Do it by DesignAn introduction to human factors in medical devices

by Oliver Schreck

Oliver Schreck

Human Factors Engineering Process

01 August 2013 2/99

Oliver Schreck

Primary usability questions• How easy is it to learn to use the device?

• How soon will the intended user feel comfortable using the device?

• Once learned, how efficiently can the device be used?

• Do users remember how to use the device after several days, weeks, or months of non-use?

• Does the device prevent users from making errors or help users recover from their errors?

• Are users satisfied with the device?

• Is the device design appropriate for the capabilities and limitations of users?

01 August 2013 3/99

Oliver Schreck

Rule of Thumb• Do not contradict the user’s expectation. • Be consistent and unambiguous in the use and design of headings,

abbreviations, symbols, and formats.• Always keep users informed about current device status.• Provide immediate and clear feedback following user entries.• Design procedures that entail easy-to-remember steps.• Use prompts, menus, etc. to cue the user regarding important steps • Do not "strand” the user.• Give users recourse in the case of an error. • Do not overload or confuse users.• Consider the use of accepted symbols, icons, colors, and

abbreviations.• Do not over use software when a simple hardware solution is available.• Consider using dedicated displays or display sectors for highly critical

information. 01 August 2013 4/99

Oliver Schreck

Seek user input• Involve users early and often• Get feedback asap from

customer

• Refine designs through usability testing

01 August 2013 5/99

Oliver Schreck

Establish design principles• Keep it simple• WWAD?

• Ensure safe use

• Ensure essential communication• E.g. Alarm design

• Anticipate device failures

• Facilitate workflow

01 August 2013 6/99

Oliver Schreck

Accommodate user characteristics and capabilities• Do not expect users to become masters

• Expect user errors

• Accommodate diverse users

•Maximize accessibility

• Consider external factors that influence task performance

01 August 2013 7/99

Oliver Schreck

Accommodate users’ needs and preferences• Prioritize user input

•Do not rely exclusively on “thought leaders”

• Let user set the pace

01 August 2013 8/99

Oliver Schreck

Establish realistic expectations of user

• Do not rely on training

• Do not rely on instructions for use

• Do not rely on warnings

• Do not rely on users’ memory

• Avoid information overload

• Do not assign users tasks that are better suited to the device

01 August 2013 9/99

Oliver Schreck

Consider real-world demands• Consider the context of use

• Consider worst-case scenarios

• Make devices as rugged as necessary

• Limit user workload

• Consider the potential for device migration into other users or use environments

01 August 2013 10/99

Oliver Schreck 01 August 2013 11/99

Oliver Schreck

Develop compatible designs• Accommodate mental models

• Establish natural or conventional mappings• People associate an action with a response

• E.g. turning a knob to turning off something

• Follow industry conventions and consensus standards

• E.g. UI guideline from Microsoft• AAMI HE75

• Reduce learning time by applying known principles

01 August 2013 12/99

Oliver Schreck

Enhance safety and effectiveness• Make devices error-tolerant and fail in a safe manner

• Avoid physical strain, repetitive motions and cumulative traumas

• Help users to anticipate future events

• Confirm important actions

• Make critical controls robust and guard them

• Clarify operational modes

• Employ redundant coding

• Design to prevent user confusion

• Don’t neglect device appeal01 August 2013 13/99

Oliver Schreck

Managing the risk of use errorUse related hazards vs. device failure hazards

• Device failures: mechanical, electrical, biochemical, or other system related problems

• Use-related hazards: Hazards initiated by users during interactions with a device and resulting in an unintended consequences

01 August 2013 14/99

Oliver Schreck

Types of use errors

01 August 2013 15/99

Oliver Schreck

Use error types and risk controls

01 August 2013 16/99

Oliver Schreck

Managing use-related hazardsa) Failures of the device for reasons other than use should generally be treated separately.

b) Mitigations for identified use errors should be validated by actual users to ensure that they are effective.

c) The risk management process should specifically consider use-error scenarios that could lead to catastrophic consequences. Any risks associated with these scenarios should be evaluated further.

d) Evaluation of use should include scenarios that cover aspects of use associated with risk and should be emphasized by priority.

e) At least 15 or more users should be involved in a summative usability study, and they should be broadly representative of the population of intended users in terms of their abilities. The quality of a study depends more on the extent and completeness of risk evaluation than on the number of participants.

f) Members of the design team should not participate in evaluations of use, especially validation (summative usability) studies. Other employees of the manufacturing company are generally not good test candidates either, because they are likely to be biased toward positive assessments.

g) Assessment tools that include ratings of “ease of use,” “intuitiveness,” and other global concepts should only be used early in the evaluation process to help identify aspects of use that should be evaluated further. Validation studies limited to this kind of measurement are incomplete.

01 August 2013 17/99

Oliver Schreck

Managing the risk of use errorCheck:

•ANIS/AAMI HE74•ANSI/AAMI HE75• ISO 14971• IEC 62366

01 August 2013 18/99

Oliver Schreck

Implementation of risk control•Most preferred• Design modifications• Safeguards• E.g. confirmation by user

• Less preferred• Modification of intended use• Training• Warnings and labeling

01 August 2013 19/99

Oliver Schreck

Basic human skills and abilities

01 August 2013

Organization of basic human skills and abilities

20/99

Oliver Schreck

Vision – Visual perception• Vision threshold - Minimum light level• Visual acuity - Accuracy of vision, static and

dynamic• Focus abilities - Myopia, hyperopia,

astigmatism• Visual angle• Distance and perceived size• True object size• Minimum type sizes• Visual illusions – parallax error• Perception of motion• Flickering lights• Photosensitive epilepsy• Display flicker• Defective color vision – Dichromacy(two

colors), Protanopia (red blindness), Deuteranopia (green blindness), …

01 August 2013 21/99

Oliver Schreck

Vision – Visual acuitya) minimum distinguishable (detection of detail in an arbitrary

test target)

b) minimum perceptible (detection of a spot, e.g., on a magnetic resonance imaging [MRI] scan)

c) minimum separable (detection of a gap between parts of a target)

d) stereoscopic acuity (detection of depth for a three-dimensional target)

e) Vernier acuity (detection of lateral displacement of one line from another).

01 August 2013 22/99

Oliver Schreck

Factors that affect visual acuity

01 August 2013 23/99

Oliver Schreck

Color choices

01 August 2013

Color choices with less then 2% misidentification rates

24/99

Oliver Schreck

Human information processing•Stimulus discrimination• Better relative judgment than absolute

•Attention•Only one source of sensor at a time

•Vigilance (sustained attention)•Decreases over time

01 August 2013 25/99

Oliver Schreck

Speed of information processing

01 August 2013

Simple reaction time for various senses

26/99

Oliver Schreck

Workstation design

01 August 2013 27/99

Oliver Schreck

Usability testing – general considerations

a) User-interface verification can ensure that design outputs meet design inputs.

b) User-interface validation can ensure that the product’s usability meets user requirements.

Acceptance criteria for the final summative usability test:1) valid usability objectives that truly reflect user requirements;2) use of risk analysis and risk management in the selection and

prioritization of usability objectives as acceptance criteria; 3) risk-mitigating controls related to use-related hazards.

01 August 2013 28/99

Oliver Schreck

Types of usability testing

01 August 2013 29/99

Oliver Schreck

Good usability testing designA usability testing plan should specify

a) the types and numbers of representative end users who will be tested (e.g., nurses, physicians, patients, technicians, therapists);

b) a sample size large enough for the test goals (5 to 8 participants for early formative testing, 15 to 20 for later summative testing against usability objectives as acceptance criteria);

c) realistic tasks based on user scenarios derived from previous task analysis and risk analysis;

d) a realistic use environment (e.g., one that accurately represents the expected environmental lighting, noise, and presence of other equipment);

e) the use of real product, sufficiently interactive simulations, or low-fidelity prototypes;

f) the recording of users’ actions and comments (usually one-on-one);

g) performance measures (e.g., task completion, time, errors, accuracy, observer ratings); and,

h) optionally, user satisfaction ratings and comments. Such measures should be used only as secondary measures to complement the primary measures of observable user performance.

01 August 2013 30/99

Oliver Schreck

User - Labs

01 August 2013 31/99

Oliver Schreck

Eye tracking of user

01 August 2013 32/99

Oliver Schreck

Usability testing in the design process

01 August 2013 33/99

Oliver Schreck

StudiesCan provide the following:• a snapshot of how healthcare professionals use devices;• a picture of how operating conditions, including the multi-device environment, affect use;• a snapshot of what problems are encountered;• anecdotal reports or comments;• a sampling of the user population with respect to individual differences;• ideas for new designs and reactions to design concepts; and• information necessary for establishing performance test protocols and performance criteria.

01 August 2013 34/99

Oliver Schreck

Further details usability testing• Test plan content• Test case selection• Test organization• Interviews•Data collection, video, logs, eye tracking, …•Data analysis•… see chapter 9. AAMI HE75

01 August 2013 35/99

Oliver Schreck

Follow the norms for symbolsa) All equipment elements that are not intuitively obvious, that have some usable

functions, or that require identification for other reasons should be labeled or marked.

b) Markings should be positioned so that they are clearly associated with the correct equipment feature and not obscured by hand positions or equipment components.

c) Appropriate markings can enhance the identification of both individual elements and their functional relationships.

d) Labels should be visible at typical viewing distances and angles.e) Labels should be resistant to wear and tear.f) Ambiguous symbols, codes, or terminology should be avoided.g) Designing for legibility requires careful analysis of ambient illumination in typical

use environments.h) Consistency of placement, terminology, and coding is critical.i) All markings should be tested with typical users. Users can be clinicians, caregivers,

patients, or maintenance personnel and can vary by age, disability, and other characteristics.

01 August 2013 36/99

Oliver Schreck

Use of symbols – AAMI TIR 60878

01 August 2013

Example: Overview X-Ray: Equipment and movement

37/99

Oliver Schreck

Cross-Cultural Design

01 August 2013 38/99

Oliver Schreck

Cross-Cultural design

a) Investigate the distinguishing characteristics of the target markets in terms of cultural and national factors.

b) Generate design adaptations that address the unique requirements for that culture or nation.

c) Develop a design that encompasses the needs of all cultures and nations.

d) Validate the design adaptations in the target markets.

01 August 2013 39/99

Oliver Schreck

Cross cultural design factors

01 August 2013 40/99

Oliver Schreck

Differences among nations and cultures

a) national issues (language, regulatory requirements, unit system and format)

b) cultural issues (technical environment, use environment, social context, professional traditions, work organization)

c) culture-specific or nation-specific user profiles (demographics, anthropometric characteristics, system of values, preferences and expectations, attention, knowledge and educational background, interpretation of colors and symbols, learning style)

01 August 2013 41/99

Oliver Schreck

Language issues among countries

01 August 2013 42/99

Oliver Schreck

Differing units and formats

01 August 2013 43/99

Oliver Schreck

Color symbolism by culture

01 August 2013 44/99

Oliver Schreck

Guidelines for cross-cultural designa) How large is the expected market for the

device?

b) What is the scope of the changes required to modify the design for the new market?

c) How complex is the user interface?

d) If device-related information is not translated into the local language, will speakers who do not speak the offered language have difficulties operating the device, particularly under high-stress conditions?

01 August 2013 45/99

Oliver Schreck

User device interfaces• Hardware

• Input of Chinese vs. western language

• Sequence of operations• Influence by the language, work habits, education

• Language• Display resolution• Text field widths (+30% for German/French)• Prioritized translation (no mixed languages)• Country requirements

• User-device interface orientation

• Format issues (dd/mm/yyyy vs. mm/dd/yyyy)

01 August 2013 46/99

Oliver Schreck

Symbols in various countries

01 August 2013

Comprehensibility of some standardized symbols, AAMI TIR 60878

47/99

Oliver Schreck

Touchscreen-control selection matrix

01 August 2013 48/99

Oliver Schreck

Touchscreens

01 August 2013 49/99

Oliver Schreck

TouchscreenParticularly appropriate for applications in which:a) menu selections are required;b) the user’s focus is on the display;c) it is time-consuming or dangerous for the user to

divert attention from the display;d) the workload can be reduced with a limited number

of inputs;e) potential users are relatively inexperienced; and/orf) the device will be used in a high-stress environment.

01 August 2013 50/99

Oliver Schreck

Touchscreen advantagesa) Touchscreens only require natural pointing and finger-sliding

gestures.b) The touchscreen provides a direct relationship between hand

and pointer movement in terms of direction, distance, and speed.

c) The possible inputs are limited by what is displayed on the screen; thus, no memorization of commands is required.

d) The limited number of possible inputs minimizes input errors.e) Users need only a minimal amount of training.f) Touchscreens do not require additional surface space like most

other pointing devices.g) Touchscreens are durable in high-use environments.

01 August 2013 51/99

Oliver Schreck

Touchscreen disadvantagesa) Touchscreens with overlays mounted in front of the display (typically a laminated film

mounted on glass or Plexiglas) can degrade the image quality and affect the display’s luminance, contrast, and effective resolution.

b) Users can experience considerable fatigue with touchscreen use over extended periods of time.

c) During selection, the user’s finger obscures part of the screen.d) The resolution constraints of touchscreens and finger size can make selecting small

items such as small symbols or single characters difficult or even impossible.e) Unless multi-touch technologies are used, data input could be slowed because only

one finger is used.f) Fingers might soil the screen.g) Touchscreens with capacitive overlays might not work reliably for users wearing

certain types of gloves.h) Touchscreens are not appropriate for some users (e.g., users with tremor or poor eye–

hand coordination).i) Users can damage touchscreens by making selections using objects other than their

fingers.j) Directional, continuous-control functions can be more difficult to implement on a

touchscreen than with sliders or rotary knobs.

01 August 2013 52/99

Oliver Schreck

Touchscreen - geometry and layouta) Size and configuration: Height and width of the actuation

areas on the screen should be at least 13 mm (0.5 inches). Spacing between adjacent areas should be at least 6 mm (0.25 inches). Errors increase as controls get smaller than approximately 23 mm (0.9 inches), but providing “dead space” between keys helps prevent errors.

b) Shape: Visually "concave" and "convex" shapes should be used to indicate button status.

c) Labeling: Placing labels in the center of touchable areas improves usability because users are drawn to and tend to touch them. Labels next to touchable areas often cause confusion and frustration, leading to use errors.

d) Parallax: A common problem with touchscreens is parallax, which is the misalignment between an object’s perceived position on a screen and the position of the object’s associated touch area.

01 August 2013 53/99

Oliver Schreck

Force, activation and feedback

• Force: • direct touch resistance should be in the range of 0.25N-1.5N

• Activation:• “Up-triggers" (activation upon release) are generally preferable to “down-triggers" (activation

upon initial touch) because activation upon release of the finger decreases errors. A good approach is to highlight an item when it is touched and then execute the choice when the finger is removed.

• Making the entire area of a button touchable facilitates use and reduces confusion. • Displaying crosshairs can be helpful when accurate target selection is needed.• Highlighting the currently selected area compensates for the lack of tactile feedback.• Coding by shape or color permits differentiation of active areas from text and background

graphics.

• Feedback: • Touchscreens should generally provide auditory feedback to indicate activation or selection

input.• Touchscreens intended for regular use should provide users with the option of muting auditory

feedback to prevent sound distraction or redundancy.• Buttons on touchscreens should provide immediate feedback with a press-and-hold repeat time

(with continuous pressing) of 0.09 seconds.• “Touch mice” are cursors (indicated by crosshairs or arrows) that are controlled by the finger.

They decrease mistakes in finger placement and provide clear feedback for positioning.01 August 2013 54/99

Oliver Schreck

Software – user interfaces• Easy to use• Untrained user should be able to accomplish

basic and critical tasks

• Focus on user task• The work on a device is task oriented• Identify the frequent, urgent and critical tasks

• Provide user guidance• Step by step advice

• Safeguard against use error• Verify entered data

01 August 2013 55/99

Oliver Schreck

Software – user interface • Optimize interaction requirements

• User normally has to master multiple devices

• Improve software and hardware integration• Touchscreen on device

• Select the interaction style• Menus• Direct manipulation• Q&A dialog• Command line• Additional interaction (e.g. spread sheets, speech, …)

• Support product evolution

01 August 2013 56/99

Oliver Schreck

Special considerations• Screen size

• Use the real estate with care

• Compatibility• One unified “look & feel”

• Information priority• Communicate critical information quickly, accurately and reliably.

• Information legibility

• User population• Take care of education background• Server multiple user groups• Flexibility by advanced and simple operating mode

• Standardization• Apply to standards for medical devices

• System integration• Share data with other devices

01 August 2013 57/99

Oliver Schreck

Design guidelinesa) Conceptual model: Provide an overarching construct for how users will think about a

user interface (i.e., the mental picture that users form about how the software–user interface works).

b) User-interface structure: Place individual screens in a logical hierarchy that complements how people prefer to approach frequent, urgent, and critical tasks.

c) Interaction style: Establish a pattern of interaction between the user and the software application that facilitates tasks and accommodates users’ capabilities.

d) Screen layout: Organize information on the screen so that users can locate specific items quickly and make appropriate associations.

e) Legibility: Present textual and graphical information clearly so that users can read it and discriminate important details.

f) Aesthetics: Present information in a visually pleasing way so that the user interface does not intimidate new users and positively influences task performance.

01 August 2013 58/99

Oliver Schreck

Bad Interaction – 4 steps to one menu

01 August 2013 59/99

Oliver Schreck

Design guidelinesf) Data entry: Establish rules for how users input data or make selections by

means of the software–user interface.

g) Color: Use color to contribute to the clarity of information in meaningful ways and draw attention to the most important information.

h) Dynamic displays: Use active graphical and textual elements to convey information more compellingly than is possible using static displays.

i) Special interactive mechanisms: Provide information related to less common control mechanisms, such as soft-key, control-wheel, and touchscreen user interfaces and on-screen keyboards and keypads.

j) User support: Give users information at the correct time and in the correct format to help them perform tasks safely, quickly, and effectively.

k) Consistency: Try to provide the same type of controls whenever possible.

01 August 2013 60/99

Oliver Schreck

How to use this controls?

01 August 2013 61/99

Oliver Schreck

Include numbers in combo boxes

01 August 2013

You can't tell the number of results and there is a scroll bar

The number of results is clearly displayed.

62/99

Oliver Schreck

Checkbox vs. Radio button

01 August 2013

Bad Example – Radio buttons are not appropriate when there are only two options

Good Example – These yes/no questions have a better representation with checkboxes

63/99

Oliver Schreck

Timely feedback

01 August 2013

This progress bar looks like it is stuck at 99%. Ideally the progress bar should be hidden when completed and replaced by a green tick

64/99

Oliver Schreck

Conceptual modela)Number of elements: The number of

basic elements should be limited, 10 or less, so that users can form a simple mental model of how the user interface is organized.

b)User task orientation: Should reflect a logical organization of user tasks rather than the device’s electromechanical functions or software modules.

01 August 2013 65/99

Oliver Schreck

User-interface structurea) Compatibility with conceptual model: The software–user interface architecture—

should reflect the underlying conceptual model.

For example, a conceptual model prescribing a software– user interface with five major elements (e.g., setup, calibration, treatment, reports, and system administration) should be complemented by user-interface architecture with five major pathways (e.g., menu options) that users follow to perform associated tasks.

b) Menu depth: People prefer relatively shallow menu systems that require users to navigate, preferably, two levels but, ideally, no more than three levels deep in a menu hierarchy to reach the desired content or options.

c) Menu breadth: People prefer menu systems that do not have an overwhelming number of options. A medical device with too many options might intimidate new users and can make it more difficult for users to form an accurate mental model of how the device works. The optimal number of options seems to fall in the broad range of three to twelve options, although five to nine options is typical.

d) Linear, branching, and networked (web-like) structures: User-interface structures can be linear, branching, or networked (web-like). Each structure has advantages and disadvantages in terms of ease of use, task speed, and overall compactness (number of screens). Many user interfaces are blends or hybrids of these types.

01 August 2013 66/99

Oliver Schreck

Interaction stylea) Pace of tasks: A device’s interaction style should allow users to perform tasks at the pace

necessary to accomplish the intended tasks safely and efficiently.

For example, users should not have to endure delays when requesting laboratory results to be displayed on a patient monitor at the patient’s bedside. Neither should they have to repeatedly navigate to a patient data entry page because the associated device timed out and automatically returned to a “resting screen” or top-level display.

b) Pointing device compatibility: Interaction style should be carefully matched to the type of pointing device so that users can accomplish tasks without hindrance.

For example, direct manipulation calls for a pointing device suited to selecting and “dragging” on-screen objects. By comparison, ATM-like soft keys facilitate a menu-based approach.

c) Consistency of interaction style: Although a large proportion of user interfaces employ more than one interaction style, developers should apply them consistently.

For example, the same user interface should not require users to select from a comparable set of options by clicking on a “radio button” in one case and typing an option into a blank form in another case unless using two approaches has benefits that outweigh the benefits of consistency.

01 August 2013 67/99

Oliver Schreck

Clear icons show the function

01 August 2013 68/99

Oliver Schreck

Corner menus guide the user

01 August 2013 69/99

Oliver Schreck

Screen layouta) Alignment grid: Screen content should be aligned to a grid to give the screen a

consistent, orderly appearance that facilitates rapid scanning with clear demarcation of functional groups and information hierarchies.

b) Content hierarchy: Screen content should ascribe to a hierarchy appropriate to the selected design schema, the tasks being supported, and the user population. Accordingly, the most important information or the first step in a multistep procedure should be presented in the highest-priority (i.e., most conspicuous) location (typically the top and/or left side of the screen for most cultures).

c) Content distribution: Screens should appear balanced because of a relatively even distribution of content (figure) and blank space (ground).

d) Gutters and padding: Readability is usually enhanced by adding a gutter or margin between the screen’s edge and content. To ensure visual separation from the screen’s edge, gutters should be at least a few pixels in width and height.

01 August 2013 70/99

Oliver Schreck

Where is the content hierarchy?

01 August 2013 71/99

Oliver Schreck

Legibility

01 August 2013

Font design influences the ability to differentiate “1” and “7”

72/99

Oliver Schreck

Legibility• Text size

• Sized to ensure readability at the maximum expected viewing distances.

• Figure to ground contrast• Text and background should have sufficient contrast to ensure readability.

• Text capitalization:• DO NOT WRITE ALL IN CAPITALS, better use mixed case letters. They are easier to

read.

• Line spacing• Lines of text should be spaced far enough apart, at least a gap of 1 pixel

• Text justification• For most languages should be left-justified with right margin ragged.

• String length

01 August 2013 73/99

Oliver Schreck

Aestheticsa) Use of color: Communicate special meaning and establish an information

hierarchy, color can give medical software a pleasing and appropriate appearance. Moderately saturated colors in a harmonious color palette tend to give software–user interfaces a softened, professional appearance.

b) Use of graphics: Graphics should have a useful function, not simply be decorative. Graphics should be rigorously tested with the end-user population to ensure that they offer a genuine gain in information processing performance.

c) Screen density: Users like screens that use modest amounts of white space (perhaps 20% to 30%) to separate major screen elements.

d) Branding: Logos and similar branding elements should not impede device function. They should not distract or impede users from accomplishing tasks.

01 August 2013 74/99

Oliver Schreck

Data entry• Completeness, accuracy and efficiency•Data entry fields•Use of labels and units of measure• Label placement and appearance

01 August 2013 75/99

Oliver Schreck

Data entry• Data justification

• Text should be left-justified• Numeric values should be right-justified

• Data arrays• Better 20 rows x 2 columns,

then 2 rows x 20 columns

• Automatic fill-in• Use when possible,• Let user verify

• Data validation and checking• User should not be able to make invalid selections• User should not be able to enter invalid data

01 August 2013 76/99

Oliver Schreck

Color codes for medical applications in the USA

01 August 2013 77/99

Oliver Schreck

Color• Non-reliance on color

• ~10% of population have color-vision defects

• Color combinations• Combine foreground and background carefully

• Color associations

• Color customizations• Should prevent the user from change as certain “message” are associated with

the color

• Use color to demarcate or indicate status• Grouping • Change in status e.g. green – red for a battery

01 August 2013 78/99

Oliver Schreck

Trend displaya) Displaying real-time data: Trend displays should

clearly differentiate current values from older (historical) values.

b) Data resolution: Trend data should be presented at a resolution that facilitates interpretation of the data’s meaning. The timeframe could be the last few minutes, hours, or even days, depending on various clinical factors.

c) Timeframe: Trend data should be presented for a period of time that facilitates clinical decisions based on that data. When the timeframe can be adjusted, the means of adjustment should be readily obvious and the timeframe being used should be clearly indicated.

01 August 2013 79/99

Oliver Schreck

Touch screensa) Activation states: Should have a range of appearances to

differentiate when they are unselected, selected but not actuated, and actuated (like a “latching” button).

b) Target size: Should be sufficiently large to facilitate rapid, error-free inputs by individuals with large fingers. A target size of at least 1.5 centimeters (cm) (0.6 inches) is preferred. Oversizing the touch area helps to prevent parallax problems.

c) Target spacing: In general, the centers of touchscreen targets should be spaced 2.0 cm (0.8 inches) apart to help users avoid pressing the wrong target. Other error-prevention methods (e.g., lift-off actuation, strong highlighting) could be employed to counteract the potential for users to touch the wrong target.

01 August 2013 80/99

Oliver Schreck

Touchscreend) Feedback: There should be no perceptible delay between touching an on-

screen target and receiving visual, and possibly audible, feedback.

e) Placement: Touch targets should be placed in the same location on every screen. To allow users to develop “muscle memory,” which can increase reliability and speed, and it will also reduce use errors. If a medical device is likely to be placed at or above the user’s normal line of sight, it is advantageous to locate touch targets on the lower portion of the screen, where they are easier to reach and where users can touch the targets without blocking other portions of the screen with their hands.

01 August 2013 81/99

Oliver Schreck

Touchscreen

01 August 2013

f) Scrolling: Avoid using scrolling lists and slider bars. Difficult to slide a finger across flat surfaces and stop it on a precise spot. Buttons are generally easier to use than a slider for scrolling up and down. A slider can be a good design option to help the user move rapidly through a set of options. A slider also offers the advantage of indicating one’s place within a scrolling list.

82/99

Oliver Schreck

On screen keyboards and keypadsa) Display: Cover significant portions of the screen. That fact is not normally an issue while

users are actively entering data, but it can be an issue in critical situations if the user is distracted and accidentally leaves the tool on the screen. Therefore, these tools should be automatically removed from the screen after some period of inactivity.

b) Data entry: Unless a physical keyboard is available, users should be able to request display of an onscreen keyboard or keypad when they need to enter alphanumeric data. If an application requires frequent entry of alphanumeric data, the on-screen keyboard and keypad should generally remain on-screen (to avoid the need for the user to request them again).

c) Key layout: Keyboards should include number keys except in cases when there is insufficient vertical space on the screen. A QWERTY key arrangement is best suited to medical workers, most of whom are familiar with computer applications requiring the use of a keyboard.

d) Number presets: Touchscreens should present users with an array of preset options rather than require them to scroll or toggle through multiple options. For example, instead of requiring users to increment a value from 10 to 50 by holding down a key, the design can provide 11 keys (from 0 to 100, in increments of 10) and allow users to precisely set the value using arrow keys.

01 August 2013 83/99

Oliver Schreck

On screen keyboards and keypads

d) Audible feedback: A soft “click” or “beep” is helpful when a user touches an on-screen target. Different sounds can be provided for valid versus invalid selections. Users should be allowed to turn off audible feedback, because added noise in clinical environments can be distracting and annoying when quiet is needed or there is already a lot of noise.

e) Actuation: Use errors can be reduced by employing a “lift off” rather than “touch down” method of actuation.

f) Graphical buttons: Rather than placing an icon on a button, it can be visually simpler and more appealing to make a graphic (icon) the actual button. This approach enables one to enlarge the graphic without using up more screen space.

01 August 2013 84/99

Oliver Schreck

User supporta) Pop-up messages: Provide user guidance or to correct a problem (e.g., the user proceeds

with a task before completing a data entry form). Should be kept to the minimum necessary to offer essential guidance and protect against actions that would compromise safety. They can physically overlay information on other screens and be annoying or distracting. Pop-ups should be used at times when a specific user action is required, be placed where they do not overlay important information on other screens, and, when possible, contain all the information needed for users.

b) Conciseness: Information intended to guide the user should be meaningful and concise. It helps to keep paragraphs and sentences short and to use terminology familiar to the user.

c) Differentiation: Supporting information should be visually distinct from primary information so that the user’s eye goes to the primary information first.

d) Style: Prompts, instructions should employ a consistent style.

e) Graphics: In most cases, a graphic can guide users more effectively than words. Graphics should be as simple as possible, focusing the user’s attention on the most important details. Simple drawings can be more effective than photographs.

f) Animations: Animations are sometimes a superior way to guide users. For example, an animation might better illustrate how to calibrate a sensor or replenish fluids.

01 August 2013 85/99

Oliver Schreck

Bad Pop-Up

01 August 2013

All popups are evil but this may be the most annoying one in history.

How ironic that the popup is informing you that IE has blocked a popup.

86/99

Oliver Schreck

Good pop-up

01 August 2013

Popup with the dimmed background is much more intuitive

87/99

Oliver Schreck

Workstations

01 August 2013

Sample of medical workstations88/99

Oliver Schreck

General rule

Adapt the workstation to the user rather than making

the user adapt to the workstation.

01 August 2013 89/99

Oliver Schreck

Safety• Protect users from hazards• Do not overload UI

• Protect against use error• Double check entered values• Analyze potential errors

01 August 2013 90/99

Oliver Schreck

Usability• Avoid excess complexity• Avoid gold plating• Adopt a minimalistic design philosophy

• Allocate functions appropriately to the user versus the workstation• Decide what should be done by user, what by the

machine

01 August 2013 91/99

Oliver Schreck

Usability• Arrange controls correctly

• Task oriented manner, based on analysis of interaction

01 August 2013 92/99

Oliver Schreck

Design guidelinesClearly always indicate mode of operation

a) Mode indication: Workstations should continuously indicate their operational mode and status. They should indicate whether they are in an automatic or manual mode. They should also indicate whether they are in an active (in-use) mode, passive (standby) mode, or turned off.

b) Training mode: Workstations should clearly indicate when they are in a demonstration or training mode.

c) Default mode: Automatic functions, particularly those that are life-critical, should default to a safe operating mode in the event of a component failure. The workstation should immediately alert the user to the mode change by initiating a suitably high-priority alarm condition.

d) Mode changes: Workstations should alert users immediately to any mode changes when an awareness of the current operational mode is critical to maintaining situational awareness or being prepared to act quickly and effectively in an emergency.

01 August 2013 93/99

Oliver Schreck

Indication of selected patient mode

01 August 2013 94/99

Oliver Schreck

Prevention of use errorsa) Workstations should incorporate features that prevent critical use errors (e.g., interlocks,

confirmation requests, and physical guards).

b) Workstations should require users to confirm critical and irreversible machine functions, giving users time to detect and correct slips and mistakes that could waste time, waste resources (e.g., a tubing set), cause property damage, and possibly harm the user or patient.

c) Workstations should preclude dangerous settings, such as high ventilator pressures or high radiation dose levels, or at least require their confirmation before the setting takes effect.

d) Some automatic functions, such as a moving boom on a C-arm fluoroscopy machine, could pose a hazard if activated at the wrong moment during therapy. Workstations should incorporate guards against the inadvertent activation (enabling) of such functions.

e) Workstations should conspicuously indicate when user input or intervention is needed. This design feature will help expedite tasks and avoid errors of omission, such as failing to restart a pump that is delivering a critical medication to a patient.

f) If a user fails to act in a timely or appropriate manner, workstations should automatically perform the functions necessary to ensure the safety of the user and patient.

g) Workstations should provide patients and users with multiple layers of protection against potential hazards. This could include two or more of the following: a mechanical interlock, a protective cover, an access code, a visual-alarm signal, a printed warning, and user training.

h) Users should be able to override automatic functions except for those associated with critical protection systems. This approach grants users ultimate control and accommodates unanticipated circumstances and needs.

01 August 2013 95/99

Oliver Schreck 98

Thanks!

01 August 2013