Editorial

Complement Med Res 2020;27:131–133 Received: November 8, 2019

Accepted: February 18, 2020
DOI: 10.1159/000506672 Published online: March 27, 2020

Quantified Complementary and Alternative

Medicine: Convergence of Digital Health
Technologies and Complementary and
Alternative Medicine
Nico Steckhan a, b Bert Arnrich b
a Institute
for Social Medicine, Epidemiology and Health Economics, Charité – Universitätsmedizin Berlin,
Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Berlin, Germany;
b Digital Health Center, Hasso Plattner Institute, University of Potsdam, Potsdam, Germany

In recent years, consumer health technologies includ- Sensor technologies are able to provide metrics by means
ing wearables, Internet-of-Things, and health apps that of active (prompted) or passive (unnoticed) measure-
support wellness and healthy lifestyle have emerged rap- ments, offering considerable flexibility in approach.
idly [1]. However, only a few of these new technologies Those high-frequency longitudinal data sets can then be
have been certificated as medical product yet (e.g., only used for characterisation of a disease [2]. This yields a so-
40 CE-certified apps in Germany as of July 2019). lution to the often acclaimed subjectivity of patient-re-
Common application scenarios include lifestyle-asso- ported outcome, that is common not only in CAM re-
ciated aspects like diet, exercise, stress management, or search. Further, mood states could be assessed using dif-
combinations of it. Interestingly, the same aspects are ad- ferent sensor technologies. Even if this may be an indirect
dressed by several fields of complementary and alternative measure of one’s psychological state, we believe we now
medicine (CAM) like integrative or Mind Body Medicine. have better tools at hand to quantify, for example, stress
This synergy has mainly two reasons, first the modern in real time [3, 4].
technologies are getting better to process lifestyle data, In order to apply such technologies in medicine, how-
and second, the regulations (Medical Device Regulation) ever, methods are needed that are also suitable for re-
are more complex if one includes non-wellness health search beyond big players and big data (e.g., Apple,
claims (e.g., dosing of medications, medical decisions). Google, and Samsung), such as in complementary medi-
In 2020, the German Medical Device Directive will in- cal research with small numbers of cases. The support
crease the risk classification for apps and health services from validated models and tools can be a beneficial ingre-
implying any medical purpose, thus many lifestyle tech- dient if not transforming complementary medicine re-
nologies need a certification comparable to other phar- search (think of the one channel ECG in Apple Watch).
macological or medical products (Federal Institute for For example, monitoring vital signs in real time and in
Drugs and Medical Devices). Hence, technologies that are daily life can dramatically enhance the field of research
“just software” should be clinically investigated compa- [5].
rable to invasive medical products. This yields several op- Thus, passive data collection using sensors in wrist-
portunities and challenges for a patient-generated data- bands has several advantages: mainly, the high frequency
driven CAM. or even continuous data acquisition, and further, better
Still medical sampling has a snapshot approach lack- objectivity (not influenced by patient reporting and learn-
ing the dynamical behaviour of a person’s physiology. ing effects) can be achieved.

karger@karger.com © 2020 S. Karger AG, Basel Nico Steckhan

www.karger.com/cmr Institute for Social Medicine, Epidemiology and Health Economics
Charité – Universitätsmedizin Berlin
Charitéplatz 1, DE–10117 Berlin (Germany)
nico.steckhan @ charite.de
 As doctors spend more time with administrational achieve the same effect as a mind body therapy or how it
work rather than talking with their patients, we have an can support Mindfulness-Based Stress Reduction has yet
information bias that could be circumvented if we had the to be evaluated. The ability to intervene and measure at
right tools to effectively summarise a patient’s health sta- the same time, making interactions in real time accessible
tus. Providing longitudinal disease-related digital bio- is of great value. One may think of the already established
markers to a healthcare practitioner to allow for objective continuous glucose monitoring systems that help to not
and continuous clinical evaluation of a person’s lifestyle only improve quality of life but further excels in optimis-
could be a necessary step towards a more holistic medical ing glucose control of diabetic patients.
understanding not only in CAM [2]. Longitudinal quantification of the individual-level hu-
Journals like Nature – Digital Medicine or Journal of man phenotype in situ using data from personal digital
Medical Internet Research have become popular in this devices has established the so-called digital phenotyping.
field. Recently, also The Lancet started a digital health sub In general, an efficient digital phenotyping can be a
journal. Published papers show that not only health ser- potential resource for a more lifestyle-concerned medi-
vices have to be established but rather research on so- cine. Especially when one realizes that Western diseases,
called digital biomarkers is needed. Finally, establishing such as coronary heart diseases, are associated with over
the basic research of digital health. 80% participation of lifestyle. These chronic non-com-
Also these digital biomarkers are more in line with the municable diseases will create a massive burden for health
phenomenological understanding of behavioural thera- systems, including insurances, in the coming years.
pies. The quantification of behavioural and conscious- Having quantifiable measures of one’s lifestyle could
ness involving processes can assist squaring the circle of transform healthcare into the direction of a preventive
qualia-related outcome [6, 7]. and lifestyle concerned medicine. What especially is of
Monitoring and modelling biomedical, healthcare, importance if we address non-communicable diseases
and wellness data from individuals and converging data and like to monitor adherence. New methods have to be
on a population scale have tremendous potential to im- developed to approach disease forecasting, detection, and
prove understanding of chronic non-communicable dis- monitoring. Those data will then enable a more quantifi-
eases [8]. This will also improve primary care and will able lifestyle-related picture as often suggested by several
help to quantify transitions from healthy to diseased state, holistic approaches like traditional medical systems (e.g.,
thus strengthening the importance of prevention strate- Traditional European Medicine, Traditional Chinese
gies already before classical biomarkers hit thresholds [9]. Medicine, and Ayurveda). The rise of interdisciplinary
If we have such monitoring systems, we are allowed to research like systems medicine and digital health bridges
analyse such data in different contexts, thus enabling lat- the way for addressing the complexity and dynamics of
er options for early diagnosis and treatment. multifactorial diseases, in such that CAM becomes more
Forecasting of health relevant events by using artificial relevant for the scientific zeitgeist of these days.
intelligence has the ability to intervene early with lifestyle We may not need to transform patients to quantified
changes or more gentle medical approaches [10]. selves but introduce a whole new dimension to modern
Whether the claimed improvements based on predic- health research, prevention, and disease management.
tions such as better prevention and therapy adherence
rates will help reduce hospital admission and readmission
Acknowledgement
levels and doctor’s visits remains to be proven.
Inevitably, these technologies have limitations that We would like to acknowledge Jonas Chromik for reading the
need to be dealt with. Some people will argue that self- final draft.
awareness will not be achieved using external devices and
more technical feedback loops/systems. On the other
hand, it could be contradictory to distress patients with Disclosure Statement
technical problems (e.g., battery life, connectivity), for ex- The authors have no conflicts of interest to declare.
ample, during stress assessments. Depart from this, data
privacy will become a major risk concerning the probably
most sensitive data of a person. Funding Sources
Cooperation between start-ups and health insurance
companies is growing, with the legal framework taking No funding sources received.
shape and a number of health apps already in Germany’s
reimbursement scheme. Most of the available mobile Author Contributions
health apps focus on relaxation, mindfulness, and medi-
tation. Whether an app like Headspace or 7Mind can N. Steckhan has written the manuscript and B. Arnrich edited it.

132 Complement Med Res 2020;27:131–133 Steckhan/Arnrich

DOI: 10.1159/000506672
 References
1 Peake JM, Kerr G, Sullivan JP. A Critical Re- 4 Smets E, Rios Velazquez E, Schiavone G, 7 Baumeister H, Montag C. Digital Phenotyp-
view of Consumer Wearables, Mobile Appli- Chakroun I, D’Hondt E, De Raedt W, et al. ing and Mobile Sensing: New Developments
cations, and Equipment for Providing Bio- Large-scale wearable data reveal digital phe- in Psychoinformatics. Springer Nature; 2019.
feedback, Monitoring Stress, and Sleep in notypes for daily-life stress detection. NPJ https://doi.org/10.1007/978-3-030-31620-4.
Physically Active Populations. Front Physiol. Digit Med. 2018 Dec;1(1):67. 8 Bhavnani SP, Narula J, Sengupta PP. Mobile
2018 Jun;9:743. 5 Shameer K, Badgeley MA, Miotto R, Glicks- technology and the digitization of healthcare.
2 Kourtis LC, Regele OB, Wright JM, Jones GB. berg BS, Morgan JW, Dudley JT. Translation- Eur Heart J. 2016 May;37(18):1428–38.
Digital biomarkers for Alzheimer’s disease: al bioinformatics in the era of real-time bio- 9 Price ND, Magis AT, Earls JC, Glusman G,
the mobile/wearable devices opportunity [In- medical, health care and wellness data Levy R, Lausted C, et al. A wellness study of
ternet]. Vol. 2, npj Digital Medicine. 2019. streams. Brief Bioinform. 2017 Jan; 18(1): 108 individuals using personal, dense, dy-
Available from: http://dx.doi.org/https://doi. 105–24. namic data clouds. Nat Biotechnol. 2017 Aug;
org/10.1038/s41746-019-0084-2. 6 Jacobson NC, Weingarden H, Wilhelm S. 35(8):747–56.
3 Goodday SM, Friend S. Unlocking stress and Digital biomarkers of mood disorders and 10 Norgeot B, Glicksberg BS, Trupin L, Lituiev
forecasting its consequences with digital tech- symptom change. NPJ Digit Med. 2019 Feb; D, Gianfrancesco M, Oskotsky B, et al. Assess-
nology. NPJ Digit Med. 2019 Jul;2(1):75. 2(1):3. ment of a Deep Learning Model Based on
Electronic Health Record Data to Forecast
Clinical Outcomes in Patients With Rheuma-
toid Arthritis. JAMA Netw Open. 2019 Mar;
2(3):e190606.

Digital Health Perspectives for CAM Complement Med Res 2020;27:131–133 133
DOI: 10.1159/000506672