64–74.

Van Hecke

Heinen

Fernández-Ortega

, et al. Systematic literature review on effectiveness of self-management support interventions in patients with chronic conditions and low socio-economic status. J Adv Nurs 2017; 73(4): 775–793.

Kay

Santos

Takane

. mHealth: new horizons for health through mobile technologies. World Health Organization, Geneva, 2011.

Matthew-Maich

Harris

Ploeg

, et al. Designing, implementing, and evaluating mobile health technologies for managing chronic conditions in older adults: a scoping review. JMIR Mhealth Uhealth 2016; 4(2): e29.

Lee

J-A

Choi

Lee

, et al. Effective behavioral intervention strategies using mobile health applications for chronic disease management: a systematic review. BMC Med Inform Decis Mak 2018; 18(1): 12.

Whitehead

Seaton

. The effectiveness of self-management mobile phone and tablet apps in long-term condition management: a systematic review. J Med Internet Res 2016; 18(5): e97.

Institute of Medicine Committee on the Crossing the Quality Chasm: Next Steps Toward a New Health Care Study, Adams

Greiner

Corrigan

. The 1st annual crossing the quality chasm summit: a focus on communities. Washington, DC: National Academies Press (US), 2004.

Monroe

Thompson

Bassett

, et al. Usability of mobile phones in physical activity–related research: a systematic review. Am J Health Educ 2015; 46(4): 196–206.

10.

McCarroll

Eyles

Ni Mhurchu

. Effectiveness of mobile health (mHealth) interventions for promoting healthy eating in adults: a systematic review. Prev Med 2017; 105: 156–168.

11.

Whittemore

Chao

Jang

, et al. Methods for knowledge synthesis: an overview. Heart Lung 2014; 43(5): 453–461.

12.

Buck

Shadmi

Topaz

, et al. Theoretical examination of chronic illness mHealth interventions. Res Nurs Health 2021; 44(47): 59.

13.

Wong

Greenhalgh

Westhorp

, et al. RAMESES publication standards: realist syntheses. BMC Med 2013; 11(1): 21.

14.

Pawson

Greenhalgh

Harvey

, et al. Realist review—a new method of systematic review designed for complex policy interventions. J Health Serv Res Policy 2005; 10(Suppl 1): 21–34.

15.

Agarwal

Mukerji

Desveaux

, et al. Mobile app for improved self-management of type 2 diabetes: multicenter pragmatic randomized controlled trial. JMIR Mhealth Uhealth 2019; 7(1): e10321.

16.

Liu

W-T

Huang

C-D

Wang

C-H

, et al. A mobile telephone-based interactive self-care system improves asthma control. Eur Respir J 2011; 37(2): 310–317.

17.

Chandler

Sox

Kellam

, et al. Impact of a culturally tailored mHealth medication regimen self-management program upon blood pressure among hypertensive Hispanic adults. Int J Environ Res Public Health 2019; 16(7): 1226.

18.

Varnfield

Karunanithi

Lee

C-K

, et al. Smartphone-based home care model improved use of cardiac rehabilitation in postmyocardial infarction patients: results from a randomised controlled trial. Heart 2014; 100(22): 1770–1779.

19.

Kim

Kwak

Jung

, et al. The effect of a smartphone-based, patient-centered diabetes care system in patients with type 2 diabetes: a randomized, controlled trial for 24 weeks. Diabetes Care 2019; 42(1): 3–9.

20.

Plow

Golding

. Using mHealth technology in a self-management intervention to promote physical activity among adults with chronic disabling conditions: randomized controlled trial. JMIR Mhealth Uhealth 2017; 5(12): e185.

21.

Lee

Chae

Kim

, et al. Evaluation of a mobile phone-based diet game for weight control. J Telemed Telecare 2010; 16(5): 270–275.

22.

Sieber

Flacke

Link

, et al. Improved glycemic control in a patient group performing 7-point profile self-monitoring of blood glucose and intensive data documentation: an open-label, multicenter, observational study. Diabetes Ther 2017; 8(5): 1079–1085.

23.

Toro-Ramos

Kim

Wood

, et al. Efficacy of a mobile hypertension prevention delivery platform with human coaching. J Hum Hypertens 2017; 31: 795–800.

24.

Cook

Modena

Simon

. Improvement in asthma control using a minimally burdensome and proactive application. J Allergy Clin Immunol Pract 2016; 4(4): 730–737.e1.

25.

Ryan

Holland

Stroulia

, et al. Improved A1C levels in type 1 diabetes with smartphone app use. Can J Diabetes 2017; 41(1): 33–40.

26.

Seto

Leonard

Cafazzo

, et al. Mobile phone-based telemonitoring for heart failure management: a randomized controlled trial. J Med Internet Res 2012; 14(1): e31.

27.

Nundy

Razi

Dick

, et al. A text messaging intervention to improve heart failure self-management after hospital discharge in a largely African-American population: before-after study. J Med Internet Res 2013; 15(3): e53.

28.

Martin

Feldman

Blumenthal

, et al. mActive: a randomized clinical trial of an automated mHealth intervention for physical activity promotion. J Am Heart Assoc 2015; 4(11): e002239.

29.

Quinn

Shardell

Terrin

, et al. Cluster-randomized trial of a mobile phone personalized behavioral intervention for blood glucose control. Diabetes Care 2011; 34(9): 1934–1942.

30.

Chow

Redfern

Hillis

, et al. Effect of lifestyle-focused text messaging on risk factor modification in patients with coronary heart disease: a randomized clinical trial. JAMA 2015; 314(12): 1255–1263.

31.

Mallow

Theeke

, et al. The effectiveness of mI SMART: a nurse practitioner led technology intervention for multiple chronic conditions in primary care. Int J Nurs Sci 2018; 5(2): 131–137.

32.

Mira

Navarro

Botella

, et al. A Spanish pillbox app for elderly patients taking multiple medications: randomized controlled trial. J Med Internet Res 2014; 16(4): e99.

33.

Beebe

. Rapid assessment process: an introduction. New York: Altamira Press, 2001.

34.

Kleinman

Shah

, et al. Improved medication adherence and frequency of blood glucose self-testing using an m-Health platform versus usual care in a multisite randomized clinical trial among people with type 2 diabetes in India. Telemed E Health 2017; 23(9): 733–740.

35.

Waki

Fujita

Uchimura

, et al. DialBetics: a novel smartphone-based self-management support system for type 2 diabetes patients. J Diabetes Sci Technol 2014; 8(2): 209–215.

36.

Kirwan

Vandelanotte

Fenning

, et al. Diabetes self-management smartphone application for adults with type 1 diabetes: randomized controlled trial. J Med Internet Res 2013; 15(11): e235.

37.

Fukuoka

Gay

Joiner

, et al. A novel diabetes prevention intervention using a mobile app. Am J Prev Med 2015; 49(2): 223–237.

38.

Isetta

Torres

González

, et al. A new mHealth application to support treatment of sleep apnoea patients. J Telemed Telecare 2017; 23(1): 14–18.

39.

Selter

Tsangouri

Ali

, et al. An mHealth app for self-management of chronic lower back pain (Limbr): pilot study. JMIR Mhealth Uhealth 2018; 6(9): e179.

40.

Adams

Sieverdes

Brunner-Jackson

, et al. Meditation smartphone application effects on prehypertensive adults’ blood pressure: dose-response feasibility trial. Health Psychol 2018; 37(9): 850–860.

41.

Lim

Kang

Shin

, et al. Improved glycemic control without hypoglycemia in elderly diabetic patients using the ubiquitous healthcare service, a new medical information system. Diabetes Care 2011; 34(2): 308–313.

42.

Torbjørnsen

Jenum

Småstuen

, et al. A low-intensity mobile health intervention with and without health counseling for persons with type 2 diabetes, part 1: baseline and short-term results from a randomized controlled trial in the Norwegian part of RENEWING HEALTH. JMIR Mhealth Uhealth 2014; 2(4): e52.

43.

Chhabra

Sharma

Verma

. Smartphone app in self-management of chronic low back pain: a randomized controlled trial. Eur Spine J 2018; 27(11): 2862–2874.

44.

Skrepnik

Spitzer

Altman

, et al. Assessing the impact of a novel smartphone application compared with standard follow-up on mobility of patients with knee osteoarthritis following treatment with Hylan G-F 20: a randomized controlled trial. JMIR Mhealth Uhealth 2017; 5(5): e64.

45.

Ong

Jassal

Miller

, et al. Integrating a smartphone-based self-management system into usual care of advanced CKD. Clin J Am Soc Nephrol 2016; 11(6): 1054–1062.

46.

Arora

Peters

Burner

, et al. Trial to examine text message-based mHealth in emergency department patients with diabetes (TExt-MED): a randomized controlled trial. Ann Emerg Med 2014; 63(6): 745–54.e6.

47.

Goh

Tan

Malhotra

, et al. Short-term trajectories of use of a caloric-monitoring mobile phone app among patients with type 2 diabetes mellitus in a primary care setting. J Med Internet Res 2015; 17(2): e33.

48.

Tinetti

Fried

Boyd

. Designing health care for the most common chronic condition—multimorbidity. J Am Med Assoc 2012; 307(23): 2493–2494.

49.

Ofori-Asenso

Chin

Curtis

, et al. Recent patterns of multimorbidity among older adults in high-income countries. Popul Health Manag 2019; 22(2): 127–137.

50.

Talmon

Ammenwerth

Geven

. The quality of reporting of health informatics evaluation studies: a pilot study. Stud Health Technol Inform 2007; 129(Pt 1): 193–197.

51.

Munos

Baker

Bot

, et al. Mobile health: the power of wearables, sensors, and apps to transform clinical trials. Ann N Y Acad Sci 2016; 1375(1): 3–18.

52.

McCurdie

Taneva

Casselman

, et al. mHealth consumer apps: the case for user-centered design. Biomed Instrum Technol 2012; 46(Suppl 2): 49–56.

53.

Edwards

Lumsden

Rivas

, et al. Gamification for health promotion: systematic review of behaviour change techniques in smartphone apps. BMJ Open 2016; 6(10): e012447.

54.

DeSmet

Van Ryckeghem

Compernolle

, et al. A meta-analysis of serious digital games for healthy lifestyle promotion. Prev Med 2014; 69: 95–107.