Table 1.
Coaching systems that were designed for and evaluated in home environment.
| Study | Healthcare domain | UI technologies | Coach-user interaction | Subject monitoring technologies | Continuous monitoring | Real-time evaluations and feedback | Number of subjects | Duration of trial | Evaluation outcomes |
|---|---|---|---|---|---|---|---|---|---|
| Salvi et al. (17) | Physical rehabilitation | Tablet | Motivational virtual coach avatar | ECG, heart rate, respiration rate, accelerometers | - | Heart rate feedback for cardiac rehabilitation | 118 adults with myocardial infarction | 21 weeks | Improved educational levels but not adherence or exercise habits significantly |
| Prvu Bettger et al. (18) | Physical rehabilitation | On board screen | Virtual coach avatar to demonstrate and guide activity | 3D pose and motion tracking | - | Feedback on exercise quality | 287 patients after total knee arthroplasty | 12 weeks | Significantly lower health-care costs while providing similar effectiveness |
| Segerståhl and Oinas-Kukkonen (19) | Physical activity | Wristband display and PC | Exercise level guidance | Heart rate | Heart rate feedback | 30 young adults | 3 weeks | Feasibility, usability and acceptance of technology study | |
| Watson et al. (20) | Physical activity | PC | Virtual embodied exercise coach | Digital pedometer | Daily number of steps | - | 70 overweight adults | 12 weeks | Effective in increasing short-term physical activity, long-term effects need further evaluation |
| Bickmore et al. (21) | Physical activity | Tablet | Virtual embodied conversational exercise coach to motivate walking | Digital pedometer | Daily number of steps | - | 263 sedentary older adults | 6 months | Effective in increasing short-term physical activity, but problematic long-term maintenance of behavior change |
| Friederichs et al. (22) | Physical activity | Online program | Motivational interviewing with coach avatar | - | - | - | 500 Dutch adults | 1 month | If avatar failed to strengthen social relationship with user the impact of the intervention was not enhanced |
| op den Akker et al. (23) | Physical activity | Smartphone and PC | Feedback by virtual human coach or text | Activity tracker, digital pillbox | Amount of physical activity | - | 43 adult office workers | 6 weeks | Virtual coach had lower effectiveness compared to text messaging |
| Gabrielli et al. (24) | Physical activity | Smartphone | Virtual coaching for behavioral change | Activity tracker | Daily number of steps | - | 6 families with overweight children | 6 weeks | Acceptable solution to support health promotion interventions in primary care |
| Fadhil et al. (25) | Physical activity | Smartphone | Text-based conversational agent (chatbot) | - | - | - | 19 sedentary adults | 4 weeks | Preference of using virtual human agent or a combination depends health domain |
| Brandenburgh et al. (26) | Physical and social activity | Tablet and PC | Virtual coach providing motivational feedback | Accelerometers | Amount of daily activity | - | 7 lonely older adults | 6 weeks | Reduced loneliness scores |
| Báez et al. (27) | Physical and social activity | Tablet | Virtual gym environments, personalized avatars, gym coach avatar | Accelerometers, barometric pressure | - | Participation and completion of exercises | 34 older adults | 10 weeks | High usability and technology acceptance of virtual fitness environment, but poor real-time user interaction |
| Espie et al. (28) | Insomnia | Online program | Animated personal therapist with personalized advice | - | - | - | 164 adults with insomnia | 6 weeks | Modest superiority over placebo on daytime sleepiness, improved sleep-wake functioning |
| Beun et al. (29) | Insomnia | Smartphone | Health coaching dialog system (chatbot) | - | - | - | 151 adults with mild insomnia | 6 weeks | Significantly more patients reached a meaningful clinical change on Insomnia severity |
| Lorenz et al. (30) | Insomnia | Online program | Animated conversational coach | - | - | - | 56 adults with insomnia | 6 weeks | Online CBT had substantial long-term effects on relevant sleep-related outcome parameters |
| An et al. (31) | Smoking cessation | Online program | Personalized virtual avatar, online peer coach | - | - | - | 1698 young adults | 12 weeks | Increased smoking abstinence and positive changes in health behaviors |
| Abdullah et al. (32) | Smoking cessation | Tablet | Virtual conversational agent | - | - | - | 6 Veterans | 2 weeks | Feasibility study, agent helped in setting a quit date. |
| Connelly et al. (33) | Diabetes | Online program | Virtual conversational coach | Accelerometers | Level of physical activity | - | 31 patients with type 2 Diabetes | 6 months | Effective in promoting physical activity, interactive features did not affect activity |
| Block et al. (34) | Prediabetes | Online program, smartphone | Automated physical activity and eating suggestions and logging | - | - | - | 339 adults at risk for developing diabetes | 12 months | Improved glycemic control, body weight, BMI, waist circumference, TG/HDL ratio, and diabetes risk |
| Ellis et al. (35) | Parkinson's disease | Tablet | Virtual embodied conversational exercise coach to motivate walking | Digital pedometer | Daily number of steps | - | 20 patients with PD | 1 month | High retention, satisfaction and adherence to daily walking, significant improvement in mobility |
| van der Kolk et al. (36) | Parkinson's disease | PC | Virtual cycling targeting heart rate zone | Static bike with heart rate | - | Heart rate feedback | 130 adults with mild PD | 6 months | Aerobic exercise at home is feasible and reduced off-state motor signs |
| Andrade et al. (37) | Overactive bladder | Online program | Virtual human avatars and self-avatar | - | - | - | 41 adult women | 12 weeks | Improved QoL and overactive bladder symptoms |
| Hartanto et al. (38) | Social anxiety disorder | Head mounted display | Virtual health agent, dialogues with virtual characters | Head tracker, heart rate, microphone | - | Natural speech recognition, heart rate feedback | 5 adults with social phobia | 10 sessions | Anxiety was evoked and over time gradually decreased due to exposure therapy, but serious technical problems occurred |
| Wu et al. (39) | Powered wheelchair seating | Smartphone (mobile app) | Virtual seating coach promoting powered seating functions | Accelerometers | Seating angle monitoring | Reposition adjustment | 5 powered wheelchair users | 5 days | Feasibility study, accelerometers were unstable for seating angle while moving |
UI, User Interface; Subject monitoring technologies: Type of sensing data/devices used, Continuous monitoring: Always-on tracking features, Real-time evaluations and feedback: Dynamic user evaluations using sensing data and provision of appropriate interventions.