In this section, we present the methods used in the design and implementation of the Patient Reported Outcome Measures (PROMs) and Shared Decision-Making (SDM) interventions in the ADLIFE architecture to empower patients suffering from COPD and CHF for taking an active role in the management of their diseases in cooperation with their healthcare professionals.
2.1. Selected PROMs for ADLIFE study
PROMs allow the measurement of outcomes in relation to clinical interventions from the patients’ perspective and represent a means of assessing clinical effectiveness and safety [8][9]. ADLIFE follows an outcome-based and patient-centered approach where PROMs [7], represent an especially valuable tool to evaluate the outcomes addressed in this project. These questionnaires are completed by patients to ascertain perceptions of their health status, level of impairment, disability and health-related quality of life [10][11]. Selected PROMs for ADLIFE will allow evaluating the most recent patients’ clinical context, constituting a supportive tool for the health status assessment, the decision-making process, and the definition of care goals and activities according to the patients’ specific needs.
The definition of the specific PROMs relevant for ADLIFE (i.e. PROMs that will be useful to measure the health outcomes described in the ADLIFE conceptual framework) has been a crucial step of this project. The process has been conducted by seven working teams, each one comprised of members of a multidisciplinary group of health professionals (hereinafter referred to as the ‘Clinical Reference Group’) and local project teams from each one of the participating ADLIFE pilot sites. These working groups have contributed to defining the PROMs that should be collected according to the ADLIFE functional requirements and targets in order to provide useful information to assist either in the evaluation of the patients’ health status and/or in the clinical decision process.
Based on the Clinical Reference Group’s expertise and after thorough research to identify the most suitable tools to measure the health outcomes addressed in ADLIFE, it was s agreed to include a list of PROMS to provide the information needed to cover ADLIFE health-related areas including: symptoms, functioning quality of life, clinical status, healthcare responsiveness and care. Subsequently, as part of the process of defining ADLIFE PROMs each pilot site reviewed the proposed list and evaluated their adequacy and coherence with the project in terms of the intended use, relevance, and feasibility for collecting and retrieving them. The pilot sites identified some key dimensions which were not fully covered by the proposed list of PROMs and suggested additional tools to be included in the ADLIFE PROMs list, (see Table 1): (i) The 5-level EQ-5D version (EQ-5D-5L)[18]; (ii) The COPD Assessment Test (CAT)[19]; (iii) The Modified Medical Research Council Dyspnea Scale (mMRC) [20]; (iv) The Shared decision-making: “ask 3 questions” [21]; (v) The Person-centered Climate Questionnaire – patient version (PCQ-P) [22]; (vi) The Zarit Burden Interview: 12-item version (ZBI-22) [23]; (vii) Wellbeing questionnaire (WEMWBS) [24]; (viii) Kansas City Cardiomyopathy Questionnaire (KCCQ) [25], (ix) Lawton Instrumental Activities of Daily Living Scale (IADL) [26], (x) Barthel Index [27], and (xi) Hospital Anxiety and Depression Scale (HADS) [28].
Table 1
The list of PROMs to be used in the ADLIFE project.
ADLIFE areas
|
ADLIFE dimensions
|
PROMs
|
Symptoms, functioning quality of life
|
Autonomy, control
|
EQ-5D-5L
|
Symptom control
|
EQ-5D-5L
|
Mood and emotional health
|
EQ-5D-5L, HADS
|
Social context
|
EQ-5D-5L
|
Activities of daily living
|
EQ-5D-5L, Lawton IADL, Barthel Index
|
Clinical status
|
Complexity
(i.e. hurdle, severity)
|
CAT, mMRC, KCCQ
|
Healthcare responsiveness
|
Participation
|
Shared decision making: “ask 3 questions”
|
Care
|
Satisfaction
|
PCQ-P
|
Carer burden
|
ZBI-22, WEMWBS
|
The challenge of identifying and selecting PROMs in ADLIFE was whether the questionnaires were available in the different languages spoken in each of the pilot sites. Most of the selected PROMs included in the final set list are available in the languages which patients are expected to speak in each site, although in some cases there is uncertainty about their validation.
2.2. Shared Decision Making and Decision aids in ADLIFE
The care model suggested in ADLIFE will facilitate a more active role of patients and caregivers in their own health and symptom management by implementing shared decision-making (SDM) and offering individualized adaptive interventions. This patient-centered approach, in which the patients’ values and preferences are incorporated, enable the definition of an individualized and personalized treatment for patients.
Despite professionals indicating that they consider it important to share decisions with patients [29], SDM seems to be applied in daily practice to a limited extent only. The primary barrier to the adoption of SDM in practice is clinical perception that SDM is not pertinent to the decisions they are making with patients [30]. In addition, implementation of SDM into daily clinical practice may seem inapplicable in the busy and highly responsible work of a doctor: the dilemma about the time consumption of the conversation and the impact on the clinical decisions [31]. As SDM cannot be successfully implemented without the goodwill of the clinicians [32], the ADLIFE projects promotes, encourages and offers the possibility of promptly integrating SDM through its digital integrated care platforms served for the use of healthcare professionals and patients.
The ADLIFE integrated care solution provides two complementary software platforms for the use of healthcare professionals and patients: (1) A Personalized Care Plan Management Platform (PCPMP) supported by clinical decision support services, which acts as a chronic disease management platform for multidisciplinary care team members, and (2) A Patient Empowerment Platform (PEP) for 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.
The PCPMP serves the multi-disciplinary care team members and facilitates the creation of personalized care plans for patients. It retrieves important parameters from the Electronic Health Records (EHR), and invokes Clinical Decision Support Systems (CDSS),to recommend personalized suggestions about care plan goals and activities. The suggestions of the CDSS are produced by automatized evidence-based clinical guidelines, that support healthcare professionals in creating a care plan for the patient.
As part of the care plan, roles and responsibilities of the patient in the management of his chronic condition are clearly defined. Once the care plan is finalized, this care plan is then shared with the multidisciplinary care team members via PCPMP, and also with the patient and his/her informal caregivers via PEP. In this way, the care plan and all its components can be accessed by the patients. PCPMP and PEP also enable shared decision-making while the personalized care plan is being created based on the preferences of the patients.
From at least 22 different definitions of SDM that exist [33], the ADLIFE project implements the "SHARE approach [34]”, a generalized SDM model that streamlines the nine essential steps of SDM identified by Makoul and Clayman, into five tasks: (1) Seek the involvement of your patient (practitioner), (2) Helping the patient explore and compare treatment options, (3) Assess the patient’s values and preferences, (4) Make a decision, (5) Evaluating the patient’s decision after a period of time (Fig. 1).
The SHARE approach proposes the involvement of patients and professionals in different tasks such as information transfer, risk communication and preferences elicitation, tailoring options or broader decision making. The five steps model adapted and defined for ADLIFE included assets and information related to implementation processes, the identification of activities to be carried out in each task of S-H-A-R-E approach (Fig. 1), skills that professionals and patients should acquire for applying SDM activities, and factors that influence the process and potential 'decision aids' enabling SDM.
The SHARE workflow will start when a decision needs to be reached as part of the patient’s care plan in ADLIFE. Those situations have been identified in the NICE evidence-based clinical guidelines for COPD, Heart Failure and other comorbidities by the ADLIFE Clinical Reference Group. Fact sheets with task definition, the triggers of the task, the aids for professionals, the aids for patients and the scenario in which aids are offered in specific situations are identified based on the evidence-based clinical guidelines utilized in ADLIFE. In Table 2, we summarize the information provided to the Clinical Reference Group (CRG) as a template for designing each task of the S-H-A-R-E approach and identifying shared decision-making processes that can be provided via the ADLIFE toolbox.
To implement shared decision-making directly into the ADLIFE platforms, trigger points from evidence-based clinical guidelines were reviewed by the CRG to identify an opportunity and the potential task for utilizing the SHARE model between the patient and the clinician. More than 40 trigger points were identified in the clinical flowcharts with slightly more than half pertaining to medication options. The shared decision-making trigger points are eminently suited for integration into the PCPMP and PEP platforms by developing them into generic decision aid tools. Each trigger point has been turned into a prompt on the PCPMP through the application of the CDSS (Clinical Decision Support System) framework, where the progression through the clinical flowcharts triggers a pop-up notification stating that there is an opportunity for initiating the SHARE model. Here, healthcare professionals are provided with information as depicted in Table 2, as information cards. For some of the trigger points PCPMP suggests that healthcare professionals assign educational materials to the patient, that can be reviewed on PEP.
Table 2
Implementation of SHARE approach in ADLIFE
|
TASK DEFINITION
|
TRIGGERS
|
AIDs FOR PROFFESIONALS
|
AIDs FOR PATIENTS
|
SCENARIO IN WHICH AIDS ARE USED
|
Seek your patient's (professional’s) participation
|
Preparation where the professional seek and start engaging the patient before introducing the choices.
Introduction, professional confirm with the patient that there is a choice, and we provide a rationale for choice.
|
• Selected decision points in the flowchart of the clinical guidelines or specific goals/activities.
• By request (patient, caregiver, or professional)
|
Info cards that can introduce supporting tools via PCPMP:
Adapt the information and treatment options to the patients’ emotional states and circumstances. Take into account the patient’s life style and a way of life to perform the choice
Long-term chronic patient may be reluctant to actively manage that condition and less likely to engage in the SDM discussion.
Anticipate patient doubts and the demand of information.
There are different levels of patient participation possible; from fully involved to not involve at all.
|
• Option to start the activity of SDM via a selected decision aid to be completed by the patient via PEP
|
• Preparation when reviewing the patient’s record and preparing the encounter.
• Introduction in an appointment via PCPMP
|
Help your patient explore and compare options
|
This iterative process aims to determine which of the alternative responses best reflects the patient’s preferences after personal reflection and feeling confident for making an informed decisions.
Tailor, evolve and integrate an option conversation with the patient through presenting initial options, preference elicitation; tailoring options; checking understanding the options presented.
|
• After the S task
• Selected decision points in the flowchart of the clinical guidelines or specific goals/activities.
|
Info cards that can introduce supporting tools via PCPMP:
• Patients’ previous knowledge about the condition or available treatment options can be an influence. It is preferable to review treatment “facts” options with the patient.
• Social and cultural circumstances can influence in the options presented
• Tailor the presentation of automatic proposed options to the patient and ensure that they are understood and address any knowledge error or gap.
• Keep the conversation active and plan next steps based on the results of the patient’s decision making needs.
|
• Educational materials available in the PEP will be used to help to better understand the condition.
|
In an appointment:
• virtually
• face-to-face
|
Assess your patients’ values and preferences
|
Patient view of the options and values are discussed.
Characterized by multiple and multi-stage decisions, which are distributed among the patient, clinicians and sometimes other services.
|
• After the H task
• Selected decision points in the flowchart of the clinical guidelines or specific goals/activities.
|
Info cards that can introduce supporting tools:
• Perceived urgency to make a decision may influence on the SDM process, specifically it reduces the patient’s level of engagement
• Review preferences in order to understand what “most matter” to the patient
• Ensure and check there are no further and negatively influence decision-making needs.
• Based on patient vales and preferences rate the benefits and risks identified for each of the options listed.
|
• Tools to carry out the activity (such as reviewing a selected decision-aid) can be offered in the PEP
• Share the resulting rates with the professional prior consultation
|
• Patients’ home via PEP
|
Reach a decision
|
Consolidate or summarize preferences and moving towards a decision or an ongoing reflective and iterative process until the point a decision needs to be made.
Assess the patient's readiness to make a decision.
Negotiation with the patient a mutually agreed upon a course action.
|
• After the A task
• Selected decision points in the flowchart of the clinical guidelines or specific goals/activities.
|
Info cards that can introduce supporting tools:
• Give “informed preferences” about what might be suitable for the patient. Recommendations should be based on the information that the patient had given about their preferences and checking always the patient's understanding and agreement.
• Ensure that the decision is taken based always in patient’s life style and preference.
• If applicable and necessary, check on a review frequency.
|
• Confirm the decision via PEP
• Make a summary of the next steps. (if applicable.)
|
• PCPMP
• PEP
• In an appointment: virtually or face-to-face.
|
Evaluate the patient's decision after some time
|
Review and revisit the decisions reach by the patient, if the decision aid has been used, whether a problem occurred and others.
Discuss follow-up.
|
Set in the review frequency.
|
Info cards that can introduce supporting tools:
• All decision aids are not created equal. Revise tools to help you evaluate the quality and usability of patient decision aids and other evidence-based resources.
• Follow-up patient and evaluate the treatment, care or choice taken. A good option is to set a reminder to carry out this activity.
|
• Reminder for reviewing decision
|
In an appointment:
• virtually
• face-to-face
|
Finally, two well-structured decision aids have been selected (‘Ask three questions’, and ‘Shared decision-making on inhalation medicine in patients with COPD’) to be implemented. These are offered to healthcare professionals at the identified trigger points in PCPMP, to be added to the care plan of the patient. Once they are added to the care plan and assigned to the patient, the patient can see, review and complete them in PEP. The feedback from the patient is seen on the PCPMP by the healthcare professional. As an example, before an inhalation device is to be prescribed to the patient in the scope of a care plan for COPD patients, the healthcare professional can assign the ‘Shared decision-making on inhalation medicine in patients with COPD’ to the patient, to determine the most suitable inhalation medication for a patient by having them prioritize different factors such as minimizing the frequency of inhalation medication intake, minimizing the number of different inhalation devices used daily, and minimizing the cost of medication. The implementation of decision aid is detailed in Section 2.3.
There are too many local, cultural, and practical variations in all the possible answers to each of the steps in the SHARE model across the pilot sites to build bespoke prompts for all the trigger points. The clinical guidelines are likely to change over time and available treatment options may also change in the pilot sites during the intervention. These can be easily reflected in the PCPMP by updating the CDSS triggers, and the information cards to be presented to the healthcare professionals via PCPMP.
2.3. Technical implementation of PROMs and Decision Aids in ADLIFE architecture
As part of the care plan management, an important feature enabled via the digital ADLIFE Toolbox (PEP and PCPMP) is the collection of feedback from the patient via standardized Patient Reported Outcome Measures (PROMs) and via symptom reporting questionnaires. PEP and PCPMP acting together also enable shared decision making about the care plan activities.
ADLIFE PEP and PCPMP are built upon international standards, and interoperability of data exchange among these components is enabled via HL7 Fast Healthcare Interoperability Resources (FHIR) standard [35]. We are using an HL7 FHIR Repository as the common data repository that enables seamless data exchange between local EHRs, PCPMP and the PEP. In our architecture we are using the open source on FHIR.io FHIR Repository [36].
In the following sections we will be focusing on presenting the details of the implementation of PROMs, symptom reporting questionnaires, shared decision-making processes and decision aids in the ADLIFE architecture. The overall details of ADLIFE PEP platform are presented in [37].
Implementation of PROMs in ADLIFE Architecture
In order to assess the care delivered to the patients via the 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 PROM questionnaires. A detailed list of these PROMs utilized in ADLIFE is presented in Section 2.1. These PROMs are assigned to patients as a part of their care plan, also indicating the time plan to complete the questionnaires.
In the ADLIFE context, patients are able to report Patient Reported Outcome Measures (PROMs) and symptoms via the questionnaires that are assigned to them as a part of their care plan.
In the ADLIFE architecture, we have followed the FHIR Patient Reported Outcomes Implementation Guide [38] and PROMs are represented as HL7 FHIR Questionnaire Resources in a machine-processable manner. An example HL7 FHIR Questionnaire Resource representing Kansas City Cardiomyopathy Questionnaire (KCCQ) is presented in Additional File 1. These are included into the care plan of the patient via Service Requests as care plan activities to be carried out by patients periodically. An example of such an assignment is presented within a care plan resource in FHIR format presented in Additional File 2. The Care plan resources defined in HL7 FHIR, including references to PROM questionnaire resources are created by PCPMP interfaces, and saved to the common data repository.
The care plan resources are retrieved as a bundle by PEP whenever a new care plan is created or updated from the common data repository. After this, the PROM assignments are presented to the patient as part of their care plan and rendered as user-friendly web-based surveys, enabling the patient to easily fill them. An example snapshot from ADLIFE Web based PEP portal, listing the questionnaires assigned to the patient is presented in Fig. 2.
PEP interfaces have been built to automatically render these machine-processable questionnaire definitions to present them to the patients and enabling patients to fill in these questionnaires easily. Depending on the content of the questionnaire we have enabled different presenting views to collect the responses via multiple choices, via Yes/No questions, via a slider, or via free text. Different views from web-based PEP Portal and mobile PEP application are depicted in Fig. 3, Fig. 4 and Fig. 5.
The responses are recorded as FHIR Questionnaire Response Resources and saved back to the FHIR Repository. An example KCCQ questionnaire response as a FHIR resource is presented in Additional File 3.
When the PROM includes a scored assessment, the resulting score is represented as an Observation Resource that is linked with the PROM. Once each PROM is completed in the PEP by the patient or the informal caregiver, the resulting questionnaire response is sent to a decision support service we have implemented to determine if there is a score attached with the PROM or not. CDS service receives the questionnaire response, determines if it is a scored assessment, calculates a score if it is and creates an Observation resource to be put in the FHIR repository. The score observation contains a reference to the Questionnaire Response resource indicating that it is created as a result of that specific response instance. The observation might also contain an interpretation of the score, if it can be interpreted by PROM definition. Additional File 4 contains a resulting score Observation resource created from a KCCQ response.
The PCPMP, which is designed to receive notifications whenever a new PROM questionnaire is fulfilled, is informed, and the responses are made available to the practitioners as depicted in Fig. 6. If the questionnaire is a scored assessment, the scores calculated are depicted with the submission dates. When the questionnaire is clicked, the full responses are also rendered and presented to the healthcare practitioners.
ADLIFE PEP aims to enable patients to report their symptoms as part of their responsibility in shared care plan management. In order to identify the particular symptoms, which need to be reported as part of the ADLIFE pilot studies, we have organized several meetings with CRG members. Possible implementation options to enable patients to report symptoms have been presented. As a result of discussions, CRG members have decided to use questionnaires as a means to collect the symptoms from the patients in a structured manner.
As a result of the discussions the following symptoms are to be collected via PEP through the selected tools:
Table 3
List of Symptoms to be collected via PEP
Conditions
|
Symptoms to be collected and Tools to be used
|
COPD
|
A specific adaptive questionnaire has been designed for COPD Symptom Reporting. This questionnaire is presented in Additional File 5. In addition to this, COPD Assessment Test (CAT) [19] will be used, which has already been implemented as a PROM.
|
Generic (For all patients)
|
We have designed the Self-Assessment Questionnaire as a means to be complementary to the care plan and to collect information about the perceived change in symptoms in an easy and simple way. This questionnaire is presented in Additional File 6. It is already validated in CareWell European project [39].
It can be seen as a set of self-checking questions to help the patient learning about the warning signs (self-control), and provide guidance about what actions need to be carried out, in case of symptoms (such as ‘You have perceived changes in your breathing and swollen legs, review your care plan and call your nurse or doctor for advice’). The responses are also saved and shared with healthcare professionals via PCPMP.
|
Diabetes
|
A short questionnaire has been designed to ask for new gastrointestinal symptoms 2–3 weeks after initiation with metformin, and provide feedback to the patient via PEP.
|
Mild Cognitive Impairment and Depression
|
Global Depression Scale (GDS), Hospital Anxiety and Depression Scale (HADS) and Montgomery and Asberg Depression Rating Scale (MADRS) has been decided to be used. The Physician will decide which one to use while s/he is preparing the care plan.
|
Hepatopathy
|
Alcohol Screening Tool (AST) and Fast Alcohol Screening Tool (FAST) have been selected to be used to be assigned to patient via PEP
|
Hepatopathy
|
A short questionnaire has been designed to ask for ‘Nausea and itching skin’ as liver disease symptoms.
|
Heart Failure
|
Modified Medical Research Council Dyspnea Scale (MMRC) has been decided to be used to record ‘decrease in physical functioning’.
|
All these additional questionnaires have also been represented as FHIR resources, and made available to PCPMP so that healthcare professionals can assign them to their patients as part of their care plan to be filled via PEP.
As a result of discussion in the CRG, it was decided to notify healthcare professionals via PEP as warnings when certain symptoms are reported via these questionnaires as follows.
-
Once the patient fills in the COPD Symptom Reporting Questionnaire, the system automatically checks whether based on the flow (see Additional File 4), a worsening in symptoms is detected. In this case a specific HL7 FHIR Observation is created to represent this as a red flag.
-
Similarly, the system checks the responses to the general self-assessment test (See Additional File 5), if any symptom is reported, a specific HL7 FHIR Observation is created to represent this as a red flag.
-
For CHF patients, if swollen legs, or increased cough is reported in the general self-assessment test, a specific HL7 FHIR Observation is created to represent this as a red flag.
-
For CHF patients, if there has been 1.5 kg change in the recorded weight in a period equal or less that one week, a specific HL7 FHIR Observation is created to represent this as a red flag.
-
Finally, if the patients have not filled in COPD Symptom Reporting Questionnaire or general self-assessment test in the last month at all, the system detects this and a specific HL7 FHIR Observation is created to represent this as a red flag.
These automatically created red-flag observations are presented to the healthcare professionals as warnings in PCPMP as presented in Fig. 6.
Implementation of Decision Aids in ADLIFE Architecture
As summarized in section 2, clinical partners have carried out an extensive review of SDM practices that can be employed in ADLIFE pilots and selected two particular decision aid tools that can be implemented as shared decision-making tools for ADLIFE pilots.
The first is the ‘Ask Three Questions [21]’. Research shows that encouraging patients to ask three simple questions that leads clinicians to provide higher-quality information about options and their benefits and harms. The three questions are: (1) What are my options?; (2) What are the pros and cons of each option for me?; and (3) How do I get support to help me make a decision that is right for me? An optional fourth question may also be asked: ‘What if I do nothing?’.
Via the PEP, the patient is always able to raise these questions. In addition to these, we have enabled the patient to add 5 additional free text questions which will be delivered to the healthcare professional via PCPMP to be discussed during the next care plan review meeting.
The second one is the ‘Shared decision-making on inhalation medicine in patients with COPD’. It is a tool that can be assigned to a patient, in order to assess which inhalation medication will best suit the patient, asking the patient to assess what is most important to them and what matters less. The patient is asked to prioritize different choices by assigning scores to each of them, while keeping the total score as 10 to indicate which option is more important for them. The options are as follows:
-
Keeping the daily ’frequency’ of inhalation medication intake (number of times you should take inhalation medication daily) as low as possible;
-
Keeping the number of different inhalation devices you need to use daily as low as possible;
-
Keeping the cost of medication as low as possible.
The complete description of this decision-making tool is available in Additional File 7. This tool is implemented in PEP as a questionnaire (see Additional File 8 for FHIR representation as a Questionnaire Resource). During the care planning session in PCPMP, the clinician can add this decision aid as an activity to be carried out by the patient via the PEP.
This questionnaire is presented to the patient via PEP interfaces, and the user is guided in a step-by-step manner about how s/he should answer the questions. A snapshot from the mobile PEP App is presented in Fig. 7. The responses of the patient are shared with the healthcare professional via PCPMP as depicted in Fig. 8.