2014 ISEK Congress Keynote LecturePersonalized Coaching Systems to support healthy behavior in people with chronic conditions
Introduction
The burden on healthcare services in the Western world is increasing substantially during the past decades. Both the quality and quantity of the care has to increase to meet the demands, especially for people with chronic conditions. The demand for an increase in quantity is due to the aging western population and an increase in prevalence of chronic diseases. Life expectancies have increased yearly with a rate of 0.24 years over the past century in the United States, Europe and Japan (Christensen et al., 2009). This increase in life expectancy however has hardly resulted in more healthy life years; in contrast, these additional life years are years with chronic conditions. This is becoming especially apparent now as the baby boom generation is reaching the age of retirement, also an age at which many chronic conditions are becoming more prevalent. These trends make that in the near future, we have to cope with a growing population in need of chronic care, during longer periods of time and on the other hand a smaller workforce that is able to pay and deliver the care. Consequently, healthcare services have to be innovated towards more automated services that are scalable and limit the need for trained healthcare professionals.
The need for such drastic innovation is most evident when considering people with chronic conditions. Their chronic conditions cannot be cured but their behavior has a major impact on further progression of the disease, quality of life and the occurrence of secondary health problems. It is becoming more evident that actual and perceived health can be influenced positively by creating awareness of adverse behavior and developing a sustainable healthy behavior. Regular physical activity reduces the risk of (chronic) diseases like coronary heart disease, type II diabetes and some cancers (Kohl et al., 2012, Lee et al., 2012). For example, in Chronic Obstructive Pulmonary Disease (COPD) patients avoid physical activities due to their symptoms, which cause a downward spiral towards lower physical condition and quality of life. Stress has also a substantial impact on health; it is a major cause of sickness absences (EC, 2010) and is strongly related to burn-out and depressions (Leka and Jain, 2010). Chronic neck/shoulder pain has often no clear cause and is often associated with a downward spiral of experienced pain and changes in behavior. Behavioral approaches for the interventions are most successful, although success rates are rather low; below 50%.
So, changing behavior into a more healthy behavior can contribute substantially to a better health, but changing behavior is for us humans quite difficult, requiring substantial feedback when adverse behavior occurs and encouragement when progress is being made. It requires also that the feedback is given during the daily activities of living, so changes can be easily integrated. These considerations make adequate coaching important but make also clear that solving this by using human coaches is not feasible and not scalable to the required level.
The recent developments of small on body sensors, powerful smartphones and knowledge on behavioral changes have opened the way to the development of an artificial coach.
The concept of such an artificial coach, a so-called Personal Coaching System (PCS) is shown in Fig. 1. This coach is able to provide intensive and timely persuasive feedback to its user. Targeted physiological variables are measured on-body that are relevant and valid for the health aspect that we want to influence. These signals are processed and used as input for the intelligent module. This module is capable of using these variables and relevant contextual information to generate a personal advice to the user for that specific moment. The third module takes care of the presentation of the actual feedback to the user in the most persuasive way.
This paper summarizes our work in the area of Personal Coaching Systems (PCSs). This will be done using three cases, illustrating research on the different aspects:
- 1.
The pain coach. This concerns a PCS for the treatment of people with chronic neck shoulder pain. It involves the use of surface EMG electrodes embedded in a garment to measure the EMG patterns of the upper trapezius muscles.
- 2.
The Activity coach. This concerns a PCS designed to stimulate people towards a more active and healthy activity behavior. Sensing is done by a dedicated activity sensor.
- 3.
The stress coach. This concerns a PCS to assess and feedback mental stress. It involves the use of several wearable physiological sensors to assess stress level.
These PCSs start from the same concept but have different approaches in terms of the sensing, the reasoning and the feedback, dedicated to the specific application area they were designed for. They are also in different stages of development. The pain coach was the first system that was developed and it has gone through all the different cycles of development. It is a relatively simple system with respect to its reasoning component, but it is also the most mature one with respect to its implementation and validation. It was extensively evaluated in a large European project (Myotel). The research concerning the activity coach has been focused on the reasoning part; how and when should feedback be provided in order to be as persuasive as possible in getting people to develop and maintain a physically active lifestyle. Several studies were carried out in different pathologies to investigate the specific activity patterns. The stress coach is most recently developed and research is especially focused on how to measure stress using multiple sensors and how to develop a personalized model from the recorded data that drives the feedback.
Section snippets
Discussion
In this paper we presented the concept of Personal Coaching Systems and our experience with three different realizations of such a system. The Pain PCS utilizes a new way of myofeedback and has gone through many interactive cycles of development and was demonstrated in a large international trial with many subjects suffering from chronic neck/shoulder pain. It was also the first PCS that involved the use of streaming EMG data and real-time feedback. The Activity PCS is illustrating how
Conflict of interest
None.
H. Hermens did his master in biomedical engineering at the University of Twente. His PhD was on surface EMG modeling, processing and clinical applications. He became Professor in Neuromuscular Control at the University of Twente and was founder of Roessingh Research and Development (RRD). In 2008, he became professor Telemedicine and head of the Telemedicine research group, at UTwente and in 2010 director Telemedicine at RRD. He is (co)-author of over 200 journal papers. He is fellow and past
References (57)
- et al.
Survey of chronic pain in Europe: prevalence, impact on daily life, and treatment
Eur J Pain
(2006) - et al.
Ageing populations: the challenges ahead
Lancet
(2009) - et al.
Deviations in daily physical activity patterns in patients with the chronic fatigue syndrome: a case control study
J Psychosom Res
(2011) - et al.
Development of recommendations for SEMG sensors and sensor placement procedures
J Electromyogr Kinesiol
(2000) - et al.
The pandemic of physical inactivity: global action for public health
Lancet
(2012) - et al.
Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy
Lancet
(2012) - et al.
Chronic neck pain disability due to an acute whiplash injury
Pain
(2003) - et al.
Cardiovascular and endocrine responses to experimental stress: effects of mental effort and controllability
Psychoneuroendocrinology
(1998) - et al.
Motivational cues as real-time feedback for changing daily activity behavior of patients with COPD
Patient Educ Couns
(2014) - et al.
Daily physical activities in chronic lower back pain patients assessed with accelerometry
Eur J Pain
(2009)
Physical activity monitoring in COPD: compliance and associations with clinical characteristics in a multicenter study
Respir Med
Towards continuous mental stress level estimation from physiological signals
Int J Psychophysiol
A triaxial accelerometer and portable data processing unit for the assessment of daily physical activity
IEEE Trans Biomed Eng
The myofeedback-based teletreatment system and its evaluation
J Telemed Telecare
American college of sports medicine position stand. Exercise and physical activity for older adults
Med Sci Sports Exerc
Does exercise training change physical activity in people with COPD? A systematic review and meta-analysis
Chron Respir Dis
Interventions to increase physical activity among healthy adults: meta-analysis of outcomes
Am J Public Health
Health systems evaluation of telemedicine: a staged approach
Telemed J
Static work loads and occupational myalgia - a new explanation model
Load pattern and pressure pain threshold in the upper trapezius muscle and psychosocial factors in medical secretaries with and without shoulder/neck disorders
Int Arch Occup Environ Health
Understanding tailoring in communicating about health
Health Educ Res
Cited by (48)
Personalized mobile technologies for lifestyle behavior change: A systematic review, meta-analysis, and meta-regression
2021, Preventive MedicineCitation Excerpt :The frequency of personalization delivery was diverse, ranging from as needed (e.g. whenever users indicated a high risk of unhealthy behavior (O'Donnell et al., 2019; Forman et al., 2018; Forman et al., 2019)) to daily, weekly, or monthly (Elbert et al., 2016). Only seven interventions explicitly stated the personalization algorithms used: two mentioned rule-based approaches (Dantzig et al., 2018; Korinek et al., 2018); five mentioned machine learning methods (Rabbi et al., 2015a; Zhou et al., 2018a; Zhou et al., 2018b; Hermens et al., 2014; Rabbi et al., 2018; Everett et al., 2018; Rabbi et al., 2015b; Forman et al., 2018; Forman et al., 2019). Twenty-one interventions collected lifestyle behavior data for personalization (Bertholet et al., 2019; Davies et al., 2017; Elbert et al., 2016; Fanning et al., 2017; Fukuoka et al., 2015; Hébert et al., 2020; Höchsmann et al., 2019a; Rabbi et al., 2015a; Recio-Rodriguez et al., 2018; Zhou et al., 2018a; Zhou et al., 2018b; Mitchell et al., 2020; Mitchell et al., 2018; Dantzig et al., 2018; Hermens et al., 2014; Khatri et al., 2016; Boerema et al., 2019; Höchsmann et al., 2019b; Klasnja et al., 2019; Korinek et al., 2018; Rabbi et al., 2018; Simons et al., 2018a; Zuckerman and Gal-Oz, 2014; Everett et al., 2018; Rabbi et al., 2015b; Businelle et al., 2016; Middelweerd et al., 2020).
Telemedicine is the way forward for the management of cystic fibrosis– the case against
2018, Paediatric Respiratory ReviewsPersonalized persuasion for behavior change
2023, Personalized Human-Computer InteractionMining Sensor Data to Assess Changes in Physical Activity Behaviors in Health Interventions: Systematic Review
2023, JMIR Medical InformaticsMachine Learning Techniques in Adaptive and Personalized Systems for Health and Wellness
2023, International Journal of Human-Computer Interaction
H. Hermens did his master in biomedical engineering at the University of Twente. His PhD was on surface EMG modeling, processing and clinical applications. He became Professor in Neuromuscular Control at the University of Twente and was founder of Roessingh Research and Development (RRD). In 2008, he became professor Telemedicine and head of the Telemedicine research group, at UTwente and in 2010 director Telemedicine at RRD. He is (co)-author of over 200 journal papers. He is fellow and past president of the Int. Society of Electrophysiology and Kinesiology (ISEK), editor in chief of the JBMR and coordinated the Seniam group leading to the first surface EMG standards. His present research is focused on combining Biomedical Engineering with ICT to create innovative Personalised Health Systems for people with chronic conditions.
H. op den Akker MSc. obtained his bachelor degree in Technical Computer Science, and master degree in Human Media Interaction at the University of Twente in Enschede, the Netherlands. He is currently working as a postgraduate Ph.D. candidate researcher at Roessingh Research and Development, an independent research centre affiliated to the Roessingh Rehabilitation Centre, and the University of Twente. The focus of his research is on tailoring to enhance compliance in personal coaching systems.
M. Tabak received her MSc degree (Cum Laude) in Biomedical Engineering from the University of Twente in Enschede, the Netherlands. Since 2009, she is working at Roessingh Research and Development (RRD), the largest research centre for rehabilitation technology in the Netherlands. She has been working on the development and evaluation of ambulant feedback applications and telemedicine programmes to promote healthy (active) behaviour. As such she received her PhD for the research “Telemedicine for patients with COPD – new treatment approaches to improve daily activity behaviour”. She is currently employed as a post-doctoral researcher at both RRD and the University of Twente. Her research interests include motivational strategies to promote healthy behaviour by means of technology, e.g. ambulant monitoring and feedback, intelligent navigation support, trusted healthcare services and gamification.
J. Wijsman received her B.Sc. and M.Sc. degrees in Biomedical Engineering from the University of Twente in Enschede, The Netherlands in 2007 and 2009 respectively. She is currently working as a Ph.D. student in a collaboration between Holst Centre, Eindhoven, The Netherlands and University of Twente, where she focuses on how mental stress can be measured in daily life from various physiological signals.
Prof. M. Vollenbroek Miriam Vollenbroek-Hutten is a human movement scientist. She received her PhD on an innovative assessment method for chronic pain in 1999. Currently she is clustermanager of the research cluster Telemedicine at RRD and Professor in technology supported treatment of patients with chronic disorders at the University of Twente. She is supervising a research staff of 20 people from various backgrounds and 15 PhD’s. She has more than 10 years of experience with large European studies including NEW (KP5), ‘Hellodoc’ (KP6), Clear ICT-PSP(ICT-PSP-224985). She was the coordinator of the European Project Myotel (Eten, KP6). Presently she is the coordinator of the European project PERSSILAA aiming at screening of frail elderly and offering them physical, cognitive and nutritional training modules. M. Vollenbroek is strongly involved in the EIP-AHA framework in A3 and B3 action lines. She is (co)-author of over 80 peer reviewed papers.