Valuable visuals

Valuable visuals

Defining a design space for presenting medical results

Linnea Öhlund

Department of informatics Magister thesis, 15 hp

Human Computer Interaction & Social Media

Abstract

Information visualization aims at creating visually concrete representations of big and abstract data. Research shows that the subject has good potential when it comes to conveying information within many various sectors. However, there is a scarce range of research defining a design space for information visualization tools with the aim of trying to convey clear and true information that is completely understandable for many different target groups with varying educational background, cognitive state, etc. To explore the gap of such design and the potential that information visualization has towards it, a study was conducted to gather the experience of healthcare practitioners with various experience of examining and presenting results to patients with a certain condition. To gather their opinions and thoughts, semi-structured interviews were used and from the results, four defined design aspects emerged. These four emerged design aspects have potential to work as a base when designing a future design concept.

Keywords: Information visualization, HCI, healthcare, design

1. Introduction and research question

In recent years, the demand for understandable visualizations of big and abstract information has risen (Tiziana & Cruz, 1996). The need for quick and easy graspable information can mean that services will be better sold, products are used by a wider audience and in extreme cases, it can be a matter of life and death. The topic of visual representations has according to some been around for centuries, linking it back to how art is a way of visualizing the world (Bailey &

Pregill, 2014). A topic within the subject of understanding and visualizing information and data is called information visualization. With the development of technology, information visualization has emerged together with human-computer interfaces, data mining, imaging, scientific visualization and graphics (Gershon, Eick & Card, 1998). It can be seen as a subdiscipline of human-computer interaction and is at the same time a big and diverse subject that can be found in many fields (Tiziana & Cruz, 1996). For example, in HCI information visualization can be used to create interfaces that visualize large data (Conti, Ahamad & Stasko, 2005; De Oliviera Barros & Bertoti, 2012). In the governmental sector large information systems are hard to navigate and can be more effective with information visualization (Ingram

& Benford, 1995). In the health care area results from tools and examinations have to be presented truly and correctly to provide adequate treatment for some patients (Sharma, et al, 2018).

Visualization can come in many forms and depending on the input and output, ergo what the data is and how it’s going to be visualized some different types of visualization categories exist. Information visualization often refers to the data input being big and abstract and

This could, for example, be big technological data. Another signifier of information visualization is that the output should be graspable for not just one specific type of user. Of course, the material should be graspable for a user type but if the data is only understandable for scientific researchers and technically trained professionals it is sometimes instead referred to as scientific visualization. Scientific visualization is specified by its different types of both output and input. These two visualization concepts, the difference between them and why one will be focused on, is further discussed in related research.

One area that is constantly changing and evolving together with digitalization and visualization is the medical and health care area (Ratia, Myllärniemi & Helander, 2018; Topol, Steinhubl, Steven, Torkamani & Ali, 2015). Hospitals and the medical sector are undergoing massive changes relating to this. Services that were before only physical can now be used digitally online and methods of measuring different variables and other tools have gone from being analogue to digital (Danna et al, 2019; Anderson, Agarwal, Mishra & Angst, 2007).

Digital solutions can be very beneficial for practitioners and patients and sometimes a digital solution is both time-efficient and cost-effective (Murray et al, 2016). These digitalized tools often contain some sort of visualization, whether it be information or scientific. Specifically, information visualization is a useful method when it comes to understanding data in a higher degree and can be of big value to the people who are to understand it (Gresh, Rabenhorst, Shabo & Slavin, 2002).

The data or information sometimes needs to be understood by more than a user group. In some cases, the information provided should appeal to both creator and customer, or practitioner or patient. Within medicine, it is of big importance that results and data from medical examinations are conveyed in such way so that the individuals who are to make important decisions regarding someone else’s or their own health can gather and understand the information (Faisal, Blandford & Potts, 2013). Another aspect of information visualization within medicine is that it is sometimes used in tools and treatment methods where the results can be helped by objective views. This is because when evaluating the status of a patient within the many fields of medicine some evaluations are based on the estimations by the practitioner (Bäcklund et al, 2017; Jain, Kuskowski & Selcuk, 2012). These subjective estimations can mean that examination results can be different depending on the practitioner and the way the examinations are done. This also means that when presenting results, it may come to vary depending on who has done the examinations but also depending on who is to understand the information. This is because not everyone may understand the entirety of the result.

In summary, information visualization aims to aid for example systems, products, tools, services and interfaces with the understanding and translation of somewhat abstract data into more comprehensible visualizations. The usage of information visualization is widespread and has many benefits to the fields where it is used. The two topics of information visualization and presenting results from medical examinations can be explored relating to each other and the possible potential it may have. This study aims to explore the value and potential that information visualization can have and wants to define a design space for future designs. This in relation to practitioners and patients suffering from a certain medical condition. The aim

- What is the potential of information visualization when presenting results from medical examinations towards various users of different backgrounds in relation to the condition of hydrocephalus and how can these findings help to define a design space for future design concepts?

A qualitative study exploring the above presented question was conducted and in the section below the case given will be presented and further explained.

1.1 Case

At a R&D department a tool aimed at measuring certain parameters of the gait and balance functionality of patients suffering from hydrocephalus has been developed. Hydrocephalus is a condition characterized by fluid accumulation in the brain and a disturbance in the hydrodynamics of the cerebrospinal fluid inside and surrounding the brain. Elderly individuals suffering from hydrocephalus may experience symptoms such as gait and balance disturbances, cognitive decline and urinary incontinence. The system developed by the

R&D-department consists of a measuring device that is put on the patients calves during an examination and measures, for example, the height of the foot while walking, the pace of the patient and the distance between the legs. The data from the examinations are sent to a digital interface that translates the data into specific parameters of interest to the clinical personnel.

In the program, the data is visualized in a certain way making it easier to see and interpret it.

This tool aims to make results from these examinations more concrete, since it offers an objective way of measuring certain aspects of the patients gait and balance ability.

In order to explore the aim and research question of this thesis, interviews with a selection of practitioners at certain hospitals in Sweden will be held. In the section below related research will be presented.

Image 1. Image of digital interface createdby department

1.2 Background

Information and data are created more now than ever before. With the rise of technology and the internet of things, in every country, in every business and almost everywhere products, tools and services are being used to create an information-rich society (Venkatraman &

Venkatraman, 2019). Sometimes information and data can be big and abstract, leading it to be hard to grasp or hard to make concrete summaries of. This is because of the data not being comprehendible enough to make sense of just on its own, and instead, it needs to be translated into data that is more graspable (Gershon, Eick & Card, 1998). By turning abstract data into informative visual representations, it is possible to further understand the information provided by the data. Information visualization is a subdiscipline of human-computer interaction and it combines aspects of human-computer interaction, graphics, scientific visualization and according to some phycology (Myers, 1996; Gershon, Eick & Card, 1998; Zuk et al, 2006). Information visualization stems from the development of digital technologies and the need of wanting to understand the massive amount of data being produced (Blythe, 2000) It emerged in the field when data started to become bigger and the need for being able to grasp the large amounts of information increased.

Gershon, Eick & Card (1998), attempts to define the parameters of information visualization against scientific visualization claiming that the two are separate and should be treated as such. Some researchers mean that information visualization and scientific visualization can be merged as one topic since they both involve the visualizing of data (Faisal, Blandford & Potts, 2013), but this thesis will treat them as two separate topics and will assume the stance of information visualization. The arguments for this and a definition of the two visualization concepts is explained below.

Knowing how to visualize either data or information can be challenging. Sometimes information does not connect to anything “natural”. For example, if something medical is being visualized, like an image of the body or something body-related, this can also be seen as the natural choice of visualization. If, for example, something less “natural”, like rows of codes are to be visualized the natural choice of visualization may be less “obvious” as in the example of the body. According to Gershon, Eick & Card (1998) depending on who the user or receiver of the data or information is, this is also where the different types of visualizations can be used. If the user group is broad, meaning that many different individuals with many various backgrounds are supposed to grasp and understand the data or if the audience is narrow and has for example collected knowledge of a certain field, this also means having different goals with the visualization. The first means creating information visualization and the latter, scientific visualization. Gershon, Eick & Card (1998) write about the advantage of knowing the difference of the two types of visualization techniques and that this gives the creator different directions in the sense of what is to be visualized and who it is for. According to some research, the difference between information visualization and scientific visualization is very marginal while according to others the difference is considerably larger. (Faisal, Blandford & Potts, 2013;

Gershon, Eick & Card, 1998) It is nonetheless a big field that is used in many varying parts of society. A few examples being in the technological and data sector (Craig, 2015; Emerson,

Churcher & Cockburn, 2013; Marco et al, 2016), the governmental sector (Ingram & Benford, 1995) and the healthcare sector (Jalote-Parmar & Badke-Schaub, 2010; Zhou et al, 2019).

Figure 1. Information visualization compared to scientific visualization (Gershon, Eick & Card, 1998.)

Despite there being somewhat widespread opinions as to what can be defined as scientific visualization and information visualization, this thesis will assume the stance of information visualization since the audience/users of the visualizations in this case are both specialized and not specialized, thus making them diverse. Having a diverse target group with varying education background is one of the criteria for information visualization. Despite the fact that physical data sometimes counts as scientific visualization the data in this case is more abstract towards one target group. For the practitioners the data is not abstract, but it is big, and for the patients the data can be both big and abstract. This, in combination with the fact that there is other research where the data input is physical and still relating to information visualization (Xu et al, 2006: Plaisant et al, 1998: Forsman, et al, 2013), creates a base to stand on when using information visualization.

Although information visualization has advanced into a big and varied field the background and roots can be found in HCI. In the section below the history of information visualization will be briefly presented and examples of how it has been used in HCI will be given.

2. Related research

Relating to the focus of this thesis are many various research fields both within information visualization, HCI and healthcare. Prior research done within these fields will be explored, related to and from this a research gap will emerge.

2.1 Information visualization and HCI

In relation to technology, human computer interaction is a discipline focusing on the interaction between people and the technological devices they use. The discipline focuses on the design, interaction and effects of computers amongst the individuals who are to operate the devices (Myers, 1996). Using and appreciating technological artefacts means also being able to understand them. Applying methods that will help with this can, amongst other reasons, be very economically beneficial (Myers, 1996). With the development of technology, the subdiscipline of HCI called information visualization emerged from a need to concretize the massive amount of data being generated (Tiziana & Cruz, 1996). Large amounts of data can be called big data and have become a phenomenon since the amount of data being generated has increased for every year and the prognosis for the future seems to only inhabit larger amounts

of data (Keim, 2002). This means that the need for methods of visualizing these large data sets will also most likely increase. Rendering visual representations of big data can be conducted in many ways. There are, for example, 2D and 3D techniques, data tables, graphs, network tree visualization and other treemap visualization techniques. These examples are all information visualization techniques with the aim of finding relationships between very large and at times abstract data (Bouthier, 2002).

When information visualization started to emerge as a field of its own (although still a subdiscipline of HCI) it was not created with a theory as its base. Therefore, information visualization does not have a framework to explain it. This can be argued for being one of its bigger disadvantages since it means that information visualization is lacking a part which some may argue is crucial for the existence of a field (Purchase, Andrienko, JankunKelly, Ward, 2008). Purchase et al (2008), write in the article Theoretical foundations of information visualization about how it may be the case that since information visualization stems from so many various fields it requires the theory of likewise as many fields. But despite being without a proper theory much work has been done trying to find an appropriate theory for the field.

Zhicheng Liu, Nersessian, & Stasko (2008) argue for why distributed cognition is the adequate theory to apply to information visualization. They argue that a large part of information visualization includes human cognition and perception, this means that the most adequate theory for information visualization should also consider this for the best possible outcome.

This means that distributed cognition provides the concepts and methods that are essential in order to analyze the large part of information visualization that includes human cognition and perception. Another theory explored concerning information visualization is human cognition (Patterson, et al, 2014). Human cognition is much like the distributed cognition framework focused on underlying cognitive processes. Patterson et al. (2014) make an argument for why the framework of human cognition means that well-designed visualizations promote the user to gain insights, reason and understand the presented information. Lund (2004) writes in his thesis “Massification of the intangible” about the possible use of embodied realism as a framework theory for information visualization and discusses the question if artefacts can express theory and if in that case information visualization expresses the theory of embodied realism. His focus lies more into the acting through the body and mind, than the previously discussed theories of distributed and human cognition. These examples show that there is much research conducted towards exploring potential theories for the topic of information visualization even if the topic was not created with a specific theory in question.

Although information visualization is not specifically theory-based, this does not seem to be a discouraging fact when using it for many various purposes. This can be assumed from the many examples of information visualization within HCI and the broad range of information visualization in general. It has for example been used to create an information visualization tool to support decision making under uncertainty and risks (Daradkeh, Churcher &

McKinnon, 2013). The study wanted to shine a light on the pressure that can come from having to make decisions possibly affecting one’s economy, relationships or work-related matters. The result of the study showed that the tool that gave the users visual information and options

relation to mobile interfaces in contrast to its normal field of usage which is computational interfaces (Craig, 2015). In the article, design guidelines were created and the results from the study showed that there is great potential in using information visualization on mobile devices and that animations and interaction can be key aspects to a smoother interface making it more used by a potential target group. In addition to these examples, there are information visualization tools focused on holding context (Cox, Grinter & Mantilla, 2000), user needs and appropriate visualizations (Conati & Maclaren, 2008) and the visualization techniques most suited for different types of user characteristics (Toker, Conati, Steichen & Carenini, 2o13).

These examples of information visualization within HCI is only a few out of many and the subject ranges further beyond the field of HCI and into other related areas.

In summary, information visualization means the visualization of large abstract data sets into more graspable representations meant for users of varying educational and professional backgrounds. It is a multidisciplinary field with many relating influences and is not theory- based per se, but there is much work is done and efforts made on why some theories are more suitable ones. Information visualization is well used within HCI for many varying purposes and can also be found in related fields. One such field is healthcare. The way information visualization is used and applied within healthcare and medicine will be presented below.

2.2 Information visualization within healthcare

In recent times more and more systems, methods, tools and processes within healthcare have become digital. With the rise of more easily accessible technology and a will to further explore digital options, healthcare has changed massively over just a few years (Rydenfält & Persson, 2020). The reasons behind information visualization which is broadly speaking turning large and/or abstract data and information into more visually graspable representations have been discussed in earlier sections. Also mentioned is that information visualization is used within many sectors such as HCI, government, economic and amongst them, the healthcare sector. In this section, more focus will be put on examples of what these types of information visualization can be and how they are applied within health care.

Information visualization is about turning complex (and sometimes abstract) data into a more comprehendible and graspable form. Within medicine, this can mean visualizing data in relation to many user groups with varying education and knowledge. The data might come to be understood by practitioners, who are the ones gathering the data or patients who are the ones affected by the data. They may see results from examinations, protocols, and other information that has been visualized. Based on the information presented, practitioners often have to make decisions regarding the health and quality of life of the patient (Faisal, Blandford

& Potts, 2013). Also, patients who are the ones being affected by the data have to be able to make sense of it in order to make important decisions. This can include managing various diseases, conditions or gathering information to be able to pass it forward. In some cases, there is a third party in the form of a caregiver or relative who is the one handling the patient and their medicines. In order to provide proper care, the caretakers need collected and adequate information (Rajwan, Y., & Kim, G. 2010). These various groups make up for a base in which information is to be understood by all of them. Therefore, much information within healthcare

Another example of how information visualization can be used in healthcare is helping with subjective and objective estimations. Many methods are based on the educated estimations of a practitioner who makes certain choices depending on how they think the treatment has gone or the state of the patient. They may discuss their estimation within their teams but as a result of this, the outcome of some medical data or some medical estimations vary depending on who is making the estimations (Bäcklund et al, 2017; Jain, Kuskowski & Selcuk, 2012).

An example of an information visualization tool existing in healthcare is patient record tools. They try to make sense of personal health information provided by different sections within healthcare. This is in order to better collect what can sometimes be a massive amount of information on one patient but made by several practitioners. This creates a challenge for both parts in trying to understand conditions in order to better the life quality of patients by overlooking the massive amount of data provided. In some instances, patients are also being encouraged to get a better view of their health and make decisions about their life. Information visualization has been proven to have potential in helping individuals, such as practitioners and patient getting a better overview and as a result of this make well-informed decisions (Faisal, Blandford & Potts, 2013).

Sometimes studies done within medicine can render big amounts of data. On its own, the data can be very complex to understand and in particular, the relationship between the data can be hard to grasp. Another attempt to use information visualization within healthcare was made by Forsman et al (2013). In the article, Integrated information visualization to support decision making for use of antibiotics in intensive care, a need for support in making decisions regarding antibiotic use in intensive care was explored in relation to a holistic information visualization program. Preventing the misuse of antibiotics and prevention of addiction is the biggest factor in preventing antibiotic resistance, which today is a big problem worldwide (Forsman, et al, 2013). A challenge in the way of making more adequate decisions regarding antibiotic distribution in intensive care is the big amount of data stored and visualized across many various platforms. This creates an essive amount of work for practitioners within intensive care to scan these platforms and themselves establish a relationship from all of the data. The results showed that the tool would be useful and appreciated in helping with the presented challenges. In addition to this research, information visualization has been used as a tool in studies about stroke patient rehabilitation. (Xu et al, 2006), designing interfaces for medical records (Plaisant et al, 1998) and many other purposes. Information visualization can be considered a useful tool within many fields of medicine. It can help with getting a better overview of patient information, both in regards to making health-related decisions, distributing medicine or analyzing big amounts of medical data. Healthcare has not always consisted of digital solutions or technology, assisting practitioners and patients through their hospital stay and in recent years many new concepts have started to emerge due to this increase in digitalization within healthcare. In the following section, a broader view on the topics of healthcare and technology will be presented in order to show where the subject of information visualization has emerged within this area.

2.3 Healthcare and technology

There are many different categories within the scope of technology and healthcare. E-health, M-health, information visualization, medical visualization and digital health are just some examples that are being used today. In parallel to regular physical clinics where patients walk in, now digital clinics with doctors and nurses calling over digital communications have emerged. These digital clinics are in many countries popular and have had both positive and negative reactions to them (Persson, Wiezell & Claeson, 2016). Tools that was before more analogue like measurement tools have also today many of them been converted into more digitally used ones. Some of the tools used today are a mix of physical measuring devices and digital interfaces translating the data resulting in more visually understandable formats (Salber

& Niksch, 2015). Within this development of technology in recent years, a concept known as E- health has emerged. It has every year grown bigger and services like health advice via the internet, digital patient records and mobile-health or m-health are only a small collection of the services that that can be found (Löhr, Sadeghi & Winandy, 2010). Sometimes the new and supposedly improved digital solutions do not live up to the hype that digital solutions can give and instead give the users more work and a heavier cognitive load when using and understanding the new systems and tools (Rydenfält & Persson, 2020). In spite of digital technologies sometimes being hard to comprehend they can also be the base of many new and exciting opportunities for both practitioner and patient. The development of technology and the increased digitalization within healthcare means that the amount of data that practitioners such as doctors, surgeons and nurses have to process has increased considerably. With the increased amount of information and data, there is a risk of overflowing the user, i.e rendering to much information for people to be able to understand and process (Rydenfält & Persson, 2020). In this case, the aim of information visualization is to help with the sometimes-large amount of data and to make it easier for users to make adequate estimations of medical examinations (Chittaro, 2001).

In order to conduct a study regarding the potential of information visualization when presenting results from medical examinations and defining a design space in relation to a certain health-condition interviews were conducted with a specific group of individuals. In the section below the method, procedure and ethical reflections will be presented.

3. Method

In this section, the applied methodological approach will be discussed in further detail. Firstly, parameters and arguments surrounding why a qualitative approach using semi-structured interviews has been used as opposed to a quantitative approach are presented. In relation to this, how the interview questions were created is also presented. Secondly, the procedure and selection of participants are presented. Arguments for the selection of participants are made and the ethical considerations of the study are presented. Lastly, the approach of analyzing the data is presented as a part of the method.

3.1 Applied method

A qualitative research study can, in very few words be described as striving towards a qualitative result that is not quantitative. This means that the result and focus of qualitative research shall not consist of numbers, amounts or focusing on the frequency of a certain aspect but the focus should instead lay in what the meaning and value are (Widerberg, 2002; Preece, Rogers & Sharp, 2002). In qualitative research, there are certain key-ways in which to explore and study phenomena. A big part is an interview. Sometimes depending on how much time an interviewer has; one type of structured interview can be chosen. The completely structured and semi-structured interview will help keep the interview within certain frames, the

semi-structured interview being more explorative using probing. Probing means asking the participant to further develop an answer or opinion (Preece, Rogers & Sharp, 2002). The varying level of structure does not exclude the fact that every interview needs an initial structure, focus or theme. Even if the interview is massively un-structured and seeks to only ask the participant to further develop, a direction simply means driving an interview forward towards the research subject (Benyon, 2014).

The interview is all about exploring subjects and phenomena together with interviewees and participants. Although it is good to have questions as a base to work from when conducting interviews, doing them in a qualitative way means exploring beyond them at any given time. It is important to take advantage of the particular individual being interviewed every time and see the difference in every person. (Widerberg, 2002). The semi-structured interviews were conducted in order to provide structure and focus into the research fields. The importance of being able to further explore answers was considered to be massive since this would mean that a greater picture could emerge surrounding the medical condition which is researched in relation to information visualization.

The interview questions were divided into three parts, they can be read in their entirety in appendix: A. An initial part focused on simply introducing the participant and the interviewer to each other. The second part explored various questions related to the examinations of individuals with the condition of hydrocephalus. These second part questions focused on the different types of examinations done, challenges and other aspects of these examinations. After this part which took about half the time, questions about the presentation of the results of the examinations were asked. These later questions were thought to have a bigger connection to the main research area of information visualization. They revolved around how the results were presented, to whom, and what type of challenges they could see. Every interview contained the same initial questions, but in accordance with semi-structured interviews, many questions were asked on the spot in regards to every participant's unique experience.

The next step in the process was to transcribe the interviews. Transcribing the data non- verbatim means converting the recorded interviews into textual documents. Non-verbatim means leaving out, for example, non-speech sounds, filler words, and corrections. The transcriptions were written in a non-verbatim way in order to clean up the text and make it easier to analyze at a later stage, leaving out excessive material.

3.2 Procedure

In this section, the procedure of gathering and contacting the participants will be presented.

Arguments as to why the selection of participants was made can be read about in the following section of selection and participants. Firstly, the participants were contacted through e-mail to make initial contact. The participants contacted were both physiotherapists, and occupational therapists working at uvarious hospitals. The study was presented, and a consent document was sent in order to give the participants a fair amount of information and to be as open as possible towards what the study was about. The consent document can be read about in its entirety in Appendix: B. If the participants agreed to an interview they were booked in for an online interview. The interview took place in zoom. Zoom was chosen because of the possibility to record the entire session. Also, another software was used to record the interviews in order to with as much means as possible secure the outcome of the interviews. 5 interviews were done, and they took approximately 30-40 minutes. Four out of the five interviews were conducted online but one participant preferred to have the interview in person and therefore it was not held online. Initially, 7 people were contacted about potentially being interviewed.

One person did not respond, and one person cancelled at the last minute due to a hectic work situation. Thus, five individuals were interviewed and arguments for why this selection was thought to be sufficient can be read about in the following section.

3.3 Selection and participants

Conducting a study with an adequate selection of participants can be complex. First off, just gathering a sufficient amount of them, can be hard just on its own. Specifically, if the topic being researched is very specific to a group of individuals that are hard to get into contact with.

Having a varied selection is important and ethnicity, work position, sexual orientation or for example religion are all factors that can be considered in relation to selection (Ryen & Torhell, 2004). The case given in this thesis was surrounding a specific patient group. This patient group is a relatively small one and the individuals working with the group and making the examinations is also them, limited, therefore a non-random sampling was used. This means that anyone who agreed to do the study was accepted (Preece, Rogers & Sharp, 2002).

Nonrandom sampling was used in combination with the fact that the asked individuals also had to fulfil very specific criteria. The participants for the interviews had to be working with individuals with the condition of hydrocephalus since this was the case given. This could include physiotherapist, doctors, nurses, and occupational therapists. The case given was to interview the practitioners working with the examinations. Despite this, it could have been for this thesis and the research question beneficial to include the patients to gain even larger insight into the subjects. Even if interviewing the patients would have made the study richer, there are many factors into why this simply was not feasible. Firstly, because the main group of individuals that suffers from this condition are elderly, and many of them suffer from cognitive decline. This presents a massive challenge to conduct interviews with them. Secondly, these individuals go through many tests such as physical and cognitive while being examined and may at some point be very tired from these examinations and procedures making it hard also to conduct interviews. The selection for this thesis was due to the above-stated reasons

limited and instead of a variation of practitioner and patient, the age and experience varied within the selected participants. The individuals who were contacted had to fall under the earlier explained group of practitioners working with hydrocephalus. If they filled these criteria, they were deemed fitting to the study.

Profession Sex Experience

Physiotherapist Male 3 years

Physiotherapist Female 20+ years

Physiotherapist Female 20+ years

Physiotherapist Male 2 years

Occupational therapist Female 6 years

Figure 2. selection participants

3.4 Ethical considerations

It is of great importance that all research, whether it is qualitative or quantitative, stand on a base of integrity and ethical considerations. Specifically, studies related to healthcare and patients can include very sensitive data. This means that the ethics of any such study should be thorough and respectful. This study has used two ways to ensure that the outcome of every interview and any opinion expressed by a participant is not in danger of exposing them or any patient they may be working with. First of the codes in The European code of conduct for research integrity was used (Drenth, 2011). These codes are guidelines on which the researcher should relate to in order to provide the study with a high level of integrity. These codes consist of four guidelines. Reliability means ensuring the quality of the research in relation to design, method, analysis and use of resources. Honesty means having a true and open communication towards both participants and presenting the results with honesty. Towards participants, it should also be completely clear what the purpose is and how their responses will be protected and presented. Respect may speak clearly for itself. Having respect towards all the parties within the study such as participants, research, colleagues, sometimes nature and other relationships between people is of big importance. The last code is accountability.

Accountability means knowing and having integrity towards for example the research question, publication of research and within the organization in which the study is done. Being accountable for what has been done and what is going to be done. These four codes were constantly reviewed during the thesis. The second ethical part of the study included having the participants sign a consent document for being a part of the study. This approval stated what the purpose of the study was, what was to be researched, how the interview was going to be conducted, and how the result was going to be handled. The participants were informed that their answers would be presented completely anonymous and that they could at any time abort their participation. The approval can in its entirety be found in appendix: B.

3.5 Data analysis approach

In the first section a brief introduction to thematic analysis is presented and it will be followed

Doing qualitative research and using interviews, can sometimes render a big amount of data.

Depending on how many interviews are conducted and how long they are, the amount of data can vary. In order to further analyze the data, a method has to be applied. One way of analyzing the data within the frame of qualitative research is by using a thematic analysis approach (Braun & Clarke, 2006). By using thematic analysis, the data will potentially be rich and detailed, and it is a flexible way of analyzing qualitative data. There is research that states that thematic analysis is not on its own a sufficient method of analysis (Ryan & Bernard, 2000;

Boyatzis, 1998). This means that thematic analysis should be treated as a tool or an aspect within other theories. The reason behind this is that thematic analysis is supposedly too vague to be treated as its own theory (Braun & Clarke, 2006). In this thesis, thematic analysis was treated as its own analysis. This on the ground of earlier research meaning that thematic analysis is free of a specific theory and epistemology, making it possible to apply it to most data (Braun & Clarke, 2006). Using thematic coding, or thematic analysis means using a method that is flexible and aims to identify, analyze and illustrate patterns. The thematic analysis often happens inductively, meaning that no pre-existing themes are made without them being found in the analysis of the data (Preece, Rogers & Sharp, 2002).

The aspect of choosing what is valuable enough to save into themes when analyzing data can be a complicated question. Saving too much means a possible information overload but saving too little can mean losing the richness of the material. The approach of saving data and creating themes that is somehow related to the overall research question is a relatively easy and concrete definition to relate to (Braun & Clarke, 2006). It is also possible to create a theme looking at the prevalence of a certain aspect. If something is said many times and is mentioned by most participants, this could be a potential theme because it seems to be of a common opinion.

This thesis has used themes, sub-themes and codes to structure the different levels of the results. Codes are used as the initial step towards categorizing a material. They can emerge due to patterns, an interest by the analyzer or many other various reasons. They are the most basic aspects of the material and is the first step when using a thematic analysis approach. Directly merging codes into themes is one way of approaching the analysis. In some instances, subthemes are used. Sub-themes are a way of creating a structure to a large and complex theme.

By applying a sub-theme, it is possible to give a more understandable picture into what the main theme includes. Lastly when codes and sub-themes are applied the final themes of the material can be set. The main themes offer a broader perspective of the material. (Braun &

Clarke, 2006).

The analysis stage started with the process of reading through the data transcriptions in order to create an initial outlook of the material. The process of getting familiar with the material was thought to be important in order to refresh the sense of every interview and with this making it easier to code. When reading, notes were taken in order to put a focus on potentially interesting parts. The material was initially put into a coding program called Atlas ti. This program was used to help get a better overview of the coding of different themes. After getting to know the material then came the process of generating initial codes. When reading the material initial thoughts of potential codes was noted, as a sort of “code-draft. After going through all of the transcripts, 20 codes had been identified, after reviewing them, some were removed and other added. In the end, 17 codes had been made. The codes were after this stage

put into clusters, these clusters consisted of the codes that had been deemed alike or somehow related to each other.

Later came searching for themes, what codes could be put into what themes and if subthemes needed to be used. This fourth stage involved getting an overview of the created themes, sub-themes and within the themes, codes. In order to get a bigger picture of the data, a thematic map was made. The fifth and sixth stage involved establishing the themes, subthemes, and finally naming them in an adequate way. In this stage, it should be clear what codes are related to what theme and what the themes involved. Four final themes were created and within them are varying subthemes and codes. The themes are called “Examinations and effects of them”, “Presentation of results”, “Value of presenting” and “Potential of visualizing”.

Figure 3. Network trees over the four final themes

These six parts of, initial reading, initial codes, coding, merging, thematizing and naming themes were used as guidelines for analyzing the results from the study in a thematic way (Braun and Clarke, 2009)

4. Results

In the following section, the results from the study will be presented relating to the respective theme created from the data analysis. It can be important to know for the sake of the overall understanding of the result that patients who are to be examined for the condition of hydrocephalus first come in for an initial test, after these tests they rest, and on the second day they do another set of tests. Some hospitals do a minor procedure where they drain fluid from the head of the patient, while some hospitals don’t do this. If the patients after the second tests, and/or the drain-procedure are considered to fulfil a certain number of criteria, they undergo an operation. After this operation, they also have a checkup. The participants in the study sometimes talk about the results presented after the first checkup or the post-operation checkup. The quotes from the participants will be written using cursive letters and anything within parentheses is added to make the quote more contextual. Extracts from the interviews will be presented in English.

4.1 Examinations and effects of them

When a patient is to have an examination for hydrocephalus they have the first encounter with a physiotherapist or occupational therapist and undergo a various set of tests. These tests can include balance, walking, cognition, standing, sitting, amongst others and often, they film a part of the examinations. It initially became clear that depending on what hospital a practitioner is working in they have different ways of measuring and examining the different aspects of hydrocephalus. Some felt that the different ways of examining had less or more amount of value and potential. This did not seem to be a problem itself, but some participants mentioned that they would have liked more standardized methods and specifically more objective methods since this would have helped them amongst each other both when conducting the examinations and when presenting the results. Participant 1 felt this way and expressed a need for wanting to have more objective ways of measuring in the future.

” I would like to have it (the ways of measuring) a bit more objective. A lot of things feel a bit biased, and that in itself is not a problem because right now I’m the only one doing this, but my hope is that in the future there is going to be more people than me doing this (the examinations)”

The many different ways to examine the patients was by most of the participants discussed as positive. The best part for some, was the mixture between being able to present hard data (numbers) and also being able to, for example, show the video footage or tell their colleagues about their own perceptions about the patients. This was for them, the most effective way of presenting the results within their medical teams. Participant 3 mentioned that for her, numbers and objective data are not always what matter the most and that the objective and subjective ways of estimating complete each other. She also discussed that a good future scenario would be if the hard data and numbers could be actualized in a better way making it easier to do the examinations and knowing that the results will be shown in a true and correct way.

“I think that they (numbers and subjective estimations) complete each other, that’s what I mean… I think that we really need to complete the tests with more subjectively experienced movement patterns”

When examining the patients some participants discussed ways of measuring and examining patients as intrusive and challenging, while other ways were seen as simple and positive.

Participant 4 talked about the positive aspects and felt that for him it was good that they had so many varying tests since this meant getting a good range of test results.

“I think that it is good that we test (the patients) for different things which means that everyone gets a perception about the various situations. The purpose is in the end to make the life better for the patients in their everyday life and if we test the balance, we get those aspects, and the different walk-tests we get those aspects. I think it is good to have a bunch of various tests”

These examinations tests are of varying difficulty for the patients. At times, some patients cannot do all the tests because their health status is too bad. Participant 3 discussed the fact that she felt that some ways of examining and measuring were too insensitive and blunt.

“I think for example that the balance test is very blunt. To have a scale for checking if you can handle standing on one leg is, I mean, it’s a rather insensitive instrument”

This theme of “Examinations and effects” of them included the different opinions the participants had towards the different ways of measuring aspects and doing the examinations.

After doing the examinations, then comes the stage of presenting the results either to a team of medically educated personnel or to the patients themselves.

4.2 Presentation of results

Within this theme of “Presentation of results”, it became clear that the way the results were presented depended on to whom and for what purpose they were to be presented. This was since the results were presented to two different groups of people. Firstly, it was presented to the individuals working with the other parts of the condition, this could for example be nurses taking care of administrative aspects, neurosurgeons who were the ones doing the operations, and other various medical staff. Within these team rounds a decision was to be made surrounding the health status of the patient. The participants discussed how they preferred to present the results, the structure and the reasons why. All participants had a protocol to follow but when presenting some felt the pressure of time, dealing with the technicalities. Others liked having visual aids to refer to in order to make things clearer. Participants 1, 2, 3 and 4 talked about this and particularly participant 2 liked having a specifically designed figure to follow

“Yes we have the team rounds every week and then I tell my part, and then we have *First name* *second name*, he has created a scale that we put up, he has made a figure, and then we put the tests I have done in there, and then it’s like a little green rhombus”

Presenting to other members of the team was done in various ways and using various technical artefacts for support. Sometimes when presenting the result to the other staff members, time was mentioned as a factor when it came to navigating the technology in a smooth and easy way.

Participants 1 and 2 mentioned the aspect of time and technology navigating being part of the presentations. Participant 1 meant that it could be time consuming depending on who was navigating since different people have different experience with technology.

“Well there is a lot of clicking in the different parts for whomever is using the computer. First you have to open the x-ray and they are to be counted on, what different measurements you are to look at, and then you have to click that. Yes, it would have been good if they could have been collected (the interfaces)…

Depending on who is by the computer it can take different amounts of time depending on computational habit, but it is definitely time demanding”

After the team round and discussing the examinations with the medical team, the practitioner meets with the patient again. Then another set of tests are done. After these tests and/or draining fluid from the head of the patient, the results are to be presented to them. How this was done also varied to a high degree. This could be because of the cognitive state of each patient or because some participants felt that different aspects mattered more to the patients.

All participants brought with them some type of protocol to use as a base when presenting to the patients. Participant 4 preferred to give the patients a broad and general result.

“I bring the paper from the journal and then I don’t tell them anything about it before, and then we do the tests and then I write it (the results) very broadly and then we look at it together afterwards, but they don’t get any document with them home, it’s more like we look at the results directly”

Generally, the way of presenting the results to the patients was done in an easy way. None of the participants felt that giving the patients the whole protocol or journal would make any sense, nor make any sense to the patients. In the following theme “Value of presenting” the participants gave their opinions and thoughts as to why presenting to the patients mattered.

4.3 Value of presenting

Even though the presentations varied depending on who was doing them and to whom they were presented, it was a common agreement that there was a great value in presenting the results to the patients. The value of presenting to the practitioners stands on its own since they have education, knowledge and experience in both viewing and reviewing results from this type

of examinations. For the practitioners a part of their job is reviewing patients results and make decisions accordingly. Because of this, this theme does not include the value that presentations have to the practitioner’s presentation amongst themselves.

It was thought to have a great value to show the patients their progress and especially the patients who did not see a change in their health. This meant that sometimes the patient does not believe that they have gotten any better from the procedures or surgeries, and in these instances, participants described it to be of great value to be able to present the patients with their results in a graspable way. Participant 4 and 5 preferred to give the patient a small summary and their own opinion since they had experienced that this was what the patients wanted. Participant 4 would approach the concerned patients calmly and try to connect their results to their everyday life and felt that this was of value.

“Then I Usually say that I can understand that you cannot feel that big of a difference but I can see, partially how you walk but also on the tests that you have gotten better….I normally connect it (the results) to the fact that this means things will work better, and specifically better in the everyday life and that’s where it is important for them to improve”

What was perceived as being valuable when presenting varied from one participant to another.

As mentioned above participant 4 preferred to connect the result to the patient’s everyday life.

Participant 1 felt that showing the video footage of some parts of the examinations was the best and most efficient way. This since he felt that there was no better visual way of showing the patients their progress and improvement.

“Sometimes they cannot see it (the difference) and then they are not happy because they still think they are experiencing big problems, but they at least get a little bit happier when they actually see (on the movie) that they have gotten better”

When presenting results there was a common agreement that presenting the patients with hard data would not really make sense in the same way as it would for the practitioners with a medical education. The value for the patients would not be the same as the value for the practitioners. The reason for this was that the patients would not be able to understand what the numbers meant even if they had changed. Participant 5 said that she understood that for the patients it can be challenging to understand the hard data and would prefer to have more figures to use.

“How many seconds counts as an improvement really? So it would have been interesting to have perhaps more staples, or if we could get like a point-system where it would have shown how many seconds really is a good improvement.”

part for many physiotherapists and some of them also said that they would have liked to use the video footage when presenting to the patients. Participants 3 and 5 discussed the visual value of showing the video footage and participant 3 thought that showing the video footage would be a good support.

“That’s when you would like to show the movie, because it would make things way clearer and concrete because the results are so abstract anyways. So that’s seven seconds faster than that, it’s like whatever? Still you don’t feel better, that’s the thing about the disease that you don’t get tip top you only get better”

In summary the general value that the participants felt that presenting results in a graspable way had, was high, both in regards to the practitioners and the patients. Furthermore, they had many thoughts on what they felt was the best way, such as using figures, showing the video footage etc. In the following theme the discussions surrounding what the aspects that should be most focused on when presenting is presented.

4.4 Potential of visualizing

This theme of “Potential of visualizing” includes the various aspects that the practitioners deemed to be good when wanting to present the results towards the patients in such way that they can truly understand the meaning of their own result. What aspects that were most important when presenting the results were discussed by some of the participants, saying that the varying cognitive state of the patients was a factor to take into consideration. Participants 3 and 5 discussed the factor of being concrete and clear as important. Participant 5 wanted the presented results to be clear and understandable in a way so that the patients could understand and grasp their progress.

“Well I guess it would be that it (the result) should be clear and it should be understandable, what I have tested and that you then get a summary of what they have done and the improvement I have seen, because sometimes it’s hard to know what counts as an improvement”

When presenting the results participant 1 felt that showing the movies to the patients was the most effective way of presenting data and could not see it getting clearer than that. Previously discussing the value of the video footage, he also felt that the qualities that a movie and video footage brings are the best possible for the situation. Translating numbers into more comprehensible figures does not seem to be as effective as showing the movie according to participant 1.

“I have a hard time understanding how things can be more visual then the movies, because translating numbers into images spontaneously means diagrams and curves….I have a hard time seeing that the results from the results could be made clearer and that it would be more clear than the movies”

When a patient is to reflect and understand what their result really means it can be hard to connect it back to their everyday lives. This is both because of their potential cognitive situation but also the fact that it can be hard to understand the true meaning of numbers. Participant 3 discussed the fact that she felt that visualizing hard data, such as numbers and seconds was good, but the most important aspects was that the patient understood and grasped what the improvements meant and how it affected their everyday life.

“The question is, if it (the result) could be done more clearly and visually and I think so too but I also think we need to think about what these tests say for the patient's real activity in life as well, because sometimes you become a little blind……we say that he walks better in 10 meters but that’s not the best result for the everyday life, which is what I think is most important”

In this theme, results surrounding the potential of visualizing have been presented. Many participants discussed factors such as understandable, clear and concrete when it came to presenting the results visually to patients. What seemed to be most important was making sure that the patients knew what their results really meant both in relation to the data and to what effect this can have on their everyday lives. Presenting results in a visual way can help with this according to the participants.

5. Discussion

The result will in this section be discussed and summarized relating to existing research and four defined design spaces will be presented.

Many studies found within the broad range of Information visualization have focused on, for example, healthcare, HCI (Gershon, Eick & Card, 1998; Craig, 2015; Keim, 2002) and there are many other various fields where it has also been used (Marci et al, 2016; Ingram & Benford, 1995). It is possible to state that information visualization is generally well used. Within HCI information visualization was proven to have good potential when making generally difficult decisions (Daradkeh, Churcher & McKinnon, 2013) and it also showed to have potential on mobile platforms when using animations and interactions (Craig,2015). Within healthcare, information visualization was proven to be of help in both making difficult decisions regarding antibiotics (Forsman et al, 2013) and when trying to make sense of the massive amount of data produced in patient journals (Faisal, Blandford & Potts, 2013). This selection of previous research shows the great potential that information visualization has in many ways in aiding users in understanding complex data. The research also to some extents shows application themes and design challenges for future designs. Furthermore, the research also presents how information visualization can be applied to various artefacts and various target groups. What this previous research fails to define, is a design space of presenting potentially complex medical results that should appeal to several user groups with massively varying education and

and healthcare by putting a focus on design and information visualization. By defining a design space for future designs aimed at conveying understandable information to every group who is to grasp the information of medical results it is possible to provide a better experience for every party involved. To make such a contribution was the aim of this study.

Four design aspects were defined based on the results from the study conducted. These design aspects make up for a base in which can be reflected upon when designing an information visualization tool aimed at helping to present the results from medical examinations. In this case the specific parameters were those of the condition of hydrocephalus, but it is possible that these design aspects can be applied to other various medical situations where the target groups differ largely. The four design aspects will be presented below starting with the first, “Contextual”.

Figure 4: Design aspect: Contextual

The subject of having and creating context was discussed as being an important part when both examining and presenting. The importance of this can also be found in research. The results of this study showed that the practitioners had very different ways of doing the examinations.

Some participants perceived their examination methods as satisfactory and others said that they lacked sufficient ways of measuring objectively. Corresponding to previous research there is a lack of objective ways to measure certain aspects of the condition hydrocephalus (Bäcklund et al, 2017; Jain, Kuskowski & Selcuk, 2012). Some participants also thought that the most preferable way of examining and as an extension of this presenting the result was to have objective and subjective ways together because this is what gave the most contextual version of the results. Providing users with context has been explored before within the field of information visualization. Cox et al (2000) make an argument for why providing context can be of big importance. They meant that their information visualization tool can provide context to the users and from this provide better user experience. Another aspect of context explored in research is to provide practitioners with enough contextual data to create a smooth and good user experience. Xu et al (2006) explored an information visualization tool for biofeedback from stroke patient rehabilitation. One of their set design goals was to create a contextual interface which meant presenting hard data at the same time as media feedback and meant that this was critical for providing good user experience. In summary, context relating to information visualization has been proven to be an important aspect when creating a good user experience. The result and analysis of the data from this thesis have also explored the contextual aspect. The second design aspect that emerged from the study can be summarized as “visually structured”, referring to how information is best presented.

VISUALLY STRUCTURED

Figure 5: Design aspect: Visually structured

When it came to presenting results to medical staff there was big variation in how this was done. All used protocols, but none did it in the same way. Some looked at the video footage of the examination first, as some felt that the technical aspect of presenting was hard since there was a lot of unnecessary clicking. According to existing research, there is a challenge in providing practitioners with too much of information for them to be able to process smoothly and concretely (Rydenfält & Persson, 2020) and thus there is a need for tools to make it easier for them to process information (Chittaro, 2001). Previous research has also identified the need for having good visuals and easily navigable interfaces as a crucial element in providing practitioners with the best possible solution when trying to make assessments from medical information. Too much of a cognitive overload which can be caused by receiving too much information can be prevented with an interface that is well structured and visually graspable (Xu et al, 2006). Plaisant and Mushlin (1998) comes to a similar conclusion when researching computerized medical records. Positive aspects when using graphics and structure can also be seen in the result of this thesis. Although the results from the examinations are processed in a way so that many various individuals with various medical education is to understand the information, which they eventually do, there is a need for the information to be more consistent and time-efficient. This can be summed up in a need for standardizing how and what to look at and an additional need for visual support. These needs are summed up in the second design aspect of “Visually structured” since having visual support and having a clear structure can help with both the time-aspect and being clear when presenting. The third design aspects relate to how the tool should be “adjusted for target groups”.

ADJUSTED FOR TARGET GROUP

Figure 6: Design aspect: Adjusted for target group

The other group of individuals who are to understand the information from the medical results are the patients. The patients, in this case, can have a varied cognitive state and this is taken into consideration when presenting to them. Within the aspect of designing towards targets groups, research shows how basing visualization on the characteristics and cognition of a user can be beneficial (Toker et al, 2013). Another aspect explored by Conti & Maclaren (2008) is understanding that a user’s cognitive ability can work as an indicator of how a system should be designed. They suggest having a system that can recognize the cognitive state of the user and adjust its visualization accordingly. Supporting earlier research, the results of this study also shows how important it is to adjust the results to each patient. Sometimes the patient does not,or is not capable of understanding the results and in some cases a caregiver such as a spouse or a child is there to help. This means that how the results are presented needs to be adjusted to the individual who is to understand it and their level of knowledge and cognition.

The fourth and final design aspect that emerged from the data analysis is called “highly visually