Interoperability Architecture of the ADLIFE Patient Empowerment Platform

Abstract

Chronic diseases introduce challenges for the patients to manage the changing requirements of the disease. Patient empowerment activities are a critical component to assist patients in their long-term care journey. In order to be effective, patient empowerment tools need to be well-integrated with the chronic disease management tools used at the clinical sites. This paper explores and analyzes the exploitation of HL7 FHIR to design and implement an interoperable patient empowerment platform that can be seamlessly integrated with external chronic disease management and Electronic Health Record (EHR) systems

Full text

Interoperability Architecture of the
ADLIFE Patient Empowerment Platform
Gokce Banu Laleci ERTURKMENa,1, Mustafa YUKSELa, Mert BASKAYAa, Bunyamin
SARIGULb, Alper TEOMANa, Gökhan YILMAZa, Esteban de MANUELd on behalf of
the ADLIFE Consortium
a SRDC Software Research Development & Consultancy Corp, Ankara, Turkey
b Informatics Institute, Middle East Technical University, Ankara, Turkey
c Kronikgune Institute for Health Services Research, Spain
Abstract. Chronic diseases introduce challenges for the patients to manage the
changing requirements of the disease. Patient empowerment activities are a critical
component to assist patients in their long-term care journey. In order to be effective,
patient empowerment tools need to be well-integrated with the chronic disease
management tools used at the clinical sites. This paper explores and analyzes the
exploitation of HL7 FHIR to design and implement an interoperable patient
empowerment platform that can be seamlessly integrated with external chronic disease
management and Electronic Health Record (EHR) systems
Keywords. Patient Empowerment, Interoperability, HL7 FHIR, Chronic Disease
Management
1. Introduction
As the prevalence of chronic diseases continuously increases, so does the burden of
management of these long-term conditions on healthcare services: chronic diseases are
reported as major causes of disability and mortality and are associated to significant
healthcare costs and, often, inadequate access to care services[1]. Chronic diseases
introduce demanding challenges for the patients as well: long-term conditions
necessitate continuous and complex management, disease and treatment requirements
change over time requiring continuous decision making and adjustments at the patient
side as well. One of the key features of healtsh policies addressing the needs of chronic
diseases is cited as exploitation of patient empowerment tools in integrated care [3][4].
Patient empowerment activities involve providing the means for active patient
participation for the delivery of care for their chronic diseases, providing education
materials related with their diseases, enabling shared decision making in the definition
of their care plans, encouraging and assisting self-management activities in their care
plans such as adherence to medications, facilitating healthy lifestyle, self-monitoring
activities to look after their health status, and enabling feedback and interaction
between patients and practitioners. Empirical evidence of a positive relationship
between patient empowerment and better healthcare outcomes has been reported[3],
1 Corresponding author, SRDC Software Research Development & Consultancy Corp, ODTU
Teknokent Silikon Blok Kat:1 No:16, 06800 Cankaya / Ankara, Turkey; E-mail: gokce@srdc.com.tr.

several studies demonstrated increasing patients’ and health professionals’ satisfaction,
better adherence to treatment activities, and improved clinical outcomes[4].
In order to be effective, patient empowerment tools need to be well-integrated with
the chronic disease management and care planning tools used at the clinical sites.
Continuous two-way information exchange is essential to convey the care plan in clear
terms to the patients and their informal care givers, to encourage them to adhere to their
care plan, be in constant communication with the patient to collect and process
preferences, feedback, symptoms and patient recorded data. This introduces an
interoperability challenge, and results in development of custom patient empowerment
tools that are tightly integrated with the chronic disease management systems mostly
via proprietary interfaces, which diminishes re-usability across sites. This paper
explores and analyzes the exploitation of HL7 FHIR to design and implement an
interoperable patient empowerment platform that can be seamlessly integrated with
external chronic disease management systems and Electronic Health Record (EHR)
systems.
This work is being carried out within the scope of ADLIFE Project, (H2020, SC1-
DTH-11-2019, 875209) [1] which aims to provide an integrated care solution targeting
early detection and assessment of deterioration, advanced and well-coordinated care
planning and supportive care for patients with advanced chronic diseases, namely, heart
failure and COPD. The technical solution is composed of a Personalized Care Plan
Management Platform (PCPMP) supported by clinical decision support services, which
acts as a chronic disease management platform served to multidisciplinary care team
members, and a Patient Empowerment Platform (PEP) served to the patients and their
informal care givers, enabling them to be informed, educated, and guided about their
active care plan and to be active participants of their care plan activities. This paper
focuses on interoperability architecture between PCPMP and PEP to seamlessly share
care plan and related artifacts enabling integrated supportive care.
2. Methods
We have positioned a standard based HL7 FHIR Repository as the common data
repository that enables seamless data exchange between local EHRs, chronic disease
management platforms such as PCPMP and the PEP. In our architecture we are using
the open source on FHIR.io FHIR Repository [6]. In the following sections we present
how we have represented core data items for patient empowerment and how we have
enabled exchange of these data between PCPMP and PEP.
2.1. Exchanging Care plans and patient feedback
Shared care plans are an indispensable component of an integrated care environment
for long-term care of chronic disease patients, presenting clear measurable goals, as
well as treatment activities jointly agreed across multi-disciplinary care team members,
by also involving patients and their informal care givers. Sharing care plans with
patients and informal care givers empowers them to be actively involved in their
treatment journey.
In our architecture, care plans are represented as HL7 FHIR CarePlan Resource
[7]. Within the care plan, the care team members are clearly described via a reference

to CareTeam Resource, and the clinical conditions that this care plan is targeting to
address are referenced via the Condition Resource. Each care plan is composed of a set
of goals represented via Goal Resource where the desired target health state, such as
‘Keeping the blood pressure (BP) measurements below 140/80 mmHg in the next 3
months’ is defined, and a set of planned activities within the scope of the care plan.
These activities can be prescriptions represented as MedicationRequest Resource,
control appointments represented as Appointment Resource, medical device
assignments as DeviceRequest Resource, referrals to specialists, lab orders, diagnostic
tests and patient activities (such as symptom reporting, diet and exercise regiments,
patient-reported outcome measure questionnaires and self-blood glucose/BP
measurements) represented as ServiceRequest Resources. CommunicationRequest
Resource is used to represent education materials assigned to patients as a part of their
care plan. For each activity, the planned period, timing or frequency upon which the
described activity is aimed to be fulfilled and also the assigned performer such as a
specific care team member as a Practitioner, Patient or Informal Care Giver are
specified.
PEP is subscribed to the HL7 FHIR Repository which supports FHIR subscription
protocol [8], to be informed whenever a new care plan is created for the specified
patients or whenever existing care plans of these patients are updated. In this way, PEP
is automatically informed about new and updated care plans that are stored to the
shared repository and care plan retrieval is triggered via standard based HL7 FHIR
RESTful API. Once the machine-processable care plan content is retrieved from the
common repository, it is displayed to the patient in a patient-friendly manner. Patients
are enabled to provide feedback about care plan goals and activities both by marking
their achievement status via emojis and by free text comments. These are noted as
Annotation Resources linked to care plan activities and goals. Through the subscription
mechanism, the subscribed local EHRs, Chronic Disease Management Platforms are
informed about patient feedback to be notified to practitioners in the care team.
2.2. Patient-Reported Outcome Measures (PROMs) as scored questionnaires
To assess the care delivered to the patients via ADLIFE integrated care approach from
the patient perspective and to measure the patients’ perceptions of their own health
status and quality of life, our clinical reference group has identified several Patient-
Reported Outcome Measures (PROMs) questionnaires. Some examples are, CAT
(COPD Assessment Test) and HADS (Hospital Anxiety and Depression Scale). In
ADLIFE architecture, we have followed the FHIR Patient-Reported Outcomes
Implementation Guide [9] and PROMs are represented as HL7 FHIR Questionnaire.
These are assigned to the care plan of the patient via ServiceRequests as care plan
activities to be carried out by patients periodically. The PROM assignments are
presented to the patient as a part of their care plan and rendered as user-friendly web-
based surveys, enabling the patient to easily fill them out. The responses are recorded
as FHIR QuestionnareResponse Resources and saved back to the FHIR Repository.
When the PROM includes a scored assessment, the resulting score is represented as an
Observation Resource that is linked with the PROM. The chronic-disease management
solutions that have already subscribed to receive notifications are informed, and the
responses are made available to the practitioners.

2.3. Collection of Patient-Recorded Observations and Medical Device Data
In ADLIFE, as a part of the care plan, patients can be requested to regularly report self-
measurements of clinical parameters and vital signs such as blood pressure (BP)
measurements, pulse, oxygen saturation, weight. These self-measurement activities are
assigned to the care plan of the patient as ServiceRequests, medical device assignments
are also recorded in the care plan as DeviceRequests.
ADLIFE PEP implements the Personal Connected Health Alliance (PCHAlliance)
Continua Design Guidelines (CDG) [10], to enable seamless integration with medical
devices in order to automatically gather device readings from assigned devices and to
record them as patient observations. The ‘Personal Health Devices Interface’ defined in
CDG is implemented between the medical sensors (BP measurement Device, Weight
scale, pulse oximeter) and ADLIFE PEP Mobile application that functions as a mobile
gateway to collect device readings via Bluetooth communications. The CDG ‘Services
Interface’ is implemented between mobile app and the common FHIR repository to
store device readings via FHIR RESTful API. As guided by CDG, we have followed
the FHIR Personal Health Device Implementation Guide (IG) [11], to represent patient-
recorded measurements: Observation Resource is used to represent values of measured
clinical parameters, and when received from a medical device via ‘Personal Health
Devices Interface’, Device Resource is used to record the device parameters linked
with the Observation Resource. ADLIFE PEP also allows the patient to manually
record measured parameters via Web and mobile interfaces that are recorded via
Observation Resource, without the link to the Device Resource.
Once these patient-recorded data are saved to common FHIR repository, the
subscribed clients are informed, and the measurements are made available to the
practitioners. Another client app that can be subscribed to receive these patient-
recorded observations is Early Warning Clinical Decision Support Services. This
enables to proactively check deterioration in the patient’s clinical condition, and based
on pre-defined thresholds, can initiate alert messages to the respective pre-assigned
care givers such as nurses, practitioners and informal care givers.
2.4. Symptom Reporting
Based on the analysis of the clinical guidelines addressing the needs of the targeted
diseases, several symptoms are identified to be closely monitored by our clinical
reference group, such as cough, shortness of breath, sputum, dyspnea. Patients are
assigned activities to periodically report these symptoms as a part of their care plan. As
a result of requirement analysis studies, we have discovered that instead of a plain
screen for asking the existence of these symptoms to the patient, it is more effective to
ask the existence and details of these symptoms via dynamic, interactive surveys.
Based on the patient’s answers, the next question to be asked may dynamically change
by adding more detailed questions if needed. For example, if the patient answers that
he is coughing more mucus than usual, he is then asked more detailed questions about
the color, quantity, and viscosity of the mucus. In our architecture, this kind of dynamic
questionnaires are represented in Adaptive Forms as defined by FHIR Structured Data
Capture (SDC) Implementation Guide[12]. SDC has extended FHIR Questionnaire
Resource to make it possible to represent adaptive questionnaires where the

questionnaire is dynamically developed based on previous answers to questions chosen
from a set of questions.
2.5. Asynchronous Messaging between care team members and Patient Forum
As part of our integrated supportive care approach, we enable asynchronous messaging
between care team members that includes not only healthcare professionals but also
patients and their informal care givers. To enable the messaging between practitioners
and patients and informal care givers, we use the same common FHIR repository to
exchange asynchronous messages. We represent messages through Communication
Resource, where the sender and receiver parties are clearly marked, and the message
can be linked with the CarePlan, Observation, Condition or DiagnosticReport
instances. The PEP and also the chronic disease management systems are subscribed
to be notified whenever a new message is received for the practitioner or the patient.
Similarly, we use the Communication Resource to represent the messages exchanged
between the patients in the implementation of the Patient Forum. The Patient Forum
enables the patients to create virtual communities involving patients suffering from
similar conditions to share their experiences.
3. Discussion and Conclusion
In this paper, we have presented the HL7 FHIR based interoperability architecture we
have designed and implemented for seamlessly integrating patient empowerment
platforms with the underlying care systems used at clinical sites. Within the scope of
the ADLIFE project, the system and its integration with care planning systems will be
validated in seven pilot sites involving 577 healthcare professionals from 75 different
hospitals, clinics and primary care services. The effectiveness of the system will be
evaluated with 882 patients and 1243 caregivers.
Acknowledgements
The research leading to these results has received funding from the European Union’s
Horizon 2020 research and innovation programme under grant agreement No 875209.
References
[1] Holman, H.R. (2020), The Relation of the Chronic Disease Epidemic to the Health Care Crisis. ACR
Open Rheumatology, 2: 167-173. https://doi.org/10.1002/acr2.11114
[2] Wagner EH, Bennett SM, Austin BT, Greene SM, Schaefer JK, Vonkorff M. Finding common ground:
patient-centeredness and evidence-based chronic illness care. J Altern Complement Med. 2005;11
Suppl 1:S7-15. doi: 10.1089/acm.2005.11.s-7. PMID: 16332190.
[3] Small, N., Bower, P., Chew-Graham, C.A. et al. Patient empowerment in long-term conditions:
development and preliminary testing of a new measure. BMC Health Serv Res 13, 263 (2013).
https://doi.org/10.1186/1472-6963-13-263
[4] Ernesto Mola (2013) Patient empowerment, an additional characteristic of the European definitions of
general practice/family medicine, The European Journal of General Practice, 19:2, 128-131, DOI:
10.3109/13814788.2012.756866
[5] ADLIFE Project, Integrated personalized care for patients with advanced chronic diseases to improve
health and quality of lifeH2020-Contract No: 875209, https://adlifeproject.com/

[6] OnFHIR., HL7 FHIR-based Secure Health Data Repository, https://onfhir.io/, available at GitHub under
GPL license: https://github.com/srdc/onfhir
[7] HL7 FHIR Care Plan Resource, https://www.hl7.org/fhir/careplan.html
[8] HL7 FHIR Subscription Protocol, https://www.hl7.org/fhir/subscription.html
[9] HL7 FHIR Patient Reported Outcomes Implementation Guide, http://build.fhir.org/ig/HL7/patient-
reported-outcomes/pro-overview.html
[10] Continua Design Guidelines, Personal Connected Health Alliance, July 2019,
https://www.pchalliance.org/continua-design-guidelines
[11] HL7 FHIR Personal Health Device Implementation Guide, http://build.fhir.org/ig/HL7/PHD/
[12] HL7 FHIR Structured Data Capture Implementation Guide, http://hl7.org/fhir/uv/sdc/2019May/