In the opening Professor John Mantas chair of the scientific committee commemorated the bicentennial of the Greek revolution (1821-2021) with a video that shared the modern history of the oldest democracy.
Then, in my role as president of EFMI for the next two years, commemorated the 40 years from the establishment of the European Federation for medical informatics, with a video developed by the EFMI Honorary Fellow, Professor Izec Masic (https://vimeo.com/554429832). In this way, my opening address connected the rich past of EFMI with its digital future. I shared the strategic initiatives of eFMI for 2021-22, renew, connect and grow, focusing on yEFMI led by Ivana Ognjanović, digital EFMI by Paris Gallos, Institutional Member Collaboration by Lars Lindsköld.
Highlights from FAIR4Health (on fair health data for research, www.fair4health.eu, led by Carlos Parra Calderón) and HOSMARTAI (www.hosmartai.eu, led by Lacramiora Stoicu Tividar).
Elske Ammenwerth received the EFMI AC2 Certificate for the biomedical informatics program of University UMIT Tirol. While we could only see each other online, thanks to past EFMI presidents, and current honorary fellows George Michalas, Christian Lovis, and Jacob Hofdijk we were able to take a stroll in the memory lane. We remembered past times of scientific and cultural interaction in person, during past MIEs.
Following the opening ceremony, Christian Lovis delivered a keynote address entitled “COVID-19 predictions : reality check” that advocated for high quality data noting that when legal, organizational and other restrictions to the reuse of health data are lifted, we are left with data that simply are not of sufficient quality. A key lesson from the pandemic said Prof Lovis, “is that the problem was not access the data, but use of the data. Moving forward, we should focus on semantics and describe the data at the source”, he reflected.
As EFMI moves into the digital world, we need to find new ways to interact. Therefore, I asked some our EFMI colleagues to reflect on the sessions of this first day of MIE2021, please enjoy and share your own observations and thoughts. Until we meet again in person and enjoy stimulating scientific debates.
Catherine Chronaki, EFMI President (2020-2022)
Session 1 Biomedical data, tools and methods (Room A, 9am), Chair Anne Moen.
By Anne Moen, EFMI Honorary Fellow, EFMI President (2015-2016)
During the session we learned about novel informatics approaches to solve logistics, diagnostic, and management problems. This picture is from “ExpetingU”, which sought to improve orientation as well as patient and visitor flow into a complex institution, seeking to meet specific challenges during COVID-19 pandemic. We also heard about analytics approaches to improve accuracy in ICD-10 coding, diagnostic accuracy, detect risk of self-harm, from mental stress or alcohol use disorder, as well as the intricacies influencing adherence to specific treatment. Discussion with the presenters focused om Good Data, patient consent to use EHR data for analytic purposes and further work to advance predictive power going forward.
What do you think of Anne’s account of session 1? Send us your thoughts at info@efmi.org mentioning the “MIE2021 Daily May 29, Session 1” subjectline. You can access session 1 recording in the video library.
Session 3 – Health and prevention (Room C, 9am), Chair: Romaric MARCILLY
by Romaric Marcilly, Chair EFMI HOFMI WG
The session explored the availability and usefulness of health databases in the contexts of vaccine certification and of profiling of HIV pre-exposure prophylaxis candidates. It also explored the meaning of “safety” for assisted living technologies and the adaptation of data visualization to the profile of people involved in the care process of smart home occupants. Finally, the development and evaluation of applications to support nurses’ shift report and to empower patients with kidney diseases were also presented.
What do you think of Romaric’s account of session 3? You can access the recording of session 3 in the video library. Send us your thoughts at info@efmi.org mentioning the “MIE2021 Daily May 29, Session 3” on the subject line.
Session 4: Precision medicine and public health – Public health informatics (Room A, 10:30), Chair Jacob Hofdijk
By Jacob Hofdijk, Honorary Fellow, EFMI President (2009-2010)
Had four presenters from four different continents and four different time zones. Melody Greer was actively participating in the discussion at 03.00 am Arkansan time. The four papers each bridged the link between a clinical issue (precision medicine) and a public health issue. Two were related to Covid, where the paper from Arkansan showed that there is an interesting difference in high rates of Covid infections in the large group of low health literacy in the rural population, while in the metropolitan area with higher rates of health literacy and lower Covid infection rates. The research showed the relationship between the Social Determinants of health (SDoH) and health, a breakthrough. In the discussion we created some perspectives for a further use of the SDoH in primary care and the improvement of the Internet accessibility in the USA.
The other COVID paper was from Arizona on Deep learning methods based on a literature search, with the intension to create a prediction model for mortality. DL subset of AI have become an important technology in health-related crisis management. The study was based on the first papers, which were based on a limited dataset. It was an interesting test, but we know that we still are learning what the relevant variables and models are before we can really predict the mortality. The use of Deep Learning technics has already been tested.
From Iran Atiehsadat Mirkhani presented his work on a Prognostic Model to Predict Mortality in Patients with Acute Bacterial Meningitis. It was interesting to realise that the new cut-off points for CSF protein level value and the age were decisive to a rather stable predicting model. It was a pity that other areas of Iran could not be involved and that the connection with Iran was lost, so we could not have a discussion on the proliferation of the model to primary care.
The last trip of this session was Indonesia, where Angelina Kurniati presented her research on UK data from Leeds on Endometrial Cancer. Angelina constructed after a long time waiting for approval to use the data Patient Pathways based on the timestamped process data from GP Referral to the First Treatment. The research gained a lot of value by the discussions between the process miners and the clinical specialists. The interaction proved a strong asset of the work. The analysis focused on the high variability of the pathway. In the discussion we identified the lack of the outcome of the different diagnostic tests and procedures could add an important additional dimension to the process mining approach. This approach which was tested on UK data in Indonesia can as principle also be used for other clinical pathways in other regions. With four presentations we had an interesting discussion and the chair offered the over 40 participants a half hour coffee break.
Session 5 – Human factors and citizen centred digital health Chair: Ulla-Mari KINNUNEN
By Ulla-Mari KINNUNEN, EFMI Community, IMIA NI SIG, Finland
The session “Human factors and citizen centred digital health” included six full paper presentations related to health informatics (HI)/telemedicine competencies, and teaching and learning HI. A very timely topic internationally is the gap between technology and seniors. We heard a presentation of user interface development. Participants (n=11) aged over 70, and the oldest about 95 years, were involved in multi-method research.
A topic of citizens’ understanding of their laboratory results was interesting. It is important to transform lab test results in a such format that citizens without knowledge of health care can understand them. All in all, e-health literacy competencies for citizens are important. An evaluation of an app was performed to find out the how this can be done to better support citizens engagement in each own health and well-being.
A vide international survey was conducted in 18 countries concerning emergency remote learning (ERL) for nursing students during COVID-19 pandemic. The questionnaire was developed by the IMIA Student and Emerging Professionals Special Interest Group (IMIA-NI SEP), it was translated in 7 languages. The result showed vide variation in online teaching methods. Even the problems faced, teachers will probably use the methods also in the future.
Two presentations highlighted the importance of medical students’ competencies in digital health. We discussed e.g. what the competencies are and in what phase of the studies should those competencies be teached. So far there are too few courses in medical curriculum or at least not what IMIA recommendations suggest.
Related to COVID-19 pandemic we heard a presentation of Open WHO Massive open online (MOOC) courses. The utilization of courses was high, and sure we all agree the importance of them is this internationally harmful situation.
Workshop Defining One Digital Health or how training the future generations of One Health professionals (Room B, 14:00): EFMI’s new “One Digital Health” WG presented at MIE2021
By Oscar Tamburis, Chair EFMI ODH WG and Arriel Benis, Executive Officer (2021-2023), co-Chair ODH WG
The new Working Group of the European Federation of Medical Informatics dedicated to One Digital Health (ODH) has been officially presented by its founders – Arriel Benis (Co-Chair), Catherine Chronaki, Anne Moen, and Oscar Tamburis (Co-Chair) – in a workshop held during the first day of the MIE 2021 Conference.
The One Digital Health framework, is articulated around the “ODH Steering Wheel” with:
· 2 enabling Keys (One Health; Digital Health);
· 3 overarching Perspectives (Individual; Population; Ecosystem);
· 5 interconnecting Dimensions (Education; Citizen engagement; Environment; Healthcare Industry 4.0; Human and Veterinary Healthcare), and;
· a 1 Technology Ring.
One Digital Health aims to digitally transform future health ecosystems, by implementing a systemic health and life sciences approach that takes into account broad digital technology perspectives on human health, animal health, and the management of the surrounding environment. This approach allows for the examination of how future generations of health informaticians can address the intrinsic complexity of novel health and care scenarios in digitally transformed health ecosystems.
The main challenges that the new EFMI’s WG “ODH” is called to deal with include facilitating and improving interactions between One Health and Digital Health communities, allowing efficient interactions and the delivery of near–real-time, data-driven contributions in systems medicine and systems ecology. One of the first objectives is to develop an information model to connect animal health, human health and the overall ecosystem.
LinkedIn community webpage: https://www.linkedin.com/company/efmi-working-group-one-digital-health/
One Digital Health framework seminal paper: https://www.jmir.org/2021/2/e22189/
Session 8 – Ethics, Legal and societal aspects, Chair Arriel Benis
By Arriel Benis, Executive Officer (2021-2023), co-Chair ODH WG
The session explored the use of transversal approaches related to various informatics approaches and tools for trying to understand the diversity both from the patients, clinicians, and researchers’ perspectives. This session allowed to point-out the short and long terms impacts of the COVID-19 pandemics on healthcare services, research, personal and professional developments.
Impressions after moderating Session 9 (room C, 15:00) Session 9 – Biomedical data, tools and methods – Artificial Intelligence, symbolic reasoning, machine learning
by Mikhail SHIFRIN, EFMI Council Member, National Representative Russian Federation.
Session 9 comprised presentations and pre-recordings of 7 accepted papers, followed by live discussion and Q&A. All presentations except #2 were delivered in the presence of about 35 attendees. Find hereafter my general impressions. Unlike face-to-face conferences, timing was excellent, and it was enough time for the discussion. Unfortunately, the audience was not active, there was a single question asked by one of the presenters. Design of presentations was at rather high level, though sometimes letters were too small. As soon as I remember, at one of the ICIMTH conferences, there was a master class devoted to design of presentations – it seems that it may become a good practice like traditional master classes “How to write a good scientific paper”.
In this session, I reinforced my impression that many researchers, especially young ones, pay too little attention to the formulation of the problems they solve. At least 3 of 6 presented works may be classified as ‘proof of concept’, and authors understand and sometimes declare it. As to my mind, ‘proof of the concept’ has to investigate some simplified or restricted but medically reasonable problem. Unfortunately, many authors using ML or NLP methods solve problems that can be satisfactorily formulated in formal terms but cannot be implemented in practice. With no doubt, such problems may be a necessary stage of the investigation, but the presence of medically reasonable and implementable problems, even at the horizon, makes every stage more significant and purposeful. There is a very clear reason for appearing of many such ‘proofs of the concept’: free access to rich libraries of ML and NLP programs and mighty computers. These circumstances provoke researchers to do a lot of computational work without deep understanding of it from the point of view of the applicability of the results in real practice. In connection with this, I have two proposals.
First, to introduce a new label at submissions to our conferences – ‘proof of the concept’. It will not be a kind of black mark; it will just indicate the stage of the work and inform listeners/readers not to wait for completed results and implementations.
Second to organize at future conferences master classes: ‘How to state formal problems, so that results would be useful and implementable’.
PANEL – Implementing HL7 FHIR: The FHIR Accelerator Program (Room B, 16:00) Chair Charles Jaffe.
By Catherine Chronaki, EFMI President (2021-22)
This panel organized by Charles Jaffe, introduced several HL7 FHIR accelerator programs presenting the scope and vision and recent developments. Accelerators presented included Argonaut (Brett Marquard), DaVinci (on payer use cases, Viet Nguyen), CodeX (on cancer models, Steven Bratt), Gravity (Socioeconomic Determinants, Walter Suarez), Vulcan (Robert Goodwin on clinical research), and CARIN (Bluebutton, patient access, Ryan Howells). Viet presented the model for use case maturity shown in the picture below, demonstrating also that different use cases are at different stage of interoperability. Robert described the use case of adverse event, and showed an impressive set of activities organized.
Session 11: Health and Prevention (16.30 – 18:00) Chair Kaija Saranto
By Kaija Saranto, Chair EFMI Education WG
The session had six interesting topics which focused on Health and Prevention from different angles eg. ePrescription, booking and communication, as well proposed digital solutions by using patient use cases (older adults, breast cancer and depression).
ePrescription is an EU level endeavor but we can also ask is the on-line technique suitable for all countries. QR-codes are used for multiple purposes and the technique could have possibilities also for cross-border exchange of ePrescriptions. However, whatever digital technique eg. for booking we propose in health care we always need to remember the users and their level of knowledge and skills. Covid-19 pandemic calls for various solutions for remote health monitoring to create and produce situational awareness and to support messaging between patient, professionals and administrators.
Covid-19 outbreak has also created needs to focus more effectively on older adults’ quality of life. Web-based cognitive training software could have potential to improve interaction and maintaining mental health among older people. The exponentially growing amount of scientific literature calls for solutions to generate knowledge synthesis frameworks that extract and organize validated information mined directly from scientific literature. Knowledge graph could be one solution to represent associations between risk factors and incidence.
Session 12 – Precision medicine and public health (May 29, 16:30) Chair: Ulla-Mari KINNUNEN
By Ulla-Mari KINNUNEN, EFMI Community, IMIA NI SIG, Finland
In session “Precision medicine and public health” three full paper and six short communication papers were presented. We had 17 participants totally around. Here are some examples of the wide variation of the excellent presentations. Because of so many rapid presentations, probably the questions for presenters were difficult to set. However, finally, we had a good conversation. Thank you, all the active presenters, from Thailand, Argentina, German, Italy, USA and Malaysia. I would like to thank also the technical, background support during all my sessions. Zoom, chat and Q&A worked very well.
A cross-sectional retrospective study was conducted in Thailand of patients (n=327) suffering from tuberculosis. Both electronic and manual data was used to find out the prevalence and expected survival time of drug-induced hepatotoxicity. The results showed differences in patients under and/or over 50 years.
A presentation of the preliminary findings of medication errors showed problems in ordering medication. The physician trainees don’t always understand the physician’s orders, and that might lead to that he/she might enter a wrong medication information. Also e.g. high workload of physicians and lack of pharmacists effect to this situation. Patient safety will be improved when these problems will be faced and fixed.
Several presentations were related to COVID-10 pandemic. We heard about the tracing app and the perceived usefulness and intention to use the app. Also, a study of digital skills of how to deal with the information related to pandemic was discussed. Differences were found between the subjective social status and educational status of citizens. Very important and timely topic had the presentation of spatial inequity of the ICU beds in different regions in Italy. Likewise, a computational solution, a reporting system for ICU capacity was developed. Two different systems were offered, and the mobile device or web portal system was preferred and the most often used.
Pandemic situation has raised demands to develop teaching methods. Problem based learning (PBL) in not a new method but can very easily be implemented also in online teaching (ePBL) for health care professionals. The presenter recommends highly to use the method.
Workshop Designing, modeling and implementing health ecosystems in transformation
By Bernd Blobel, EFMI Honorary Fellow
HL7 Germany, represented by Bernd Blobel, organized the Workshop presenting contributions from 5 HL7 Affiliates. Ecosystems in general and transformed health ecosystems in special require communication and cooperation of stakeholders from many different domains using their specific methodologies, terminologies, ontologies and individuals skills. Therefore, we have to represent the concepts guiding the stakeholders involved in the business case in their specific context through domain-specific ontologies and to harmonize them by top level ontologies. Thereafter, we must transform the resulting business viewpoint into the enterprise, the information, the computational, the engineering and finally the technology viewpoint according to the development process specification of ISO 42010 and ISO 10746. Bernd Blobel (HL7 Germany) introduced the principles and challenges of designing, modelling and implementing transformed health ecosystems and the architecture-centric, ontology-based, policy-driven solution standardized in ISO 23903 and deployed by all workshop speakers. Stefan Schulz (HL7 Austria) supported that approach by highlighting, why ontological thinking matters for representing and harmonizing entities and relations in the biological domain. Ken Rubin (HL7 USA) explained the value of healthcare architectures for business and process modeling and the approach of Business Process Management +, when modeling the enterprise viewpoint. William Goossen (HL7 The Netherlands) demonstrated the transformation of ontology-based clinical concepts into clinical information models and related applications. Frank Oemig (HL7 Germany) discussed in detail the model transformation by addressing relations and dependencies of information models, data models and implementation artifacts. Mauro Giacomini (HL7 Italy) finally presented modeling and model transformation up to the corresponding FHIR resources for integrated, disease-specific regional health systems in Italy. In their presentations, but also in the discussion, all attendees emphasized the necessities and opportunities provided by ISO 23903 for enabling knowledge-level interoperability as well as the integration of independently developed specifications.
What do you think of Bernd’s report from the ecosystem workshop? You can access the workshop recording in the video library. Send us your thoughts at info@efmi.org with “MIE2021 Daily May 29, ecosystem transformation workshop” on the subject
You must be logged in to post a comment.