<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Dr Dave Pao on Medium]]></title>
        <description><![CDATA[Stories by Dr Dave Pao on Medium]]></description>
        <link>https://medium.com/@davepao?source=rss-5a8314fb7309------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*1AMvtHqLAVXae4JB8sNI3A.jpeg</url>
            <title>Stories by Dr Dave Pao on Medium</title>
            <link>https://medium.com/@davepao?source=rss-5a8314fb7309------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 13:17:43 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@davepao/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[What if medical records were designed like a game of chess?]]></title>
            <link>https://uxdesign.cc/what-if-medical-records-were-designed-like-a-game-of-chess-c3b6782c3df2?source=rss-5a8314fb7309------2</link>
            <guid isPermaLink="false">https://medium.com/p/c3b6782c3df2</guid>
            <category><![CDATA[healthcare]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[ehr]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Dr Dave Pao]]></dc:creator>
            <pubDate>Mon, 01 Feb 2021 11:46:44 GMT</pubDate>
            <atom:updated>2021-02-14T16:52:09.260Z</atom:updated>
            <content:encoded><![CDATA[<h4><em>Demystifying the ‘wicked’ problem of clinical usability through the ‘tame’ game of chess.</em></h4><figure><img alt="A cluster of medicinal glass vials against a gradated grey backdrop, mostly stood upright,  like abstract chess pieces" src="https://cdn-images-1.medium.com/max/1024/1*m6TA2dXtDj8JVn2g4YcieQ.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@nci?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">National Cancer Institute</a> on <a href="https://unsplash.com/s/photos/medical--record?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>The <a href="https://www.nature.com/articles/d41586-019-02876-y">Electronic Health Record (EHR)</a> is arguably the most important tool in modern medicine, and it is <a href="https://med.stanford.edu/content/dam/sm/ehr/documents/EHR-Poll-Presentation.pdf">not playing its best game</a>. It is literally the <a href="https://bmcmedinformdecismak.biomedcentral.com/track/pdf/10.1186/1472-6947-10-67.pdf">3rd actor</a> in the consulting room, and it is <a href="https://www.newyorker.com/magazine/2018/11/12/why-doctors-hate-their-computers">fluffing its lines</a>.</p><p>Whatever analogy we choose to describe the current state of EHR usability, it is hard to think of a worse design dystopia where fragmentation has trumped collaboration. Clinician performance (and consequently patient care) falls victim to technological determinism, by which:</p><blockquote>‘The machine throws out anthropomorphic [clinical] habits of thought. It compels the adaptation of the workman to his work, rather than the adaptation of the work to the workman. The machine technology rests on a knowledge of impersonal, material cause and effect, not on the dexterity, diligence, or personal force of the workman…’ (Veblen, 1904).</blockquote><figure><img alt="The current form that most EHR interfaces take, with dropdowns, tabs and text. And no graphics." src="https://cdn-images-1.medium.com/max/1024/1*8GKfrxlnQGRpKFGYvyrCew.png" /><figcaption>Edmonds, K., et al., De-identified screenshot of Electronic Medical Record (EMR) interface, Journal of Palliative Medicine (2015)</figcaption></figure><p>More than a century later, <a href="https://www.newyorker.com/magazine/2018/11/12/why-doctors-hate-their-computers">Atul Gawande</a> tells us the medical profession ‘<em>actively, viscerally, volubly hate their computers’.</em></p><p>And Dr Elwyn Elias, in the <a href="https://www.bmj.com/content/348/bmj.g2078"><em>British Medical Journal</em></a><em>,</em> paints a picture of a grinding reality where ‘<em>We have medical records, but they are like the shadows on the wall of a cave, punctuated by codes and jargon</em>’.</p><p>I estimate that I have ‘done’ around 30,000 consultations in my career. With some patients, I could have used a tablet of stone and managed them perfectly well. For most, I need an EHR; the more usable the better.</p><p>Why has the Digital Age failed clinicians? Indeed, why hasn’t it transformed clinician performance as it has for users in most other activities of modern life? To understand why we have to look at how clinicians think and why clinical data is not just any data. Only then can one look at the landscape of EHR design — in this case as a chess game — and work out how to make EHR better.</p><p>In her book <a href="https://global.oup.com/academic/product/how-doctors-think-9780195187120?cc=gb&amp;lang=en&amp;"><em>How Doctors Think</em></a><em>,</em> anthropologist Kathryn Montgomery reminds us that, contrary to popular belief, clinicians are not scientists. Rather, they are <em>practitioners</em> who interpret science by juxtaposing it against clinical experience and findings elicited from the patient consultation.</p><p>The clinician’s special skillset, therefore, is <em>reasoning </em>with data, from a myriad of sources. They need the cleanest, richest data flowing in real-time because they are in liminal conversation with the patient.</p><p><a href="https://abetternhs.net/2015/08/23/patients-as-people-and-liminal-medicine/">Liminal</a> can be defined as an unstable, impermanent space of transition. The patient, balanced either side of the equilibrium between sick and healthy, can be tipped over in an instant by one or any number of symptoms, findings, lab results or radiological findings.</p><p>At its most basic, the EHR is a tool that supports the clinician in collecting, communicating and <em>curating</em> patient data, and it in this third role the true power of digital lies. The key to supporting the ‘clinician as curator’ is recognising that we are not dealing with any old data but ‘data collected or verified at the point of clinical care’, or DCVPCC.</p><p>DCVPCC includes data collected during the consultation (<em>e.g. </em>the patient’s past medical history, symptoms, and clinical findings such as blood pressure and weight) and external data (<em>e.g.</em> laboratory results, scan results and other investigations) that is verified with the patient during the consultation.</p><p>The richness, precision, and context of DCVPCC directly impact the richness, precision, and context of clinical care. The suboptimal flow of DCVPCC into and out of the clinical conversation — a consequence of poor clinical usability — is how the EHR most damages clinician performance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SdtB-jUsrSZjiqnNDghegA.jpeg" /><figcaption>Pao D, Poor EHR usability interrupts the flow of the clinical consultation (2019)</figcaption></figure><figure><img alt="A coloured wheel with segments showing the different areas impacted by DCVPCC" src="https://cdn-images-1.medium.com/max/1024/1*I2YTng6p8ZDtb3f4By2zxQ.png" /><figcaption>‘Data collected or verified at the point of clinical care’ (DCVPCC)</figcaption></figure><p>As the scope of the <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5171496/">EHR’s role</a> expands, the consequences of poor quality DCVPCC ripple through the entire health ecosystem. Because digital data can be shared infinitely and without degradation, inaccuracies get amplified.</p><p>This affects individual patient care, of course, and service planning, <a href="https://www.sciencedirect.com/science/article/pii/S1532046407000603">population health</a>, and health policy, not to mention our fundamental understanding of disease and disease processes.</p><p>To add to the complexity, DCVPCC do not exist solely in the timeframe of a single consultation. They build a history, cumulatively influencing contemporaneous decision-making.</p><p>Take the current global Covid-19 pandemic. How can we reliably know, for example, if anosmia (loss of sense of smell) remains a key symptom of infection with new variant Covid-19 and not the coincidence of seasonal rhinitis (hayfever) in an individual who happens to be asymptomatically infected with the new variant virus?</p><p>The only way to know is from high-quality DCVPCC from as many patients’ consultations as possible. Any corrupted (missed, false or inaccurate) diagnoses of seasonal rhinitis would confuse any attempt to understand the diagnostic relevance of anosmia.</p><p>With the unique characteristics of the user and data explained, can we use the analogy of chess to demystify the EHR design landscape?</p><p><strong>The Game of Chess</strong></p><figure><img alt="An oblique view of a crystal glass chess set with pieces set up in starting formation" src="https://cdn-images-1.medium.com/max/1024/1*g7R3U2cE-SP_Aoq_ajenYg.jpeg" /><figcaption>bnPhoto by <a href="https://unsplash.com/@polarmermaid?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Anne Nygård</a> on <a href="https://unsplash.com/s/photos/pieces-on-a-chess-board?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>Chess is played on a square <a href="https://www.britannica.com/topic/chessboard">board</a> of 8 vertical rows called files, and 8 horizontal rows called ranks. These 64 squares alternate between two colours: one light and the other dark.</p><p>There are 6 different types of pieces with unique, sometimes iconic, ways of moving on the board, with some common moves between different pieces. They vary in number: 1 king, 1 queen, 2 bishops, 2 knights (or horses), 2 rooks (or castles) and 8 pawns.</p><p>By analogy, the individual disciplines or ‘pieces’ in EHR design have their own unique and proven ways of working, yet they often share methodologies.</p><p>How these disciplines are defined, how they operate and what they contribute to the EHR design literature sheds light on how they might collectively engage to bring medicine the best of the digital era.</p><p><strong>The Patient (The King)</strong></p><figure><img alt="Close up of white queen toppling a black king on a Hand and fingers and chess board with vintage look" src="https://cdn-images-1.medium.com/max/700/1*pjNya4avM5dG4Rd-C8dNsA.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@grstocks?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">GR Stocks</a> on <a href="https://unsplash.com/s/photos/chess?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p><strong><em>Definition: </em></strong>A patient is any recipient of health care services that are performed by healthcare professionals</p><p><strong><em>On the board:</em> </strong>The<strong> </strong>King is the most important but least powerful piece in chess. It is uniquely vulnerable with severely restricted freedom and is unable to challenge other pieces. If not protected, the game is lost. The King is cared for at all costs and the <em>raison d’être</em> of the game.</p><p><strong><em>In the literature:</em></strong> Patients historically have not been included in the EHR design process. This is beginning to change, for example at <a href="https://bmcmedinformdecismak.biomedcentral.com/articles/10.1186/s12911-019-0993-7">Imperial College London</a>.</p><p>It comes as no surprise that patients play a much stronger role in designing the <a href="https://en.wikipedia.org/wiki/Personal_health_record">Personal Health Record (PHR)</a>.</p><figure><img alt="A matrix and telephone-dial style visualisation dashboard of a PHR" src="https://cdn-images-1.medium.com/max/1024/1*_u9RNKZl5BnJou0wX7Z40g.png" /><figcaption>d+collab submission: HealthEd Innovation (2013)</figcaption></figure><p>With the inevitable merging of EHR and PHR, this can only be a good thing — as evidenced as far back as 2013 by a multitude of submissions to the d+collab <a href="http://healthdesignchallenge.com/">health design challenge</a>.</p><p>Here, the patient’s key physical, emotional and lifestyle factors are visually represented, and health status is also represented as a single number derived from these parameters.</p><figure><img alt="A coloured, visual timeline of a patient’s symptoms" src="https://cdn-images-1.medium.com/max/1024/1*mHMJT65aVB2x2HOCfduV1g.png" /><figcaption>Kate McCurdy, ‘Pictal Health’</figcaption></figure><p>Another example is <a href="https://www.katiemccurdy.com/work/pictal">Katie McCurdy’s</a> visualisation of the patient’s perspective of their own history. This technique means that hours spent around the kitchen table, discussing health concerns with friends and family, can be translated within seconds into the consulting room.</p><figure><img alt="5 body shadows depicting where emotions are felt by people: anger, joy, fear, sadness and love." src="https://cdn-images-1.medium.com/max/1024/1*oBM640yUfU0XY9x2v5jZDA.png" /><figcaption>Orlagh O’Brien, ‘Emotionally}Vague’</figcaption></figure><p><a href="https://orlaghobrien.com/emotions-in-the-body-research-project">Orlagh O’Brien’s</a> visual research of emotions in the body, ‘Emotionally}Vague’, offers a breathtaking insight into a person&#39;s emotional states, which words would struggle to express.</p><p><strong>User Experience and User Interface Design, UX and UI (The Queen)</strong></p><p><strong><em>Definition:</em></strong> User experience can be defined as ‘<em>the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use</em>.’</p><p><a href="https://en.wikipedia.org/wiki/ISO_9241#ISO_9241-210">ISO 9241–210</a> further defines user experience as ‘<em>a person’s perceptions and responses that result from the use or anticipated use of a product, system or service… This includes all the users’ emotions, beliefs, preferences, perceptions, physical and psychological responses, behaviors and accomplishments that occur before, during, and after use’</em>.</p><p>The ISO also lists three factors that influence user experience: ‘t<em>he system, the user, and the context of use’</em>.</p><p><strong><em>On the board: </em></strong>The Queen has the largest body (in size and publications) and authority over the entire board. It has the freedom to move in any direction, in any number of steps. Powerful and almost unassailable, it can dominate without needing other pieces.</p><p>A well-played Queen will often hold still in a strategic position, undertaking its key responsibility (historically under-appreciated) to support and coordinate the other pieces.</p><p>The Queen excels in this supporting role because it can move in the same way as every other piece on the board (except the mentally agile Knight). Over-confident players who are unaware of this responsibility usually lose the game soon after their Queen is taken out.</p><p><strong><em>In the literature:</em></strong><em> </em>User Experience design champions the principle of user-centredness, but the academic literature tells a different story.</p><p>[<em>I ask that the UX or UI experts do not feel judged and do not take offence that I have subsumed their skillsets into the restricted version (not their fault) of their discipline that reigns supreme in the literature.</em>]</p><p>I am forced to begin with the mildly derogatory moniker of ‘Usability Engineering’ because it dominates EHR usability publications. If there is evidence that EHR design lags years behind the ‘real’ world, it can be found in these papers.</p><p>Usability Engineering experts talk to clinicians, but unlike in the ‘real world’, rarely in a back-to-basics way. Until recently, the starting point of most EHR usability research was an existing EHR platform, which most clinicians know would be best scrapped and redesigned. For this reason, their main contribution has historically been restricted to evaluation rather than design.</p><p>Usability Engineering draws on one particularly strong technique, based on Nielsen’s 1994 <a href="https://web.cs.wpi.edu/~kal/courses/hci/module6/neilsen10heuristics.htm">Ten Usability Heuristics</a>. To understand the limitations of Usability Engineering, it is helpful to know how these <a href="https://en.wikipedia.org/wiki/Heuristic">heuristics</a>, which are defined as ‘rules of thumb’, have often been modified and updated for healthcare use.</p><figure><img alt="A list of the 14 clinical usability heuristics with (3) Match and (5) Memory highlighted" src="https://cdn-images-1.medium.com/max/865/1*bagOtyN0pDEIpV77GwRA1A.png" /><figcaption>Zhang, J., et al., Fourteen Principles Customized for the Health Domain (2003)</figcaption></figure><p>The most widely used set of EHR usability heuristics (<a href="https://www.sciencedirect.com/science/article/pii/S1532046403000601">Zhang, 2003</a>) are a hybridisation of Nielsen’s Ten Usability Heuristics and Shneiderman’s <a href="https://www.cs.umd.edu/~ben/goldenrules.html">8 Golden Rules</a>.</p><p>Not such a bad idea, until you know that they were modified (not even by clinicians) to evaluate a medical infusion pump (not even for a data interface).</p><figure><img alt="A graphical representation of the TURF framework" src="https://cdn-images-1.medium.com/max/630/1*_OwJeYTwX8YcwQTKxNaS9g.png" /><figcaption>Zhang, J., et al., TURF: Toward a unified framework of EHR usability (2011)</figcaption></figure><p>It gets worse. This set of 14 heuristics then became embedded into perhaps the most commonly cited (594 times) theoretical framework for EHR usability, <a href="https://reader.elsevier.com/reader/sd/pii/S1532046411001328?token=42FAED62D4686F9C9D71B42006AB4144F72533DE284DC7859DB9246BE0180E69D480E62F65EE05BCBC39466F8E4D9BE7">TURF (2011)</a>.</p><p>(The TURF framework itself is pretty useful.)</p><p>To understand the evolution of this heritage more clearly, an excellent review of usability heuristics can be found in Tarrell &amp; Fruhling’s 2015 paper, <a href="https://www.academia.edu/10720441/Toward_improved_heuristic_evaluation_of_EHRs">Toward improved heuristic evaluation of EHRs</a>.</p><p>An even more detailed review of the specific research that has informed guideline creation can be found in Zahabi’s 2015 <a href="https://www.researchgate.net/publication/274641779_Usability_and_Safety_in_Electronic_Medical_Records_Interface_Design_A_Review_of_Recent_Literature_and_Guideline_Formulation">Usability and Safety in Electronic Medical Records Interface Design: A Review of Recent Literature and Guideline Formulation</a>.</p><p>The elephant in the room of this particular Queen is that heuristic evaluations are carried out by usability experts, not clinicians. A sure bet for a revolving-door mentality, where a discipline explores its own theories in isolation from actual users.</p><p>A glance at Zhang’s 14 usability heuristics through a clinician’s lens will invoke the response, ‘OK, there are the bones, where’s the clinical meat?’</p><p>For example, take the 3rd heuristic, ‘Match between system and world’. The guidance on this is ‘<em>The system should speak the users’ language, with words, phrases, and concepts familiar to the user, rather than system-oriented terms. Follow real-world conventions, making information appear in a natural and logical order</em>’. No one advises on <em>how</em> to do this for the world of medicine.</p><p>Or my other favourite, the 5th heuristic, ‘Minimise memory load’. The guidance reads ‘<em>Presenting all information needed for the task at hand reduces cognitive load. Displaying information organized by meaningful relationships is one method for providing cognitive support’</em>. Again, this is a big ask, and how do you know how to do this without going to medical school?</p><p>What about design rather than just evaluation?</p><p>If there is an absence of UX and UI (in the user-centred sense that we know and value them) in the academic papers, their presence is felt much more keenly in <a href="https://uxdesign.cc/healthcare-ux-a-journey-just-begun-73eebbde3e8d">digital media</a> and EHR <a href="https://uxdesign.cc/how-user-research-helps-improve-the-electronic-health-record-90ca793c037e">startup</a> products.</p><p>As <a href="https://chriskiess.medium.com/">@Chris Kiess</a> asserts in his 2020 (the year and the vision) <a href="https://medium.com/swlh/designing-beyond-2020-44fd644d7e5c">article</a>, UX is a discipline that is both growing up and looking forward.</p><figure><img alt="A screenshot of , with an anatomical human body and clinically salient issues to its right" src="https://cdn-images-1.medium.com/max/1024/1*kqrLlCnUgAwY-NIH7EgE4Q.png" /><figcaption>RonDesignLab EHR (www.dribbble.com)</figcaption></figure><p>Free from the constraints of a Jurassic platform, the single example of RonDesignLab’s <a href="https://dribbble.com/shots/13994226-EHR-Electronic-Health-Record-System-Animation">EHR animation</a> is all that’s needed to show what can be achieved.</p><p>It embodies the very essence of Shneiderman’s 1996 <a href="http://www.cs.umd.edu/~ben/papers/Shneiderman1996eyes.pdf">visual information seeking mantra</a>, offering ‘<em>Overview first, zoom and filter, then details-on-demand</em>’ and successfully tackles Norman’s <a href="https://www.routledge.com/User-Centered-System-Design-New-Perspectives-on-Human-computer-Interaction/Norman-Draper/p/book/9780898598728">gulfs of evaluation and execution</a>.</p><p>The EHR becoming a clinician’s dream, and an example of UX and UI design in its most modern EHR incarnation. The value of one discipline understanding the ‘moves’ of another can be seen in the overlap with Information Visualisation (below).</p><p><strong>Information Visualisation (The Bishop)</strong></p><p><strong><em>Definition:</em></strong> Information visualisation can be defined as ‘t<em>he use of computer-supported (interactive) visual representations of abstract data to amplify cognition</em>.’</p><p><strong><em>On the board:</em></strong> The Bishop does its best work coming into play early and fast. It possesses a divine line of sight, giving it the ability to slice across the board and almost invisibly cover other pieces. A game without Bishops is slow and clunky. Often working the same diagonal lines, there is a natural synergy between Bishops and Pawns (clinicians).</p><p><strong><em>In the literature: </em></strong>Perhaps the key contributor to clinical usability because without visualisation there can be no decent <a href="https://www.interaction-design.org/literature/article/a-very-useful-work-of-fiction-mental-models-in-design">mental models</a>, which in turn allow for dual-processing cognition — or ‘<a href="https://medium.com/swlh/every-chapter-of-thinking-fast-and-slow-in-7-minutes-5e6adf89cf39"><em>Thinking, fast and slow</em></a><em>’</em>.</p><p>You can attempt mental models with dropdowns, tabs and text, but not with any degree of intuition <em>and</em> sophistication.</p><p>Information visualisation excels at design as well as evaluation.</p><figure><img alt="A variable-scale timeline representation of 50+ glucose measurements, showing values, normal range, date ranges" src="https://cdn-images-1.medium.com/max/1024/1*-EzFoCVR7UsfgwmTnPHAzQ.png" /><figcaption>Powsner, S., et al, Graphical summary of patient status (1994)</figcaption></figure><p><a href="https://www.cs.tufts.edu/~nr/cs257/archive/edward-tufte/Graphical%20Summary%20of%20Patient.pdf">Powsner &amp; Tufte’s</a> 1994 summary visualisation of blood glucose is an early representation of a single data type over time. It charts 50+ glucose readings against a variable timescale whilst showing reference ranges and salient dates.</p><p>Anyone can interpret it, from a professor to a patient, gaining an understanding of the data relevant to their personal level of knowledge.</p><figure><img alt="Block-like colour infographic showing ICU measurements of fluid, cardiovascular, respiratory, ventilation, inotropic support" src="https://cdn-images-1.medium.com/max/640/1*P5aOkQtaYYu4fYzcE3rN9g.jpeg" /><figcaption>Horn, W., et al., Metaphor graphics to visualize ICU data over time (1998)</figcaption></figure><p><a href="https://www.researchgate.net/publication/2300982_Metaphor_graphics_to_visualize_ICU_data_over_time">Werner Horn’s</a> dynamic metaphor graphic of intensive care data looks complex, but the user rapidly understands data groups, values, changes, patterns and surprise.<em> </em>The density of diverse but relevant, visually intuitive data is a clinician’s delight.</p><p>See West’s 2015 review, <a href="https://pdfs.semanticscholar.org/937d/42a0ab3f703e7041e277804e0b34bbe6bacb.pdf?_ga=2.139119155.1605627999.1610225828-1446165551.1604852829&amp;_gac=1.255809786.1606919481.Cj0KCQiAk53-BRD0ARIsAJuNhpv9CbSo3zSrc1Yo8oCmeUG7ELgml4Ism2b0cHTVKg8hjXSqpBzBkP8aApYNEALw_wcB">Innovative information visualization of electronic health record data: a systematic review</a>.</p><p>For a definitive description of medical visual taxonomy, read Starren’s 2000 overview, <a href="https://www.semanticscholar.org/paper/Review%3A-An-Object-oriented-Taxonomy-of-Medical-Data-Starren-Johnson/095fd43a1306ca5ad241a76f44311b420ea3620b">An Object-oriented Taxonomy of Medical Data Presentations</a>.</p><p>To back up my claim that we need a coordinated strategy, <em>visualisation</em> heuristics are distinct from <em>usability</em> heuristics — see Forsell’s 2010 paper, <a href="https://www.researchgate.net/publication/220944700_An_heuristic_set_for_evaluation_in_Information_Visualization">An heuristic set for evaluation in Information Visualization</a>.</p><p><strong>Visual Analytics, VA (The Knight, or Horse)</strong></p><p><strong><em>Definition:</em></strong> Visual Analytics is ‘<em>the science of [computational] analytical reasoning facilitated by interactive visual interfaces.</em>’</p><p><strong><em>On the board:</em></strong> The knight is a unique mover, oozing mental agility. Able to jump and rotate over other pieces (within its limited scope), it is the most sophisticated pincer operator on the board. Sometimes called ‘the traveller’, it can appear aloof with competing interests. A formidably effective piece if used wisely, the knight undoubtedly elevates the game, but it is susceptible to getting trapped, unable even to retreat.</p><p><strong><em>In the literature:</em></strong> The mathematical heroes of interface design, with data analysis techniques that can process data to infinity and back again. VA’s linear, planar, volumetric, multi-dimensional, network, temporal and hierarchical visualisation can find data relationships no one ever thought of before.</p><p>However, outputs are not always clinically intuitive. The caution, as on the chessboard, is to use VA wisely and not get trapped into irrelevance.</p><figure><img alt="VisuExplore, heterogeneous data overview, using well-known chart types (2011)" src="https://cdn-images-1.medium.com/max/1024/1*WqUeu7_We_aaZ09LJf9Slw.png" /><figcaption>VisuExplore, heterogeneous data overview, using well-known chart types (2011)</figcaption></figure><p>Rind’s 2017 review, <a href="https://link.springer.com/chapter/10.1007/978-3-319-28661-7_5">Visual Analytics of Electronic Health Records with a Focus on Time</a>, gives a good idea of VA’s cross-pollination with other disciplines without getting too mathematical.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bbUcyd0z-QBsN2wCZSusRA.png" /><figcaption><em>Gnaeus</em>, a guideline knowledge-assisted EHR visualization for cohorts (2015)</figcaption></figure><p>VA’s second superpower is <a href="https://www.frontiersin.org/articles/10.3389/fict.2018.00030/full#:~:text=Better%20Patient%20Outcomes%20Through%20Mining%20of%20Biomedical%20Big%20Data,-Clemens%20Suter%2DCrazzolara&amp;text=Digitalization%20is%20changing%20healthcare%20today,patient%20outcomes%2C%20while%20containing%20costs.">mining big data</a>, across huge patient cohorts rather than individual cases, where the maths comes into its own. Big data analytics of medical information empowers diagnostics, prevention and development of personalised medicines, leading to better patient outcomes while containing costs.</p><p>Its third lies with its intimate relationship with machine learning (ML) and artificial intelligence (AI). In 2018, a form of machine learning known as a <a href="https://www.sciencedaily.com/releases/2018/05/180528190839.htm">deep learning convolutional neural network (CNN)</a> was found to miss fewer melanomas (skin cancers) than a group of 58 dermatologists. What’s more, CNN misdiagnosed non-cancerous moles as cancerous less often than the specialists.</p><p>Numerous other <a href="https://www.healthcareitnews.com/projects/across-healthcare-ai-use-cases-expanding-and-evolving">applications of ML and AI</a> continue to be unleashed. For example, using Natural Language Processing to make sense of large volumes of clinical text, as can be seen in the <a href="https://www.linkedin.com/posts/jesuscaban_nlp-ml-datavis-activity-6761240800627044352-unny">CDR Explore</a> project.</p><p><strong>Interaction Design, IxD (The Rook, or Castle)</strong></p><p><strong><em>Definition: </em></strong>IxD has been defined as the<strong><em> </em></strong>‘<em>design of the interaction between users and products’.</em></p><p>In his <a href="https://www.elsevier.com/books/thoughts-on-interaction-design/kolko/978-0-12-380930-8">2011 book</a>, John Kolko gives IxD a richer definition:</p><blockquote>‘the creation of a dialogue between a person and a product, system, or service. This dialogue is both physical and emotional in nature and is manifested in the interplay between form, function, and technology as experienced over time.’</blockquote><figure><img alt="Tom Cruise operating the interactive gesture-based interface for analysing ‘pre-crimes’ in the film Minority Report" src="https://cdn-images-1.medium.com/max/666/1*X_BFtRMsQqEw1iYvGKl3tA.jpeg" /><figcaption>Spielberg, S., et al., <em>Minority report (2</em>003)</figcaption></figure><p><strong><em>On the board:</em></strong> The Rook moves in a no-nonsense but powerful way but is easily blocked by pieces positioned on the board with different tactical intentions.</p><p>The Rook almost always comes into play late because it is hemmed in by other pieces. The Bishop (Information Visualisation) and Knight (Visual Analytics) and sometimes the Queen (User Experience) initially sit between the Rook and the King, but as they move out into the board the Rook’s distant relationship with the King becomes increasingly intimate. In most games, the Rook is doggedly loyal to the King to the very end.</p><p>The Rook is devastatingly effective when given clear freedom of movement and a long-range view. Just as the Bishop works well with Pawns in their shared diagonals, the Rook works well with Pawns on the shared straight lines.</p><p><strong><em>In the literature: </em></strong>IxD has a strong heritage of working with clinicians and patients.</p><p>An evocative paper setting out the general principle of ‘fluid interaction’ was written by Emqvist in 2014, <a href="https://www.researchgate.net/publication/220586676_Fluid_interaction_for_information_visualization">Fluid interaction for information visualization</a>. These principles applied to the EHR interface will also apply to the clinical consultation itself.</p><figure><img alt="A screenshot of Twinlist, highlighting changes in medication" src="https://cdn-images-1.medium.com/max/935/1*QArtUq4j3Ux6sh8CHKMosQ.png" /><figcaption>Plaisant, C., et al., Twinlist (2013)</figcaption></figure><p><a href="https://www.youtube.com/watch?v=YoSxlKl0pCo&amp;t=0s">TwinList</a> is a user interface prototype for medication reconciliation developed at the University of Maryland.</p><p>In layman terms, it helps a clinician review two lists of medications from different sources (for example when a patient is admitted into hospital) and prepares a new list that tracks changes in medication, doses and timings.</p><figure><img alt="Annotated photographs of LAB-IN-A-BOX" src="https://cdn-images-1.medium.com/max/915/1*18w3pIb4dsS5U42VBOevDw.png" /><figcaption>Weibel, N., et al., LAB-IN-A-BOX (2015)</figcaption></figure><p>IxD also excels at <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6606140/pdf/ojphi-11-e320.pdf">evaluation</a> and has created tools that capture clinician activity and behaviour, pioneered by Weibel’s 2015 <a href="https://phys.org/news/2015-02-lab-in-a-box-aim-doctors.html">LAB-IN-A-BOX</a>.</p><p>This system records synchronised data collected during the consultation, including keyboard and software activity, eye tracking and directional microphone inputs.</p><p>These research data can then be used to inform sequential iterations in the design and development cycle.</p><p><strong>Clinicians (The Pawns)</strong></p><p><strong><em>Definition:</em></strong> ‘<em>a person qualified in the </em><a href="https://www.merriam-webster.com/dictionary/clinical"><em>clinical</em></a><em> practice of medicine, psychiatry, or psychology as distinguished from one specialising in laboratory or research techniques or in theory</em>.’</p><p><strong><em>On the board:</em></strong> Numerous, single-step movements (apart from an optional initial double-step). Uni-directional (apart from sideways diagonal capture movement). The pawn has yet more idiosyncratic moves, such as passing capture (<em>en passant</em>) and the ability to transform into any piece of choice on reaching the final rank (Pawn promotion).</p><p>Widely considered the weakest piece, but the Pawn is on the frontline and more useful than people think. A mature player values the multitude of Pawns; a weak one dismisses them at his peril.</p><p><strong><em>In the literature:</em></strong> The patient-facing users of EHR. The only people who have first-hand experience of using the EHR day in, day out, typically for <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5593724/pdf/0150419.pdf">6 hours a day</a>. Only clinicians possess the authentic, real-world knowledge and skillsets the EHR should be supporting.</p><figure><img alt="A single tan-white pawn stands upright in a sea of fallen pieces of both colours." src="https://cdn-images-1.medium.com/max/1024/1*NqB6ppcska4yM844Odlm-A.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@randyfath?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Randy Fath</a> on <a href="https://unsplash.com/s/photos/playing-chess?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>Clinicians dream of embedding the foundations of clinical reasoning into their EHR, of bringing competence and delight to clinical practice.</p><p>Clinicians are frustrated and feel resigned because they find what they need from their EHR difficult to convey in any meaningful way.</p><p>Some clinicians have developed their <a href="https://www.kingsfund.org.uk/sites/default/files/media/Geoff_Hall.pdf">own EHR systems</a>, scaled them and tackled interoperability hurdles. But whilst they have functionality, they remain behind the curve from a usability perspective.</p><p><strong>Off the board.</strong></p><p>If these pieces are already in place, what is the hold-up?</p><p>The barriers and frustrations of EHR usability do not lie with this global design collective but with external stakeholders, off the board but tyrannically influential.</p><p><a href="http://www.gregoryschmidt.ca/">Gregory Schmidt M.D</a>. presents the most comprehensive overview of EHR design in <a href="http://www.gregoryschmidt.ca/writing/the-role-of-design-in-ehrs"><em>The Role of Design in Electronic Health Records: Past, Present, &amp; Future</em></a>.</p><figure><img alt="Slide from Schmidt’s presentation showing the 4 milestones of the evolution of the EHR" src="https://cdn-images-1.medium.com/max/1024/1*nFh-Mtm6-B7cQ6bMmDnyuw.jpeg" /><figcaption>Schmidt, G., Evolution of the medical record: Milestones (2019)</figcaption></figure><p>He offers the analogy of the EHR as the conductor or quarterback, or the nervous system or road network. Not as a filing cabinet. Fundamentally, he asserts, EHR design is really always about healthcare system design.</p><p>In his compelling presentation, he offers a detailed overview of the history, purpose, challenges, idiosyncrasies, policies, obstacles, stakeholders, frustrations and dreams of EHR. It is essential reading for anyone interested in the field.</p><p><strong>We are decades behind</strong></p><figure><img alt="The chess master clock showing 2 clock faces, each with 2 additional red dials and a push button on top of each" src="https://cdn-images-1.medium.com/max/820/1*Cwn7zOVilqA5dXbGfSzTcQ.jpeg" /><figcaption>chess master® c. 1970, TAG Heuer, Switzerland</figcaption></figure><p>Every year the dissonance grows between the digital tools clinicians use at work and those they use at home and on their smartphone. And every year, design of the EHR interface evolves towards stalemate rather than innovation.</p><p>In 1997, information visualisation guru Edward Tufte presciently cautioned:</p><blockquote><em>‘…</em>a graphical summary should encourage doctors and nurses to reshape, perhaps re-invent, the medical record before computer programmers cast institutional convenience into silicon<em>’.</em></blockquote><p>In that same year, Andreas Broeckmann, likely unwittingly, described digital as the perfect partner to the complexity of the clinical consultation, in his essay <a href="https://v2.nl/events/machine-aesthetics"><em>Remove the Controls</em></a>:</p><blockquote>‘As an aesthetic principle, the machinic [interface] is associated with process rather than object, with dynamics rather than finality, with instability rather than permanence, with communication rather than representation, with action and with play.’</blockquote><figure><img alt="Beth Harmon as a young child, standing, and being invited to a chessboard by her first mentor, the orphanage janitor Shaibel" src="https://cdn-images-1.medium.com/max/624/1*gAWpAKzep8CjbQP1i9j4cA.jpeg" /><figcaption>The Queen’s Gambit, Netflix (2020)</figcaption></figure><p><strong>What do we need to begin the game?</strong></p><p>First, the traditional EHR vendors would do well to acknowledge that beyond <a href="https://www.himss.org/resources/interoperability-healthcare">interoperability</a> lies clinical usability, and that improving clinician performance brings efficiency and safety gains in its slipstream. New metrics can be developed that measure the quality aspect of efficiency rather than simple time and function aspects.</p><p>Secondly, this talented design collective needs their skills to be appreciated and the freedom to excel at what they are good at. They could create clinical delight as if by alchemy, rather than clinical irrelevance because they are siloed. They are not waiting for a grandmaster, just someone who can play the game.</p><p>Thirdly, EHR design needs more authentic engagement with clinician users, which means more time and a more agile platform, but also a shared language. Medicine may be complex but like many other disciplines, its building blocks are not. Like the Pawn transformation in chess, <a href="https://joyclee.medium.com/">clinician-designers</a> are rare but design schools increasingly offer <a href="https://www.rca.ac.uk/study/schools/school-of-design/mres-healthcare-design/mres-programme-description/">courses</a> to healthcare doctors and nurses.</p><p>These are key catalysts to demystifying clinical usability, and a new shared language could elevate a <a href="https://lalgudi.medium.com/on-understanding-wicked-and-tame-problems-6d30c41b542d">wicked problem</a> into a rewarding collaborative challenge.</p><p>Above all, patients need this game to begin in earnest because the best EHR will bring out their best clinician. Chess is, lest we forget, the ‘game of kings’.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/94/0*LZtUC2U-H9Ibu1Gc" /><figcaption>The UX Collective donates US$1 for each article published on our platform. This story contributed to <a href="https://www.bayareablackdesigners.com/">Bay Area Black Designers</a>: a professional development community for Black people who are digital designers and researchers in the San Francisco Bay Area. By joining together in community, members share inspiration, connection, peer mentorship, professional development, resources, feedback, support, and resilience. Silence against systemic racism is not an option. Build the design community you believe in.</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c3b6782c3df2" width="1" height="1" alt=""><hr><p><a href="https://uxdesign.cc/what-if-medical-records-were-designed-like-a-game-of-chess-c3b6782c3df2">What if medical records were designed like a game of chess?</a> was originally published in <a href="https://uxdesign.cc/">UX Collective</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>