Sage Journals: Discover world-class research

Abstract

Objective

To summarise the effectiveness of digital health maintenance interventions for subjectively and objectively measured physical activity level (PA) and physical function, fitness and health-related quality of life (HRQoL) after completion of phase II cardiac rehabilitation (CR).

Methods

We conducted a search for studies in MEDLINE, Embase, CENTRAL, and CINAHL (inception to May 2024). Independent reviewers selected and included randomised controlled trials (RCTs) using digital health interventions to maintain PA in patients with cardiovascular disease after phase II CR. Independent reviewers conducted data extraction, assessed the risk of bias using the Cochrane Risk of Bias 2 tool and rated the certainty of evidence using Grading of Recommendations Assessment, Development and Evaluation – registered at PROSPERO: CRD42023396629.

Results

From 17,455 hits, 20 RCTs with 1801 participants were included. Risk of bias for most studies reporting on PA outcomes was rated as ‘some concerns’. Comparing digital health intervention with minimal intervention/usual care showed no effect of objective PA (standardised mean difference (SMD) 0.85, 95% CI: −0.07 to 1.77) and a small effect of subjective PA (SMD 0.37, 95% CI: 0.05 to 0.69) at the end of intervention both graded as very low certainty of evidence. We found very low certainty of evidence for moderate effects on physical function (SMD 0.63, 95% CI: 0.03 to 1.24), and low certainty of evidence for no effect on physical fitness (SMD 0.19, 95% CI: 0.05 to 0.34) and HRQoL (SMD 0.13, 95% CI: −0.02 to 0.28), I²= 0.00%) at the end of intervention.

Conclusion

Digital health interventions showed a small effect on subjectively measured PA, and no effect on objective PA, physical fitness and HRQoL, but may increase physical function after CR completion. Yet, the certainty of evidence is low and higher quality studies with longer follow-up duration are needed to guide this area further.

Keywords

Cardiac rehabilitation maintenance digital health eHealth systematic review meta-analysis

Introduction

Regular physical activity (PA) is vital for the secondary prevention of cardiovascular disease.^1,2 PA plays a key role in exercise-based cardiac rehabilitation (CR), to improve aerobic fitness and health-related quality of life (HRQoL), and reduce symptoms and risk of post-cardiac events.^3,4

CR typically consists of three phases⁵: Phase I CR begins after an acute cardiovascular event and takes place in an inpatient setting aiming to stabilise and treat the patient. Phase II CR provides structured exercise-based rehabilitation, assessing risk factors and treating patients through exercise. A plan is initiated towards the end to prepare for a physically active, independent life at home. Globally, phase II CR consists of a median of three sessions per week with a session median of 60 minutes for a median of 8 weeks.⁶ CR III is the lifelong maintenance phase, during which the patient independently maintains exercise recommendations from CR II by self-monitoring their PA levels.

Unfortunately, many patients do not maintain the recommended level of PA in their daily lives after completing CR phase II.^7–10 Lack of PA engagement could be attributed to personal, environmental and organisational barriers, such as lack of time and lack of motivation.^11,12 There are no standardised definitions for maintenance.^13–15 Hence, we define maintenance as a period of at least 6 months with the goal of keeping the benefits gained and preventing relapse.¹⁶

The use of digital health has increased rapidly in recent years.^17,18 Digital health, often referred to as eHealth, serves as an umbrella term for eHealth, mHealth and telehealth, which all cover the use of various types of technology in healthcare.¹⁸ eHealth involves internet-based solutions such as websites and emails. mHealth includes phone calls, text messages and apps. Digital devices encompass wearables (with accelerometer or pedometer capabilities), activity trackers or Bluetooth-connected devices. Digital health interventions are easily accessible to patients and can be tailored to the individual patient, encouraging them to become more physically active, thereby potentially reducing barriers and to improve their disease management and improve PA.^19–21

Of previous systematic reviews^22–25 investigating maintenance interventions after CR II, only Hannan et al.²⁴ and Heimer et al.²⁵ investigated the effectiveness of digital health interventions exclusively. Yet, Hannan et al.²⁴ only assessed the effects of wearables after CR II and found a small effect on step count. Heimer et al.²⁵ found a small overall effect on physical exercise and activity that favoured digital health interventions on maintenance after CR phase II (standardised mean difference (SMD): 0.35 (95% CI: 0.02 to 0.70), I²= 72.3%). However, both reviews did not distinguish between subjective and objective PA outcomes, and both excluded interventions that used phone calls or emails exclusively. Subjective (self-reported) and objective assessments of PA capture different aspects, emphasising the need for separate analyses.²⁶ Objective measurements provide a more precise evaluation of PA levels and intensity compared to subjective measures,²⁶ while subjective measurements are not affected by the type of activity or placement of a wearable.²⁷

Hence, an up-to-date overview is needed to assess the effectiveness across all digital health interventions, separating subjective and objective PA assessments and contributing knowledge to support patients after CR II completion.

Objective

The purpose of this systematic review and meta-analysis is to summarise the effectiveness of digital health interventions in maintaining objectively and subjectively measured PA (primary outcomes) and physical function, physical fitness, HRQoL and adverse events after completion of phase II CR compared to usual care or minimal intervention.

Methods

Study design

This study was conducted according to the recommendations from the Cochrane Handbook and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement.^28,29 The completed PRISMA checklist is available in Appendix A in the online supplemental materials. The review was performed according to the protocol pre-registered with PROSPERO (CRD42023396629).

Eligibility criteria

Population

The study population of each study had to have a mean age of ≥18 years and had to have finished phase II CR after being medically diagnosed with cardiovascular disease or undergoing a heart-related procedure, e.g., coronary artery bypass graft or percutaneous coronary intervention.

Intervention

The studies had to evaluate the effectiveness of digital health interventions, either exclusively or as a multicomponent (e.g., in-person intervention + digital component). The intervention had to aim to increase, maintain or promote PA or physical fitness after CR II completion. Unlike previous reviews by Hannan et al.²⁴ and Heimer et al.,²⁵ to be included as digital health interventions, interventions had to begin after CR II and not during CR II. This was done to ensure more consistency, homogeneity and comparability across studies.

Comparison

The intervention had to be compared to usual care or minimal intervention (e.g., blinded wearable, sham website or education).

Outcomes

For a study to be included, it had to have at least one outcome on the primary outcome of interest, PA (objectively or subjectively measured), or at least one outcome on secondary outcomes of interest, physical function (objectively or subjectively), physical fitness, HRQoL or adverse events (serious and non-serious). Eligible time points were at the end of intervention, 6 months, and 12-month follow-up after the digital health intervention.

Study types

Eligible study designs had to include randomisation, i.e., randomised controlled trials (RCTs), randomised pilot and feasibility studies, factorial design studies, cluster-randomised trials and cross-over studies (only the period until cross-over was included).

Search strategy, data sources and study selection

Embase (Ovid), MEDLINE (Ovid), CINAHL (Ebsco) and Cochrane Central Register of Controlled Trials (CENTRAL) were searched from inception to the 2nd of August 2023. ClinicalTrials.gov was searched for any ongoing studies. The search was updated 22nd of May 2024. References of eligible studies were hand-searched (‘pearl growing’) by two independent reviewers (AL and TGH). The ‘pearl growing’ approach was used to identify relevant studies, with each study (pearl) containing relevant keywords or citations, which were used to identify and track more relevant studies.³⁰ The search was validated against included studies from Graham et al.²² and Martinello et al.²³

The search matrix comprised relevant keywords, MeSH and Thesaurus terms for cardiovascular disease, digital health, maintenance and PA. However, RCT filters^31–33 were added to the search strings to reduce the number of non-randomised studies from the search. No limitations or restrictions in relation to the publication date or language were applied. The full search strategy for each database is available in Appendix B in the online supplemental materials.

Studies not reported in English were translated into English via Google Translate.³⁴ Two reviewers (AL reviewed all studies, while the studies were equally distributed between SRM, TGH and JRA) independently screened the studies. Reviewer disagreements were discussed, and if no consensus could be reached, then a third reviewer (LHT) was consulted. In addition, authors were contacted if full-text articles were inaccessible and had missing data, or if data from their analysis were not reported. A maximum of four email attempts were made: two emails to the corresponding author and two emails to the first, second, or last author.³⁵

Data extraction

Data extraction was done by the same reviewers as for study selection. The following data were extracted using a priory-defined and pilot-tested extraction template. Data reported only in graphs or figures were extracted using the pilot-tested Web-Based Plot Digitizer.³⁶ Extracted data included study, author, year of publication, the country where the trial took place, study design, duration of preceding CR, weekly frequency of preceding CR, mean age, the proportion of females, number of participants enrolled, number of participants completing the trial, heart conditions, non-serious adverse events and serious adverse events. Adverse events were extracted if reported explicitly in text or figures/tables, as recommended by The Consolidated Standards of Reporting Trials.³⁷ The adverse events reported in the flow diagram were not considered explicit, even if there were indications of adverse events. Flow diagrams often lack sufficient details to determine whether an event is directly linked to the intervention and was therefore deemed inappropriate for extraction. The Food and Drug Administration's definition of non-serious (e.g., pain, fatigue or falls) and serious adverse events (e.g., hospitalisation, disability or permanent damage) was used.³⁸ Primary digital health component (mHealth, eHealth, digital device or a combination), whether the type of digital health intervention was exclusive or combined with in-person intervention, was also extracted. We also extracted the duration, weekly frequency of the intervention, number of personal encounters during the intervention, use of theoretical model or framework, and usual care or minimal intervention (e.g., sham website). Mean and standard deviations (SDs) were extracted for outcomes in the following prioritised order: the primary outcome of interest was PA objectively (e.g., measurements by using an accelerometer) or subjectively (e.g., International Physical Activity Questionnaire) measured. Physical function objectively (e.g., 6-minute walk test) or subjectively (Short-Form 36 physical function score) measured, physical fitness (e.g., peak oxygen uptake (VO₂ peak)) and HRQoL (e.g., European Quality of life - 5 Dimensions) were secondary outcomes of interest. If reported, both objective and subjective outcomes were extracted and sought. A pre-specified hierarchy for outcomes was used for data extraction (Table 1).

Table 1.

Outcome prioritisation for data extraction.

Outcome	Primary or secondary outcome of interest	Hierarchical prioritisation
Physical activity (objective)	Primary	Moderate-to-vigorous physical activity measured with accelerometer, daily amount of energy expenditure measured in calories, daily MET (minutes or hours), number of daily steps
Physical activity (subjective)	Primary	No hierarchical prioritisation was planned. Subjectively measured physical activity (e.g., Global Physical Activity Questionnaire, International Physical Activity Questionnaire, Seven-Day Physical Activity Recall Exercise)
Physical function (objective)	Secondary	6-minute-walk-test, Shuttle walk test, Sit-to-stand or Timed Up and Go
Physical function (subjective)	Secondary	No hierarchical prioritisation was planned. Subjectively measured physical function (e.g., Short Form 36 – physical component)
Physical fitness	Secondary	VO₂ max, VO₂ peak, daily MET (minutes, hours, or watts)
Health-related quality of life	Secondary	A generic questionnaire (e.g., EQ-5D or Short Form 36), disease specific questionnaire (e.g., the Minnesota Living with Heart Failure Questionnaire)

Note. EQ-5D: EuroQol-5 domain; MET: metabolic equivalent of task; VO₂ max: maximal oxygen uptake; VO₂ peak: peak oxygen uptake.

In addition, data were extracted for each outcome of interest at each assessment point, end-of-intervention, closest to 6-month follow-up and closest to 12-month follow-up. Furthermore, the type of instrument used to measure the outcome and the placement were also extracted. The data extraction form and data are available in Appendix C in the online supplemental materials.

Quality assessment

Risk of bias was assessed for each outcome of interest using the Cochrane Risk of Bias Tool version 2 (RoB 2) for each study.³⁹ All assessments were made in Excel (RoB 2, 22nd of August 2019, version 9).⁴⁰ Studies were assessed by two independent reviewers (AL assessed all studies, while the studies were equally distributed between SRM, TGH and JRA). A third reviewer (LHT) was consulted if a consensus was not reached.

Certainty of evidence

Grading of Recommendations Assessment, Development and Evaluation (GRADE) was used to assess all outcomes at the end of the intervention (assessed by AL and JRA). A summary of findings table was used for reporting the results, along with a separate evaluation of the certainty in evidence for each outcome. Assessments and the summary of findings were made using the online software GRADEpro GDT.⁴¹

Statistical analysis

The effectiveness of the interventions on objectively measured PA, subjectively measured PA, objectively measured physical function, subjectively measured physical function, physical fitness, and HRQoL was evaluated separately. The priory-defined prioritisation was used if more than one relevant outcome was reported within a trial (Table 1). Missing SDs were calculated using reported 95% confidence intervals, p-values or standard errors.⁴² If two or more studies reported data on the same outcome of interest using different outcome measures, a random-effects model, adjusting to Hedges’ g using endpoint scores, was used to calculate the effect size. If more than one study reported the same outcome using the same measures, an effect size was calculated as the weighted mean difference (WMD) of the final scores using a random-effects model meta-analysis adjusting to Hedges’ g. If only one study reported data for a specific outcome, a narrative approach was used in accordance with the guidelines from the Cochrane Handbook.⁴²

Standardised mean difference (SMD) with a 95% confidence interval was calculated for outcome measures of continuous data. Effect sizes pooled as SMD were interpreted as a small effect (0.2), moderate effect (0.5) or large effect (0.8).⁴³ WMD was calculated for the outcome reported in the same unit. Dichotomous outcomes, e.g., dropouts and adverse events, were analysed and expressed as risk ratios with 95% confidence intervals.

The heterogeneity of the results was examined using the Cochrane Q test and quantified as I² values and the between-study variance τ². Arbitrary interpretation of I²: 0% to 40% represents that it might not be important, 30% to 60% may represent moderate heterogeneity, 50% to 90% may represent substantial heterogeneity and 75% to 90% represents considerable heterogeneity.⁴⁴ Small study bias was examined by using Eggers's test score (p ≤ 0.05 indicates a small study bias) and visually inspected funnel plots for each outcome.

The statistical software Stata 17 (StataCorp. 2021. Stata Statistical Software: Release 17. College Station, TX: StataCorp LLC) was used for all statistical analyses.

As protocolised, heterogeneity was explored on nominal variables (cardiac diagnoses, primary digital health component, intervention type, type of comparator and risk of bias) by sub-group analyses. For continuous data, heterogeneity was explored by sub-analyses using univariate meta-regression on mean age, sex distribution, intervention length and mean baseline PA.

Results

Study characteristics

In 2023, the search identified 16,012 studies, ultimately including 19 studies.^45–63 In 2024, the search was updated and identified 1443 studies, ultimately resulting in the inclusion of one study.⁶⁴ Among the included studies, there were 18 RCTs and two studies with secondary reporting,^51,59 involving a total of 1801 unique participants for analysis. The study flow diagram is visually presented in Figure 1. A summary of the included studies is shown in Table 2. Excluded studies and ongoing studies are available in Appendix D in the online supplemental materials. The mean age of the study participants was 61.2 years, ranging from 57 ⁶³ to 66.8.⁵⁸ Two of the included studies only included participants with coronary heart disease.^46,55 The rest of the studies included a mixed cardiac population.^{45,47–54,56–64} The length of CR II ranged from 4 weeks⁵⁴ to 16 weeks,⁵⁰ and the weekly frequency of CR sessions was reported as two times per week^48–50,61 or three times per week,^52,57,59 while the remaining studies did not report the length of CR II.

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) flow diagram.

Table 2.

Study characteristics of the included studies.

Study:First author, (year of publication), country where the trial took place, study design, reference number	Participants:Number of participants randomised in intervention/control group, heart disease, participants completing study in intervention/control group,risk ratio for dropout (95% confidence interval)	Participants:Mean age, proportion of females	Comparator:Usual care or minimal intervention (type of usual care or minimal intervention)	Description of intervention:(intervention based on model/theory)	Digital health intervention:Intervention type, digital solution(s), (primary digital health category) (overall digital health category)	Intervention:Intervention duration in mean weeks (follow-up duration in mean weeks)	Outcomes:	Adverse events:Adverse event reported (types of events).Risk ratio for adverse event (95% confidence interval)
Antypas & Wangberg. (2014), Norway, RCT, ⁵⁴	29/40, CVD 7/12 RR = 1.1 (0.8; 1,4)	59.2. 22.3%	Minimal intervention (access to website and discussion forum, received no tailored text messages)	Access to website with general health information, discussion forums and extra questions to answer (every other week). Tailored SMS messages, SMS and email reminders were sent. The participants were asked to plan training activities or set weekly goals. (Integrative models, e.g., Health Action Process Approach)	Only digital, website + text message, (eHealth) [eHealth + mHealth]	13 (0)	Subj PA (total PA; MET-minutes/week, International Physical Activity Questionnaire)	*
Avila et al. (2018), Belgium, three-arm RCT, ⁵³	30/30, CAD, CABG, PCI 28/26 RR = 0.5 (0.1; 2.5)	60.2. 11.7%	UC (PA advice)	First three exercise sessions, supervised by the investigator. Received individualised exercise prescriptions. Exercise data collected by Garmin Forerunner (worn on non-dominant hand) was manually logged. Once a week feedback by email or phone. Both groups were recommended to exercise ≥150 min/week at a target heart rate of 70–80% of heart rate reserve. *	As component, wearable + email or text message, (digital device) [digital device + mHealth + eHealth]	12 (0)	Obj PA (Moderate PA duration, > 3 METs; min/day. Unable to extract or calculate SD) Subj PF (SF-36; physical function subscale)) PFIT (VO₂ peak) HRQoL (SF-36; total score)	SAE (no events related to exercise)
Avila et al. (2020), Belgium, three-arm RCT, ⁵¹	30/30, CAD, CABG, PCI 26/25 RR = 0.8 (0.2; 2.7)	63. 9.8%	UC (PA advice)	First three exercise sessions, supervised by the investigator. Received individualised exercise prescriptions. Exercise data collected by Garmin Forerunner (worn on non-dominant hand) was manually logged. Once a week feedback by email or phone. Both groups were recommended to exercise ≥150 min/week at a target heart rate of 70–80% of heart rate reserve. *	As component, wearable + email or text message, (digital device) [digital device + mHealth + eHealth]	12 (39)	Obj PA (moderate PA; min/day) accelerometer Obj PF (sit and rising test) Subj PF (SF-36; physical function subscale) PFIT (VO₂ peak; ml/kg/min)	*
Brubaker et al. (2000), USA, RCT, ⁵²	16/15, MI, CABG, AP, PCI, CHF No dropouts	60. *	UC (lifestyle and PA advice)	One home visit from exercise physiologist discussing home exercise program. Exercise physiologist phoned every other week, entire duration of the study. Exercise heart rates were monitored by self-palpation and from intermittent electrocardiogram monitoring (defibrillator paddles). Instructed to exercise for 30–40 min., 3–5x/week at 50–75% of the heart rate reserve from the exercise test. *	As component, phone call + defibrillator paddles (mHealth) [mHealth + digital device]	52 (0)	PFIT (MET)	*
Butler et al. (2009), Australia, RCT, ⁴⁸	62/60, CABG, PCI, MI, ACS 51/51 RR = 1.2 (0.5; 2.7)	63.8. 25.5%	UC (given two general PA brochures)	Given two general PA brochures. Wore a pedometer (Yamax Digiwalker 700b), noted steps/day in a step calendar. Received a total of four phone calls which consisted of verbal reinforcement, individualised goal setting, and objective feedback based on measurements from the pedometer. *	Only digital, wearble + phone call (digital device) [digital device + mHealth]	6 (26)	Subj PA (total physical activity; weekly/min; Active Australia survey)	*
Duscha et al. (2018), USA, RCT, ⁴⁷	21/11, MI, PCI, CABG, HF 16/9 RR = 1.3 (0.3; 5.7)	63.3. 24%	UC (PA advice)	Wore a Fitbit activity tracker, received feedback through weekly phone calls from a coach, and text messages. Exercise prescribed based on the performance from the two previous weeks and were encouraged to be active, eat healthy and take their medicine. Coaches called 1–2x/week for 30–60 min. *	Only digital, wearable + phone call (digital device) [digital device + mHealth]	13 (0)	Obj PA (moderate-high activity; min/week; accelerometer) PFIT (VO₂ peak; ml/kg/min)	*
Foccardi et al. (2021), Italy, pilot-RCT, ⁴⁶	16/16, PTCA, CABG 15/15 RR = 1.0 (0.1; 14.6)	61.3. 15.6%	UC (exercise prescription)	Received a daily text message through phone app. Both groups received a written exercise prescription and recommended participants to do aerobic exercise up to 5x/week, and to do strength training up to 3x/week. Intensity based on Borg rating 10. *	Only digital, text message, (mHealth) [mHealth]	13 (0)	Subj PA (moderate recreational activity; min; Global Physical Activity Questionnaire) Obj PF (30-s-sit-to-stand) PFIT (submaximal iso-watt test)	Minor adverse events, SAE (no events during iso-watt testing)
Guiraud et al. (2012), France, RCT, ⁶²	19/10, angioplasty, CABG, HF No dropouts	58.7. 17.2%	UC (PA advice)	Wore an accelerometer, single axis, at the waist, received a phone call every 15 day for 2 months. The phone call consisted of feedback based on prior week's performance, providing instructions, to encourage exercise with other persons, identifying obstacles, overcoming obstacles, goal setting and providing strategies for adherence. *	Only digital, wearable + phone call (digital device) [digital device + mHealth)	8 (0)	Obj PA (active energy expenditure; kcal/week; accelerometer)	*
Labrador et al. (2023), Spain, RCT, ⁶⁴	40/40, CAD No dropouts	55.85. 16.25%	Minimal intervention (prescribed 12-month long exercise program, chest strap connected to a phone for monitoring heart rate, monthly educational sessions on risk factors over 12 months)	Once a month 90 min of supervised exercise booster sessions at outdoor park using exercise stations, advised to exercise 2x/week at the park, with exercise compliance monitored by phone calls, chest strap connected to phone app for monitoring heart rate, 3x/week SMS messages for education on PA, diet, disease knowledge, and the need to continue pharmacological treatment *	As component, phone call + digital device + app + SMS (mHealth) [mHealth + digital device]	52 (0)	Subj PA (leisure; minutes/week; International Physical Activity Questionnaire) Obj PF (6-min-walk-test) Obj PFIT (Bruce Protocol; METs)	Adverse events (no events)
Lear et al. (2006), Canada, RCT, ⁵⁰	151/151, MI, CABG, PCI 130/119 RR = 0.7 (0.4; 1,1)	64.1. 17.6%	UC (under the care of the family physician)	First year: four cardiac rehabilitation sessions in the first month, followed by monthly sessions in the second and third months, six telephone follow-ups, and three lifestyle and risk factor counselling sessions (PA, diet, weight, smoking cessation, diabetes) throughout the year. The second, third and fourth year each consisted of four telephone follow-ups and two lifestyle and risk factor counselling sessions. Telephoned by case manager to identify new or change in symptoms, follow-up on goals, and asses and counsel on lifestyle. (Transtheoretical Model of Change & Social Cognitive Theory)	As component, phone calls (mHealth) [mHealth]	208 (0)	Subj PA (Physical Activity; kcal/week, the 4-week modified Minnesota Leisure Time) PFIT (Symptom limited treadmill exercise stress test; METs) HRQoL (Perceived stress scale)	*
Lunde et al. (2020), Norway, RCT, ⁵⁶	57/56, ACS, CAD, valve, cardiac arrest, AF Two dropouts in intervention group, no dropouts in control group. RR could not be calculated.	59. 22.1%	UC (*)	Access to an app, individual goals could be set, notes could be written, automated evaluations and reminders of tasks and goals would be sent. Feedback was given through mail by the same supervisor. Tailored feedback from the app would be sent 1–3x/week depending on the individual's app settings, and emails would be sent once a week for 12 weeks. Hereafter, every fourth week. *	Only digital, app + email (mHealth) [mHealth + eHealth)	52 (0)	PFIT (VO₂ peak ml/kg/min) HRQoL (EuroQol-5 Domain)	SAE (no events during cardiopulmonary exercise test
Park et al. (2021), USA, pilot RCT, ⁵⁸	32/28, CHD, HF, valvular heart disease, structural heart disease 26/25 RR = 1.8 (0.5; 6.4)	66.8. 21.6%	Minimal intervention (non-blinded pedometer, walking 3D, Icefox, worn at waist, daily steps recorded in a diary, received two phone calls)	Wore an accelerometer, Charge 2 (Fitbit Inc.), at the wrist to record steps, had access to app used to record exercise data, send push-through motivational and educational messages. Another app could be connected to the Charge 2 to view step counts. Messages were sent from the study team 3x/week. In the app, participants could record daily weight, blood pressure, heart rate, medication use, and exercise not recorded. Both groups received 2x phone calls to verify adherence and to answer questions regarding the study. (Social Cognitive Theory)	Only digital, wearable + app + phone calls (digital device) [digital device + mHealth]	8 (0)	Obj PA (Steps/day) pedometer Obj PF (6-min-walk-test)	*
Pinto et al. (2011), USA, RCT, ⁵⁷	64/66, * 44/52 RR = 1.5 (0.8; 2.7)	63.6. 20.8%	Minimal intervention (received phone calls)	Wore a pedometer (Digiwalker; Yamax Corporation, Tokyo, Japan) during exercise activities that involved walking and received phone calls from the intervention coordinator. Weekly phone calls in the first 2 months, bi-weekly for the next 2 months, and monthly for the last 2 months. A total of 14 phone calls. (Transtheoretical Model & Social Cognitive Theory)	Only digital, wearable + phone calls (digital device) [digital device + mHealth]	52 (0)	Subj PA (Seven-Day Physical Activity Recall Exercise; minutes/week) Subj PF (SF-36 physical function) HRQoL (SF-36 (physical function subscale) not included in HRQoL analysis)	Adverse effects (no events)
Pinto et al. (2013), USA, RCT, ⁵⁹	64/66, * 44/52 RR = 1.5 (0.8; 2.7)	63.6. 20.8%	Minimal intervention (received phone calls)	Wore a pedometer (Digiwalker; Yamax Corporation, Tokyo, Japan) during exercise activities that involved walking and received phone calls from the intervention coordinator. Weekly phone calls in the first 2 months, bi-weekly for the next 2 months, and monthly for the last 2 months. A total of 14 phone calls. (Transtheoretical Model & Social Cognitive Theory)	Only digital, wearable + phone calls (digital device) [digital device + mHealth]	52 (0)	HRQoL (SF-36 (Mental Health subscale))	*
Reid et al. (2021), Canada, RCT, ⁶⁰	226/223, PCI, CABG, MI, angina, HF 146/160 RR = 1.3 (1.0; 1.7)	64.4. 30.1%	UC (exercise prescription based on exit test, shared with primary care provider. Information on community exercise locations)	Wore an accelerometer (Actigraph GT3X+) over the right hip. Combination of face-to-face, group teleconference, and individual telephone counselling was offered. During the sessions, participants reviewed their activities, identified experienced barriers regarding exercise maintenance, and brainstormed a solution with the group. Physical activity of moderate-vigorous intensity was recommended for ≥150 min/week. A facilitator helped with developing plans for adhering to the exercise standard. (Used factors known to influence exercise behaviours)	As component, wearables + teleconference + phone call (digital device) [digital device + mHealth + eHealth]	50 (2)	Obj PA (weekly MVPA; non-bouted minutes) accelerometer PFIT (VO₂ max (ml/kg/min)) HRQoL (EuroQol-5 Domain)	*
Skobel et al. (2016), Germany, Great Britain, Spain, RCT, ⁵⁵	55/63, CAD 19/42 RR = 2.0 (1.3; 2.9)	59. 11%	UC (national standard; writing daily PA in a diary)	Portable station (vest and an electrocardiography sensor to record vitals and a smartphone to view vitals). The smartphone app based the individual training on the vitals and provided immediate training intensity feedback. The healthcare professionals had access to a website which showed each participants medical history, exercise prescription, and vitals. (Frequency, intensity, time and type of activity principles model).	Only digital, wearable + feedback in app (digital device) [digital device + mHealth)	26 (0)	PFIT (VO₂ peak (ml/kg/min)) HRQoL (EuroQol-5 Domain)	SAE (chest pain; hospitalised; atrial fibrillation; angina; pseudo aneurysm) RR = 3.6 (1.0; 13.2)
Snoek et al. (2021), the Netherlands, RCT, ⁶¹	61/61, ACS, PCI, CABG 58/60 RR = 3.0 (0.3; 28.0)	59.5. 18%	Minimal intervention (received phone calls)	Wore a Bluetooth connected heart rate belt (Zephyr, Annapolis, USA) and a smartphone (Samsung Galaxy Ace GT-s5830i, Korea). Training history available on webpage and on their phone. Received phone calls weekly first month, bi-weekly in the second month and then monthly. Recommended 30 min of exercise 5x/week of moderate intensity. Both groups received phone calls every month to register additional care utilisation and adverse events. Both groups were seen by a cardiologist by three and 12 months post cardiac rehabilitation, which was considered usual care. *	Only digital, smart phone + phone call + wearable + individual webpage (digital device) [digital device + mHealth + eHealth]	26 (26)	Subj PA (total PA; MET-minutes/week, International Physical Activity Questionnaire) PFIT (VO₂ peak (ml/kg/min)) HRQoL (Quality of Life after Myocardial Infarction questionnaire)	NSAE, SAE (outpatient visit; general practitioner visit, hospitalisation; procedure; adverse event) *
Ter Hoeve et al. (2017), the Netherlands, three-arm RCT, ⁴⁹	240/245, MI, angina, PCI, CABG, stroke, transient ischemic attack 183/183 RR = 0.9 (0.7; 1.3)	58.7. 17.1%	UC (health information on PA)	Received information about risk factors and received 5–6x individual coaching sessions over the phone over a period of 9 months by a nurse. During the sessions, psychosocial problems were discussed, and a plan was developed for a heart-healthy lifestyle (PA, diet, smoking cessation, medicine adherence). After each session, the participant would receive a written summary of the subjects covered and the agreements made. *	Only digital, phone calls + written overview of discussion + wearable (mHealth) [mHealth + digital device]	39 (26)	Obj PA (MVPA (% of wear time)) accelerometer	*
Vieira et al. (2017), Portugal, three-arm RCT, ⁶³	15/16, ACS, AP, post-angioplasty 11/11 RR = 0.9 (0.3; 2.6)	57. *	Minimal intervention (received pamphlets on PA, diet, smoking; education about risk factors and received phone calls)	Received pamphlets on PA, diet, smoking; virtual reality headset (Kinect, Microsoft) which was connected to each participant's computer. A watch (Polar Wearlink) measured heart rate and repetitions during training. The virtual reality headset was worn during exercise and performed the Kinect-RehabPlay. Received weekly text messages and emails to encourage program adherence. All groups received phone calls at week 4, 10, and 22. The intervention group received an additional phone call at week 6 and 18. Home visits or in-person meetings at week 6 and 18 to adjust exercise. *	As component, virtual reality headset + wearable + phone calls + emails (digital device) [digital device + mHealth + eHealth]	26 (0)	Subj PA (the MacNew Heart Disease Health-related Quality of Life questionnaire (physical subscale)) HRQoL (the MacNew Heart Disease Health-related Quality of Life questionnaire (total score))	*
Yates et al. (2005), USA, three-arm RCT, ⁴⁵	24/20, CABG, MI, PCI 23/18 RR = 0.4 (0.01; 4.3)	66.7. 31%	Minimal intervention (received phone calls)	The intervention group received structured phone calls at week three and nine post cardiac rehabilitation. Goals set during cardiac rehabilitation was used as basis for the session. Those reporting progress towards their goal were praised and encouraged. (Bandura's self-efficacy theory).	Only digital, structured phone call (mHealth) [mHealth]	9 (17)	Subj PF (SF-36 (physical function subscale))	*

Note. ACS: acute coronary syndrome; AF: atrial fibrillation; AP: angina pectoris; CABG: coronary artery bypass graft surgery; CAD: coronary artery disease; CHD: coronary heart disease; CHF: congestive heart failure; CVD: cardiovascular disease; HF: heart failure; HRQoL: health-related quality of life; MET: metabolic equivalent; MI: myocardial infarction; MVPA: moderate–vigorous physical activity; Obj PA: objective physical activity; Obj PF: objective physical function; PA: physical activity; PCI: percutaneous coronary intervention; PFIT: physical fitness; PTCA: percutaneous transluminal coronary angioplasty; SF-36: Short Form 36; Subj PA: subjective physical activity; Subj PF: subjective physical function; UC: usual care; *: no available information or data.

The median digital health intervention duration was 26 weeks ranging from 6 weeks⁴⁸ to 208 weeks.⁵⁰ All studies included a mHealth component in their interventions, utilised through SMS,^47,54,63,64 phone calls^{45,47–50,52,57–62} or phone apps.^{46,55,56,58,64} In the intervention by Avila et al.,^51,53 it was not stated whether feedback given via phone utilised SMS or phone calls. The intervention was solely digital in 13 studies,^{45–49,54–59,61,62} where the interventions primarily involved apps,^{46,55,56,58,61} wearables (e.g., accelerometers)^{47,48,57–59,61,62} and web-based tools.^54,61 The remaining seven studies^{50–53,60,63,64} integrated digital health interventions with in-person components, including face-to-face counselling,^52,60,63 supervised exercise sessions^50,51,53 and exercise booster sessions.^50,64 Seven studies based their intervention on a model or a theory^{45,50,54,55,57,58,60} while the remainder did not. Only six studies reported follow-up after completion of the digital health intervention^{45,48,49,51,60,61}; follow-up durations ranged from 2 weeks⁶⁰ to 39 weeks,⁵¹ and no studies reported at the 12-month follow-up.

The primary outcome, objective PA, was measured using accelerometers^{47,49,51,60,62} and pedometers,⁵⁸ primarily reported as moderate PA,⁵¹ moderate to high PA⁴⁷ and moderate to vigorous PA.^49,60 Furthermore, subjective PA was measured using questionnaires,^{46,48,50,54,57,61,63,64} primarily reported as leisure PA^50,64 and total PA.^48,54,61 Of note, secondary outcomes for objective physical function^46,51,64 was primarily measured using a 6-min-walk-test;^58,64 physical fitness was primarily measured using a VO₂ peak,^{47,51,53,55,56,60,61} and HRQoL was primarily measured using European Quality of life - 5 Dimensions.^55,60

Study quality

Risk of bias assessments is shown in Figure 2.

Figure 2.

Risk of bias on all included studies and categorised based on the assessed outcome.

Risk of bias assessments for objectively measured PA: 16.5% (1/6)⁴⁷ of the RCTs was rated as ‘low risk of bias’, 67% (4/6)^49,51,60,62 as ‘some concerns’ and 16.5% (1/6)⁵⁸ as ‘high risk of bias’. Subjectively measured PA was rated as ‘some concerns’ in 75% (6/8) of the RCTs^{46,48,57,61,63,64} and ‘high risk of bias’ in 25% (2/8) of the RCTs.^50,54 All (4/4)^46,51,58,64 of the RCTs that measured objectively and subjectively (4/4):^45,51,53,57 physical function were rated as ‘some concerns’. For physical fitness, 18% (2/11) of RCTs^52,55 was rated as ‘high risk of bias’, while the remaining 82% (9/11)^{46,47,50,51,53,56,60,61,64} was rated as ‘some concerns’. For HRQoL, one RCT – 11% (1/9)⁵⁶ was rated as ‘low risk of bias’, 67% (6/9) as ‘some concerns’^{51,53,59–61,63} and 22% (2/9) as ‘high risk of bias’.^50,55

The effects of digital health interventions

Objectively measured PA

Four studies (357 participants) assessed the effect of digital health interventions on PA.^47,49,58,62 We found no effect on objective PA, when digital health interventions were compared to usual care or minimal intervention at the end of intervention (K = 4, SMD 0.85, 95% CI: −0.07 to 1.77, I²= 90.37%) (Figure 3). After completing the digital health intervention, at 6-month follow-up, no effect was found between groups (K = 2, SMD 0.02, 95% CI: −0.13 to 0.17, I²= 0.01%) (Fig. E.1).

Figure 3.

Forest plots of standardised mean difference for (a) objectively measured and (b) subjectively measured physical activity in digital health interventions at the end of intervention.

Subjectively measured PA

Eight studies (711 participants) investigated the effect of digital health interventions on subjectively measured PA,.^{46,48,50,54,57,61,63,64} The results showed a small effect on subjective PA when comparing digital health interventions to usual care or minimal intervention at the end of intervention (K = 8, SMD 0.37, (95% CI: 0.05 to 0.69), I²= 73.42%) (Figure 3).

WMD for two studies (136 participants)^54,61 using International Physical Activity Questionnaire (IPAQ) showed no effect between groups (K = 2, WMD 1384.55 MET-min/week, (95% CI: −1273.76 to 4042.85), I²= 77.81%) (Fig. E.2). After the digital health intervention, at 6-month follow-up, no effect was found between groups (K = 2, SMD 0.11, (95% CI: −0.25 to 0.48), I²= 43.65%) (Fig. E.3).

Effects of digital health on physical function

Objectively measured physical function

Three studies (161 participants)^46,58,64 were included for analysis. A moderate effect for objectively measured physical function was found favouring digital health interventions compared to usual care or minimal intervention at the end of intervention (K = 3, SMD 0.63, (95% CI: 0.03 to 1.24), I² = 70.73%) (Figure 4). WMD for two studies (131 participants)^58,64 using a 6-minute walk test showed no effect between groups (K = 2, WMD 40.57 meter, (95% CI: −17.65 to 98.78), I²= 77.81%) (Fig. E.4). After the digital health intervention, one study⁵¹ measured physical function objectively at 9-month follow-up, reporting no effect on the Sit and Rising test (intervention group: N = 16, Sit and Rising test score 7.9, SD 1.3; control group: N = 18, score 7.3, SD 1.3 (between group p = 0.53)).

Figure 4.

Forest plot of standardised mean difference for (a) objectively measured and weighted mean difference for (b) subjectively measured physical function in digital health interventions at the end of intervention.

Subjectively measured physical function

Three studies (191 participants)^45,53,57 assessed the effect of digital health interventions on subjective physical function all using the Short Form 36 physical function subscale, showing a WMD of 8.47 points (K = 3, (95% CI: 3.83 to 13.11), I²= 0.00%), favouring digital health interventions compared to usual care or minimal interventions at the end of intervention (Figure 4).

Effects of digital health on physical fitness

Nine studies (748 participants)^{46,47,50,52,53,55,56,61,64} assessed the effect of digital health interventions on physical fitness. Little to no effect for physical fitness was found, favouring digital health interventions compared to usual care or minimal intervention at the end of intervention (K = 9, SMD 0.19, (95% CI: 0.05 to 0.34), I²= 0.00%) (Figure 5). WMD for the VO₂ peak at the end of intervention showed an effect favouring the digital health interventions (K = 5, WMD 1.75 ml/kg/min, (95% CI: 0.47 to 3.02), I²= 0.00%) (Fig. E.5). After the digital health intervention, at 6-month follow-up, this was not maintained (K = 2, WMD 0.67 ml/kg/min, (95% CI: −0.79 to 2.13), I²= 0.00%) (Fig. E.6). The same conclusion was found in one study [51] measuring the VO₂ peak at 9-month follow-up (intervention group: N = 26, VO₂ peak: 27.1, SD 6.5; control group: N = 25, VO₂ peak: 26.2, SD 7.6 (between group p = 0.54)).

Figure 5.

Forest plot of standardised mean difference for (a) physical fitness and (b) health-related quality of life in digital health interventions at the end of intervention. In (a), heterogeneity is reported as I².

Effects of digital health on health-related quality of life

Seven studies (702 participants)^{50,53,55,56,59,61,63} investigated the effect on HRQoL. Little to no effect was found for digital health interventions compared to usual care or minimal intervention at the end of intervention (K = 7, SMD 0.13, (95% CI: −0.02 to 0.28), I²= 0.00%) (Figure 5). WMD for two studies (163 participants)^55,56 using the EuroQol-5 VAS at the end of intervention showed no effect between groups (K = 2, WMD 2.57 VAS score, (95% CI: −1.81 to 6.95), I²= 00.0%) (Fig. E.7). Two studies^60,61 (567 participants) reported data on HRQoL at 6-month follow-up (Fig. E.8), which showed no effect between groups (K = 2, SMD −0.04, (95% CI: −0.20 to 0.13), I²= 0.00%).

Adverse events

In total, seven studies^{46,53,55–57,61,64} reported on adverse events within their respective study periods, and adverse events were not reported solely in relation to the applied digital health solution. More details are shown in Table 2. The varied reporting methods for adverse events made quantitative analysis unfeasible. A narrative approach was therefore taken instead.

One study⁵⁵ reported that 24% (6/25 participants) in the digital health group experienced a serious adverse event (chest pain, infection after coronary artery bypass grafting and chest pain before training) compared to 7% (3/45) of the control group (atrial fibrillation, new angina and pseudo-aneurysm). One study⁶¹ reported the number of study participants presenting with at least one serious adverse event for each of the following categorisations (the same patient could be present in several categories): cardiac-related hospitalisations (intervention: 7; control group: 10), procedures (intervention: 19; control group: 19) and adverse events (intervention: 13; control group: 16). Non-serious adverse events were reported as outpatient visits (intervention: 18; control group: 22) and general practitioner visits (intervention: 43; control group: 46). Lunde et al.,⁵⁶ Foccardi et al.⁴⁶ and Avila et al. (2018)⁵³ reported adverse events but only if it occurred during or in relation to exercise⁵³ or exercise testing.^46,56 Lunde et al.⁵⁶ and Pinto et al. (2011)⁵⁷ did not observe any adverse events; however, the flow diagrams mentioned a case of death (intervention: 1; control group: 0 ⁵⁶; intervention: 1; control group: 1 ⁵⁷) and medical issues (intervention: 8; control group: 6 ⁵⁷). Avila et al. (2018)⁵³ observed no serious adverse events but reported two dropouts in the control group due to percutaneous coronary intervention, indicating a serious adverse event. Labrador et al.⁶⁴ observed no adverse events and reported no dropouts.

Heterogeneity

Heterogeneity was only found for objectively, subjectively measured PA and objectively measured physical function. None of the meta-regressions may explain the heterogeneity for objectively or subjectively measured PA, but age and proportion female may explain the heterogeneity for the objectively measured physical function (p = 0.01) (Table F.1–3). Only the pre-specified sub-group analyses (Fig. F.1–15) for the primary digital health component (mobile, internet based or digital device) (p = 0.03) may explain some of the heterogeneity for objective PA, and the type of cardiac diagnosis (p = 0.01) and the type of digital health use (exclusively or as component) may explain some of the heterogeneity for subjective PA (p = 0.05). For objective physical function, only the primary digital health component (p = 0.01) may explain the heterogeneity.

Evidence level and small study bias

The certainty of evidence was very low for objective and subjective PA and objective physical function and low for subjective physical function, physical fitness and HRQoL. The certainty of evidence was (Table 3) evaluated accordingly to GRADE.

Table 3.

Summary of the findings – the effectiveness of digital health interventions after CR phase II completion compared to minimal intervention or usual care.

Certainty assessment							Number of patients		Effect	Certainty of evidence
Number of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	Digital health interventions	Usual care or minimal intervention	Absolute(95% CI)	Certainty of evidence
Physical activity objectively measured
4	RCTs	Serious^a	Very serious^b	Not serious	Serious^c	Publication bias strongly suspected^d	180	177	SMD 0.85 SD (−0.07 to 1.77)	⨁◯ ◯ ◯ Very low^f
Physical activity subjectively measured
8	RCTs	Serious^a	Serious^e	Not serious	Serious^c	Publication bias strongly suspected^a,d	354	357	SMD 0.37 SD (0.05 to 0.69)	⨁ ◯ ◯ ◯ Very low^f
Physical function objectively measured
3	RCTs	Serious^a	Serious^e	Not serious	Serious^c	None	81	80	SMD 0.63 SD (0.03 to 1.24)	⨁ ◯ ◯ ◯ Very low^f
Physical function subjectively measured
3	RCTs	Serious^a	Not serious	Not serious	Serious^c	None	95	96	MD 8.47 score (3.83 to 13.11)	⨁ ⨁ ◯ ◯ Low^f
Physical fitness
9	RCTs	Serious^a	Not serious	Not serious	Serious^c	None	370	378	SMD 0.19 SD (0.05 to 0.34)	⨁ ⨁ ◯ ◯ Low^f
Health-related quality of life
7	RCTs	Serious^a	Not serious	Not serious	Serious^c	None	339	363	SMD 0.13 SD (−0.02 to 0.28)	⨁ ⨁ ◯ ◯ Low^f

Note. CI: confidence interval; MD: mean difference; RCTs: randomised controlled trials; SMD: standardised mean difference.

Majority of studies had some concerns.

I² is considerable >90%.

CI overlaps the line of no effect.

Indicated by the funnel plot, Egger's test score and the number of studies.

I² is substantial >50%.

Important outcome for patients with cardiovascular disease.

Visual inspection of funnel plots (Fig. F.16–21) only indicated publication bias in the outcome domain for objective and subjective PA, which was confirmed by the p-values calculated by Egger's test: subjective PA (p = 0.0001); objective PA (p = 0.004).

Discussion

We assessed the effectiveness of digital health interventions on PA, other physical outcomes, HRQoL and adverse events after CR II completion. Digital health interventions supporting PA maintenance showed a small effect on subjectively assessed PA but no effect on objectively assessed PA at the end of the digital health intervention or at 6-month follow-up. However, the evidence was very uncertain. For physical function, digital health interventions showed a moderate effect, whereas digital health interventions may have little to no effect on physical fitness and HRQoL at the end of intervention. The quality of evidence was low to very low, with few studies reporting results at 6-month follow-up, and none at 12-month follow-up. Furthermore, adverse events were poorly reported. High-quality studies on digital health interventions with longer follow-up periods are needed to investigate which type of digital component can provide the most optimal support to maintain PA health behaviour.

PA is a core component in rehabilitation for patients with cardiovascular disease as it positively affects prognosis and HRQoL, and is crucial for reducing mortality.^1,2,65 Furthermore, PA is an important secondary prevention component essential for managing and prolonging a healthier life.^1,66 Based on the current evidence, we cannot confirm or deny findings from previous systematic reviews investigating the effect of digital health intervention to support patients’ maintenance of PA after CR II completion. Martinello et al.²³ (2019) reported that either in-person or digital health (i.e., pedometer, telemonitoring, online or mobile component) maintenance interventions had a large effect on PA levels. Unfortunately, they did not conduct a sub-analysis on the mode of delivery of digital health interventions and in-person interventions in their PA analysis, which is why a conclusion of an isolated effect on digital health interventions cannot be drawn from this study.²³ Heimer et al.²⁵ (2023) reported a small effect on PA levels from eHealth interventions that aimed to maintain PA after CR comparison to usual care. However, a direct comparison with Heimer et al.²⁵ is not appropriate, as they combined subjective and objective measurements in their analysis. In contrast to our review, all previously mentioned reviews either combined the subjective and objective outcome data in their PA analysis or reported exclusively on objective PA. Objective measurements provide a more precise evaluation of PA levels and intensity compared to subjective measures, which tend to overestimate these parameters.¹⁹ Subjective measurements, however, capture the individual's self-perceived activity level.²⁶ Hence, subjective and objective assessment tools on PA assess two different aspects of PA, which is why it is important to analyse them separately.²⁶ Hence, by separating subjective and objective PA measurements and adding a broader spectrum of digital health interventions, our findings add important knowledge to the existing literature.

In our PA analysis, we observed a high level of heterogeneity, which affected our ability to conclude the overall effects. Despite our planned sub-analyses, the underlying cause remained unidentified. However, some of the heterogeneity could be explained by how PA was measured, e.g., type and placement of the PA assessment wearables, i.e., on the wrist⁵⁸ or the waist.^49,58,62 In the study by Park et al.,⁵⁸ PA was assessed using a Fitbit device for the intervention group (worn on the waist) and a pedometer for the control group (worn on the wrist), introducing measurement bias into the meta-analysis of objective PA. The placement of the wearable may affect the ability to detect certain activities.²⁷ Different wearables may have different cut-points for activity-related intensity levels, and may, therefore, have different abilities in measuring PA.⁶⁷ Some of the heterogeneity may also be explained by the type of cardiac diagnosis, type of primary used digital component (mobile, internet based or digital device), and type of intervention (exclusively digital health or digital health + in-person). Heterogeneity may also be explained by the complexity of the digital health interventions, which vary across studies. Some interventions only include phone calls,⁴⁵ while others incorporate virtual reality headsets.⁶³ The number of digital health components employed also varies; for instance, some studies include only a single digital health component, such as a phone call,⁴⁹ whereas others involve two components (wearable devices and phone calls),^47,62 and some incorporate three components (wearable devices, apps and phone calls).⁵⁸ The impact of all these varying aspects should be considered and explored in future studies.

Our findings of little to no effect for physical fitness are similar to the overall findings by Heimer et al.²⁵ The WMD analysis for the VO₂ peak illustrates an increase of 1.75 ml/kg/min favouring digital health interventions. Typically, an increase of 3.5 ml/kg/min (1 MET) is considered clinically relevant as it has been shown to improve the survival rate by 12% in patients with cardiovascular disease.^68,69 However, an increase of 1 ml/kg/min has been associated with a 15% decrease in risk for all-cause and cardiovascular disease-specific mortality.⁷⁰ Therefore, the increase of the 1.75 ml/kg/min may have some clinical relevance. At 6-month follow-up, the effect was not sustained (WMD of 0.67 ml/kg/min).

We found three studies indicating a moderate effect of digital health intervention on objective physical function. In contrast, Hannan et al.²⁴ found only one study that assessed physical function using the 6-minute walk test, but they found no between-group difference at 6 weeks or 6-month follow-up. Thus, the limited evidence on digital health maintenance interventions makes the effectiveness of physical function uncertain. Qualitative research indicates that digital health could potentially support patients and may improve their HRQoL.⁷¹ Yet, it is not supported by our findings. Improving well-being and quality of life for individuals with cardiovascular diseases is the overarching aim of cardiac rehabilitation.⁵ Exploring various ways in which digital health interventions can contribute to enhancing patients’ HRQoL should be investigated.

For the control group, measurements of PA were only considered minimal intervention if participants in the comparator group were not blinded to their PA levels; e.g., Park et al.⁵⁸ was considered minimal intervention as the control group monitored PA using a pedometer and recorded their steps in a diary. If both groups were unaware of their PA levels, the Hawthorne effect⁷² should be present in both groups. In some studies,^{47,48,51,53,62} intervention group participants were aware of their PA levels during measurements and received feedback on the prior week's performance. In the intervention group, this feedback or awareness of PA levels was expected to have an additional effect beyond the Hawthorne effect.

We define maintenance as a period of at least 6 months with the goal of keeping the benefits gained and preventing relapse.¹⁶ Our analyses show the trajectory of effectiveness of digital health analysed at the end of the intervention, closest to 6-month follow-up, and closest to 12-month follow-up. However, our findings show that most of the included studies assessed outcomes only at the end of the digital health intervention.^{46,47,50,52,54–56,58,62–64} Despite this, all of the studies used the maintenance term, except two,^55,63 to describe their intervention while having no follow-up^{46,47,50,52,54,56,58,62,64} or a short follow-up period (under 6 months).^{45,48,49,60,61} One could question whether the included studies truly investigated maintenance. No studies clearly defined when maintenance was considered successful. We recommend future studies to include longer follow-up periods and clearly describe their definition for maintenance.

An interesting observation was that in studies where exercise was prescribed at the start of the digital health intervention followed by weekly^{47,53,54,58,64} or bi-weekly⁶² contact through text messages,^47,53,54,64 messages in app,⁵⁸ phone calls^47,62,64 or emails,⁵³ the effect size tends to be higher. This is contrary to studies where digital contact with the patients was only made through phone calls, averaging throughout the intervention duration every 2 weeks,⁶¹ 4–5 weeks,⁴⁸ 8 weeks⁴⁹ and 11 weeks.⁵⁰ Hence, associations between intervals of digital interactions and patient-related outcomes are worth exploring more in depth.

Implications for clinical practice and future research

The aim of our review was not to understand the deeper effective mechanisms when applying digital health interventions to patients with cardiac diagnoses. However, to tailor effective future digital interventions and help explain some of the current heterogeneity, which is an important research area. Digital health is accessible for many and can be tailored to patients’ individual needs and preferences, encouraging healthy behaviour.^19,20 Current evidence highlights that digital health interventions applied to support PA after phase II CR are heterogeneous in type and often combine several types of behavioural change techniques.²⁵ Incorporating behaviour change theory and behavioural change techniques into an intervention has been found to encourage changes in health-related behaviour, especially social support and action planning have been found effective in improving PA.⁷³ Still, the research field needs to explore the optimal dose and combination of digital health to promote or maintain PA, as numerous combinations and scant evidence in the field of digital health interventions compromise any solid clinical recommendation. Nevertheless, given that our data suggest a favourable effect on PA and physical function, and that digital health solutions were previously reported to be cost-effective for improving health-outcomes⁷⁴ and could improve accessibility in healthcare for rural or remote populations,⁷⁵ healthcare providers could consider to implement digital health into clinical practice. Yet, it is important to tailor interventions based on patients’ preferences and needs, and healthcare providers should consider that digital health might not be the optimal solution for all patients.⁷⁶ The evidence includes limited reporting on adverse events, and the possibility of adverse events during a digital health intervention cannot be dismissed. Still, given the risks associated with inactivity, the benefits of increasing or maintaining exercise seem to outweigh any potential associated risks.^1,2 However, future studies need to enhance their reporting on adverse events following the recommendations presented by The Consolidated Standards of Reporting Trials.³⁷ Future studies should consider the complexity, the number of digital health components used, the intervals between digital interactions with the study participants, the duration of the digital health intervention and using a follow-up period of at least 6 months after completing the digital health intervention.

Limitations

The present systematic review has several strengths. First, our review was pre-registered at PROSPERO and adhered to PRISMA guidelines, ensuring transparency and completeness in reporting. Second, only RCTs were included. Third, the methodology followed the recommendations from Cochrane.⁷⁷ Last, our search was broader than previous systematic reviews, identifying seven additional relevant studies^{46,52,58–61,64} compared to existing reviews within the same field,^22–25 and we divided outcomes in objective and subjective PA assessment types.²⁶

Nevertheless, our review holds some limitations. First, our PA analyses had considerable heterogeneity. The Cochrane Handbook advises against performing subgroup analyses, meta-regressions or funnel plots when the number of studies is fewer than ten.⁴⁴ Therefore, until further research explores the heterogeneity further, our results should be interpreted with caution. Second, digital health interventions were primarily investigated among males, and the effect was not reported separately for males and females. Third, although we tried to mitigate the risk by manually searching references; there is a possibility that eligible studies might have been missed due to the use of RCT filters^31–33 in a search block. Fourth, the exclusion of studies that introduced digital health interventions during phase II CR limits our understanding of how early digital health integration affects patients during the transition between CR phases and its impact on long-term maintenance. Last, we only covered physical outcomes and HRQoL. Hence, the effectiveness of digital health maintenance interventions on other important outcomes such as return-to-work, anxiety, depression, re-hospitalisation and death were not investigated.

Conclusion and perspective

Current evidence shows that digital health interventions in patients with cardiovascular disease has a small effect on subjectively measured PA and a moderate effect on objective physical function but no effect on objectively measured PA and little to no effect on physical fitness and HRQoL compared to usual care or minimal intervention after completing phase II CR. However, the evidence is very uncertain across all outcomes, with few available studies, and potential harms poorly reported. Despite uncertain evidence, the health risks induced by physical inactivity, coupled with the potential health benefits of digital health interventions, may still justify the use of digital health in clinical practice to maintain PA until further evidence is available.

Supplemental Material

sj-docx-1-dhj-10.1177_20552076241286641 - Supplemental material for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis

Supplemental material, sj-docx-1-dhj-10.1177_20552076241286641 for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis by Alexander Luijk, Sofie Rath Mortensen, Trine Grønbek Hamborg, Graziella Zangger, Jonas Risum Ahler, Jan Christensen, Søren T Skou and Lars Hermann Tang in DIGITAL HEALTH

Supplemental Material

sj-docx-2-dhj-10.1177_20552076241286641 - Supplemental material for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis

Supplemental material, sj-docx-2-dhj-10.1177_20552076241286641 for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis by Alexander Luijk, Sofie Rath Mortensen, Trine Grønbek Hamborg, Graziella Zangger, Jonas Risum Ahler, Jan Christensen, Søren T Skou and Lars Hermann Tang in DIGITAL HEALTH

Supplemental Material

sj-xlsx-3-dhj-10.1177_20552076241286641 - Supplemental material for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis

Supplemental material, sj-xlsx-3-dhj-10.1177_20552076241286641 for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis by Alexander Luijk, Sofie Rath Mortensen, Trine Grønbek Hamborg, Graziella Zangger, Jonas Risum Ahler, Jan Christensen, Søren T Skou and Lars Hermann Tang in DIGITAL HEALTH

Supplemental Material

sj-docx-4-dhj-10.1177_20552076241286641 - Supplemental material for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis

Supplemental material, sj-docx-4-dhj-10.1177_20552076241286641 for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis by Alexander Luijk, Sofie Rath Mortensen, Trine Grønbek Hamborg, Graziella Zangger, Jonas Risum Ahler, Jan Christensen, Søren T Skou and Lars Hermann Tang in DIGITAL HEALTH

Supplemental Material

sj-docx-5-dhj-10.1177_20552076241286641 - Supplemental material for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis

Supplemental material, sj-docx-5-dhj-10.1177_20552076241286641 for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis by Alexander Luijk, Sofie Rath Mortensen, Trine Grønbek Hamborg, Graziella Zangger, Jonas Risum Ahler, Jan Christensen, Søren T Skou and Lars Hermann Tang in DIGITAL HEALTH

Supplemental Material

sj-docx-6-dhj-10.1177_20552076241286641 - Supplemental material for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis

Supplemental material, sj-docx-6-dhj-10.1177_20552076241286641 for The effectiveness of digital health interventions for the maintenance of physical activity following cardiac rehabilitation: A systematic review and meta-analysis by Alexander Luijk, Sofie Rath Mortensen, Trine Grønbek Hamborg, Graziella Zangger, Jonas Risum Ahler, Jan Christensen, Søren T Skou and Lars Hermann Tang in DIGITAL HEALTH

Footnotes

List of abbreviations

Acknowledgements

We want to express gratitude to Konstantinos Antypas, Sherry L. Grace and Jomme Claus for providing us with additional data and information. The authors used a large language model (ChatGPT, version 3.5) solely to improve grammar and structure. The authors verified the accuracy and appropriateness of any suggested corrections proposed by ChatGPT. All content, analyses, and conclusions presented in this review were solely the work of the authors and were not influenced by large language models or artificial intelligence.

Contributorship

All authors contributed to the development of the protocol. AL, GZ, JC, STS and LHT contributed to the design. AL performed the systematic literature search, and AL, SRM, TGH and JRA screened the studies, extracted the data and assessed the studies using RoB 2. AL and JRA assessed the certainty of evidence using GRADE. AL, GZ and JC performed the meta-analysis, and AL, JC, STS and LHT interpreted the results. AL and LHT wrote the manuscript. All authors contributed to the revision of the manuscript and approved the final version of the manuscript.

Declaration of conflicting interests

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

Ethical approval and consent to participate

This study is a systematic review and meta-analysis of existing literature, and no ethics approval or consent to participate was required.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Region Zealand funded the project (Exercise First research program), but was not involved in any part of designing, conducting or reporting the study. Luijk, Dr Tang and Dr Skou were funded by a grant from Region Zealand (Project: Exercise First). Dr Tang was further funded by The Danish Health Confederation through the Development and Research Fund (project no. 2703) and the Næstved-Slagelse-Ringsted Hospitals research fund, Denmark (project no. A1277). Dr Skou was further funded by two grants from the European Union's Horizon 2020 research and innovation program – one from the European Research Council (MOBILIZE, grant agreement no. 801790) and the other under grant agreement no. 945377 (ESCAPE).

ORCID iDs

Alexander Luijk

Trine Grønbek Hamborg

Supplemental Material

Supplemental material for this article is available online.

References

Alves

Viana

Cavalcante

, et al. Physical activity in primary and secondary prevention of cardiovascular disease: overview updated. World J Cardiol 2016; 8: 575–583.

Tucker

Fegers-Wustrow

Halle

, et al. Exercise for primary and secondary prevention of cardiovascular disease: JACC focus seminar 1/4. J Am Coll Cardiol 2022; 80: 1091–1106.

Heran

Chen

Ebrahim

, et al. Exercise-based cardiac rehabilitation for coronary heart disease. Cochrane Database Syst Rev 2011: Cd001800. 2011/07/08.

Dibben

Faulkner

Oldridge

, et al. Exercise-based cardiac rehabilitation for coronary heart disease. Cochrane Database Syst Rev 2021; 11: Cd001800. 20211106.

Tessler

Bordoni

. Cardiac rehabilitation. Available at: https://pubmed.ncbi.nlm.nih.gov/30725881/ (2023, accessed 16 November 2023).

Chaves

Turk-Adawi

Supervia

, et al. Cardiac rehabilitation dose around the world. Circ Cardiovasc Qual Outcomes 2020; 13: e005453.

Ramadi

Buijs

Threlfall

, et al. Long-term physical activity behavior after completion of traditional versus fast-track cardiac rehabilitation. J Cardiovasc Nurs 2016; 31: E1–e7. 2016/10/19.

Bock

Carmona-Barros

Esler

, et al. Program participation and physical activity maintenance after cardiac rehabilitation. Behav Modif 2003; 27: 37–53.

Moore

Charvat

Gordon

, et al. Effects of a CHANGE intervention to increase exercise maintenance following cardiac events. Ann Behav Med 2006; 31: 53–62. 2006/02/14.

10.

Dolansky

Stepanczuk

Charvat

, et al. Women's and men's exercise adherence after a cardiac event. Res Gerontol Nurs 2010; 3: 30–38. 2010/02/05.

11.

Bray

Brawley

Millen

. Relationship of proxy efficacy and reliance to home-based physical activity after cardiac rehabilitation. Rehabil Psychol 2006; 51: 224–231.

12.

Fleury

Lee

Matteson

, et al. Barriers to physical activity maintenance after cardiac rehabilitation. J Cardiopulm Rehabil 2004; 24: 296–305; quiz 306-297.

13.

Kwasnicka

Dombrowski

White

, et al. Theoretical explanations for maintenance of behaviour change: a systematic review of behaviour theories. Health Psychol Rev 2016; 10: 277–296. 20160307.

14.

Murray

Brennan

French

, et al. Mediators of behavior change maintenance in physical activity interventions for young and middle-aged adults: a systematic review. Ann Behav Med 2018; 52: 513–529.

15.

Kahlert

. Maintenance of physical activity: do we know what we are talking about? Prev Med Rep 2015; 2: 178–180.

16.

Norcross

Krebs

Prochaska

. Stages of change. J Clin Psychol 2011; 67: 143–154.

17.

Jiang

Ming

W-K

You

. The cost-effectiveness of digital health interventions on the management of cardiovascular diseases: systematic review. J Med Internet Res 2019; 21: e13166.

18.

Burrell

Zrubka

Champion

, et al. How useful are digital health terms for outcomes research? An ISPOR special interest group report. Value Health 2022; 25: 1469–1479.

19.

Fanning

Mullen

McAuley

. Increasing physical activity with mobile devices: a meta-analysis. J Med Internet Res 2012; 14: e161. 2012/11/23.

20.

Lettieri

Fumagalli

Radaelli

, et al. Empowering patients through eHealth: a case report of a pan-European project. BMC Health Serv Res 2015; 15: 309. 2015/08/06.

21.

Zangger

Bricca

Liaghat

, et al. Benefits and harms of digital health interventions promoting physical activity in people with chronic conditions: systematic review and meta-analysis. J Med Internet Res 2023; 25: e46439. 20230706.

22.

Graham

Prue-Owens

Kirby

, et al. Systematic review of interventions designed to maintain or increase physical activity post-cardiac rehabilitation phase II. Rehabilitation Process and Outcome 2020; 9: 1179572720941833.

23.

Martinello

Saunders

Reid

. The effectiveness of interventions to maintain exercise and physical activity in post-cardiac rehabilitation populations: a systematic review and meta-analysis of randomized controlled trials. J Cardiopulm Rehabil Prev 2019; 39: 161–167. 2019/04/26.

24.

Hannan

Harders

Hing

, et al. Impact of wearable physical activity monitoring devices with exercise prescription or advice in the maintenance phase of cardiac rehabilitation: systematic review and meta-analysis. BMC Sports Sci Med Rehabil 2019; 11: 14. 20190730.

25.

Heimer

Schmitz

Teschler

, et al. eHealth for maintenance cardiovascular rehabilitation: a systematic review and meta-analysis. Eur J Prev Cardiol 2023; 30: 1634–1651.

26.

Hagstromer

Ainsworth

Oja

, et al. Comparison of a subjective and an objective measure of physical activity in a population sample. J Phys Act Health 2010; 7: 541–550.

27.

Mannini

Intille

Rosenberger

, et al. Activity recognition using a single accelerometer placed at the wrist or ankle. Med Sci Sports Exercise 2013; 45: 2193–2203.

28.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Br Med J 2021; 372: 71.

29.

Higgins J, Thomas J, Chandler J, et al. (eds). Cochrane handbook for systematic reviews of interventions (version 6.4). Cochrane, 2023.

30.

Schlosser

Wendt

Bhavnani

, et al. Use of information-seeking strategies for developing systematic reviews and engaging in evidence-based practice: the application of traditional and comprehensive pearl growing. A review. Int J Lang Commun Disord 2006; 41: 567–582.

31.

Wong

Wilczynski

Haynes

. Developing optimal search strategies for detecting clinically sound treatment studies in EMBASE. J Med Libr Assoc 2006; 94: 41–47.

32.

Haynes

McKibbon

Wilczynski

, et al. Optimal search strategies for retrieving scientifically strong studies of treatment from Medline: analytical survey. Br Med J 2005; 330: 1179. 20050513.

33.

Glanville

Dooley

Wisniewski

, et al. Development of a search filter to identify reports of controlled clinical trials within CINAHL Plus. Health Info Libr J 2019; 36: 73–90. 20190208.

34.

Jackson

Kuriyama

Anton

, et al. The accuracy of google translate for abstracting data from non-English-language trials for systematic reviews. Ann Intern Med 2019; 171: 677–679. 20190730.

35.

Goossen

Rombey

Kugler

, et al. Author queries via email text elicited high response and took less reviewer time than data forms - a randomised study within a review. J Clin Epidemiol 2021; 135: 1–9. 20210209.

36.

WebPlotDigitizer, https://automeris.io/WebPlotDigitizer/

37.

Junqueira

Zorzela

Golder

, et al. CONSORT Harms 2022 statement, explanation, and elaboration: updated guideline for the reporting of harms in randomized trials. J Clin Epidemiol 2023; 158: 149–165. 20230424.

38.

FDA. What is a serious adverse event? https://www.fda.gov/safety/reporting-serious-problems-fda/what-serious-adverse-event (2022, accessed 15 December 2022).

39.

Sterne

JAC

Savović

Page

, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. Br Med J 2019; 366: l4898.

40.

Current version of RoB 2. https://www.riskofbias.info/welcome/rob-2-0-tool/current-version-of-rob-2 (2019, accessed 17 January 2024).

41.

GRADEpro GDT: GRADEpro Guideline Development Tool. https://www.gradepro.org/ (2022, accessed 24 April 2023).

42.

Higgins

Thomas

. Cochrane handbook for systematic reviews of interventions version 6.3. https://training.cochrane.org/handbook (2022).

43.

Cohen

. Statistical power analysis for the behavioral sciences. 2nd ed. USA: Lawrence Erlbaum Associates, 1988.

44.

Deeks

Higgins

Altman

. Chapter 10: analysing data and undertaking meta-analyses. www.training.cochrane.org/handbook (2022).

45.

Yates

Anderson

Hertzog

, et al. Effectiveness of follow-up booster sessions in improving physical status after cardiac rehabilitation: health, behavioral, and clinical outcomes. Appl Nurs Res : ANR 2005; 18: 59–62.

46.

Foccardi

Vecchiato

Neunhaeuserer

, et al. Effectiveness of text messaging as an incentive to maintain physical activity after cardiac rehabilitation: a randomized controlled pilot study. Int J Environ Res Public Health 2021; 18: 6645.

47.

Duscha

Piner

Patel

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

48.

Butler

Furber

Phongsavan

, et al. Effects of a pedometer-based intervention on physical activity levels after cardiac rehabilitation: a randomized controlled trial. J Cardiopulm Rehabil Prev 2009; 29: 105–114.

49.

ter Hoeve

Sunamura

Stam

, et al. Effects of two behavioral cardiac rehabilitation interventions on physical activity: a randomized controlled trial. Int J Cardiol 2018; 255: 221–228.

50.

Lear

Spinelli

Linden

, et al. The extensive lifestyle management intervention (ELMI) after cardiac rehabilitation: a 4-year randomized controlled trial. Am Heart J 2006; 152: 333–339.

51.

Avila

Claes

Buys

, et al.

Home-based exercise with telemonitoring guidance in patients with coronary artery disease: does it improve long-term physical fitness?

Eur J Prev Cardiol 2020; 27: 367–377.

52.

Brubaker

Rejeski

Smith

, et al. A home-based maintenance exercise program after center-based cardiac rehabilitation: effects on blood lipids, body composition, and functional capacity. J Cardiopulm Rehabil 2000; 20: 50–56.

53.

Avila

Claes

Goetschalckx

, et al. Home-based rehabilitation with telemonitoring guidance for patients with coronary artery disease (short-term results of the TRiCH study): randomized controlled trial. J Med Internet Res 2018; 20: e225. 20180622.

54.

Antypas

Wangberg

. An internet- and mobile-based tailored intervention to enhance maintenance of physical activity after cardiac rehabilitation: short-term results of a randomized controlled trial. J Med Internet Res 2014; 16: 77.

55.

Skobel

Knackstedt

Martinez-Romero

, et al. Internet-based training of coronary artery patients: The heart cycle trial. Heart Vessels 2017; 32: 408–418.

56.

Lunde

Bye

Bergland

, et al. Long-term follow-up with a smartphone application improves exercise capacity post cardiac rehabilitation: a randomized controlled trial. Eur J Prev Cardiol 2020; 27: 1782–1792.

57.

Pinto

Goldstein

Papandonatos

, et al. Maintenance of exercise after phase II cardiac rehabilitation: a randomized controlled trial. Am J Prev Med 2011; 41: 274–283.

58.

Park

Elnaggar

Lee

, et al. Mobile health intervention promoting physical activity in adults post cardiac rehabilitation: pilot randomized controlled trial. JMIR Formative Research 2021; 5: e20468.

59.

Pinto

Dunsiger

Farrell

, et al. Psychosocial outcomes of an exercise maintenance intervention after phase II cardiac rehabilitation. J Cardiopulm Rehabil Prev 2013; 33: 91–98.

60.

Reid

Wooding

Blanchard

, et al. A randomized controlled trial of an exercise maintenance intervention in men and women after cardiac rehabilitation (ECO-PCR trial). Can J Cardiol 2021; 37: 794–802.

61.

Snoek

Meindersma

Prins

, et al. The sustained effects of extending cardiac rehabilitation with a six-month telemonitoring and telecoaching programme on fitness, quality of life, cardiovascular risk factors and care utilisation in CAD patients: the TeleCaRe study. J of Telemed Telecare 2021; 27: 473–483.

62.

Guiraud

Granger

Gremeaux

, et al. Telephone support oriented by accelerometric measurements enhances adherence to physical activity recommendations in noncompliant patients after a cardiac rehabilitation program. Arch of Phys Med Rehabil 2012; 93: 2141–2147.

63.

Vieira

Melo

Machado

, et al. Virtual reality exercise on a home-based phase III cardiac rehabilitation program, effect on executive function, quality of life and depression, anxiety and stress: a randomized controlled trial. Disabil Rehabil Assist Technol 2018; 13: 112–123. 20170313.

64.

Arias Labrador

Vilaró Casamitjana

Blanco Díaz

, et al. Effects of a phase III cardiac rehabilitation program on functional capacity and body composition for coronary heart disease. Rehabilitacion (Madr) 2023; 57: 100764. 20221124.

65.

Jeong

S-W

Kim

S-H

Kang

S-H

, et al. Mortality reduction with physical activity in patients with and without cardiovascular disease. Eur Heart J 2019; 40: 3547–3555.

66.

Roth

Mensah

Johnson

, et al. Global burden of cardiovascular diseases and risk factors, 1990-2019: update from the GBD 2019 study. J Am Coll Cardiol 2020; 76: 2982–3021.

67.

Mikkelsen

Berg-Beckhoff

Frederiksen

, et al. Estimating physical activity and sedentary behaviour in a free-living environment: a comparative study between Fitbit charge 2 and Actigraph GT3X. PLoS One 2020; 15: e0234426. 20200611.

68.

Kokkinos

Myers

. Exercise and physical activity: clinical outcomes and applications. Circulation 2010; 122: 1637–1648.

69.

Ambrosetti

Abreu

Corrà

, et al. Secondary prevention through comprehensive cardiovascular rehabilitation: from knowledge to implementation. 2020 update. A position paper from the secondary prevention and rehabilitation section of the European Association of Preventive Cardiology. Eur J Prev Cardiol 2020; 28: 460–495.

70.

Keteyian

Brawner

Savage

, et al. Peak aerobic capacity predicts prognosis in patients with coronary heart disease. Am Heart J 2008; 156: 292–300. 20080522.

71.

Hamborg

Tang

Andersen

, et al. It is like someone holding your hand when you need it - lived experiences of patients with cardiovascular disease participating in a digital health intervention focusing on the maintenance of physical activity after cardiac rehabilitation. Disabil Rehabil Assist Technol 2023: 1–11. 20230722. DOI: https://doi.org/10.1080/17483107.2023.2228839.

72.

McCambridge

Witton

Elbourne

. Systematic review of the Hawthorne effect: new concepts are needed to study research participation effects. J Clin Epidemiol 2014; 67: 267–277. 20131122.

73.

Bricca

Jäger

Johnston

, et al. Effect of in-person delivered behavioural interventions in people with multimorbidity: systematic review and meta-analysis. Int J Behav Med 2023; 30: 167–189. 20220428.

74.

Gentili

Failla

Melnyk

, et al. The cost-effectiveness of digital health interventions: a systematic review of the literature. Front Public Health 2022; 10: 787135. 20220811.

75.

Banbury

Roots

Nancarrow

. Rapid review of applications of e-Health and remote monitoring for rural residents. Aust J Rural Health 2014; 22: 211–222.

76.

Whitelaw

Pellegrini

Mamas

, et al. Barriers and facilitators of the uptake of digital health technology in cardiovascular care: a systematic scoping review. Eur Heart J Digit Health 2021; 2: 62–74. 20210204.

77.

Higgins

Thomas

. Cochrane Handbook for Systematic Reviews of Interventions Version 6.3. https://training.cochrane.org/handbook (2022, accessed 17 November 2022).

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.04 MB

0.03 MB

4.67 MB

0.37 MB

0.04 MB