The Feasibility and Effects of Smartphone-Based Application on Cardiac Rehabilitation for Patients After Percutaneous Coronary Intervention

A Randomized Controlled Trial

Abstract

Background 

Cardiac rehabilitation (CR) referral and participation are generally found to be underused worldwide. The data in Asian regions suggest significant underutilization compared with western countries.

Objectives 

To promote the continuity and effects of CR, this study aimed at supporting self-care and improving CR effects in phase II CR with mobile health application for Macao patients who have undergone percutaneous coronary intervention.

Methods 

The study was a randomized controlled trial conducted between January 2018 and June 2020 in a nongovernmental Hospital. Social cognitive theory was applied in the intervention for an experimental group. Measurements including anxiety and depression level, medication adherence, exercise capacity, physical activity level, cardiovascular risk modification, self-efficacy, quality of life, use of healthcare resources, mHealth CR utility, and satisfaction were analyzed.

Results 

One hundred forty eligible patients were recruited, and 110 patients completed the study. Significant interventional effects of the experimental group were found on the reduction of total cholesterol, low-density lipoprotein, sitting time, anxiety level, and total anxiety and depression level (P < .05). In addition, significant improvement on the 6-minute walk test, regular exercise performance, self-efficacy of cardiac exercise and diet, and quality of life were found as interventional effects in the experimental group (P < .05). The feasibility of intervention was achieved in a satisfactory level.

Conclusions 

This theoretically grounded mobile health CR study was a unique program tailor-made to meet the care needs of patients post percutaneous coronary intervention. It was feasible and effective to support CR effects and adherence and would inform further design of clinical practice and public policies.

Coronary heart disease (CHD) kills an estimated 7.8 million people each year globally—in 2020, it was the first leading cause (41.2%) of death attributable to cardiovascular disease.¹ In China, the mortality rate of cardiovascular disease was 364.5 per 100 000 population in 2019 and cardiovascular disease accounted for over 40% of deaths.² In Hong Kong, heart disease is the third leading cause of death, with CHD being the dominant component holding 66.6% of deaths in 2019, with nearly 10.2 persons dying from CHD each day.^3,4 Similarly, heart disease was the second leading cause of death in 2019, and CHD mortality accounted for 75% of total heart disease mortality in Macao.^5,6 With the continuous advances in myocardial revascularization, percutaneous coronary intervention (PCI) with coronary stenting has become one of the most frequently used therapeutic procedures to improve coronary blood flow and restore myocardial reperfusion.^7,8

Cardiac rehabilitation (CR) has been recognized as a quality indicator of post-PCI care.^9,10 Electronic alternative CR, supported by electronic health or mobile health (mHealth), provides a new strategy to promote access, adherence, and effectiveness. The use of mHealth has been advocated to be more effective than center-based CR delivery.¹¹ However, important program effects on self-efficacy for CHD risk modification and cardiac-related clinical outcomes have been weakly addressed in literature for patients with PCI. Few study authors have applied a theoretical framework to guide the development and understanding of the mechanism of CR effects, such as the social cognitive theory and behavior change theory.^12,13 Regarding the local service gap in Macao, cardiac clinicians should provide seamless service to support patients with PCI for transitional care during phase II CR.

The Role of Cardiac Rehabilitation After Percutaneous Coronary Intervention

It is regarded that CR is a class I–level A recommendation for patients with stable angina, with previous myocardial infarction, after PCI, and after coronary artery bypass grafting, which should be initiated in the hospital and continued after a stability change in the acute phase of cardiac condition within 1 to 3 weeks.^7,9,10 Individualized and interprofessional care includes physical exercise, optimization of pharmacotherapy, patient education, psychological support, CHD risk factor reduction, and lifestyle modification.^9,10 The benefits of CR after PCI are recognized in reducing the risk of reinfarction and cardiac mortality, improving quality of life and exercise capacity, and modifying risk factors.^14,15

Application of Electronic Technology During Cardiac Rehabilitation

Electronic health and mHealth in CR are advanced and innovative strategies aimed at enhancing CR utilization, compliance, and effectiveness. The World Health Organization defines electronic health as electronic communication and health information technology in healthcare. A component of electronic health, mHealth, is specifically used in mobile phones, monitoring devices, personal digital assistants, and other wireless devices.¹⁶

Three core features of electronic health and mHealth CR (mCR) have been identified in the literature: patient education, task reminders, and health data self-entry.^{12,13,17–20} A significant amount of literature has implemented patient education as a key component of electronic CR, including physical exercise, healthy diet, cardiovascular risk modification, and psychosocial support.^{12,13,17–20} Patient data self-entry has been identified as a method of motivating self-care monitoring, and some studies contained task reminders for healthy lifestyle modification as well as self-care monitoring.^12,17,18 In addition, some programs specifically applied portable or remote devices to facilitate the safety and motivation of exercise, because improvement of exercise is crucial for achieving the positive effects of CR.^12,20

The CR studies, integrated Internet and mobile phones, reflected 40% of their favorable outcomes.^12,18 Conversely, smartphone-based mCR studies achieved more than three-fifths of outcomes as positive.^13,17,19 The mCR, which has the advantages of being more accessible and convenient, would be more effective than an Internet-based CR.¹¹ The mCR was supported by interactions via text messages, remote monitoring in healthy behavior change, and self-monitoring feedback in personal health outcomes.

Gaps in the literature indicated that few studies were guided by the theoretical framework, while social cognitive theory and behavior change theory were applied.^12,13 Outcome measures of self-efficacy in CHD risk modification, healthy lifestyle changes, intervention compliance, and cardiac-related adverse events have rarely been evaluated. To fill the gap in local services whereby no governmental or private community healthcare center is able to support phase II CR in Macao, it leaves room for clinical consideration to conduct a theoretically grounded experimental mHealth study in the early stage of phase II CR for Chinese patients with PCI.

Purpose

The aim of this study was to assess the effectiveness of mCR in mobile applications (apps), compared with usual care, on anxiety and depression levels, exercise capacity, physical activity level, cardiovascular risk modification, self-efficacy, quality of life, medication compliance, and healthcare utility among patients with PCI in Macao. Intervention fidelity and acceptability were evaluated by the mCR app utility and satisfaction level of the experimental group participants.

Materials and Methods

Design

This was a single-blinded, 2-arm, randomized controlled trial following the Consolidated Standards of Reporting Trials guidelines.

Setting

The study was conducted between January 2018 and June 2020 at a Macao nongovernmental charitable hospital.

Participants

Sampling Criteria

The inclusion criteria were as follows: (1) adult Chinese patients given a diagnosis of CHD who underwent PCI, (2) eligible to initiate phase I CR during hospitalization and continue phase II CR at the outpatient department, and (3) possession of a smartphone. The exclusion criteria were as follows: (1) diagnosis of a mental or psychiatric disorder, (2) having repeated revascularization or post coronary artery bypass grafting, (3) remarkable physical exercise limitations other than heart disease, and (4) postcardiac device implantation.

Randomization

The study participants were referred by cardiologists for enrollment according to the sampling criteria, and the researcher assessed the eligibility of sample recruitment. Eligible patients were randomized with a block size of 4 in an opaque envelope and allocated to either the experimental group or control group according to their entry sequence of allocation.

Sample Size Determination

Lower self-efficacy in cardiac exercise is associated with higher levels of depression among Chinese patients with CHD.²¹ Anxiety and depression were considered the primary outcomes in this study. A sample size of 140 was expected to provide at least 80% power at a 5% significance level (2-sided) to detect differences. A medium effect size of 0.44 achieved a significant reduction in anxiety level by the Hospital Anxiety and Depression Scale.¹² Considering an attrition rate of 20%,^17,19 the target sample size of 70 in each group was determined.

Development of Mobile Health Cardiac Rehabilitation Program

The CR team was composed of cardiologists, cardiac advanced practice nurses, physiotherapists, dietitians, and clinical psychologists. The CR educational material was developed by the authors and validated by 5 clinical CR experts.

The use of the mCR app for the experimental group and CR booklets for the control group was allocated according to the group assignment. The study intervenors were 2 cardiac advanced practice nurses, and intervention consistency was achieved through training in intervention delivery. At the fourth and ninth weeks, 2 telephone calls were given to all participants to assess and encourage the use of the assigned mCR app or CR booklet throughout the study. For midterm and final evaluation at the 6th and 12th weeks, all participants attended 2 sessions of individual face-to-face nurse-led clinic follow-ups (Figure 1). To improve data collection quality, face-to-face interviews were conducted by 2 valid research assistants who were blinded to the study allocation.

FIGURE 1: Intervention protocol.

Interventional Care: Experimental Group With the Mobile Health Cardiac Rehabilitation Application

Social cognitive theory was adopted for experimental care to guide the mechanism of intervention. It posits a triadic interaction of environmental, personal, and behavioral factors, in which self-efficacy works together with goals, outcome expectations, and perceived environmental impediments, and facilitators in the regulation of human motivation, behavior, and well-being.²² Human function was explained, whereas perceived self-efficacy in adhering to CR recommendations influences health functioning and personal behavioral change.¹²

The mCR app was installed on the smartphone of experimental group participants by study intervenors during the first week after hospital discharge. Cloud hosting of data storage was implemented to ensure data confidentiality. The 4 major features developed in the mCR app are the following:

• Educational page: all CR teaching materials with text, pictures, photos, and videos, covering 5 teaching topics: (1) CHD knowledge and risk factor modification; (2) nursing care for CHD self-care, medication adherence skills, and chest pain management; (3) physical exercise for cardiac health; (4) CHD healthy diet; and (5) stress relation skills and psychological support.
• Recording page: the experimental group participants were instructed to record their health and behavior modification, including blood pressure, pulse, and fasting blood glucose every day. Weight, smoking status, alcohol consumption, medication status, and exercise status were recorded once a week. Push notifications of medication use and health data entry were also supported.
• Result page: all data collected from experimental group participants' entries were summarized and analyzed. The health data presented as “weekly health status review” can be displayed as a table or chart in the mCR app for health status tracking and can be printed by participants.
• Question-and-answer page: the one-on-one chartroom facilitated interaction and communication between the patients and cardiac advanced practice nurses.

Usual Care: Control Group With Booklet

Patients in the control group received a CR booklet for their usual care. It contained the same educational content and written logbooks as the mCR app. A written logbook requiring the same items and frequency for health recording was attached to the booklet. A hotline number was provided for nonemergency inquiries.

Data Collection

Nine outcome indicators were collected at different time points: time 0, baseline (at hospital discharge); time 1, first follow-up at the nurse-led clinic (6 weeks after hospital discharge); and time 2, the last follow-up at the nurse-led clinic (12 weeks after hospital discharge). Background demographic and baseline data were collected at time 0. Medication compliance and healthcare utilization were obtained at time 1 and time 2, respectively. The utility and satisfaction of the mCR app were assessed at time 2. All other outcomes were measured at 3 different time points. Hospital-necessary services, including CR, continued during the beginning of the COVID-19 pandemic in Macao.

To assess anxiety and depression levels, the Chinese-Cantonese version of the Hospital Anxiety and Depression Scale, containing 14 items, was used.²³ Good internal consistency (Cronbach α = 0.86) and concurrent validity (r = 0.63–0.67) were obtained. This scale was evaluated as an indicator of CR in Chinese patients with CHD in Hong Kong.²⁴

Six-minute walk test is a sensitive assessment of functional exercise capacity that is responsive to clinical changes after CR.^7,25 It is a simple, economical, and reliable standardized field test that assesses exercise intolerance by measuring the distance that a patient walks on a flat, hard surface for 6 minutes.²⁵ The short form of the Chinese version of the International Physical Activity Questionnaire (IPAQ-C) was used to assess participants' physical activity level. The test-retest reliability was 0.76, concurrent validity was 0.58, and criterion validity was 0.30.²⁶ The short IPAQ-C was determined to be reliable with an intraclass correlation coefficient of 0.79 and acceptable concurrent validity (r = 0.29) among the Chinese population in Hong Kong.²⁷

Cardiovascular risk factors, including resting blood pressure, pulse, body mass index, smoking status, alcohol consumption, lipid profile, fasting blood glucose, and hemoglobin A_1C supported by literature review and international guidelines,²⁸ were measured at the nurse-led clinic. Fasting blood glucose, hemoglobin A_1C, and lipid profiles, including total cholesterol, high-density lipoprotein, low-density lipoprotein, and triglycerides, were examined at the 12th week. Data regarding weekly smoking status and alcohol consumption were collected via the mCR app or a written logbook.

To examine the self-efficacy of cardiac exercise and diet among CR participants, the Cardiac Exercise Self-Efficacy Instrument–Chinese version and the Cardiac Diet Self-Efficacy Instrument–Chinese version were used.²⁹ The Cardiovascular Limitations and Symptoms Profile is a reliable and valid instrument for assessing quality of life in patients with CHD and comprises 37 items, including 4 symptom subscales (angina, shortness of breath, ankle swelling, and tiredness), as well as 5 limitation subscales (mobility, social life and leisure activities, home activities, concerns and worries, and sex). The Chinese version of Cardiovascular Limitations and Symptoms Profile has been validated among Hong Kong Chinese patients with CHD.³⁰

According to the post-PCI international treatment guidelines,⁷ dual antiplatelet therapy, β-blockers, angiotensin II receptor blockers, angiotensin-converting enzyme inhibitors, and statins are the principles of post-PCI medication treatment. The pill count was used as objective data, which was efficient and effective in evaluating medication compliance.¹⁷

Evidence indicates that mCR benefits cardiac-related healthcare utilization from reducing cardiac-related rehospitalization and emergency visits.^13,20 Therefore, 30- and 90-day cardiac-related rehospitalizations and emergency visits were collected through the “electronic health record system” supported by the Macao Health Bureau.

The frequency of mCR app utility was evaluated using a backend administration platform. The satisfaction level of using the mCR app was measured through a horizontal 100-mm visual analog scale.³¹ The contents of the satisfaction evaluation of the mCR app were modified according to the recommendation.³²

Ethical Consideration

Ethical approval was obtained following the principles of medical research ethics. A full description of the study was provided to each study participant, and they could decide to voluntarily participate in the study or terminate their participation at any time without affecting their care. When participants fully understood their rights and data confidentiality, they were asked to provide written informed consent. This study was registered in the Chinese Clinical Trial Registry (no. ChiCTR1800014521).

Data Analysis

Descriptive statistics were used to describe participants' demographic data and clinical characteristics. The independent sample t test, Mann-Whitney U test, or χ² test was used to examine the heterogeneity of data. The generalized estimating equation model was used to examine the effects of the intervention across different time points.³³ The benefits of generalized estimating equation model include accounting for within-subject correlations, allowing for time-varying covariates and time-invariant covariates, and irregularly timed or mistimed measurements.³³

Results

Sample Recruitment and Attrition

Figure 2 indicated that 180 patients with CR were screened, and 140 eligible patients were randomized into either the experimental group (n = 70) and the control group (n = 70). The recruitment rate was 77.8%, and 125 patients completed the study. The overall attrition rate was 10.7%. There were no differences in the characteristics of the total sample that was lost to follow-up (Supplemental Appendix I, https://links.lww.com/JCN/A204).

FIGURE 2: Consolidated Standards of Reporting Trials flowchart of participant recruitment and group allocation in the mobile health cardiac rehabilitation study.

Characteristics of Participants

The median age of participants was 68.0 (interquartile range, 65.0–72.0) years in the experimental group and 70.0 (interquartile range, 65.0–73.0) years in the control group. Most participants were male (78%) and married (98%), had a monthly income of <USD $1000.00 (72%), and received government-subsidized medical payments (88.8%). Hypertension (70%) and type 2 diabetes mellitus (38.4%) were the 2 most common CHD risk comorbidities among the participants. Ischemic heart disease (56%) and acute myocardial infarction (29.6%) were the primary diagnoses of CHD. More than half of the participants (70%) underwent PCI with ≤2 stents. All participants received dual antiplatelet therapy and statin medications after the procedure. The median ejection fraction was 66.5 (interquartile range, 62.0–74.0) in the experimental group and 68.0 (interquartile range, 63.0–75.0) in the control group. There was no statistically significant difference (P > .05) in any of the background data between the groups (Table 1).

TABLE 1 - Demographic Background Data of Participants by Study Group (N = 140)

Variables	Experimental Group (n = 70)	Control Group (n = 70)	P
Age, ya	68.0 (65.0–72.0)	70.0 (65.0–73.0)	.137
Gender			.865
Male	49 (79.0%)	49 (77.8%)
Female	13 (21.0%)	14 (22.2%)
Marital status			.991
Married	61 (98.4%)	62 (98.4%)
Single or others	1 (1.6%)	1 (1.6%)
Educational level			.228
Primary or lower	27 (43.5%)	35 (55.6%)
Secondary	31 (50.0%)	22 (34.9%)
Diploma or higher	4 (6.5%)	6 (9.5%)
Monthly income, USD			.322
$100–$500	17 (27.4%)	27 (42.9%)
$500–$1000	26 (41.9%)	20 (31.7%)
$1000–$2500	13 (21.0%)	12 (19.0%)
≥$2500	6 (9.7%)	4 (6.3%)
Medical payment method			.126
Government subsidy	52 (83.9%)	59 (93.7%)
Self-pay	4 (6.5%)	3 (4.8%)
Insurance	6 (9.7%)	1 (1.6%)
CHD risk comorbidity
Hyperlipidemia	16 (25.8%)	15 (23.8%)	.796
Hypertension	42 (67.7%)	45 (71.4%)	.654
Type 2 diabetes	19 (30.6%)	29 (46.0%)	.051
CHD primary diagnosis
Unstable angina/angina	9 (14.5%)	14 (22.2%)	.266
Myocardial infarction	22 (35.5%)	15 (23.8%)	.153
Ischemic heart disease	32 (51.6%)	38 (60.3%)	.327
Percutaneous coronary intervention with stenting			.780
1 stent	25 (40.3%)	27 (42.9%)
2 stents	18 (29.0%)	17 (27.0%)
3 stents	13 (21.0%)	10 (15.9%)
≥4 stents	6 (9.7%)	9 (14.3%)
Medication treatment
Dual antiplatelet therapy	62 (100%)	63 (100%)	—
β-Blockers	37 (59.7%)	39 (61.9%)	.799
ACEI/ARBs	34 (54.8%)	34 (54.0%)	.922
Statins	62 (100%)	63 (100%)	—
PCSK9 inhibitors	16 (25.8%)	19 (30.2%)	.588
Ejection fractiona	66.5 (62.0–74.0)	68.0 (63.0–75.0)	.289

All data are presented in count (percentages) and compared by χ² or Fisher exact test.

Abbreviations: ACEI, angiotensin-converting enzyme inhibitor; ARBs, angiotensin II receptor blockers; CHD, coronary heart disease; PCSK9, proprotein convertase subtilisin/kexin type 9; USD, United States dollar.

^aPresented in median (interquartile range) and compared by Mann-Whitney U test.

Clinical Characteristics and Outcomes Compared by Study Groups

Table 2 shows the descriptive statistics of the clinical, physio-psychological, behavioral, and quality of life outcomes across the time points between groups. There was no significant difference at baseline between the experimental and control groups. Significant between-group differences favoring the experimental group were observed at the sixth week in improving the total IPAQ-C (P = .046), regular exercise performance (P = .044), the score of Cardiac Exercise Self-Efficacy Instrument–Chinese version (P = .015), and self-care compliance (P = .007). Significant improvements on the six-minute walk test (P = .046), moderate activity in IPAQ-C (P < .001), the total IPAQ-C (P = .027), performance of regular exercise (P < .001), the score of Cardiac Exercise Self-Efficacy Instrument–Chinese version (P < .001), the score of Cardiac Diet Self-Efficacy Instrument–Chinese version (P < .001), self-care compliance (P = .003), and reductions in the Chinese version of the Hospital Anxiety and Depression Scale: depression (P = .002) and total score (P = .020) were shown at the 12th week between the groups. However, the remaining outcomes were not significantly different between the 6th and 12th week in the between-group comparison.

TABLE 2 - Comparison of Clinical Outcomes by Study Group (N = 140)

Clinical Characteristics	Experimental Group (n = 70)	Control Group (n = 70)	P
Systolic blood pressure, mm Hga
Baseline	131.87 (15.59)	129.92 (16.26)	.495
Sixth week (time 1)	129.18 (17.89)	132.82 (16.10)	.279
12th week (time 2)	129.89 (16.20)	131.09 (17.98)	.702
Diastolic blood pressure, mm Hga
Baseline	74.06 (10.43)	70.90 (10.30)	.091
Sixth week (time 1)b	75.27 (13.56)	73.43 (8.78)	.406
12th week (time 2)	74.80 (9.93)	72.62 (9.75)	.229
Pulse, beats/min
Baselineb	71.60 (13.32)	70.32 (11.05)	.560
Sixth week (time 1)a	72.75 (9.65)	74.16 (10.89)	.482
12th week (time 2)a	71.13 (9.58)	72.57 (10.17)	.429
Total cholesterol, mmol/La
Baseline	4.25 (1.06)	4.10 (1.06)	.467
12th week (time 2)	3.23 (0.68)	3.51 (1.10)	.104
Triglycerides, mmol/Lb
Baseline	1.39 (1.05–2.06)	1.31 (1.05–1.99)	.978
12th week (time 2)	1.14 (0.94–1.61)	1.23 (0.96–1.65)	.510
High-density lipoprotein, mmol/La
Baseline	1.08 (0.26)	1.11 (0.26)	.455
12th week (time 2)	1.18 (0.26)	1.22 (0.31)	.525
Low-density lipoprotein, mmol/La
Baseline	2.75 (1.03)	2.58 (1.03)	.355
12th week (time 2)	1.65 (0.58)	1.87 (0.74)	.066
Fasting blood glucose, mmol/Lb
Baseline	5.80 (5.30–6.90)	6.00 (5.30–7.00)	.662
12th week (time 2)	6.00 (5.50–6.70)	6.20 (5.50–6.80)	.825
Hemoglobin A1C, %b
Baseline	5.90 (5.60–6.60)	6.00 (5.60–6.80)	.567
12th week (time 2)	6.00 (5.70–6.30)	6.00 (5.70–6.40)	.882
Smoking status: nonsmoking
Baseline	56 (90.3%)	51 (81.0%)	.136
Sixth week (time 1)	56 (91.8%)	48 (84.2%)	.202
12th week (time 2)	55 (90.2%)	49 (83.1%)	.252
Alcohol consumption status: 1–2 standard drinks/d
Baseline	70 (100%)	70 (100%)	1.000
Sixth week (time 1)	67 (100%)	66 (100%)	1.000
12th week (time 2)	62 (100%)	63 (100%)	1.000
Weight, kga
Baseline	66.36 (8.75)	67.91 (10.76)	.380
Sixth week (time 1)	65.71 (9.08)	66.71 (10.47)	.768
12th week (time 2)	65.40 (8.83)	67.17 (9.94)	.307
Body mass indexa
Baseline	24.63 (2.95)	25.39 (3.42)	.188
Sixth week (time 1)	24.48 (3.03)	25.36 (3.60)	.175
12th week (time 2)	24.28 (2.94)	25.27 (3.36)	.091
Six-minute walk testa
Baseline	355.40 (80.18)	350.29 (92.62)	.746
Sixth week (time 1)	442.72 (80.08)	414.43 (89.68)	.091
12th week (time 2)	458.93 (85.28)	425.61 (90.81)	.046c
IPAQ-C: vigorous (MET-min/wk)b
Baseline	0	0	1.000
Sixth week (time 1)	0	0	1.000
12th week (time 2)	0	0	.309
IPAQ-C: moderate (MET-min/wk)b
Baseline	0 (0–40)	0 (0–240)	.686
Sixth week (time 1)	300 (160–840)	160 (0–640)	.089
12th week (time 2)	560 (40–1120)	0 (0–400)	.000c
IPAQ-C: walking (MET-min/wk)b
Baseline	1386 (693–2079)	1155 (693–2772)	.893
Sixth week (time 1)	1386 (1386–2376)	1386 (792–2772)	.268
12th week (time 2)	1386 (1155–2772)	1386 (693–2772)	.303
IPAQ-C: sitting (MET-min/wk)b
Baseline	2835 (2100–4200)	2520 (2100–3360)	.135
Sixth week (time 1)	2520 (2100–2940)	2170 (1680–2940)	.349
12th week (time 2)	2100 (1890–2940)	2520 (1890–2940)	.603
IPAQ-C: total (MET-min/wk)b
Baseline	4571 (3864–5733)	4059 (3213–5460)	.166
Sixth week (time 1)	4713 (4146–6006)	4326 (3486–5346)	.046c
12th week (time 2)	4746 (3886–6143)	4253 (3486–5340)	.027c
Regular exerciser
Baseline	16 (25.8%)	17 (27.0%)	.881
Sixth week (time 1)	46 (80.7%)	33 (63.5%)	.044c
12th week (time 2)	45 (75.0%)	21 (36.2%)	.000c
C-HADS: anxietyb
Baseline	2.5 (1–5)	2 (0–3)	.070
Sixth week (time 1)	1 (0–4)	1 (0–3)	.559
12th week (time 2)	1 (0–2)	1 (0–3)	.239
C-HADS: depressionb
Baseline	3 (1–4)	2 (1–5)	.832
Sixth week (time 1)	2 (0–3)	2 (1–3)	.537
12th week (time 2)	0.5 (0–2)	2 (1–3)	.002c
C-HADS: totalb
Baseline	4 (3–11)	5 (1–8)	.508
Sixth week (time 1)	3 (0–6)	4 (1–6)	.380
12th week (time 2)	1 (0–4.5)	3 (1–6)	.020c
CESEI-Ca
Baseline	40.76 (11.77)	37.05 (12.47)	.091
Sixth week (time 1)	54.72 (9.99)	49.16 (13.25)	.015c
12th week (time 2)	67.12 (9.78)	53.78 (13.21)	.000c
CDSEI-Ca
Baseline	50.61 (11.86)	47.74 (11.71)	.177
Sixth week (time 1)	62.23 (8.35)	60.00 (9.56)	.199
12th week (time 2)	71.33 (7.30)	62.64 (11.02)	.000c
C-CLASP: totala
Baseline	44.61 (17.58)	39.87 (16.28)	.120
Sixth week (time 1)	33.98 (15.80)	35.27 (15.12)	.666
12th week (time 2)	30.65 (14.51)	34.50 (16.42)	.179
Medication complianceb
Sixth week (time 1)	100 (100)	100 (100)	.342
12th week (time 2)	100 (100)	100 (100)	.164
Self-care complianceb
Sixth week (time 1)	100 (80–100)	80 (50–100)	.007c
12th week (time 2)	100 (80–100)	80 (50–100)	.003c
Readmission
Sixth week (time 1)	1 (1.7%)	3 (5.3%)	.619
12th week (time 2)	1 (1.6%)	0	.496

All values are presented in count (percentages), unless otherwise indicated (with footnote symbols), and compared by χ² or Fisher exact test. The nonsignificant statistical difference in P values was obtained by Mann-Whitney U test for medication compliance as the data were skewed and the value of data was different.

Abbreviations: C-CLASP, Chinese version of the Cardiovascular Limitations and Symptoms Profile; CDSEI-C, Cardiac Diet Self-Efficacy Instrument–Chinese version; CESEI, Cardiac Exercise Self-Efficacy Instrument–Chinese version; C-HADS, Chinese-Cantonese version of the Hospital Anxiety and Depression Scale; IPAQ-C, Chinese version of the International Physical Activity Questionnaire; MET, metabolic equivalent task.

^aPresented in mean (SD) and compared by independent t test.

^bPresented in median (interquartile range) and compared by Mann-Whitney U test.

^cPresented the significant differences in P value < .05.

Intervention Effects on Study Outcomes

Table 3 shows the effects of intervention on the study outcomes. In terms of outcomes related to cardiovascular risk factors, a significant interventional effect on the overall decline in total cholesterol and low-density lipoprotein level from baseline to the 12th week was achieved in both groups. The experimental group showed significant reductions of 0.114 mmol/L in total cholesterol (P = .041) and 0.183 mmol/L in low-density lipoprotein (P = .033). However, the intervention resulted in nonsignificant differences in blood pressure, pulse, weight, body mass index, and smoking rate reduction at both the 6th and 12th weeks, as well as changes in triglycerides, high-density lipoprotein, fasting blood glucose, and hemoglobin A_1C at the 12th week.

TABLE 3 - The Effect of Intervention on Study Outcomes Across Study Time Points (N = 140)

Outcome Measures	Regression Coefficients of the GEE Models
	ß (95% Confidence Interval)	P
Systolic blood pressure
Group	1.950 (−3.587 to 7.488)	.490
Time 1	2.856 (−1.397 to 7.109)	.188
Time 2	1.174 (−4.024 to 6.372)	.658
Group × Time 1	−5.149 (−12.107 to 1.809)	.147
Group × Time 2	−3.135 (−10.518 to 4.249)	.405
Diastolic blood pressure
Group	3.160 (−0.445 to 6.765)	.086
Time 1	3.006 (0.721–5.291)	.010
Time 2	1.708 (−1.339 to 4.754)	.272
Group × Time 1	−1.352 (−6.071 to 3.368)	.575
Group × Time 2	−0.963 (−5.384 to 3.458)	.669
Pulse
Group	1.279 (−2.980 to 5.538)	.556
Time 1	4.683 (1.771–7.595)	.002
Time 2	2.202 (−0.994 to 5.397)	.177
Group × Time 1	−3.744 (−8.210 to 0.723)	.100
Group × Time 2	−2.644 (−6.963 to 1.676)	.230
Total cholesterol
Group	0.034 (−0.055 to 0.122)	.455
Time 2	−0.158 (−0.233 to −0.083)	.000
Group × Time 2	−0.114 (−0.224 to −0.004)	.041a
Triglycerides
Group	−0.067 (−0.248 to 0.114)	.470
Time 2	−0.109 (−0.278 to 0.059)	.203
Group × Time 2	−0.051 (−0.254 to 0.151)	.619
High-density lipoprotein
Group	−0.035 (−0.124 to 0.055)	.450
Time 2	0.100 (0.055–0.146)	.000
Group × Time 2	−0.001 (−0.074 to 0.072)	.979
Low-density lipoprotein
Group	0.064 (−0.070 to 0.199)	.349
Time 2	−0.323 (−0.432 to −0.215)	.000
Group × Time 2	−0.183 (−0.350 to −0.015)	.033a
Fasting blood glucose
Group	0.044 (−0.068 to 0.155)	.443
Time 2	0.030 (−0.057 to 0.118)	.495
Group × Time 2	−0.099 (−0.227 to 0.029)	.130
HbA1C
Group	−0.014 (−0.079 to 0.051)	.678
Time 2	−0.008 (−0.031 to 0.016)	.524
Group × Time 2	−0.023 (−0.067 to 0.022)	.317
Weight
Group	−1.547 (−4.955 to 1.862)	.374
Time 1	−0.995 (−2.172 to 0.183)	.098
Time 2	−0.825 (−1.920 to 0.270)	.140
Group × Time 1	0.666 (−0.603 to 1.934)	.304
Group × Time 2	−0.081 (−1.344 to 1.181)	.900
Body mass index
Group	−0.758 (−1.869 to 0.353)	.181
Time 1	−0.155 (−0.351 to 0.041)	.121
Time 2	−0.109 (−0.324 to 0.106)	.320
Group × Time 1	0.025 (−0.236 to 0.287)	.849
Group × Time 2	−0.218 (−0.529 to 0.094)	.170
Six-minute walk test
Group	4.199 (−25.985 to 34.223)	.789
Time 1	62.740 (47.998–77.481)	.000
Time 2	71.703 (56.098–87.309)	.000
Group × Time 1	28.388 (7.194–49.582)	.009a
Group × Time 2	31.194 (6.972–55.417)	.012a
IPAQ-C: moderate
Group	−0.258 (−1.432 to 0.916)	.666
Time 1	−0.159 (−0.592 to 0.275)	.473
Time 2	−0.005 (−0.614 to 0.604)	.987
Group × Time 1	0.429 (−0.680 to 1.538)	.448
Group × Time 2	0.737 (−0.460 to 1.934)	.227
IPAQ-C: walking
Group	−0.049 (−0.336 to 0.238)	.739
Time 1	0.096 (−0.109 to 0.302)	.358
Time 2	0.146 (−0.063 to 0.356)	.171
Group × Time 1	0.239 (−0.153 to 0.63)	.232
Group × Time 2	0.129 (−0.125 to 0.383)	.319
IPAQ-C: sitting
Group	0.122 (−0.018 to 0.263)	.089
Time 1	−0.082 (−0.172 to 0.009)	.077
Time 2	−0.088 (−0.170 to −0.005)	.037a
Group × Time 1	−0.112 (−0.224 to 0.000)	.051
Group × Time 2	−0.165 (−0.293 to −0.038)	.011a
IPAQ-C: total
Group	0.052 (−0.083 to 0.188)	.449
Time 1	0.017 (−0.072 to 0.106)	.710
Time 2	0.022 (−0.074 to 0.118)	.653
Group × Time 1	0.076 (−0.094 to 0.246)	.381
Group × Time 2	0.075 (−0.051 to 0.201)	.241
C-HADS: anxiety
Group	0.276 (−0.035 to 0.587)	.082
Time 1	−0.174 (−0.442 to 0.094)	.204
Time 2	−0.222 (−0.452 to −0.009)	.060
Group × Time 1	−0.144 (−0.489 to 0.200)	.411
Group × Time 2	−0.452 (−0.801 to −0.102)	.011a
C-HADS: depression
Group	−0.156 (−0.402 to 0.091)	.217
Time 1	−0.288 (−0.461 to −0.115)	.001
Time 2	−0.281 (−0.511 to −0.052)	.016
Group × Time 1	0.107 (−0.122 to 0.336)	.360
Group × Time 2	−0.080 (−0.414 to 0.255)	.641
C-HADS: total
Group	0.095 (−0.176 to 0.367)	.491
Time 1	−0.213 (−0.370 to −0.055)	.008
Time 2	−0.254 (−0.451 to −0.057)	.011
Group × Time 1	−0.089 (−0.327 to 0.149)	.463
Group × Time 2	−0.381 (−0.689 to −0.073)	.015a
CESEI-C
Group	3.710 (−0.525 to 7.944)	.086
Time 1	11.612 (8.894–14.330)	.000
Time 2	16.649 (13.999–19.298)	.000
Group × Time 1	2.006 (−2.054 to 6.066)	.333
Group × Time 2	9.499 (5.252–13.747)	.000a
CDSEI-C
Group	2.871 (−1.243 to 6.985)	.171
Time 1	11.636 (9.243–14.030)	.000
Time 2	14.892 (12.334–17.450)	.000
Group × Time 1	−0.415 (−3.780 to 2.950)	.809
Group × Time 2	5.901 (2.009–9.794)	.003a
C-CLASP
Group	4.740 (−1.153 to 10.633)	.115
Time 1	−5.527 (−7.983 to −3.070)	.000
Time 2	−5.682 (−8.314 to −3.049)	.000
Group × Time 1	−6.007 (−10.107 to −1.906)	.004a
Group × Time 2	−8.397 (−12.317 to −4.478)	.000a
Medication compliance
Group	0.005 (−0.004 to 0.015)	.275
Time 2	−0.006 (−0.013 to 0.002)	.136
Group × Time 2	0.007 (−0.001 to 0.015)	.107
Regular exercise performance
Group	0.833 (0.356–1.950)	.881
Time 1	4.300 (2.148–8.605)	.000
Time 2	1.779 (0.960–3.300)	.120
Group × Time 1	3.564 (1.159–10.965)	.031a
Group × Time 2	7.141 (2.408–21.175)	.000a
Smoking status: nonsmoking
Group	0.787 (−0.264 to 1.838)	.142
Time 1	0.248 (−0.225 to 0.722)	.305
Time 2	0.125 (−0.264 to 0.515)	.528
Group × Time 1	−0.050 (−1.058 to 0.957)	.922
Group × Time 2	−0.132 (−0.960 to 0.696)	.755

“ß” indicates regression coefficient. “Group” indicates the group differences at baseline between the intervention and control groups. “Time 1” and “Time 2” indicate the time effects on the control group at the 6th and 12th week, respectively, compared with the baseline. “Group × Time” gives the differential change of the outcome in the intervention group on top of the change in the control group at the 6th and 12th week with respect to the baseline (change in the experimental group − change in the control group).

^aPresented the significant differences in P value < .05.

Abbreviations: C-CLASP, Chinese version of the Cardiovascular Limitations and Symptoms Profile; CDSEI-C, Cardiac Diet Self-Efficacy Instrument–Chinese version; CESEI, Cardiac Exercise Self-Efficacy Instrument–Chinese version; C-HADS, Chinese-Cantonese version of the Hospital Anxiety and Depression Scale; GEE, generalized estimating equation; HbA_1c, hemoglobin A_1c; IPAQ-C, Chinese version of the International Physical Activity Questionnaire.

For exercise capacity and activity level (Table 3), the intervention provided a significant improvement in the six-minute walk test (ß = 31.194, P = .012) at the 12th week and a reduction in sitting in the IPAQ-C (ß = −0.165, P = .011). In addition, regular exercise performance in favor of the experimental group significantly improved at both the 6th and 12th weeks (P < .005). The effects of intervention on psychological well-being, self-efficacy, and quality of life are shown in Table 3. Significant reductions at the 12th week were observed in the anxiety level (ß = −0.452, P = .011) and the total score of the Chinese version of the Hospital Anxiety and Depression Scale (ß = −0.381, P = .015). For self-efficacy evaluation, significant improvements were reported at the 12th week on the score of Cardiac Exercise Self-Efficacy Instrument–Chinese version (ß = 9.499, P < .001) and Cardiac Diet Self-Efficacy Instrument–Chinese version (ß = 5.901, P = .003). The score of Cardiovascular Limitations and Symptoms Profile–Chinese version improved at both the 6th (ß = −6.007, P = .004) and 12th (ß = −8.397, P < .001) weeks. Table 4 presents the effects of the intervention on cardiac-related readmission and shows a nonsignificant difference between the groups (P > .05). No adverse event associated with the intervention was reported in emergency visits.

TABLE 4 - The Effect of Intervention on Cardiac-Related Readmission Across Study Time Points (N = 140)

Outcome Measures	Logistic Regression
	Exp (ß) (95% Confidence Interval)	P
Readmission
Group	0.935 (0.126–6.914)	.947
Time	0.322 (0.033–3.188)	.333

“ß” indicates regression coefficient. “Group” indicates the group differences at baseline between the intervention and control groups. “Time” indicates the time effects on the control group at 90 days, compared with 30 days.

Intervention Fidelity and Acceptability

Mobile Health Cardiac Rehabilitation Fidelity

In the process evaluation, nearly half of experimental group participants (n = 32) were randomly selected to review the utility of the mCR app. All evaluators (n = 32, 100%) reported high fidelity in the education provided by the CR app. All evaluators reported reading all the mCR educational information, and Table 5 shows the total number of views on the educational page. The total number of views for educational text content was 3271 (102.22 views/participant), and the total number of views for demonstration videos was 332 (10.38 views/participant). The task completion page for self-monitoring is presented in Table 2. Compliance with self-care in favor of the experimental group was significantly different across the 2 time points (P < .05) (Table 2). In addition, all the mCR participants (n = 32, 100%) viewed the health data summary. Over 90% of participants (n = 30, 93.8%) sent questions or comments on the inquiry page, in a median of 4 (interquartile range, 2–5.5) count replies per participant over the study period.

TABLE 5 - Number of Views in Mobile Health Cardiac Rehabilitation Education (N = 32)

Educational Sections	Text Content (Views)	Demonstration Video (Views)
Medical	793	111
Nursing	806	54
Physical exercise section I	516	97
Physical exercise section II			50
Diet	815	Not designed
Psychological support	341	20
Total views	3271	332
Average views	102.22	10.38

Mobile Health Cardiac Rehabilitation Acceptability

The mCR satisfaction measured using the visual analog scale achieved a mean score of 91.7% (±9.04%) (Table 6). The top 3 satisfied comments for the mCR app were as follows: “providing valuable, necessary and relevant information,” “is regarded as a good, smart and beneficial idea to apply in healthcare,” and “as a bridge of CR transitional care to extend hospital care.” However, “quick to download picture, text, or video and convenient to use” scored the lowest satisfaction level at 88.0% (±13.46%) (Table 6).

TABLE 6 - Satisfaction Evaluation of Mobile Health Cardiac Rehabilitation Program (N = 32)

Evaluation Items	Score, %	±SD
1. mCR app can be quick to download text contents, pictures, photo, and video, which is convenient.	88.0	13.46
2. mCR app provides valuable, necessary, and relevant information.	96.1	10.52
3. mCR program extends hospital care after hospital discharge.	95.0	7.78
4. mCR app helps patients adhere to treatment recommendation.	92.5	9.93
5. mCR app is easy to use.	90.7	15.50
6. mCR program is a good, smart, and beneficial idea for healthcare.	95.5	15.90
7. I will continue to use mCR app.	94.3	12.04
8. mCR app is reliable.	92.7	13.69
9. The overall mCR satisfaction is	94.6	7.23
Mean score	91.7	9.04

Abbreviation: mCR, mobile health cardiac rehabilitation.

Discussion

This study demonstrated the effectiveness and feasibility of mHealth application on CR for Chinese patients with PCI in Macao. The reduction of total cholesterol and low-density lipoprotein has been proven to be a major predictor of cardiovascular risk prevention. Significant improvements in total cholesterol and low-density lipoprotein reduction were found in this study, which was congruent with previous studies supporting the interventional effect of mCR on cholesterol control among patients with CHD and post-PCI patients.^12,13,17,19 Long-term expected outcomes of CR diabetes management are suggested to attain fasting plasma glucose of 5.0 to 7.2 mmol/L and hemoglobin A_1C < 7%.^9,10,34 There was no significant difference in glycemic control by fasting plasma glucose and hemoglobin A_1C levels among experimental group participants in this study, which was similar to a 3-month study recording an inconclusive intervention effect on glycemic control.¹³ One of the reasons might be the short-term follow-up of 3 months to detect a long-term effect, and baseline levels of fasting plasma glucose and hemoglobin A_1C were within a desirable range, which might have led to a ceiling effect. Weight management, one of the core components of CR, is suggested by strategies for weight maintenance that promote gradual and sustainable weight loss (5%–10%) in 3 to 6 months to maintain a healthy body mass index.^9,10 This study was completed in 3 months and was unable to demonstrate a significant effect on weight and body mass index change, which is consistent with previous studies.^12,17,19

Functional capacity measured using the six-minute walk test is regarded as a prognostic marker for predicting the therapeutic effect of CR.⁹ Participants who underwent mCR demonstrated a significant improvement in the six-minute walk test. In terms of physical activity level, a continuous reduction in sitting energy expenditure at the 12th week was found in the experimental group, whereas the control group showed no change (P > .005). This finding was among the limited mCR studies showing a favorable intervention effect on sitting time reduction.^18,35 A statistically significant improvement in total IPAQ-C was demonstrated (P = .027) at the 12th week, which was compatible with the literature.³⁵

Psychological distress and anxiety were observed among Hong Kong Chinese men who underwent cardiac catheterization.²⁴ A relationship between more depressive symptoms and lower self-efficacy in physical activity was reported, which induced a decrease in CR effectiveness.²¹ The anxiety level and total score of the Chinese version of the Hospital Anxiety and Depression Scale were significantly reduced at the 12th week. However, this study did not demonstrate a statistically significant reduction in depression, which is consistent with the literature.³⁵

This study showed the positive effects of the intervention on improving self-efficacy scores for cardiac exercise and diet. These findings were not consistent with previous studies that found no improvement in self-efficacy for regular physical exercise with intervention delivery by the Internet and mobile text,³⁵ and in the Diet Habit Questionnaire in another mCR study.¹⁹ However, the results of self-efficacy in cardiac exercise in this study support the findings,¹⁶ which indicated that the interventional group displayed a greater belief in the effectiveness of exercise. It supported mCR, manipulated the key components of social cognitive theory, and provided some insights into the potential mechanism of mCR in promoting physical activity.¹⁶ These results revealed some benefits of behavior change and were integral to optimizing cardiovascular risk management.¹⁰

Quality of life, one of the key quality indicators of successful CR and associated with survival and subsequent episodes of acute coronary syndrome, is recommended for evaluation.^9,10 The quality of life of patients with CHD has been proven as a multidimensional indicator, which is affected by sociodemographic status, medical history, level of psychological distress, and perceived social support.³⁶ This study showed a favorable and sustainable effect of the intervention by reducing the score of Cardiovascular Limitations and Symptoms Profile–Chinese version (ß = −8.397, P < .001) at 12 weeks. Despite the different instruments used for quality of life measurement among the reviewed studies, the results of this study were positively supported by literature.^13,18,19 This effect was considered to be the result of positive findings of self-efficacy on healthy behavioral change.^9,10

The results of the process evaluation indicated a desirable level of satisfaction and utility, which is supported by other research.¹² In this study, social cognitive theory was applied in its implementation (Figure 3), with the mechanism of core determinants guiding the optimal methods to translate knowledge into healthcare practice. The mCR app was regarded as an environmental factor that facilitated personal factors in phase II CR participation and motivated the complexity of changes in behavioral and health outcomes.

FIGURE 3: Summarization: effects and interventions of the mCR study with the application of social cognitive theory.

Limitations

Some study outcomes were measured using self-reported data, and these results may reflect recall bias. A limitation of this real-world trial was the single-blinded trial design, in which only the outcome assessors were masked. The study participants were not blinded, which could have caused social desirability bias to contaminate the study findings, as self-reported data were collected in this study.^37,38 The transferability of the study findings was limited to patients with PCI in phase II CR with a mild to moderate CR risk. In addition, this study is generalizable to patients with CR in Chinese urban cities due to the similarity of regional healthcare systems, patient backgrounds, and staff education.

Implications for Nursing Practice and Research

The findings of this study support the recognition of the role of mHealth in chronic disease management and suggest its transfer to other chronic diseases that involve self-efficacy in disease management.³⁹ Second, this study provides knowledge for designing future care practice, research, and policy related to cardiac nursing care, such as blood pressure management for hypertension, rhythm monitoring for atrial fibrillation, and symptom management for chronic heart failure. In addition, it implied the possibility of using other wireless advanced devices for monitoring blood pressure, blood glucose, body weight, and physical exercise, which are expected to synchronize patient data automatically with the mHealth system to support health analysis. Third, a 6-month or 1-year study is suggested to confirm the study's long-term effects. Moreover, the cost-effectiveness, manpower utility, and duration of nurse-led clinic consultation in mCR implementation is worthy of evaluation, which could inform mHealth application policies for hospitals and local governments.

Conclusion

This mCR study, underpinned by social cognitive theory, was a unique program tailored to meet the care needs of Chinese patients with PCI. It is feasible and effective in supporting the major goals of CR and adherence to cardiovascular health recommendations. It is important for the mCR app to include key features related to CR knowledge education, health data entry with reminder notifications, health status tracking, and interactive communication with patients. In addition, mCR implementation facilitates improvements in self-efficacy and health behavior changes, and maximizes the effects of CR. It provides scientific support to inform the future design and application of mHealth in clinical practice related to cardiac nursing care, chronic disease management, research, and policies.

What’s New and Important

• Mobile health CR, underpinned by social cognitive theory, was feasible and effective for patients who have undergone PCI.
• Significant interventional effects were found in improvements of physical, psychological, and behavioral change, as well as quality of life. Study feasibility was achieved satisfactorily.
• Further understanding about the role of mHealth on cardiac nursing and chronic disease management is identified.