I have a new post on EY.com discussing the unique user experience challenges of COVID-19 exposure notification apps.
| |
|
I was granted a patent as co-inventor of “System and Method for Preparing a Pharmaceutical Compound”. This is my 12th (dozenth?) UX-related patent, and certainly one of the more complex ones.
Officially filed in 2015, the method combines information from an image verification system and a gravimetric verification system to ensure that a drug has been appropriately compounded.
In addition, the system displays both image information and gravimetric measurements obtained during drug preparation steps to a reviewing pharmacist in a clear and easily readable manner, such that the pharmacist can quickly approve or reject a particular drug preparation.
| |
|
In 2008 Bresslergroup presented FieldCREW - a concept tablet and software system empowering user researchers and designers conducting field research. It included features for real-time video capture and tagging of observed activities, allowing a single researcher to have the observation capabilities of a team, and a team of observers to synchronize their observations as though they were a single researcher. FieldCREW recevied widespread coverage on sites including SlashGear, the Awesomer, and Core77.
Fast forward five years, and to the Core77/Windows Phone App to the Future Design Challenge. Now we have reconceptualized FieldCREW as an app, providing the same core features as the original concept, in a more realizable and usable format (but, still a concept).
The Bresslergroup app team, led by myself and interaction designer Matt Thomas, created the information architecture and visual design for the app version based on the Windows Phone guidelines.
Watch this the video for more information and screen examples for the Windows 8 phone concept version of FieldCREW:
| |
|
Recently the FDA has been making an evener strong push to have medical device manufacturer's design their products more effectively with regards to human factors. By requiring manufactuers to addess ergonomics, usability and user neer needs in their product development process, we can expect a reduction in errors and improvements in ease of learning and use.
Towards that end, the FDA recognized HE75 as a standard human factors guideline in 2010. This guideline for the design of medical devices provides over 500 pages of human factors principles and recommendations on anthropometrics, user interface design and more.
HE75, like many other consolidated human factors references, is an ambitious effort to consolidate a diverse body of knowledge into a usable resource. I admire it's scope, but it sometimes fall short on clarity. More importantly, the document violates a basic tenet of human factors design - know your user. Much of the content is very technical and ambiguous, which makes it a challenging tool for someone looking for a direct answer. I present a couple of representative examples for consideration.
Sizing a Font
Suppose I am designing a medical device and wish to find out how large to make the type on the screen so that it is readable. This is actually a non-trivial question because readability depends on a number of variables including distance, illumination, and angle, not to mention the acuity of the viewer. But all I want to know is what font size to use. Fortunately, HE75 provides a reference table (see below) of recommended font sizes for various reading distances. The table provides recommendations for six different conditions, including:
but no rationale or definition for what each of these categories mean. Should I assume that if I meet the minimum threshold then I will have created a sufficient design? If I have the space, should I go to adequate, or preferred? Is upper bound for adequate better than lower bound for preferred?
Making matters more complicated, there is another section on type display (hundreds of pages later into the guidelines), that includes this statement:
The minimum character height should be 16 minutes of visual angle.
It would have been great to put that closer to original table, but at least I now know that I should use a minimum visual angle of 16 minutes, which is actually equivalent to "Adequate (lower bound)" in the table, and not "Minimum threshold" - got that? Accordingly, at a reading distance of 36 inches away, the font size should be 12.
Or should it? If I continue reading, I also learn that:
the recommended character size of important but noncritical information is 1/300th of the viewing distance, so a key parameter value on a patient monitor viewed from 3 feet away should be at least 1/8-inch high (a 9-point font: 1 point equals 1/72.27 inch).
So now 9 points is okay from 3 feet (36 inches) even though I previously understood that 12 points would be the miniumum (based on 16 minutes of visual angle from 36 inches)? I have a PhD in human factors and find this challenging - am I missing the subtleties? For such a fundamental question about font size, the guideline is confusing at best, and hypocritical at worst.
Viewing an Alarm
A second example is to find out the visibility requirements of an alarm indicator. Alarms are a critical design feature for obvious reasons, and even have their own standard - HE75 provides a dedicated chapter to alarm guidelines in addition to referencing IEC 60601-1-8.
One question for a device designer relates to placement of an alarm indicator - does it need to be seen from just the front or from all around the device? HE75 provides the following guidance:
IEC 60601-1-8 mandates that point-source visual alarm signals be visible at a viewing distance of 4 meters and from 30 inches off the perpendicular axis.
This is a specific and clear recommendation. So when I went to IEC 60601-1-8 to view the original reference in more detail, I was surprised that there is no mention of "30 inches", but there is a reference to 30 degrees regarding alarm visibility:
the viewpoint is at the OPERATOR'S POSITION or at any point within the base of a cone subtended by an angle of 30° to the axis horizontal to or normal to the centre of the plane of display of the monitoring display or visual indication
Disregarding whether either of these statements is understandable to a lay person in written terms (no diagram or visualization is provided), are these two guidelines equivalent? Is "30 inches" a mistype of "30 degrees" when going from one reference source to the other? If anyone knows, please let me know.
A User-Friendly Alternative
Given the challenges in understanding and interpreting the FDA's recommended documentation, it is not surprising that many medical device manufacturers struggle for guidance and hire consultants to sheperhed them through the human factors process. And it's all the more impressive and refreshing to see a guidelines document that is user-centered and well designed.
The United Kingdom's NHS (National Health Service) has been putting out a series of patient safety guides that are effective for two reasons. First, they focus on a narrow set of topics that are device specific, such as intravenous connectors or urinary catheters. Second, and more importantly, the documents are presented in a clear, direct and visual manner. Specific issues are succinctly described and illustrated, followed by succinctly described and illustrated recommendations. Here's how text size on medication packaging is addressed (note that the recommendation for 12 point size is clearly communicated):
One might make the case that these NHS guidelines are over-simplifications; that they generalize too much and don't accomodate for the variability of real-world conditions. This may be true in some cases, but I would counter that the simplification is much more in the communication, not the content.
Ultimately, the NHS documents come from a customer-facing design approach, while the U.S. documents are derived from engineering, technical standards. Ironically, it is when you provide this greater level of detail and complexity, that you need to provide visualizations to help communicate information more effectively. The cliches abound, but when it comes to design and usability, the FDA (and AAMI) needs to change from a perspective of "do what I say, not what I do", to one of "show me, don't tell me."
| |
|
IDSA.NYC and IDSA's Medical Section present
Innovation Methods for Medical Device Development
Thursday, June 11 — 6:30-8:30 pm ET
Location:
Live: Parsons - 2 West 13th Street, New York, NY
Online: Webinar
Moderator:
Meredith Goodwin Bajaj, IDSA
Chapter Chair, IDSA.NYC
Speaker:
Sean Hägen, IDSA, Principal, Director of Research & Synthesis
BlackHägen Design
In the early 90s the FDA recognized that the majority of deaths and injuries in the healthcare industry associated with adverse effects were due to user error and not device performance or failure. This prompted medical device manufacturers to look more closely at their device definition and development processes to insure that they were doing their due diligence to insure optimum usability.
Sean Hägen will present user research, product planning and development methodologies that meet the need of medical device companies for design practices that optimize ease of use. The presentation will review generative and evaluative research and synthesis methodologies including a case study. The presentation format welcomes audience interaction.
Speaker Bio:
Sean's current focus at BH is in the management of the user research and synthesis phase for product development, which includes ethnography, human factors, usability testing, innovation charrette workshops, translation of research insights into concepts and establishing robust device requirements as they apply to user preferences. Since 1999, Sean has been the Principal Investigator in global, user research projects for clients such as Northrop Grumman, Sunbeam, GE Healthcare, Ethicon Endo-Surgery, Medtronic and Baxter Healthcare. He has lectured on User Research and Applying Human Factors to Systems Design at Embry Riddle Aeronautical University and the Sarasota International Design Summit. Since 1999 he has served on the curriculum Steering Committee and conducted workshops at the Fort Lauderdale Art Institute Department of Industrial Design. He is currently the IDSA's Medical Section Vice Chair and a member on the AAMI Human Factors committee for drafting and reviewing guidelines that are incorporated by regulatory bodies such as the FDA, ANSI and ISO.
Fees (for in-person attendance or webinar) :
IDSA Member: $15
Nonmember: $20
Student: $10.00
To register, please visit IDSA's member portal (even if you're not an IDSA member).
| |
|
What is Evidence-Based Design?
The medical field is a useful resource for consideration by product researchers and designers. The combination of rigorous, but varied procedures, and focus on measurable outcomes lends itself to efficient and effective practices. For instance, a few weeks ago I discussed the success of checklists in surgical contexts and their potential application as a format for design recommendations & guidelines.
Evidence-based design (EBD) is another important methodology that stems, not directly from the medical field, but from architecture and the design of healthcare environments. While one might think that any design process should be, or at least could be, based on evidence, the term "evidence-based design" is specific to the "process of basing decisions about the built environment on credible research to achieve the best possible outcomes."
The goals and methods of EBD should be quite familiar to anyone experienced with the scientific method, but because the approach accounts for qualitative as well as quantitative research and data analysis, it is highly relevant to user research in product design. The basic approach of EBD includes the following:
A simple example in the context of a healthcare environment design (e.g. a new outpatient clinic), might begin with a review of published research on outpatient clinic design as well as reviewing decisions made on similar past projects. The Center for Health Design provides references to extensive resources for evidence-based design. Then conducting interviews with the staff (e.g. doctors, nurses, administrators) and consumers (e.g. patients, family members). The results from this research would drive the design decisions - for example, to provide sufficient collaborative working space in the waiting room for patients and their families. Outcome factors, such as patient satisfaction ratings and waiting time, would also be established and subsequently measured.
Evidence-Based Design and Qualitative Data
One reason I consider EBD valuable to product designers is that the EBD research approach accounts for many of the methodological concerns that arise in qualitative, small-sample research. Evidence-based design makes the valuable point of considering any research method from the perspectives of both objectivity and context. That is, the most objective, typically quantitative methods, such as controlled laboratory studies or surveys, also tend to be the most removed from the actual design context. They provide scientific credibility, but may not account for the specifics of the particular situation.
On the other hand, interviews and ethnographic observation, while qualitative, can be performed contextually, and provide deeper detail and relevance, albiet with less scientific rigor. EBD recommends a balance of both kinds of research to provide the best data set, and also discusses how to blend the two approaches - for example gathering quantitative data about patient movement while conducting observational research.
Similarly, EBD discusses how to handle apparently conflicting or contradictory research findings. Such situations are an opportunity to examine the root causes of the differences - for example did two similar studies provide contrasting results due to differences in the populations studied, or the particulars of data collection?
In other words, the EBD research approach is a realistic and pragmatic one. The mindset of gathering different types of data from different sources and then looking at the findings across the data sources is comparable to the triangulation approach that I employ in product design research.
Evidence-Based Design Vs. User-Centered Design
Presumably any product or interface design practitioner who has read this far would be wondering how evidence-based design relates to user-centered design. In both methodologies, the goals are very similar - apply appropriate design principles to create effective, usable results. And while there is overlap between the two fields, their are also significant differences in their details. In fact, there are at least two critical differences between how evidence-based and user-centered design are practiced:
Adapting Evidence-Based Design Methods to Product Design
It's a given that increasing the integration of published research and outcomes measurement would benefit the product design field. But I expect the reality of that will vary greatly with the particular types of products. Consumer-focused areas like electronics and computers will likely remain relatively closed for competitive reasons.
Medical product design would be an appropriate area to apply EBD methods. It already has the obvious connection with the medical field, and with that come some of the necessary resources. For instance, both ergonomic/human factors periodicals (example here) as well as specialized medical journals (example here) address the effectiveness of tools for the growing field of tool design for laparoscopic surgery. These articles address the effectiveness of tools from both the design and clinical outcome perspective, although require some learning & effort to understand the specialized vocabularies.
But finding appropriate reference information is only half the story. Research is a cyclical process where those taking guidance from previous research must disseminate their own findings. This may be done formally, through the journals and conferences, or informally via blogging or trade group meetings. For example, the interaction designers association, while focused on interface design, is a great model of an online community sharing best practices and guidance for design on an as-needed basis. This bottom-up information distribution is also see in social/professional networking sites such as the Medical Devices Group on LinkedIn.
More generally, the thoughtful planning and balancing of qualitative and quantitative methods advocated for in evidence-based design may be the strongest takeaway for product designers.
In the future, I expect that environmental designers/architects and product designers will work more closely in shaping the entire user experience. In such an integrated approach essentially every detail of the healthcare provider's and patient's experience will be considered - from the layout of the room to the ergonomics of a medical instruments to the usability of healthcare information systems - rather than designed as independent objects that must co-exist within the same system.
For more information on Evidence-Based Design, I highly recommend the succinct and readable Practitioner's Guide to Evidence-Based Design (pictured above).
| |
|
The humble checklist has been receiving a good deal of attention in recent years, particularly in the context of medical error avoidance. For example, earlier this year the New England Journal of Medicine published A Surgical Safety Checklist to Reduce Morbidity and Mortality in a Global Population. The international study reported a nearly 50% decline in patient deaths when a combined written/verbal checklist was used in conjunction with surgical procedures. The authors diligently point out several possible sources of error that may have contributed to the results, including the Hawthorne effect, a "short-term improvement caused by observing worker performance."
But rather than seeing the Hawthorne effect as a source of error, I see it as a source of performance. That is, the checklist process serves as a continuing form of performance observation - by having to refer to an external reference, the surgical teams were consistently self-aware and performed better as a consequence.
Self-awareness is an important issue in successful design as well. A major challenge in user research is communicating recommendations or guidelines to a design and/or engineering team, with the goal of affecting the team's behavior towards a desired design outcome. Design guidelines are typically communicated in a number of ways - reports, presentations, working sessions, collaborative discussions, etc. - but the checklist format might be an even more effective communication tool. This is because of both the way checklists are created, and how they are used.
The checklist format requires that information is presented in specific, succinct and actionable terms. For example detailing the size of touch points, the labeling of a control, or the diameter of a handle. This forces the checklist creator (i.e. the researcher) to come to specific recommendations. Those items that cannot be boiled down to a single point should be further discussed or researched until they can be, and if not, considered for removal.
For the checklist reader, the benefits of a structured list over a more vague presentation or lengthy report should be clear. But its also important to keep in mind the verbal component of checklist usage in the medical study. The check-and-balance approach of having open communication is probably as important, if not more, than the checklist itself.
Comparing the benefits of a checklist in the highly structured, rigorous surgical domain with the creative, diverse world of design may be a stretch, but I expect that the value of clear communication is consistent across all human endeavors.
| |
|
In November AAMI is presenting three online seminars on Human Factors Approaches to Ensuring Safe Medical Devices. The seminars are offered over three days and cover topics including FDA validation usability testing, designing error resistant devices, and identifying human-factors related hazards. Instructors include Michael Wiklund (Wiklund Research & Design) and Ron Kaye of the FDA.
Meanhwhile MDDI (pictured) continues Preaching the Value of Human Factors, with an interview with Peter Carstensen, formerly the FDA's #1 human factors guru. Carstensen talks about the pros and cons of human factors application in the current medical device industry, and discusses the (still) forthcoming AAMI HE75 guidelines:
"HE75 is a very comprehensive handbook describing almost everything a designer needs to know. It’s a one-stop shopping text with most all the information a designer would need to design a good user interface and validate it. But it still requires intelligent interpretation. It’s like someone could write a detailed text on how to perform brain surgery, but careful study and practice will be needed to pull it off. HE75 is a very good start but it’s not a substitute for expertise in the field. There is an even more comprehensive text based on HE75 which should be published before the end of the year. It’s around 700 pages and will sell for the remarkably cheap price of around $125."
| |
|
How do you overcome these note-taking habits? When I provide training on "Minimally Invasive User Research", I emphasize a team-based approach where multiple observers take on distinct, but overlapping roles. For example, one observer may track high-level task flows while another focuses on the detail interactions between a user and a medical instrument. But even when attention is focused to a particular set of user interactions, one can fall back in the write everything/write what's interesting habit.
An effective way to break away from those observational note-taking traps is to use guidelines. Guidelines fall between having no structure and an overly-constrained template, by giving a set of elements to pay attention to, but the flexibility to document them as the observer sees fit.
For example, in the case of observing a medical instrument interaction, I created the guideline of FoRCePS. Forceps are a common medical instrument, making the term a memorable acronym for medical product designers. The acronym represents five ergonomic areas to consider during observations, and is a loosely-based expansion of Stephen Pheasant's cardinal rules of anthropometrics. The guidelines are:
Observers are encouraged to consider each of these guidelines individually for both macro and mico ergonomic issues, but also to understand how they interact with each other. For example, if there is limited visual access, compromising feedback, then a user may change his or her body and limb postures to accommodatean improved field-of-view, but in doing so, increase the extent of reach and reduce the effective transfer strength. I recommend watching a brief segment of a medical procedure (or other task where ergonomic compromises are common) to practice paying attention to these 5 issues. So even with a set of five key principles, there's a lot to pay attention to during live observations and in follow-up video review. Fortunately, FoRCePS and similar mental "tools" give an observer guidance and provide a consistent way to track issues that can be shared with other observers who are focusing on different aspects of the observed task.
| |
|
Exploring Ethnography for Design Research in the February issue of MD&DI is more than a follow-up to the classic article Ethnographic Methods for New Product Development.
According to author Stephen Wilcox (Chair of IDSA Human Factors section), ethnographic research is now common in medical device development. The majority of this research is of course, qualitative, and primarily focused on identifying opportunities:
"Much so-called ethnographic research—perhaps most of it—is designed simply to generate ideas, that is, to stimulate creativity. Inevitably, when members of device-design teams go into the field and see directly how their devices and other devices are used, it generates insight and stimulates new ideas."
But there is another type of ethnographic research that is as much about the validity of findings as it is about generating ideas (Sidebar: simply put, in research, validity refers to the degree that you are actually measuring what you are intending to measure). Typically validity is associated with quantitative measurement based methods such as performance testing. But Wilcox suggests several ways to increase validity in ethnographic research including careful sample selection, quantitative measurement and objective data recording. This more robust approach to ethnographic research comes with a price:
"conducting such research is difficult, time-consuming, and, frankly, expensive, in comparison with the idea-generation type of ethnographic research."
It's unlikely that most organizations will be able to accommodate all of the steps necessary to conduct highly valid ethnographic research - especially since many are just getting into the practice of doing any field research regularly. But Wilcox's recommendations should really be taken as best practices for conducting any type of user research effort (whether validity is an explicit intention or not). For example, making sure that the "sample accurately reflects the population of interest" is a fundamental research planning step. The deeper challenge when addressing validity is knowing what you know - for example determining whether your sample is truly representative.
The February MD&DI issue also contains an article on considerations for designing medical devices for home use, and another article on integrating human factors into the medical device development process.
Finally, for a less technical, down to basics overview of ethnography, see Design Meets Research, from GAIN, AIGA's journal of business and design.
| |
|
One of the most challenging aspects in conducting medical ethnography/observation for design research is constructing a complete and accurate task analysis. Breaking down a complex surgical procedure into logical sub-tasks typically requires rigorous observations over multiple sessions, detailed video review and validation from subject matter experts.
Constructing hierarchical task analysis in surgery*, which was published in the January 2008 issue of Surgical Endoscopy, provides a high-level description and concrete example of the process for creating a hierarchical task analysis in a surgical context. Click on inset image for a process diagram from the article.
*Note - accessing the full article requires a paid subscription, or the individual article may be purchased for $32 - then again there's always the library.
On a somewhat related note, New Scientist reports on the development of surgical training simulations for the Nintendo Wii:
"Now they are designing Wii software that will accurately simulate surgical procedures. A training platform based on the console, which costs about $250, might be more practical for trainee surgeons in the developing world..."
For designer/researchers, lower cost training simulators may provide a way to simulate or test prototype surgical tools in the design process.
| |
|
In his most recent article in Medical Device and Design Industry, Michael Wiklund discusses the need to thoughtfully design every interaction, or touchpoint. Refined Touchpoints Drive Quality Perceptions suggests that medical product purchasers are influenced by effective, high quality design as much as consumers might be when shopping for a new household kitchen appliance.
Wiklund emphasizes the industrial design aspect of touchpoints with a clear understanding of the influence of materials and mechanics on experience. This makes the article a useful introduction for designers and usability specialists who might lack exposure to ID. For example, his recommendations include:
Key issues covered in the article include the obvious, like buttons and handles, but also less considered, but essential components such as wheels and connectors that can have a strong influence on the perception of design quality. And while the focus is on medical devices, the suggestions are applicable to other product design fields.
| |
|
| |
|
2007 has witnessed the continued maturity of user research practices in product design/development organizations. As this continues, we look to 2008 and key areas of growth and change in user research technologies and methodologies. What many of these themes have in common reflects a shift from how to conduct research, to how to manage all of the research findings and results – clearly a positive trend and a nice problem to have. Stay tuned into 2008 as these themes are tracked in further detail.
Even a casual reader of this web log will have observed the ever-growing options in data gathering technologies available for a variety of research applications. For 2008, the themes in technology are diverse – from high definition video to a new resource of anthropometric head measurements of the Chinese population. But the more compelling tools address needs in organizing and analyzing qualitative data:
Design research methods will continue to adapt for studying the wider range of user experiences, beyond the primary product. Frameworks and techniques for mapping out user touch points will assist research planning, which will become specialized to particular domains (e.g. medical vs. consumer). Threading across all of this is the need for guidelines for effective research communication and presentation:
| |
|
The November 12 issue of Businessweek features two healthcare design related articles:
| |
|
A brief article in Wired online about the development of medical instruments features the work of Mary Beth Privitera of the University of Cincinnati:
"Privitera said designers could cut training time and anticipate problems by getting involved earlier in the medical design process. But to do that, they have to increase their knowledge of anatomy and physiology by working in the field with doctors."
Mary Beth also presented at last week's IDSA Connecting '07 conference.
(Thanks to Core77 for calling attention to this article).
| |
|
A new article at Medical Device Link, Catching the Human Factors Fever, discusses the current state of human factors in the medical device industry. The author, Erik Swain, who is the editor-in-chief of MD&DI, discusses successful examples of integrating HF processes into product development. Industry trends, such as increased FDA enforcement and a shortage of healthcare providers have increased the need for usable medical products. Human factors is also playing a role at the supplier end, moving usable design further upstream in the product development process.
Swain concludes that:
"More device companies are committing to human factors programs, and that will benefit caregivers and patients in the long run. But mere acceptance of human factors programs won’t be enough. The next step is to integrate usability findings with the risk management process. "
| |
|
| |
|
I'm posting the latest newsletter for the Human Factors & Ergonomics Society - Healthcare Technical Group. The newsletter includes information on FDA Critical Paths Opportunities, as well as the program for healthcare sessions at the HFES 50th meeting in San Francisco this fall.
The file is downloadable: Download spring_2006_newsletter.pdf
| |
|
The latest issue of the Journal of the Human Factors and Ergonomics Society (Volume 48, Number 1, Spring 2006), features a special section dedicated to Patient Safety. There are ten articles covering a range of associated topics including medication labeling, information display and simulation. Of particular note are two articles:
| |
|
An article from earlier this year by By Patricia A. Patterson and Robert A. North discusses the necessity of applying Human Factors methods in the medical device design process. It's a nice overview of how and why Human Factors can fit within the product design and FDA processes.
The article can be viewed at the Medical Device & Diagnostic Industry (MD&DI) site:
| |
|
MDDI's site has an index of Design-related articles that covers a range of topics encompassing human factors, design, error avoidance, methods and case studies.
Articles are available online from 1996 to the present:
| |
|
Understanding the usability and design limitations of current medication bottles led to the invention of an innovative medication bottle that provides improvements in ergonomics, readability and safety.
| |
|
| |
|
