Background
• multifactorial individualized telehealth delivery providing individualized assessment and risk factor modification with patient-provider contact primarily by telephone; |
• internet-based delivery of programs where the majority of patient-provider contact is via the internet; |
• exercise telehealth interventions where patient-provider contact is primarily by telephone; |
• telehealth interventions focused on psycho-social recovery where patient-provider contact is primarily by telephone; |
• community or home-based CR involving patient-provider contact during home visits or patient visits to a community centre; |
• program delivery to diverse population groups including rural and remote settings; |
• multifaceted models of care incorporating interventions across these categories; |
• models utilizing complementary or alternative medicine |
Methods
Search strategy
Study selection
P | |
I | A model of care that utilized smartphone functionality (either app or Wireless Application Protocol (WAP) capabilities) for comprehensive Cardiac Rehabilitation and Secondary Prevention or heart failure rehabilitation |
C | None, traditional cardiac rehabilitation or usual care |
O | Feasibility, utility, and uptake of mHealth; service outcomes (patient engagement, acceptance, adherence and completion, provider engagement and acceptance, and cost effectiveness); patient outcomes (clinical, exercise capacity, knowledge, social and emotional, QOL); health service utilisation. |
E | Retrospective studies; non-intervention studies; systematic reviews; study protocols; conference abstracts and non-cardiac rehabilitation or heart failure programs. |
Data extraction and analysis
Results
Worringham, 2011 | Forman, 2014 | Varnfield, 2011 | Blasco, 2012 | Varnfield, 2014 | |
---|---|---|---|---|---|
Country of Origin | Australia | USA | Australia | Spain | Australia |
Levels of evidence | |||||
Design | FUU | FUU | FUU | RCT | RCT |
Level of Evidence* | IV | IV | IV | II | II |
CASP score | n/a | n/a | n/a | 7 | 8 |
Theoretical framework | ✓ | ✓ | |||
Outcome measures | |||||
Technology | |||||
Feasibility | ✓ | ✓ | ✓ | ✓ | |
Usability | ✓ | ✓ | ✓ | ||
Technical problems | ✓ | ✓ | ✓ | ||
Acceptability | ✓ | ✓ | ✓ | ✓ | |
Engagement | ✓ | ✓ | |||
Adherence | ✓ | ✓ | ✓ | ✓ | ✓ |
Usage | ✓ | ||||
Task completion | ✓ | ||||
Patient | |||||
Uptake | ✓ | ✓ | ✓ | ||
PROs | ✓ | ||||
Program completion | ✓ | ✓ | |||
Qualitative feedback | ✓ | ✓ | ✓ | ||
CV risk improvement | ✓ | ||||
Physical activity | ✓ | ✓ | ✓ | ||
Step counter | ✓ | ✓ | |||
6MWT | ✓ | ✓ | |||
Nutrition | ✓ | ✓ | ✓ | ||
Smoking status | ✓ | ||||
Psychological distress | ✓ | ||||
Depression | ✓ | ✓ | |||
Anxiety | ✓ | ✓ | |||
QOL | ✓ | ✓ | ✓ | ||
Self-efficacy | |||||
BP and HR | ✓ | ✓ | ✓ | ||
Weight | ✓ | ✓ | |||
BMI | ✓ | ✓ | |||
Waist circumference | ✓ | ||||
HbA1c | ✓ | ✓ | |||
Plasma lipid level | ✓ | ✓ | |||
PVO2 | |||||
Medication adherence | |||||
Economic evaluation | ✓ |
Worringham, 2011 (FUU) | Forman, 2014 (FUU) | Varnfield, 2011 (FUU) | Blasco, 2012 (RCT) | Varnfield, 2014 (RCT) | |
---|---|---|---|---|---|
Patient numbers n = | 6 | 26 | 15 | 102 TMG; 101CG | 60 CAP; 60 TCR |
Program completion | 100% | NR | Internal feasibility study | 87% TMG: 88% (n = 90 of 102) CG: 87% (n = 88 of 101, includes 4 deaths) or 91% (n 92 of 101, excludes 4 deaths) | ++ TMG: 80% (n 48 = of 60) vs CG: 47% (n = 28 of 60) |
Mean age (years) | 53.6 (42–67) | 59 (43–76) 33% > 65 years | 59 | 60.6 ± 1.3 TMG 61.0 ± 12.1 CG | 54.9 ± 9.6 CAP 56.2 ± 10.1 TCR |
Technology (reported inconsistently) | |||||
Feasibility | ✓ | ✓ | NR | ||
Usability/Acceptability | 4.8/5.0 Ease of use | 83% positive experience | Easy (qualitative data) | ||
Technical | 80% of completed sessions had no technical problems | NR | 36% use of Wellness Diary Connected internet -limited by computer and internet access | 42% of TMG withdrawals due to technical issues (n = 5 of 12) | 7% of CAP withdrawals reported technical difficulties |
Engagement/Adherence/Usage/Task completion | 87% completed sessions | 90% daily engagement 78% task completion | 91.5% wellness diary 100% step counter | 98% completed > 50% sessions 83% completed > 75% of sessions | ++ (94% vs 68% adherence) |
Patient | |||||
Uptake | 86% of referred patient | Internal feasibility study | 83.5% | ++ (80% vs 62%) | |
Mentorship | 91% motivational | ||||
CV risk improvement | ++ (ITT) | ||||
Physical activity | ns | ||||
6 Minute Walk Test | + | ǂ † ns | |||
Nutrition | ǂ † ns | ||||
Smoking cessation | ns (ITT) | ||||
Psychological distress | ǂ ns | ||||
Depression | + | ǂ † ns | |||
Anxiety | ǂ ns | ||||
Quality of Life | + − physical health ns - mental health (SF36) | ++ Physical health ns Mental health (SF36) | ǂ ++ (EQ5D-Index) | ||
Blood pressure (BP) and Heart rate (HR) | ++ BP (ITT) ++ SBP ≠ DBP | ++ DBP ns HR ns SBP | |||
Weight | ǂ ns | ||||
Body Mass Index | ++ | ||||
Waist circumference | ǂ ns | ||||
Haemoglobin A1c (HbA1c) | ++ (ITT) ++ | ns | |||
Plasma lipid level | ns LDL-c (ITT) ++ LDL-c ++ TG | † ns TC ǂ † ≠ TG ns LDL-c |
Publication (Author, Year, Country) | Participants, Sample Size, Rurality and Theoretical Model | mHealth and Non-mHealth Components | Intervention and Comparison | Outcomes |
---|---|---|---|---|
Worringham, 2011 Australia | Patients with an acute coronary event or revascularisation procedure unable to attend traditional CR Referred patients: n = 7 (1 withdrew prior to initial exercise testing) Intervention: n = 6 (83% M) Metropolitan and rural participants Theoretical model: None |
mHealth
Programmed smartphone External heart and activity monitor with GPS and Bluetooth connectivity Real-time monitoring of location, speed, heart rate and single lead ECG. Mobile phone contact pre- and post-exercise sessions Emergency mobile phone contact | Non-randomised feasibility trial Comparison group: None 6 week intervention Long-term engagement: recognised as a limitation and need for larger-scale and longer-term studies. Examination of outcomes of an outdoor walking-based exercise program
Statistical analysis:
Paired t-tests. | Mean age 53.6 (42–67) years Uptake: 86% of referred patients Completion/attrition: 100% completed the 6 week exercise program Usability and adherence: 87% of sessions completed (80% without technical problems), 13% cancelled. Ease of use rated 4.8/5 (95%CI 4.6–5.0) Physical function: 6MWT improved from 524 m to 637 m (p = 0.009). SF36 QOL: Physical Health score increased from 50.0 to 78.4 (p = 0.03)
Mental Health improved:
Cardiac depression scale reduced from 54.0 to 44.6 (p- = 0.007). SF36 QOL Mental Health score: NS |
Forman, 2014 USA | Patients currently enrolled in Phase 2 CR or recently completed (within 1 month) and continuing with Phase 3 CR. n = 26 (77% M) Metropolitan Hospital Rural patients not reported Theoretical model: None |
mHealth
iPhone, iPad or iPod touch Heart Coach Application: • Daily messages and tasks • Educational material and videos • Medication reminders • Physical activity prompts • Screenings and surveys
Non-mHealth
Traditional centre-based phase 2 CR or phase 3 CR (not defined) Qualitative patient and clinician feedback | Observational feasibility and utility study Comparison group: None Exercise and education-based mHealth program to augment traditional centre-based CR effectiveness 30 day intervention period Long-term engagement: no engagement or follow-up reported beyond the 30 day intervention period. Primary outcomes: qualitative feedback and engagement with technology Statistical analysis: Data collected and structured automatically by the application and presented as reports. Qualitative data assessed through surveys. | Mean age 59 (43–76) years; 33% aged > 65 years Study completion/attrition: not reported
Usability (mean):
90% daily engagement
Utility:
Task completion • 78% overall • 88% educational • 82% survey • 79% medication reminder • 70% physical activity (30 min) on 3 days/week Heart Coach had 42% lower visit cancellations vs no Heart Coach Staff typically spent about 20 min a day reviewing all patients’ progress and sending patient messages. (32 messages per patient over the study) Positive impact and reinforced impact of clinical-based sessions |
Varnfield, 2011 Australia | Post-MI patients eligible for CR n = 15% M/F not reported Metropolitan CR centres Rural patients not included Theoretical model: Self-management |
mHealth
Smartphone: • Integrated accelerometer sensor • step counter • Wellness Diary • Relaxation audio files • Educational multimedia • Weekly telephone mentoring sessions • Text messages
Non-mHealth
Clinical Review at CR centre (baseline and 6 weeks). Face-to-face training on CAP BP monitor and Weight scales (data entered into Wellness Diary) ‘My Heart My Life’ manual Participant and Mentor questionnaires on usability and uptake of technology and adherence. Internet web portal for viewing of patient data by mentors | Internal feasibility study: preliminary analysis of CAP data from RCT (CAP vs TCR). Comparison group: None Comprehensive delivery of core components of CR: Exercise and education-based intervention 6 week CAP CR program Long-term engagement: not reported Statistical analysis: Uptake and use determined through data uploaded daily to remote web portal. |
Patients:
Mean age 59 years Study completion/attrition: Internal feasibility study of 15 CAP participants
Uptake, Usability and adherence:
Average usage rate: Wellness Diary 91.5%; Step Counter 97%; Wellness Diary Connected internet application use was 36% due to lack of computer or internet connection; 91% reported that phone contact with mentor was motivational.
Mentors:
Usability and safety:
CAP practical and easy to use with benefits to patients (reduced travel, return to work). Concerns over lack of exercise supervision, individual motivation levels and group support. |
Blasco, 2012 Spain | Patients with ACS and one CV risk factor (tobacco smoking, LDLc ≤100 mg/dl (2.6 mmol/L), hypertension; or diabetes mellitus) n = 203 (80% M) TMG: n = 102 (30 T2D; 81% M) CG: n = 101 (26 T2D; 79% M) Metropolitan tertiary hospital Rural patients not reported. Theoretical model: None |
mHealth
Nokia mobile phone with Wireless Application Protocol (WAP) technology and secure Web Portal
Patient:
BP, HR, weight (weekly), glucose and lipids (monthly) levels sent through mobile phone via structured questionnaire.
Cardiologist:
Secure Web Portal for access of results. Individualized recommendations via short text messages
Non-mHealth
TMG
Omron automatic blood pressure monitor, CardioChek glucose and lipid meter. Patient Satisfaction Questionnaire at exit visit
All
Baseline and exit visit: clinical assessment, blood samples and SF-36 and State-Trait Anxiety Inventory for adults. 3 clinical visits with cardiologist | Single-blind RCT Telemonitoring Group (TMG) vs Control group (CG). All patients received lifestyle counselling and usual care 12 month follow-up Long-term engagement: no engagement or follow-up reported beyond the 12 month intervention period. Statistical analysis: Intention to treat. Independent and paired t tests; X2 test and relative risk | Mean age (years) ± SD: 60.6 ± 11.3 (TMG) vs 61.0 ± 12.1 (CG)
Primary Outcome:
CV Risk improvement: TMG (n = 87) 69.6% vs CG (n = 83) 50.5% P = 0.01 TMG vs CG meeting treatment goals for BP < 140/90 mmHg (62.1% vs 42.9%, p = 0.012); HbA1c < 7% (86.4% vs 54.2%, p = 0.018); smoking cessation (p = 0.964) and LDL-C (p = 0.948) Completion/attrition: 87% (n = 177) completion rate with a 13% attrition - 4 participants were lost to follow-up and 5 died (all in the CG). 17 participants left the study (12 TMG and 5 CG): Reasons were stress of telemonitoring (n = 3 TMG), personal reasons (n = 7 TMG, n = 5 CG) and inability to operate equipment (n = 2 TMG)
Adherence to protocol:
98% completed > 50% of sessions; 83% completed > 75% of sessions. |
Varnfield, Nov 2014 Australia | Post-MI patients n = 94 (82 M; 12 F) CAP: n = 53 (91% M) TCR: n = 41 (83% M) Metropolitan CR centres Rural patients not included Theoretical model: Self-management |
mHealth
CAP-CR Patient: Smartphone
• Integrated accelerometer sensor • Step counter • Wellness Diary • Relaxation audio files • Educational multimedia
Community Care Team
• Internet web portal for viewing of patient data • Text messages • Video and Telephone mentoring
Non-mHealth
All: ‘My Heart My Life’ manual
CAP-CR
Clinical Review at CR centre. Face-to-face training in CAP. Blood pressure monitor and Weight scales
TCR program
Two supervised exercise and 1 h education sessions weekly for 6 weeks at CR centre. | RCT of CAP compared with TCR. Comprehensive CR program 6 week CR intervention Long-term engagement: included a Self-management phase. CAP-CR participants kept smartphone and monitoring devices for this phase. Participants were encouraged to maintain lifestyle changes. Secondary outcomes, activity monitoring and perception of using a smartphone to monitor exercise was measured at 6 months
Statistical analysis:
ITT basis; Chi squared; Independent t test; Wilcoxon rank-sum test; ANCOVA adjusted for age and gender; Linear mixed model regression | Mean age (years) ± SD: 54.9 ± 9.6 (CAP) vs 56.2 ± 10.1 (TCR)
Primary outcome:
Uptake: 80% CAP vs 62% TCR, P < 0.05 Adherence: 94% CAP vs 68% TCR, P < 0.05 Study completion: 80% CAP vs 47% TCR completed, p < 0.05 Attrition: N = 44 dropouts, 70% from TCR (non-uptake / non-completion). Life demands: TCR - 10% Work, 4% stress; CAP – 0% Logistics: TCR - 16% Time, 7% location, 2% transport; CAP - 2% time Change in circumstances: TCR - 14% health, 2% criteria; CAP - 9% health, 7% smartphone Study design: TCR - 10%; CAP - 0% Motivation: TCR – 4%; CAP – 2% Improved health: TCR – 0%; CAP – 2% Privacy: TCR – 2%; CAP -)% Other reasons: TCR – 2%; CAP – 5% Technology: 7% (n = 3) reported difficulty with mHealth tools Secondary outcomes: CAP was as effective as TCR in improving: dietary intake; depression; 6MWT and triglycerides (p < 0.05). CAP effectively reduced psychological distress; anxiety levels; weight, WC and HRQOL (p < 0.05). TCR effectively reduced TC (P = 0.04) Between-group differences: DBP and HRQOL < 0.05 for CAP and Tgs < 0.05 for TCR Cost analysis: Based on 2010 Australian health economics data, CAP CR may result in AU$16.6 million readmission cost savings |
Scherr, 2006 | Scherr, 2009 | Seto, 2012 | Vuorinen 214 | |
---|---|---|---|---|
Country of Origin | Austria | Austria | Canada | Finland |
Levels of evidence | ||||
Design | FUU | RCT | RCT | RCT |
Level of Evidence* | IV | II | II | II |
CASP score** | n/a | 6 | 8 | 8 |
Theoretical framework | ✓ | |||
Outcome measures | ||||
Technology | ||||
Reliability | ✓ | |||
Feasibility | ✓ | |||
Clinical utility | ✓ | ✓ | ||
Usability | ✓ | ✓ | ✓ | |
Acceptability | ✓ | ✓ | ||
Adherence | ✓ | ✓ | ✓ | ✓ |
Usage | ✓ | ✓ | ✓ | ✓ |
Task completion | ✓ | |||
Patient | ||||
Patient satisfaction | ✓ | ✓ | ||
Qualitative feedback | ✓ | ✓ | ||
QOL | ✓ | |||
Self-care | ✓ | ✓ | ||
NYHA class | ✓ | ✓ | ✓ | |
LVEF | ✓ | ✓ | ✓ | ✓ |
BP and HR | ✓ | ✓ | ✓ | |
Weight | ✓ | ✓ | ✓ | |
ECG | ✓ | |||
Medication use | ✓ | ✓ | ✓ | |
Biochemistry | ✓ | ✓ | ||
BNP | ✓ | ✓ | ||
Mortality | ✓ | ✓ | ✓ | |
Health service utilization | ✓ | ✓ | ✓ | |
Economic evaluation | None of these smartphone heart failure studies included a health economic analysis |
Scherr, 2006 (FUU) | Scherr, 2009 (RCT) | Seto, 2012 (RCT) | Vuorinen 2014 (RCT) | |
---|---|---|---|---|
n = | 20 14 CHF; 6 HTN | 120 66 TMG (54 + 12 never beginners) 54 CG | 100 50 TMG; 50 SCG | 94 (47 TMG; 47 CG) |
Program completion | 95% (n = 19 of 20) CHF: 93% (n = 13 of 14) HTN: 100% | 87% (n = 104 of 120) TMG: 76% (n = 50 of 66, includes never beginners) or 93% (n = 50 of 54) CG: 100% | 97% (n = 97 of 100) TMG: 88% (n = 44 of 47, includes 3 deaths) or 94% (n = 47 of 50, excludes 3 deaths) SCG: 100% | 99% (n = 1 of 94) TMG: 98% (n = 1 of 47) CG: 100% |
Mean age (years) | 50 (SD14) CHF: 53 (13); HTN: 42 (16) | 66 (IQR 64–74) TMG: 66 years (IQR 62–73) CG: 67 years (IQR 61–72) | TMG: 55.1 ± 13.7 SCG: 52.3 ± 13.7 | TMG: 58.3 ± 11.6 CG: 57.9 ± 11.9 |
Technology (Reported inconsistently) | ||||
Feasibility | ✓high | |||
Usability/Acceptability | 80% did not report any problems with data entry | 98% system availability | 10–20 min initial education on use of mobile phone app. TMG: 1 patient withdrew due to increased anxiety from monitoring his condition. | |
Technical | 98% data transmission and website availability One (5%) withdrawl (poor vision) | 12 never beginners (median age 68 years (IQR 64–74) were unable to begin transmission of data (reasons NR) | TMG: 2 participants withdrew due to technical difficulties | TMG: 6 telephone calls re technical problems. 3 nurse initiated calls for start-up support; 3 patient calls initiated for failed internet connection. |
Engagement/Adherence/ Usage/Task completion | 94% (CHF) and 84% (HTN) self-measurement and data entry | 95% patient adherence | Completion of daily readings: 84% completed 50%; 66% completed 80%; 32% completed 95%. | Proportion of weekly submitted self-measurements by TMG: 86% weight (median = 28 (IQ 23–33); 89% BP, HR, and symptoms (median BP and symptoms = 32 (IQR 27–43). |
Patient | ||||
Patient satisfaction | 85% of patients continued telemonitoring at study completion | 96% responded to user experience questionnaire 95% - measures very or quite useful 91% - automatic feedback very or quite useful (9% no benefit) 66% - feedback drew attention to essential issues of disease 91% - feedback was motivational | ||
Quality of Life | ǂ overall MLHFQ ǂ Physical ǂ Emotional ++ Overall change (p = 0.05) | |||
Self-care | ǂ † Maintenance ǂ † Management ++ Maintenance | ns ǂ † | ||
New York Heart Association class | Study completion vs baseline Class I: n = 3 vs 0 Class II: n = 11 vs 10 Class III: n = 0 vs 4 | ++ (PPA) | ǂ † | |
Left Ventricular Ejection Fraction | ↑ in mean to 35% at study completion (vs 32% at baseline) | PPA: ns improvement TMG: 25% (IQR 20–38) to 35% (IQR 25–45) CG: 29% (IQR 21–36) to 35% (IQR 24–40) | ǂ † | ns ǂ † |
Blood Pressure | HTN: mean study completion SBP 135 (SD18); DPB 78 (7) vs baseline SBP 134 (21); DBP 80 (8) | |||
Medication | CHF: 71% had beta-blocker therapy initiated with a titrated increase HTN: antihypertensive medication stable | ǂ Aldosterone antagonists | ++ Medication change, both increases and decreases. | |
Biochemistry | ns Serum potassium, creatinine, sodium. | |||
Brain Natriuretic Peptide | ǂ † | ns ǂ | ||
Mortality/Health service utilization | ITT: ns TMG: 17% (0 deaths/11) hospitalizations CG: 31% (1 death/17 hospitalizations) PPA: ++ TMG: 15% (0 deaths/8 hospitalizations) PPA: ++ shorter length of hospital stay | TMG: 6% (n = 3) deaths (2 non-heart related) SCG: 0 deaths ns hospital admissions; nights in hospital; ED visits ++ Heart Function Clinic visits (TMG: 3.5 (SD 3.6); SCG 2.5 (2.5) | No mortality in TMG or CG ns HF related hospital days ++ TMG nurse time, telephone contact and visits ++ TMG unplanned clinic visits ++ patient initiated telephone contact ns physician time and visits |
Publication (Author, Year, Country) | Participants, Sample Size, Rurality and Theoretical Model | mHealth and Non-mHealth Components | Intervention | Outcomes |
---|---|---|---|---|
Scherr, 2006 Austria | Patients with chronic Heart Failure CHF or hypertension (HTN) n = 20 (95% M) CHF: n = 14 (93% M) HTN: n = 6 (83% M) Metropolitan and rural patients Theoretical model: None Comparison group: None |
mHealth
Patient terminal: Mobile phone with Wireless Application Protocol (WAP) technology Physician’s terminal: Personal Computer with internet access
Non-mHealth
Automatic BP monitor Digital weight scales Doctor–patient relationship Patient completed questionnaires on the technical aspects of the telemonitoring system | Observational study to evaluate acceptability, feasibility and reliability of a telemonitoring system. 90 day follow-up Long-term engagement: no engagement or follow-up reported beyond the 90 day intervention period. Patients measured BP, HR and weight daily and transferred data via mobile phone. Physician automatically notified by SMS of any parameters outside a pre-set range. Study physician accessed data and phoned patient as necessary for therapeutic adjustments Automatic reminders set by study physician Statistical analysis: Descriptive statistics reported. | Mean Age: All, 50 (SD14) years; CHF, 53 (SD13) years; HTN, 42 (SD 16) years.
Study completion/attrition:
95% (n = 19) completed CHF: 93% (n = 13) completed HTN: 100% (n = 6 completed) 5% (n = 1) withdrew from TMG due to poor vision
Reliability
98% data transmission success 98% website availability for physicians. Feasibility and Acceptability: Implausible data entry: 5 per CHF patient; 4 per patients with HTN Successful transmissions: 83% CHF and 84% HTN Self-measurement and data entry: CHF - 85 data transfer sessions over 90 days; HTN - two BP and HR measures on an average of 453 out of 540 cumulative days
Patient acceptance
High, 17 patients continued with telemonitoring at study end. Study completion/attrition: 19 participants completed. One participant withdrew due to inability to operate the mobile phone because of poor vision.
Clinical utility
CHF patients: stable or improved: mean LVEF improved from 32% to 35%; Beta-blocker initiation supported: commenced and up titrated successfully in 10 of the 14 CHF patients Patients with HTN: BP stable; 134/80 mmHg at baseline vs 135/78 mmHg at completion |
Scherr, 2009 Austria | Patients with heart failure and a hospital admission of > 24 h in the last 4 weeks. n = 120 (66 TMG; 54 SCG) TMG: n = 12 never beginners (50% M) TMG: n = 54 (74% M) CG: n = 54 (72% M) Metropolitan centres Rural patients not reported. Theoretical model: None |
mHealth
Mobile phone with Wireless Application Protocol (WAP) technology, Weight scale and automated BP monitor Secure web-based CRF at monitoring centre. Patients measured BP, HR and weight daily. Data entered and automatically sent to the remote server at remote Email alerts to study physician Study physician accessed data and could phone patient on mobile Study physician could set automatic reminders.
Non-mHealth
CG: Pharmacological care | Prospective, open-label RCT TMG: Pharmacological treatment with telemedical surveillance CG: pharmacological treatment alone and no planned interaction with study site. 6 months follow-up Long-term engagement: no engagement or follow-up reported beyond the 6 month intervention period. Statistical analysis: Per protocol principle and intention to treat analysis. Log-rank test, Kaplan-Meier estimation and relative risk reduction utilized for primary endpoint. Secondary endpoints: t-test, chi-square test, Wilcoxon rank sum test and Wilcoxon signed rank test utilized. |
Median age:
12 never beginners 68 years (IQR 64–74); TMG 65 years (IQR 62–72); CG 67 years (IQR 61–72) Study completion/attrition: 104 participants completed. 12 participants were unable to transmit data - classified as never beginners. 4 TMG participants withdrew early (included in intention-to-treat and per-protocol analysis). Participant adherence: 95% Intention-to-treat analysis: TMG vs CG: 0 deaths and 11 hospitalizations (17%) vs 1 death and 17 hospitalizations (33%), a RRR of 50% (95% CI 3–74%), p = 0.06 TMG: majority of re-hospitalizations occurred in first month of follow-up.
Per protocol analysis:
TMG: 0 deaths and 8 hospitalizations (15%), a RRR of 54% (95%CI 7–79%), p = 0.04 NYHA class improved (III to II) in TMG, P < 0.001 vs CG and TMG baseline Median length of stay: TMG 6.5 vs SCG 10.0 days (IQR 7.0–13), p = 0.04 LVEF: ns improvement in both TMG and CG. TMG 25% (IQR 20–38) to 35% (IQR 25–45) and CG 29% (IQR 21–36) to 35% (IQR 24–40) 375 alerts, 170 contacts, 55 adjustments to heart failure medications. |
Seto, 2012 Canada | Heart Failure patients with LVEF < 40% n = 100 TMG: n = 50 (82% M) SCG: n = 50 (76% M) Metropolitan centre (possible patients from rural or remote settings) Theoretical model: Self-care |
mHealth
TMG: Smartphone with blue tooth capability, BP monitor and scales; ECG recorder provided to 17 TMG participants – data automatically sent wirelessly to data repository. Daily morning symptom questions Email and text messages Website viewing of results by clinicians and patients.
Non-mHealth
All
Pre and post study: Demographic, clinical data, SCHFI and MLHFQ questionnaires
SCG:
Clinic visits Optimization of medication Heart Failure education Telephone contact
TMG:
Standard care as per SCG | Non-blinded RCT TMG vs Standard Care Group (SCG) Stratified 4 block randomization based on NYHA classification. 6 month follow-up: post-study questionnaire; 22 semi-structured interviews with TMG participants; 5 semi-structured interviews with clinicians Long-term engagement: no engagement or follow-up reported beyond the 6 month intervention period. SCG participants were not contacted by the study site until study end. Statistical analysis: Between group analysis: Student t tests and Mann-Whitney tests. Within group analysis: Paired Student t tests and Wilcoxon signed rank tests. | Mean age: TMG 55.1 years (SD 13.7); SCG 52.3 years (13.7) Completion/attrition: 97 participants completed. 3 participants withdrew from TMG (1 was incapacitated after a fall; 2 because of technical difficulties). No participant withdrew from SCG. Patient adherence: 84%, 66% and 32% completed at least 50%, 80% and 95% of possible daily readings. Health service utilization: no significant differences in hospital admissions; nights in hospital; and ED visits. Number of Heart Function Clinic visits increased in TMG (p = 0.04) due to unplanned cardiologist recalls in response to telemonitoring system alerts. Improvement post-study for TMG and SCG - BNP values (p = 0.001 and p = 0.002); NYHA class (p = 0.000 and 0.001); LVEF (p = 0.001); and self-care (p = 0.004 and P = 0.006). QOL improved only in the TMG (p = 0.02). Between group post-study - only self-care maintenance (SCHFI) was significant (p = 0.03). Between group change - only overall QOL (MLHFQ) (p = 0.05) |
Vuorinen, 2014, Finland | Heart Failure patients with LVEF ≤35%, NYHA class ≥2 n = 94 TMG: n = 47 (83% M) CG: n = 47 (83% M) Metropolitan centre (possible patients from rural or remote settings) Theoretical model: none | mHealth Patient: Mobile phone with preinstalled software app Provided with weight scale, blood pressure monitor, mobile phone with app and self-care instructions. Non-mHealth Multidisciplinary clinic visits and nurse feedback by telephone. | Prospective RCT TMG vs usual care (CG) 6-month follow-up Long term engagement: no engagement or follow-up reported beyond the 6-month intervention period
TMG:
Patients evaluated BP, HR, weight, symptoms and change in overall condition, weekly and transferred data via mobile phone application Received by secure remote patient monitoring server Patients received automated feedback about whether reported data was within personal targets set by nurse. Nurses accessed data and phoned patient weekly or as necessary for out-of-target parameters or failure to upload data
CG:
Patients encouraged to measure weight, blood pressure and heart rate at home Cardiac team monitor and interpret symptoms, optimize medication and provide education Statistical analysis: ZIP regression used for outcome variables that expressed counts, contiguous variables analysed within and between study groups | Mean age: TMG 58.3 (SD 11.6) CG 57.9 (11.9) Completion/attrition: 1 patient from TMG lost to follow-up. Patient adherence = proportion of weekly submitted self-measurements by TMG: 86% weight (median = 28 (IQR 23–33), 89% BP, HR, and symptoms (median BP and symptoms = 32 (IQR 27–43) Feasibility and Acceptability: 96% (44/46) from TMG responded to survey, 42/44 found making/reporting measurements with mobile app “useful” or “very useful.” 91% automatic feedback very or quite useful (9% no benefit), 66% feedback drew attention to essential issues of disease, 91% feedback was motivational. Primary Outcome: Mean HF-related hospital days: 0.7 (TMG) vs 1.4 (CG) (p = 0.351)
Secondary Outcomes:
Clinical: change in NT-proBNP, LVEF %, EHFSBS score, serum creatinine, potassium and sodium not significantly different between groups. Mortality: 0 (control), 0 (TMG). Within group changes were significant for: LVEF increased 5.0%, p = 0.003 TMG and 4.2%, p = 0.001 CG; EHFSBS (− 5.0 points, p < 0.001 TMG and − 3.8, p < 0.001); NT-proBNP decreased in the TMG (− 198 ng/l, p = 0.01) Use of health care resources: Mean nurse time, telephone contacts and visits higher in TMG (p < 0.001); TMG unplanned visits to Cardiac Outpatient Clinic higher (p < 0.001); TMG patient initiated telephone contact higher (p < 0.049); No statistical difference between groups for physician time and visits. |