Cognition,Care,and Usability: Applying Cognitive Concepts to User Experience Design in Health and Medical Contexts

Usability and Healthcare

Effective healthcare is often a matter of usability (Melonçon, 2016, 2017). If individuals cannot effectively use something to achieve a healthcare objective, then successful care often cannot be provided. Similarly, an inability to use an item correctly in a healthcare context can have adverse results ranging from minor injuries to the death of a patient (Clark & Israelski, 2012; Zhang et al., 2003). In fact, failure to address usability in healthcare has contributed to medical errors becoming a leading cause of death in the United States (Cha, 2016). For these reasons, many healthcare organizations advocate for a usability focus when developing medical products (U.S. Food and Drug Administration, 2016). Certain groups have even created guidelines for integrating usability into the development of healthcare materials, and even regulatory agencies encourage usability in the planning and development of different healthcare products (Callan & Gwynne, 1993; NWX-FDA OC, 2016). Interest in this area has also grown within technical communication where scholars such as Melonçon (2016, 2017) and Batova (2010) note the need to integrate user experience design practices into health communication contexts.

Central factors driving these dynamics are the increase in non-healthcare professionals performing caregiving activities and the location where such activities occur. Over the last decade, the time dedicated to providing patients with dedicated care in contexts like hospitals has decreased (Kalra et al., 2010). Moreover, the amount of time healthcare providers spend interacting with patients remains limited on average (Heath, 2016; Rabin, 2014). Such situations increasingly leave individuals to perform healthcare activities on their own and without the supervision of a trained healthcare professional. This situation means the contexts where such activities occur can be quite broad.

In addition, the number of mass-market technologies for monitoring personal health or treating one’s own medical conditions is growing (Mamtani, 2017; Phaneuf, 2020). These technologies include everything from websites explaining how to identify illnesses to apps that monitor biochemical processes to video tutorials on how to administer treatments. The processes these technologies cover is also expanding to include a range of medical situations and a variety of treatments (Mamtani, 2017; Phaneuf, 2020). As such, the number of non-healthcare professionals engaged in healthcare activities will likely increase over time.

Further complicating this situation is the adoption of electronic health records and information communication technologies for telehealth activities (Agency for Healthcare and Research Quality, 2018; Hecht, 2019). While these systems allow patients to provide information or interact with healthcare workers, they often require patients to input crucial medical data via different interfaces. Yet expectations for what information to collect, how, and the method of collection, can vary across systems and affect results. The contexts where individuals enter such data and interact with systems can affect the kind and quality of data provided. Such contexts can also affect the comprehensibility and usability of the interfaces individuals rely on to provide this input. As artificial intelligence (AI) technologies increasingly review such data and prescribe care, usability factors affecting how patients input information can markedly affect the resulting care received (Artificial Intelligence in Medicine, 2019; Hall, 2016; Page, 2018).

Central to these situations is knowing how patient experiences affect interactions with healthcare technologies (Melonçon, 2016, 2017). Yet individuals are increasingly using such technologies to perform care processes—activities associated with maintaining or restoring the health and wellness of individuals—outside of formal healthcare contexts. The creators of healthcare content therefore need mechanisms for understanding and addressing the contexts where individuals perform healthcare processes or use health-related technologies.

Cognition, Expectations, and Usability

We learn what items are and how to use them based on our experiences (Norman, 2002; Otto & Smith, 2013; Petroski, 1992). Either someone provides us with instruction for using an object, or we mimic the behaviors we observe others performing when they use an item (Berry et al., 2002; Cronk, 1999). The more often we are instructed in or observe the uses of an item, the easier such processes become until they seem almost reflexive. In these ways, our experiences teach us what items are and how they are used. Essentially, our minds use repeated exposure to create a mental model that represents what an object is, what it does, and how it is used. Whenever we encounter an object, we access that model to determine what at item is and does. This is how experience creates usability expectations.

None of us, for example, intuitively knows what a stethoscope is or how it is used to check a heartbeat. Rather, we observed or were provided with explanations of what a stethoscope is and does. The more often we are exposed to such factors, the more our mental model develops until we can see a stethoscope and instantly recognize what it is and how it is used. These factors apply regardless of if one uses a pen to take notes to a sphygmomanometer to check blood pressure. As a result, without experiences to identify what items are and how to use them, they remain unknown and risk being unused or misused.

The design of items is central to this process. The first time I learn what an object is, my mind creates a mental model, or representation, of it. Specifically, my mind notes the features of that item (Aitchison, 1994; St.Amant, 2005). So, the first time I encounter a stethoscope, my mind creates an entry for its features in a mental database of models I use to identify objects. Every time I encounter something identified as a “stethoscope,” I compare it to my mental model for “stethoscope” to see if that object has features in common with that model. Through repeated exposure/experiences, common features become reinforced as central to identifying what that item is. Non-common features, conversely, are omitted from this mental model, for they are not essential to effective identification. In this way, experiences shape expectations of what features something should have for me to recognize it.

We learn the uses of items in a similar way. Either someone instructs us in how to use an object, or we observe how others operate an item to determine its use (Berry et al., 2002; St.Amant, 2018). The more we see similar uses of that item, the more that use becomes established in our mental model for the item. So, our initial observations of using an adhesive bandage (e.g., a Band-Aid™) provide us with an initial mental model for how to use (apply) an adhesive bandage. The more we see this process repeated, the stronger this association becomes for how to use an item.

The associations we have for mental models can also inhibit how we recognize and use objects (Aitchison, 1994; St.Amant, 2005). Imagine your prior experiences taking your blood pressure involved using an analog sphygmomanometer (blood pressure cuff). You then encounter a fully digital one. Would you recognize what it is, or would you think it is something else—say a strange digital watch? Would you overlook it completely? Similarly, if someone told you what that item was (“It’s a digital blood pressure cuff.”), would you know how to use it? Probably not. In this example, you know how to perform a healthcare process, so the knowledge of the care-related activity—what that activity is and how to do it—is not the issue. Rather, you do not know how to identify or use the item available to perform that process. Per the blood pressure example: You know the process for taking blood pressure (restricting then restarting the flow of blood and taking a measurement at key points), but you don’t know how to use the items available to perform this process. In these ways, usability issues can greatly affect actions in healthcare contexts.

Interestingly, the relationship of mental models to usability involve where one learns to identify and use items (St.Amant, 2017b, 2018; Tomkins, 1978, 1987). This is because many objects can be used in different ways, but the context where individuals encounter them determines the related use. In the context of a construction site, one readily recognizes a hammer as a tool for pounding nails and knows the force needed to use a hammer to achieve this objective. In a hospital examining room, one would associate a hammer with testing a patient’s reflexes and related use involving gently tapping a patient’s kneecap to achieve that objective.

Spaces, Scripts, and Scenarios

Scripts are the mental models we use to manage how we behave—and use items—in familiar contexts. They aggregate the details we associate with performing activities in a context (who is there, what is there, what is used, how it is used, how to enter and exit that setting) into a single unit for cognitive processing (Norman, 2002; St.Amant, 2017b, 2018; Tomkins, 1978, 1987). This is why many of us seem to operate on reflex when we perform seemingly complex activities in familiar contexts. Many of us, for example, can enter a location, know where we are, and automatically begin to use certain items to perform certain tasks. It is also why we can easily engage in cognitively challenging activities—such as having a conversation—while easily performing familiar tasks in a known context.

Like prototypes, we form scripts based on repeated experiences we’ve had in certain contexts (St.Amant, 2017b; Tomkins, 1978, 1987). The more I visit a physician’s office, the more accustomed I become to certain elements, individuals, and activities occurring in that location. Through such exposure, I combine factors into a single script for what occurs in that setting (Tomkins, 1978, 1987). Such factors influence usability in terms of what activities I expect to be performed in a location, by whom, with what, and how (St.Amant, 2017b, 2018).

These scripts are why so many of us can automatically move through care-related activities in certain contexts, like hospitals (St.Amant, 2017a). We know our experiences will consist of certain activities in that context (e.g., a hospital) and involve certain objects (e.g., computers for check in, stethoscopes for monitoring heartbeat, etc.) and individuals (e.g., clerk who checks us in, physicians who assesses wellness, etc.). Each object used and every person who interacts with us is an independent prototype we associate with activities in this context (St.Amant, 2017b, 2018). Each of these prototypes is a combination of characteristics (e.g., shape and size for objects, dress for persons) we use to recognize items and identify associated use in that context. Scripts, thus, combine multiple, context-related prototypes into a single unit and minimize the information our brains process when interacting in locations (St.Amant, 2018).

For these reasons, scripts are essential to usability and to successful care across medical and health contexts. We expect healthcare providers to readily recognize and use certain tools, reflexively perform essential processes, quickly identify conditions, and readily provide associated care. This confluence of activities only occurs when the contexts where we receive or administer care match our scripts for care-related activities in a contexts of care, or the setting where care activities occur (St.Amant, 2017a).

These factors mean effective usability in contexts of care involves understanding the scripts individuals use to guide their expectations, activities, and uses of items in a setting (Schank & Abelson, 1977; Tomkins, 1978, 1987). Conversely, when our scripts do not align with a context of care, problems occur (St.Amant, 2017a). In some cases, these problems reflect an inability to recognize or use items in a setting. In others, they involve a failure to use items correctly based on the parameters of a setting. Occasionally, they reveal an inability to know how to provide care or a refusal to accept care based on a perceived lack of credibility associated with an individual or a process. For these reasons, technical communicators need to map such scripts in order to identify what key variables are and determine how individuals recognize them (St.Amant, 2017b, 2018). Technical communicators can then follow these maps to create products that address an individual's usability expectations for contexts of care.

Mapping Scripts and Categorizing Contexts of Care

The first and largest level of organization within scripts are the major activities that occur in a context. These are the acts of the overall script (St.Amant, 2018). The first act of the script (Act 1) for “annual physical,” for example, is likely “checking in at the reception area”—a complex activity that is part of a greater process and that occurs in a particular setting.

Each act consists of different actions, or scenes, that contribute to that overall process within acts (Tomkins, 1978, 1987). For the check-in process, the specific actions occurring could be entering at the reception area, checking in at reception desk, and waiting to be seen. Each activity represents a scene our minds associate with Act 1 of our overall script for “annual physical exam.” The related mental model would look like Figure 1.

OPEN IN VIEWER

Figure 1.

Script for Visiting Emergency Room to Have Condition Assessed.

Each scene is composed of key elements. First, one must enter the scene—a process expected to occur via a particular method of entry into a context. This element is entry conditions and is why we are often surprised when individuals enter a location in unexpected ways (Schank & Abelson, 1977). Upon entry, each process performed involves persons (roles) who perform certain activities using specific things (props) (Tomkins, 1978, 1987).

For the scene of “checking in,” the roles are patient and receptionist, and both are expected to use certain props to complete the check-in activity (St.Amant, 2017b; Tomkins, 1978, 1987). For the patient, the prop might be an insurance card with specific information. For the receptionist, the props would likely be a computer for recording patient data—including insurance card information. Individual might also be expected to dress a certain way that identifies their role in this scene (e.g., receptionist wears a nametag with the hospital’s logo), and the parties might be expected to use certain phrases, or dialogue, to engage in this process (clerk saying “Can I help you?” to indicate the start of the check-in process) (St.Amant, 2017b, 2018). In this sequence of events, individuals expect certain objects and persons to be present, and use prototypes of what these items or persons resemble and what the person should use to guide their behavior in this scene (St.Amant, 2018). The related script map might resemble Figure 2.

OPEN IN VIEWER

Figure 2.

Components of an Act in the Script for Visiting Emergency Room to Have Condition Assessed.

The scenes in the act proceed until all expected activity in that setting is completed and the individual needs to move to a new context, or act, to perform the next series of activities (scenes) in the overall process. At this point, individuals must exit their current context—and the related act—in a certain way in order to move to the next context in the process. Per the physical exam scenario, the patient often needs to remain in the waiting area until called by another individual – often a nurse. The patient then follows that nurse (generally through a doorway near the check-in desk) to the location where the next activities in the overall process occur. The elements associated with how individuals leave one setting and enter another are the exit conditions governing how we expect to move from point to point in a script (Schank & Abelson, 1977). They also explain why, in healthcare contexts, we seem comfortable following a relative stranger to an unknown location without question.

In the next act of our script, the elements and general format repeat themselves. We enter the next context a certain way (entry conditions) and interact with particular individuals (roles). These individuals perform certain activities (scenes) using particular objects (props) and often guide us via particular phrases (dialogue) (St.Amant, 2017b, 2018). Once the activities/scenes for that act are complete, we follow a particular protocol to move to the next context in the script (exit conditions) where these elements appear again and guide our activities in that new space.

These repeated variables, of acts, scenes, roles, props, and dialogue, entry conditions, and exit conditions are central to how we engage in care-related practices and use items to perform healthcare activities in a context. They reflect our prior experiences receiving care in certain settings, and they can be central to usability issues in contexts of care.

Scripts, Disconnects, and Usability Issues in Contexts of Care

While scripts can facilitate our use of items, they can also make us unaware of the usability expectations of others. This situation arises because we often use our own scripts for performing a process to guide how we design materials for others (Eyal, 2014; Lindstrom, 2010; St.Amant, 2017b). In doing so, we assume individuals will be in the same context we usually are when performing a process and will have access to the same items and individuals we consider central to that process. These assumptions can cause usability-related errors in healthcare situations.

Imagine my experiences having blood pressure taken involved going to my physician’s office and sitting passively as my physician (or a nurse) uses a stethoscope and a manual sphygmomanometer to assess my blood pressure, evaluate the results, and suggest a course of care based on these readings (e.g., suggests doing more exercise or prescribes a medication). This script has me playing a passive role, and I depend on other persons (roles) to provide and use the elements (props) involved in the process. In addition, I know how to proceed upon leaving that location in terms of what to do next (e.g., exercise more) or where to go next (e.g., to a pharmacy to obtain a prescription)—the exit conditions for that event.

Let’s imagine your experiences are different. You take your own blood pressure in the kitchen of your home while both watching your small children and preparing dinner. You use an automatic blood pressure cuff you plug into a wall outlet before pressing certain buttons and allowing the cuff to inflate, deflate, and register a reading. You must then determine what your systolic/diastolic readings mean via pamphlets from a local clinic, and your treatment is based on what you find on websites identified during a Google search.

In your script, not only is the setting different, but so are the roles and props involved in the process, entry conditions for bringing information into that location (e.g., pamphlets on assessing blood pressure), and the exit conditions for where to go and what to do after the process is complete. Moreover, your script contains additional roles (children home from school) and props (pans on the stove) that can distract you while performing activities. Likewise, certain environmental factors (e.g., an outlet to access power for the device) are essential entry conditions (bring a needed resource into the environment) as are one’s internet connection for determining a course of care.

Both scenarios encompass the same healthcare process, but each represent different experiences shaping care expectations. Should individuals from the first scenario create instructions for how to monitor blood pressure, they will use their experiences in a physician’s office and design materials for trained healthcare workers to use in a healthcare setting. The content designers will likely also assume the user’s knowledge of the process, including understanding what measurements indicate low, normal, and high blood pressure as well as prospective care for such conditions. Per documentation created for the patient, materials designed for such contexts would likely focus on how to prepare for and what to do so someone else can perform an activity on them vs. do it themselves.

Persons whose experiences correspond to the second scenario (checking one’s blood pressure at home) would likely find instructional documentation designed for healthcare providers ineffective. This second audience instead needs content that reflect their experiences and provide information on how to use a technology to perform the activity themselves. This audience also needs information on how to interpret results and how to respond to these interpretations—including rubrics they can understand and suggestions for researching care options.

Not only can script disconnects lead to different expectations of the kind of content needed, but also the mechanism for delivering the same content. Consider the situation where I need to provide you with instructions on how to use a newly developed technology to check your blood sugar levels. You and I will both use the same technology and perform the same process in the same way. The location in which we undertake this common activity, however, is different. I will use the technology while sitting quietly at my kitchen table at the start of the day or sitting quietly at my desk during my lunch break at work. In this situation, printed documentation—such as a booklet—could serve as an effective method for conveying content I can readily use based on my script expectations for where I perform this process.

You, in contrast, need to perform the related process on crowded commuter train on the way to work. In this situation, you may or may not be able to sit down when using the device, and your ability to open and leaf through a printed text—like a booklet—is limited. The same situation is the case when you attempt to user the monitoring device during your lunch break on a busy construction site where place to sit is limited as is the space available to open and read through a printed booklet. In your script, what you need—and have come to rely on—for content delivery are instructional videos you can watch on your mobile phone as you hold it in one hand and perform the related process with the other. In these cases, it is not the content of the printed manual that renders it unusable in your context of care. Rather, it is the modality used to convey that content that creates script-related usability issues.

These example situations increasingly represent today’s healthcare context where individuals use technologies to perform healthcare activities on themselves or a loved one (Gittel, 2009; Gouge, 2016; Woods, 2019). For these reasons, technical communicator need to become involved in creating health and medical materials to address audience usability expectations and avoid assumptions that could affect usability. Technical communicators can use an understanding of scripts to identify such expectations and develop materials that are usable in relation to audience experiences for a context of care.

Researching Scripts and Mapping Contexts of Use

Understanding scripts is central to creating materials audiences can recognize and use. Doing so involves identifying two factors:

Factor 1: Script Elements—The items individuals associate with performing a process in a context. These include acts, scenes, roles, dialogue, props, entry conditions, and exit conditions (St.Amant, 2017b, 2018).

OPEN IN VIEWER

Table 1.

Map of a Script Sequence.

Script for [Activity in Location]Act 1
Scene 1	Scene 2	Scene 3	Scene 4
Roles	Roles	Roles	Roles
*Props (used in role)	*Props (used in role)	*Props (used in role)	*Props (used in role)
Entry conditions	Entry conditions	Entry conditions	Entry conditions
Exit conditions	Exit conditions	Exit conditions	Exit conditions

Note. See St.Amant (2017b, 2018).

Questions to Guide Data Collection

Understanding an audience’s script expectations for a context of care requires technical communicators to interact directly with members of the intended audience to collect information on these expectations. Doing so involves collecting script-related information in a particular, systematic way that prompts audiences to access the correct script for a context of care. Such a systematic questioning can help technical communicators identify the core variables affecting usability expectations in the script for a context of care. For this reason, identification of an audience’s script expectations requires technical communicators to ask certain questions in a particular order (St.Amant, 2020). This progression of questioning, however, also requires technical communicators to assess the responses they receive to earlier questions to guide the nature of the subsequent questions they ask in this process. Failure to do so can result in decontextualized information that does not effectively or completely reflect audience expectations of a script for a context of care.

Technical communicators can address these factors via a scaffolded approach to questioning in which the responses to one set of questions lays the foundation for the next set of questions. They key is to slowly and progressively identify the script for the actual context of care an audience associates with a healthcare activity. Technical communicators can then identify the different script-related variables for expected script activities and associated usability in this context. This approach should resemble the sequence of questioning described here.

Creating Drafts and Testing Ideas

After collecting script and prototype data, technical communicators should review that information to identify patterns for the following expectations across participants:

Script elements (e.g., acts, scenes, entry conditions, roles, props, exit conditions) audiences associate with the context of care where a healthcare activity occurs

While experiences in a context of care will differ, technical communicators should identify more general trends for care expectations. Ideally, technical communicators will have the time and funding to map different scripts—and develop different materials—for the various audiences identified through this process. Pragmatically, technical communicators need to determine what factors are common across the greatest number of individuals to produce script and prototype maps for a relatively broad audience. The objective is to create

A design checklist of features audiences used to recognize the context where they perform a care-related process

Testing Draft Depictions

Testing draft depictions (e.g., sample designs) should involve interviews and focus groups to determine how effectively audiences recognize contexts, objects, people, and the care-related processes associated with such factors. Testing these draft depictions involves asking subjects questions as follows:

For depictions of contexts

Question 1: What is this place? Can you identify it?

For persons

Question 1: Who is this? How do you know?

Technical communicators should also determine if subjects recognize draft objects (in terms of prototype expectations) and the uses of these objects as related to a context of care.

For objects

Question 1: What is this item? How do you know?

Technical communicators could use the answers to these testing questions in order determine if the initial design checklists and draft depictions for contexts, persons, and objects could be recognized and used by that audience to perform a healthcare activity. Audience feedback could also be used to modify checklists and create new depictions that better meet audience expectations for a context of care.

Technical communicators would use these questions to test the effectiveness of revised materials with new members of the same audience. Data would be collected and reviewed and modifications made depending on resulting feedback. Technical communicators would then test these newly revised items in a similar fashion to assess efficacy and collect comments. This process of revise and re-test would continue until designs effectively met audience expectations, or time and funding for the process ran out. Ideally, materials (maps, checklists, and depictions) resulting from this process would guide the development of care content and care technologies to create items audiences can readily recognize and effectively use in a context of care.

Testing Initial Models

If time and resources permit, one final stage step should occur before a product or content is released for general use. This step involves creating a physical model of the item based on the final map, checklist, and depictions resulting from iterative testing. Members of the intended audience would use this model in the actual context of care location to perform a task in that setting (e.g., “Use this draft manual for operating a digital blood pressure cuff to check your blood pressure in your kitchen.”). Technical communicators would observe and record what audience members do when performing the process and ask individuals to talk aloud during the process. Specifically, audience members should note

What they are doing

The objective of this process is to assess usability in the actual context of care, and technical communicators would need to

Bring audience members to the location where they perform the care activity (if done outside of one’s home)

Final Thoughts

Healthcare contexts are complex, and designing materials for them requires researching where individuals receive care. An understanding the psychology affecting how individuals use items in contexts can facilitate creating products for such environments. Cognitive concepts, such as scripts and prototypes, can be instrumental to these activities. The approach described here helps technical communicators determine factors audiences associate with healthcare practices. Such information can contribute to the development of healthcare materials audiences can use effectively.