MedTech Science & Innovation

Wednesday last week, and as a beginning of the Swedish MedTech week 2017, was the inauguration of MedTech Science & Innovation which is a new medical research and innovation centre in Uppsala. The centre is a long term collaboration between the Uppsala University Hospital and Uppsala University.

The day started with a welcome from Fredrik Nikolajeff and Marika Edoff from MedTech Science & Innovation. It was a busy schedule with many good presentations. Magnus Larsson, the head of the Digital Development Unit at the Uppsala University Hospital, talked about the digitalization within healthcare. Anna Attefall from Innovation Akademiska talked about how they support innovations, and she stressed the importance of user tests!

Further the program included many short presentations from researchers working with a broad range of MedTech applications. One example is Robin Strand from CBA and the division of Visual Information and Interaction at the IT department at Uppsala University (same division as the HTO group) who presented their work with advanced image analysis as a support for surgery. I was last out among the research presentations and talked about how important it is that the MedTech systems are usable, and how we work with including the user perspective.

The event ended with industry presentations, with for example Carl Bennet from the Getinge Group who stressed the importance to measure other values than costs to stimulate new innovations for better healthcare.

Listen to the presentations (in Swedish) here

HTO and DISA planning activities

Last Monday, two workshops aimed at framing and planning future work activities were conducted within the HTO group. The first one concerned different aspects of the work environment within the HTO group and the second activity aimed specifically at planning the DISA project.

During the HTO workshop, we used the affinity diagram technique to map out aspects of the work environment that we liked and aspects where we felt improvements were needed. We started out by writing down our thoughts about good and not so good aspects on post-its for a few minutes whereafter we gathered by a whiteboard on which we arranged our positive remarks in columns with related notes. After all notes had been added to the whiteboard each column was labeled to make it clear which areas worked well. Among the identified positive aspects were; good support and organization, good athmosphere and good ability to communicate to the public. When we were done with the positive side we did the same for the negative aspects that needed some degree of improvement. Among the negative aspects we found; somewhat unclear boundaries between pojects, hard to get an overview of what everyone is doing and sometimes too much information in the HTO slack channels.

Later on the same day we had the workshop for planning the DISA project. Diane, Ida and I planned the workshop and invited the other DISA members to the two hour activity. Everyone started out by writing down 2-3 studies they would like to perform within the scope of the project (some of these studies had already started). This was to make sure that every participant got the chance to express what they wanted from the project.  Those who could not attend sent their ideas to one of the participating colleagues before the workshop. After about 15 minutes everyone presented their ideas shortly and put their notes on the whiteboard. Again, the affinity diagram technique was used to cluster ideas from different participants into categories. On the picture above Diane has just started the process of assigning a label to each of the categories. On the poster to the left of the post-its the main parts of the DISA project are mapped out. The next step was to match the proposed studies to the different parts of DISA shown on the poster. This exercise resulted in a study being added – this was needed in order to make sure that the last year of the project was sufficiently covered.

The second hour of the workshop was devoted to placing the proposed and already ongoing studies on a timeline, drawn on another whiteboard, which contained relevant deadlines (like conference submission dates, special issue deadlines and dates when individual project members’ contracts with the University went out). After we had placed the studies we were conducting, or wanted to conduct, during the first year on the timeline we added information about who should lead the different studies. The end result of this workshop activity was the timeline which clearly showed all the important dates, studies and responsibilities.

Some notes from EUNIS 2017

In the SISU project we study a major system implementation in a university context. This year we got accepted for the EUNIS 2017 Conference, which is an excellent opportunity to present our results outside the academic community (no pun intended).

My presentation was one of the last but we had a full seminar room and the response was positive. (The presentation will ba available from the EUNIS web site.) Many universities all over Europe are in the same stage, transitioning from old student information systems to new, so many shared the same concerns. One example was this year’s host Münster who presented their transition to a new student management system, a project which felt quite familiar to us studying Ladok.

Of course, for me it was really interesting to listen to the other presentations and to get a feel for what is important for the community and what is on the agenda.

The three largest conference tracks were e-learning, infrastructure and management respectivly. Reocurring themes were mobility and analytics. The new EU regulations on privacy were given special attention.

The conference ended with a brilliant presentation by Nikolas Guggenberger on trust in blockchains. Guggenberger discussed the relation between blockchain and law. He concluded by pointing out how reliance on blockchain will invert the field of data protection. Not by solving the problem but rather by switching focus from known individuals with secure transactions to anonymous users with open transactions. A main point was also how the lack of a central authority is misaligned with most jurisdiction as the idea of someone ultimately responsible is often critical.

To me, the keynote on open education by Sheila Macneill was nevertheless the most inspiering. Macneill is a long term advocate for open education and made a convincing case for openness being a core value of higher education. (Her presentation is available on her web site as a recording.)

Taking a somewhat longer perspective she also noted how hard it is to make reliable predictions. Just some years ago MOOCs were seen as the next big thing, ready to make all but a few HEIs obsolete. Now they are just a part of life but instead alternate facts are shaking the wery foundation of education. “That’s just an expert, what does s/he know?”. Sheila’s answer was a plea that universities must create their own strong narratives to remain relevant.

This year all but one member of the Ladok consortium were absent due to heavy workload. Pray next year’s conference will be dominated by lessons learnt from the Ladok project.

On Teaching and Four Papers Accepted for Frontiers in Education 2017

A majority of the researchers in the HTO group are also very interested in computer science education.  As a HCI senior working at the department of Information Technology you do teaching in HCI around 30-60% of your time depending on what research projects you are involved in, and depending on what you are interested in. Most PhD students do 20% of teaching in HCI courses. So, as a student you would meet our faculty in a large variety of different HCI courses!

Many of us are also a part of Uppsala Computing Education Research Group (UpCERG) which is another research group at our department which is led by Arnold Pears. This research group does research on teaching and learning of computer science education.

This year the HTO group submitted four papers to the Frontiers in Education Conference, and all of them were accepted!

1) The first paper was a joint effort with many authors, of which Åsa Cajander, Jonas Moll and Diane Golay from HTO were a few. The paper is about student behavior and makes use of the theory of planned behavior for analyzing and understanding unexpected behavior in an HCI course. Jonas Moll has written some about this paper in his blogs. The paper is called “Unexpected Student Behaviour and Learning Opportunities: Using the Theory of Planned Behaviour to Analyse a Critical Incident”. This paper is the first one in a row of papers on the same theme.

2) The second paper presents an interview study with computing instructors who were teachers in a summer camp for children. The summer camp is an example of a maker community effort targeting girls who are interested in computer science, and hence has a gender perspective. The main author of the paper is Tina Vrieler who is a PhD student in the UpCERG group. Åsa Cajander is one of the co-authors of the paper together with Aletta Nylén. The paper reports on the experiences made and what lessons there are to learn from the summer course and is called: “What Computing Instructors Did Last Summer – Experiences & Lessons Learned”. There will be more papers published from this summer course and they will make use of social capital theory.

3) The third paper is a paper where Aletta Nylén is the main author, and Åsa Cajander is one of the co-authors. The paper discusses students and their thinking related to higher education learning, and the paper is called: “Why are we here? Student perspectives on the goal of STEM higher education”.

4) The fourth paper presents a new method for scaffolding teamwork competencies through the use of a role play and the personas method. The main author of the paper is Arnold Pears, and Åsa Cajander from HTO is one of the co-authors and writes about her experiences using the personas method for discussing strategies to motivate peers in teamwork. The paper is called “The Archetype Learning Method – Scaffolding Teamwork Competences in the Engineering Classroom”.

The Frontiers in Education conference will be held in Indianapolis in the US, October 18-21, 2017. This is one of the core conferences in engineering education and includes research on a large variety of areas such as gender and IT, programming courses and professional competencies. Usually a large group of people from the UpCERG research group are present at the conference. We’ll see if some from the HTO group will come too this year J

Patient Accessible Electronic Health Records: Germany on its way to join the club?

Doing research in DOME and DISA is very exciting for me; especially seeing the Patient Accessible Electronic Health Records (PAEHR) in practice. Every time one of my colleagues logs into the portal and shows a feature in the Swedish Portal Journalen I think: “I want this, too!” I have a background in Media Informatics and worked as a medical assistant, where I myself wrote many medical notes into the EHR. Thus, I do not consider it to be rocket science to graphically visualize lab results or show an overview of patient’s visits to the doctor. However, as I am living in Germany, it feels a bit like coming from the Stone Age seeing Journalen in action, because we are currently so far behind in relation to eHealth. For instance, our prescriptions are still on paper; as are referrals and sick notes.

End of 2003 it was determined by the German government to modernize statutory health insurance, including the implementation of an electronic health insurance card. This caused lots of critique and opposition, not only from physicians but also from citizens. The development and rollout is ongoing for more than ten years now; and the cost are already amount to 1.7 billion Euro. In the future, the card may contain some medical information in case of an emergency (like allergies, allergies, implants, chronic diseases or medication). However, claims that putting medical information on a physical card would lead to “Better quality of medical care thanks to better information on medical data” are in my opinion questionable and the very idea of the need for a physical card not very future-oriented! I really want to have access to my (whole!) EHR, but at the same time I do not want to carry it around with me in my wallet. Besides: very recent information such as lab results would not be on the card anyway. Today, the card carries the same basic information about the insured person as before (name, date of birth, address, personal health insurance number and insurance status) and additionally a photograph. Until now, the promised additional services have not been implemented. Physicians cannot even update your address yet in case you moved house, which has been one of the selling points. So in sum: Germany is way behind and it seems to me that we put millions of Euros into a technology that is probably already outdated when it’s finally implemented.

But there may be a light at the end of the tunnel:
I recently watched a talk in the re:health track of the re:publica 2017, which is a European conference about digital culture in the world. The talk was called “Play it safe – the electronic health record and data security” (my translation; available on Youtube in German, but with English subtitles). The title made me curious and I was surprised to learn that a statutory health insurance company (Techniker Krankenkasse) aims to develop and implement a PAEHR in collaboration with IBM. Immediately concerns popped into my head: what about privacy and data protection? For sure, the insurance company would like to collect more data about their insured people and consequently will change the premiums accordingly (which is already happening with car insurance). This would however go against the very idea of the „Solidargemeinschaft“ (collective body of the insured) which is the case in Germany, i.e. people insured with a statutory health insurance all pay the same premium, regardless of having a pre-existing condition, engaging in high-risk sports, or whether they don’t exercise at all. Private insurers already calculate premiums based on certain risk factors. Therefore, I was surprised and relieved to hear from the CEO of the insurance company, Jens Baas, that behaviour-based insurance is neither the motivation nor is planned in the future. Furthermore, he mentioned that he is currently in discussion with the Federal Minister of Health, Hermann Gröhe, that this should be expressly stipulated in the law to prevent this also for the future.

The main points I took from the talk:
  • In collaboration with IBM, the insurance company wants to take a lead to prevent foreign companies to take over, as it is clear that citizens want to have access to their EHR. In addition: people already collect many data that are currently stored overseas on servers from Apple, Garmin, etc. According to Jens Baas, they would like to provide a “safe heaven” for these types of data as well.
  • The insured person owns her data, not the insurance company! Thus,  people themselves decide, who will have access. As the CEO points out, not even the insurance company will have access. Neither would anonymized data be given to other interest group (such as research), as they often can be de-anonymized.
  • Being a statutory health insurance, the Techniker Krankenkasse does not aim to provide different insurance options. In his talk Jens Baas emphasized that people should have the “right to irrationality” (Recht auf Unvernunft), i.e. you do not have to pay higher premiums when you do not live up to certain rules (e.g. fail to walk 10.000 steps a day, smoke or engage in sport activities that may cause injuries).
  • Data will be stored centrally and secure – the insurance card will not be necessary for this. Maybe as an access point, but the EHR will not be stored on the card. The central data storage is of concern for many, as it could offer a central entry point for hacker. As Jens Baas pointed out: there is no such thing a 100 % data protection. This is by the way also the case with paper-based records, as was seen recently in Melbourne. However, to address the security aspect, they aimed to work together with a competent partner to make it as secure as possible, while at the same time make sure that the data is still usable. Data is to be stored in Germany and thus German data privacy acts are to be applied.
  • Although this development starts as an initiation from Techniker Krankenkasse, the idea is to make this also available for other insurance companies. This means, that we hopefully won’t have a separate patient portal for each insurance company.
  • Who gains? If the insurance company will not have access to the data, why do they invest in this? According to the CEO, they want to be able to offer additional and individualized services that attract more customers.

I think, this is very exciting! Whenever we are talking about patient portals in DOME or DISA, I was rather pessimistic that Germany will have anything close to that within the next 10-15 years. The concerns especially with regard to a central data storage is valid, but that should not prevent the development of web-based access for patients to their data in general. As Baas pointed out, we have to balance risks and opportunities; saying you don’t want to take any risks means in the end the opportunities are next to zero.

On the recent German congress of physicians, Sascha Lobo (Blogger and so-called internet expert) called on the physicians, to actively co-create the process of digitalization. Hopefully, the project by Techniker Krankenkasse and IBM is more successful to present healthcare professionals the benefits PAEHR can have – not only for the patients but also for healthcare professionals – as experiences in other countries (e.g. OpenNotes in US, Journalen in Sweden) have shown.

A holistic perspective on designing for people: service design

During his short visit two weeks ago, José Abdelnour Nocera from the University of West London held a presentation on service design. I was very curious to learn more about the topic since it was a term I had stumbled upon not only throughout my Master’s studies in relation to user-centered design, but also in countless job advertisements back when I was looking for work in the industry. I had always wondered in what way service design differed from “traditional” user-centered design, and whether my skills as a user-centered interaction designer could be extrapolated to the field of service design.

As I have understood it, the main difference between user-centered design as it is understood within human-computer interaction and service design resides in the concept of “service” as opposed to that of “product”. Service design aims at considering a product’s usage flow from a holistic perspective, from acquisition of the product (and corresponding service) to “liquidation” or end of service subscription. The product is seen as only one mean to access the service, as a mediator between the user and the service – and one that only gets its value from the service it grants access to. One of José’s examples I found very telling is that of Apple’s iPod: when one buys an iPod, one does not buy it because the device in itself is better than other MP3-players on the market, but rather because it enables us to enjoy iTunes’ offers. The iPod’s value thus does not reside in the device itself, but in the service it is associated with – the cheap and almost unlimited access to music through the iTunes store.

An interaction designer would focus on how a product is to be used, answering such questions as: what are the features the user needs, what does the user need to be able to do with the device? How will she interact with and control the device? However, a service designer would take a much broader perspective and seek to answers questions such as: how will the user learn about the product and the corresponding service? How will she set-up the device and activate the service? How will she routinely access the service? And even, how will the user terminate the service / get rid of the device?

Service design is not new and re-use many different concepts from other fields, most notably user-centered design and system design. Nonetheless, I appreciated seeing how a more holistic approach can lead to the creation of a better user experience that is not limited to the use of a product, but which comprises everything that is related to it (informing oneself about the product, getting familiar with it etc.). Service design fundamentally consists in taking a step back and considering the prerequisites and context of use of a product, a mindset that I think may be helpful in many other domains as well, including healthcare.

A View from the Top

Recently I had the opportunity to listen to a full day of presentations from what could possibly be Swedens largest software development project, the New Ladok project. The New Ladok is Sweden’s new national student information system, which will serve about half a million users when fully implemented.

In the SISU project we follow the preparations for the implementation of the new system, from what could be called a bottom up perspective. It was thus very interesting to hear about this huge project from a top down perspective–from the actual development project.

Project Leader Johan Sjödin stated that the initial development backlog was estimated to have started at 60000 hours of pure coding/development. While one major effort has been the prioritizing of the backlog a project of this size is not just coding. Adding requirements elicitation, testing, validation, documentation, learning material etc. the project so far amounts to over 400000 hours of work. That is the equivalent to more than 50 PhDs.

The presentations included high level perspectives from Mauritz Danielsson and Johan Sjödin. There were also more in depth presentations: Catherine Zetterqvist commented on backlog, the technical setup was described by Staffan Ekstedt and Anna Åhnberg talked about the concepts behind the learning material being developed. Malin Zingmark said a few words about success factors for local implementations and the new web site was presented by Rebecka Guzman.

Interesting were also Jan Winkle’s presentation from Malmö University, which has formed the avant-garde by being the first larger university to implement the new system. This was nicely contrasted with the presentation by Karim Andersson from Lund University, which will form the derrière-grade and implement the system in 2018. So far the implementation at Malmö University seems to have been successful and that must indeed have been very positive for the project as a whole.

Still this huge project no doubt has many challenges ahead, both on the project level but also on the level of local implementations. Starting with the latter there is a huge undertaking at each university relating to the conversion of legacy data and the adoption of routines.

On a project level, there are of course the usual challenges related to running huge IT-development projects. I could note two principal challenges that the project has to deal with. One is the fragmented customer base. There are many differences between universities and colleges, but also within institution there are many structural differences. As the new Ladok will have a monopoly there will have to be a one size fits all for the end product, and no doubt this will be better for some then for others.

A more intricate problem relates to the overall orchestration. The project is now in a critical phase where it is simultaneously implementing the system, keeping implementations stable and developing new versions to meet the more complex requirements that have to be supported. At the same time the higher education system is very cyclic, and the project has to adapt to the academic year leaving little room for quick adjustments.

All in all, it would seem that the new Ladok would be a perfect example for the hype curve. At the beginning there seems to have been rather unrealistic expectations. The project has since worked hard creating a focus and in a sense lowering the expectations (at least for the first version). There will no doubt be a rise towards the plateau of productivity. The question that remains is whether the project already has hit rock bottom and is working towards that or if the large scale implementations ahead will be a cause of concern. It will definitely be interesting to follow this huge project as it proceeds and I hope we could gain further insight into the development project as such.