Guidance for Designing and Purchasing Accessible Medical Instrumentation

General Guidance 1.5: Accessible Device Controls and Displays, Including Alarm Systems

Overview

This general guidance addresses the design of accessible device controls and displays, which are often critical components of medical devices. The specific guidance provided here is based on the expertise and experience of RERC-AMI staff as well as sections of a national standard for medical devices that was developed by AAMI’s Human Factors Engineering Committee (ANSI/AAMI, 2008).

This guidance is intended to promote the design of accessible control and display interfaces for medical devices that can be used effectively by users with diverse abilities and disabilities. Controls are defined as any device component used by operators for altering some function of the device (ANSI/AAMI, 2008). The direction of information flow of a control is from the user to the device. Operations for individual controls can be classified as discrete and continuous, with different classes of alternatives for each. Examples of discrete controls include: push buttons, toggle switches, some types of rotary knobs, levers, and rocker switches. Examples of controls commonly used to operate continuous (analog) signals include wheels, pedals, sliders and cranks.

For displays, the direction of information flow is the opposite of controls, i.e. from the device to the user. Displays can take many forms, including visual, auditory and tactile. As denoted in General Guidance 1.1.1 to 1.1.10, there are advantages to multimodal displays. Thus, the manner in which information on the state of a control is communicated to the user can be viewed as a display of information. Examples of displays include labels on a device, cathode ray tube (CRT) displays, projection systems, LEDs (light-emitting diodes), audio tones and/or synthesized speech from a device, Braille, small embedded monitor screens and larger computer monitors.

The interface for most devices requiring operation include both control and display components, integrated in a way that ideally enables safe and effective use by all possible users who could benefit from access to the device’s functionality. However, the RERC-AMI has identified many access barriers for different types of medical products. Examples of these devices range from medical vital sign monitors to infusion pumps to glucometers to ergometers to powered examination chairs and tables. Alarms systems are also an important part of many medical devices, and represent a type of display that often also include a control component (e.g., turning off an alarm).

Because controls and displays often are used to perform critical device functions, it is especially important to consider users with disabilities and accessibility features when designing such device features. Because alarms often are used to alert users of critical information or conditions needing attention from a medical device (e.g., system failures, unsafe conditions), it is especially important to consider users with disabilities and accessibility features when designing alarm systems. Of note is that in part because of the societal emphasis on alarm systems that are accessible (e.g., building codes), medical alarms generally tend to be accessible, and often are the only part of medical device interfaces that are inherently multimodal and that follow the type of guidance provided in General Guidance 1.1.

The following guidance can be viewed as an augment to other RERC-AMI general accessibility guidance for medical devices, and it is especially relevant for devices with complex control-display interfaces. While control-display features must often be integrated (e.g., display of information near and/or on the associated control), guidance will be provided in separate subsections: controls, displays, and alarms. This guidance is not intended to be exhaustive, but rather to hit on some key points that relate to accessible interfaces. Alarms can be viewed as a special case of integrated control-display systems that are often exemplary in the degree of accessibility; there is a special section on alarms, which are an example of integrated control-display systems. For alarm systems, the AAMI guidance defines three main responsibilities: (1) to accurately and reliably detect dangerous conditions, (2) to immediately obtain the operator’s attention, and (3) to clearly and accurately tell the operator what the problem is and, if feasible, what to do in response to the problem. Other resources to consult for guidance in designing alarm systems include: IEC 60601-1-8: 2004, AAMI HE-48: 1993 and Neville Stanton’s Human Factors in Alarm Design: 1994.

This guidance is divided into three subsections: controls, displays, and alarms. It is not intended to be exhaustive, but rather hits on some key points that relate to accessibility. Alarms can be viewed as a special case of integrated control-display systems that may be exemplary in the degree of accessibility.

Guidance on Controls (1.5.1 to 1.5.11)

1.5.1 Users with disabilities should be included in the testing of controls.

Discussion: As seen in Table 1, there are many options available for both discrete and continuous controls. One might think there is a small collection of “best” controls for meeting the needs of users with diverse abilities, but the reality is that there are many alternatives and different classes of devices can be expected to have different optimal strategies. There has been limited testing of the various types of controls for persons with disabilities. While Table 1 provides some accessibility strengths and limitations of various controls that can be used on medical devices, the reality is that there remains a need for more research on various controls for diverse users. Thus this general principle (see also 1.1.X) is reasserted here when designing collections of controls for a given interface, as is often the case. This is especially important when designing medical devices used by lay users in the home environment, but is applicable to all devices with controls. This guidance builds on basic principles for using users with diverse abilities to design accessible products, as described in several chapters of the AMI book edited by Winters and Story (e.g., Story, Mueller, Smith et al., Winters et al.).

Table 1. Accessibility Advantages of Continuous and Discrete Controls

Control Type

Accessibility Advantages

Accessibility Disadvantages

Pushbuttons

Can be activated with various body parts of assistive technologies (e.g., finger, palm, pointer stick)

Requires a force component in a single direction, can be too small, low friction, or low deflection

Toggle Switches

Does not require fine motor control, can be activated with various body parts

Requires a force component in a certain direction

Continuous Thumb Wheels

Often can rest hand/arm during use

Requires fine motor control

Discrete Thumb Wheels

Often can rest hand/arm during use

Requires fine motor control

Rotary Knobs

None

Requires fine motor control and simultaneous actions (e.g., grasping and turning)

Levers

Requires minimal fine motor control

Can be accidentally activated, requires force in a certain direction

Rocker Switches

Can be activated with various body parts or assistive technologies (e.g., finger, palm, pointer stick)

 

Sliders

Good visual and tactile feedback of control state

Requires some fine motor control, including sliding in a single direction

Key-operated Controls

None

Requires simultaneous grasping and turning motions

Membrane Controls

If tactile feedback, provides systematic grid of controls, integrated signage

Minimal tactile feedback for control placement and operation

Pedals

Provides means for using feet for discrete or continuous controls

Requires coordinated leg movement

Large Levers

Good visual and tactile feedback of control state

May require larger forces/motions to operate, can be accidentally activated

Wheels

Bimanual or either hand can be an advantage

May require larger forces/motions to operate

Palm Buttons

Can be activated with various body parts or assistive technologies (e.g., finger, palm, pointer stick)

 

1.5.2 The force required to activate controls should be as low as possible, but high enough to provide tactile feedback and prevent inadvertent activation.

Discussion: Although low-force controls are advantageous for people with limited strength, it is important to require enough force, typically around 0.7 pounds for finger-operated controls, to prevent users from moving the control farther than intended as well as to provide adequate feedback. Overshoot is a bigger problem for people with limited control of hand movement. For rotary controls, a torque of 1 in-lb is appropriate for preventing overshoot. See also Guidance 1.1.9 on designing for limitations in strength or stamina, and Guidance 1.1.25 on anticipating limitations in mechanically operated controls. Motivated in part by AAMI/ANSI HE-75: Controls-16.2.1.2.1.

1.5.3 Controls that use a pushing motion are preferred over rotating controls for people with disabilities and elderly users.

Discussion: Rotating controls require users to make more refined movements (e.g., the ability to create a twisting moment, normally requiring the use of fingers and a thumb, and wrist pronation-supination) than push type controls (e.g., which are commonly activated with various user motions). See also Guidance 1.1.21 for further reasons for avoiding rotary controls. Motivated in part by AAMI/ANSI HE-75: Controls-16.2.1.2.1.

1.5.4 Controls should have redundant feedback mechanisms with different sensory modalities.

Discussion: Controls that provide visual, auditory and tactile feedback concurrently are more accessible for users with disabilities, and as noted in Guidance 1.1.19, can provide performance advantages. Redundant feedback of a change in state of a control is generally advantageous, especially for discrete controls, and there are often simple strategies for incorporating such features. An audible noise (such as a click or snap) should be used for auditory feedback of control activation, and controls should have tactile features that help capture the state of the control. It is important for the feedback to occur soon after activation of a change in state to prevent users from activating a control more than once. Note that such feedback is a simple form of a display, and principles within guidance below for displays should also be considered, as appropriate. See also Guidance 1.1.19 on performance advantages of adding audio redundancy. Motivated in part by AAMI/ANSI HE-75: Controls-16.2.1.2.2.

1.5.5 Controls should have adequate space surrounding frequently used controls so users can rest their fingers or palms when activating the control, especially for horizontally oriented displays.

Discussion: Users should be able to use various grips to operate a control, and users should also be able to use assistive devices such as pointing sticks to operate controls. Users should not have to assume awkward postures to operate a control. For frequently used controls, there are also advantages to horizontal interface panels that plan for resting locations for the user’s fingers or palms or wrists. Motivated in part by AAMI/ANSI HE-75: Controls-16.2.1.2.3.

1.5.6 Controls should be labeled with icons and/or text.

Discussion: Some icons can be more universal than text labeling because many symbols can be understood across several languages. Text labels are most effective if they are placed in the center of a control with the font size as large as possible, with text-plus-icon often more effective than either alone. The best location when placement directly on the control is prohibitive is a location unlikely to be blocked by hand operation, such as directly above the control. When controls are software-based, see also Guidance 1.1.14 on providing text descriptions. Motivated in part by AAMI/ANSI HE-75: Controls-16.2.1.2.4.

1.5.7 Concave surfaces with surface texture should be used on controls requiring pushing to help users keep their fingers from sliding off the control surface.

Discussion: Concave surfaces with a small convex radius around the edges can help keep a finger from sliding off the control. Surface textures increase the friction of the control surface. In some cases tactile feedback in the form of concave dimples can help orient the user to the specific control (or be used to provide Braille). See also Guidance 1.1.23 on integrating in tactile features to make an interface more accessible. Motivated in part by AAMI/ANSI HE-75: Controls-16.3.1.1.1.

1.5.8 Rotary knobs and toggle switches should have increments/travel distance of at least 30 degrees so that control activation is easier for users with disabilities.

Discussion: This is especially helpful for users with manual or visual limitations in function. However, of note is that rotary knobs are best avoided (see Guidance 1.5.3). Motivated in part by AAMI/ANSI HE-75: Controls-16.3.1.3.5.

1.5.9 Controls should be mounted on a vertical or horizontal surface, or on a diagonal surface that is adjustable. For controls on interfaces that require sustained use, the control should be mounted on a horizontal surface.

Discussion: Diagonal surface orientations can be difficult or impossible for some elderly users and arthritis-impaired users because such orientations may require uncomfortable angular flexion or extension of the elbow and shoulder. However, surface orientations that can be easily adjusted by the user to a personal preference are encouraged. For controls that require sustained motions for more than a few seconds, the control should be mounted on a horizontal surface, or one slightly angled from horizontal towards the user (e.g., like common computer keyboards), and accommodations should be made for resting the arm, Motivated in part by AAMI/ANSI HE-75: Controls-16.3.1.6.4.

1.5.10 Tactile control/display cues should be provided on control surfaces, especially membrane control surfaces, to make it easier for users to navigate the surface.

Discussion: This is especially important for users with visual impairment. For such users, ridges, raised surfaces, or texture changes are critical for indicating the shape of buttons. Shape coding can also be used redundantly with color, position, texture, or labeling to help users locate and differentiate different controls. See also Guidance 1.1.23 on strategies for integrating tactile features into the interface surface. Motivated in part by AAMI/ANSI HE-75: Controls-16.3.1.10.2.

1.5.11 Hands free alternative controls such as eye tracking, sip and puff, mouth sticks, head controls, and voice controls, should be available when possible for users with disabilities or other users who need to multi-task.

Discussion: Hands free controls are especially important for users with impairments to their hands or other extremities that make it difficult or impossible to use a manual control interface. See also Guidance 1.1.8 on minimizing need for fine motor control and simultaneous actions, and Guidance 1.1.23 on not requiring a timed response). Motivated in part by AAMI/ANSI HE-75: Controls-16.3.2.8.7.

Guidance on Displays (1.5.12 to 1.5.15)

Displays take many forms, including visual, auditory and tactile. The purpose of display components depends on the context, and can range from displays that are directly associated with controls to informational displays that provide information about health status. The effectiveness of a display is affected by factors such as how well it supports the user’s task requirements, how well it matches the user’s characteristics and capabilities, and how compatible it is with the demands of the use environment.
The core guidance for displays are provided in the multimodal section, including Guidance 1.1.1 to 1.1.6 for accommodating various sensory modes, Guidance 1.1.14 and 1.1.16 on having text descriptions for software displays and captioning, Guidance 1.1.15 on making use of standard computer monitors as displays when possible, 1.1.17 on the benefits of using standards-compliant multimedia formats whenever possible, and 1.18 through 1.1.20 on the benefits of and strategies for providing an alternative audio display. This section is very selective, and only hits a few high points that relate directly to accessibility.

1.5.12 Displays should accommodate large ranges of use postures for the user relative to the device.

Discussion: It is common to assume that a user can locate and orient their torso and head into a certain posture, but for persons with disabilities, this should not be assumed. For instance, it is common to design a display so that it accommodates population “extremes” of 5% to 95%. However, not only do people come in a variety of anthropomorphic sizes and shapes, but many also have limitations in the postures that are possible, including the position and orientation of their head. While this can obviously affect the user’s operation of controls, it is also important to recognize that the distance and orientation of the user from the device also affects the display. For instance, information should be retrievable from a large range of angles, and when feasible, it should be possible to reorient (e.g., rotate) the displayed information for easier viewing. For general concepts that are applicable here, see Guidance 1.1.9 on considerations for limited reach, strength or stamina, 1.1.10 on positioning and orienting for full access, and 1.1.24 on reach considerations for seated persons and users of wheelchairs and scooters.

For visual displays, an example of a type of display that is to be avoided is a passive-matrix liquid crystal display, because of its narrow angle of effective view when compared to alternative visual display technologies. This guidance also applies to audio and tactile displays.

An effective display should also minimize the effects of varying environmental conditions, such as lighting conditions, audio noise, and distractions that can be anticipated such as background movement of persons. Evaluation of displays should consider “worst-case” environments.

1.5.13 For visual displays, temporal video byproducts such as flicker and jitter should be minimized, and indeed through use of modern technology should no longer provide challenges for persons with disabilities.

Discussion: Both flicker (rapid fluctuations in brightness levels) and jitter (fluctuations in screen images) can cause challenges for persons with disabilities. Flicker perception limits curves as a function of luminance suggest that a refresh rate of 70-75 Hz, which is higher than what is used for main-stream TV and video (60 Hz), is recommended. Jitter is most visible at frequencies under 3 Hz, where the perception of speed can affect smooth pursuit tracking eye movements (whether desirable or not); at higher frequencies it is perceived as imaging blurring. The BSR/HFES 100:2002 standard suggests no more than 2 mm of jitter per cm of design viewing distance (or 12 deg of visual angle), in the frequency range of 0.5 to 30 Hz. However, given that the design viewing distance may vary (sometimes involuntarily), it is recommended that there be no more than 1 mm per cm of “design viewing distance” (or 6 deg). Of note is that because technologies are readily available that easily meet these flicker and jitter criteria, there is no reason that this should ever be a problem for medical devices. Motivated in part by AAMI/ANSI HE-75: Displays-17.4.3.

1.5.14 For visual displays, adequate luminance, contrast and use of color characteristics should be designed into the presentation of content on the display.

Discussion: Luminance refers to the light intensity emitted from a display, and relates to the perceptual “brightness” of a display. The AAMI/ANSI HE-75 standard uses 35 candelas per square meter (cd/m^2) as the minimum, and for displays that need to accommodate high ambient lighting conditions, 100 cd/m^2 (see also ISO 92413:1992). To accommodate persons with visual limitations, it is important that all displays have the capacity for 100 cd/m2. Furthermore, it is suggested that when feasible, there be a luminance control for adjusting the setting. The standard minimum contrast intensity of a display (whether color or monochrome) for all anticipated viewing angles should be no less than 7:1. Colors can be used as one means of providing information, but as noted in Guidance 1.1.3, there must be a display mode that does not depend on color vision. Motivated in part by AAMI/ANSI HE-75: Displays-17.4.4.

1.5.15 The character height of text on displays should be of adequate size, or provide accommodations for users’ assistive technologies.

Discussion: Guidance 1.1.2 states that there should be least one mode of operation and information retrieval that does not require visual acuity greater than 20/70. There has been a tendency for character heights on products to be too small, which is problematic for this population of users, and sometimes other users. Current human factors guidance has the designer estimate the desired location for the user, and then select the optimal character height based on visual arcs. The commonly preferred height of characters is about 20-22 minutes of arc, with less important information being as low as 16 minutes of arc and essential readouts being 24-30 minutes of arc. Even this larger size can be difficult to see for persons with visual acuity of 20/70, but the reality is that many of these persons can gain access by moving closer to the monitor or using a magnifier. This guidance recommends that each of these shift to a larger size. Thus, the standard size should be about 24-30 minutes of arc, smaller information should be about 20-22 minutes of arc, and essential readouts should be 30-40 minutes of arc. This is especially important for users in the home environment. Finally, it is recommended that essential information also be available in an alternative display mode, such as auditory or tactile. Motivated in part by AAMI/ANSI HE-75: Displays-17.4.5.1.2.

Guidance on Alarm Systems (1.5.16 to 1.5.22)

Alarm systems that are associated with medical devices typically include control and display components. Display components are commonly multi-modal, due to the nature of alarms, and as such, in many cases alarms are one of the more accessible components of medical devices. There is considerable guidance for alarms in buildings (e.g., ADAAG 4.28) and medical devices (e.g., AAMI/ANSI HE 75: Alarms, Chapter 14), and this guidance provides key highlights that should be considered in designing accessible alarm systems.

1.5.16 This is especially important when designing medical devices used by lay users in the home environment. Motivated in part by AAMI/ANSI HE-75: Alarm Design-13.3.10.

Discussion: This is especially important when designing medical devices used by lay users in the home environment. Motivated in part by AAMI/ANSI HE-75: Alarm Design-13.3.10.

1.5.17 The alarm signal should be designed to be compatible with the full range of human perceptual/cognitive capabilities.

Discussion: When designing accessible alarm systems, it is especially important to consider users with diverse abilities. There is a long history of guidance in this area, for instance alarm systems for buildings, that are helpful (e.g., ADAAG 4.28). Disabilities to consider include cognitive, visual, and hearing impairments such as limited memory, blindness, low vision, and deafness. Available signaling modalities include auditory, visual, olfactory, vibratory, and other tactile/haptic modes of stimulation. Alarm systems should have, at minimum, both visual and auditory modes, for example a common design practice is to get the user’s attention with an auditory alarm signal and convey specific information about an alarm condition with a visual signal. Furthermore, both modes should be of sufficient magnitude to get the attention of the persons who need to be aware of the alarm. Motivated in part by AAMI/ANSI HE-75: Alarm Design-13.2.1.

1.5.18 Designers should assume that users of medical devices may have a physical limitation such as impaired hearing or vision when selecting alarm-signaling modalities, unless there is strong evidence to the contrary for the full population of possible users.

Discussion: Users with vision impairment may get specific information via auditory signals or tactilely by reading Braille. It is not sufficient for users with hearing impairment to use an auditory alarm signal as their attention-getting signal, and thus a visual signal (e.g., flashing light) must be provided as well. Motivated in part by AAMI/ANSI HE-75: Alarm Design-13.3.5.

1.5.19 Alarms should be perceivable by users with some hearing loss.

Discussion: Hearing loss frequently occurs with normal aging, so it is important to assure that auditory-alarm signals will be perceivable by users with hearing loss. It is known that people with hearing loss tend to lose it from the “top down” so that higher frequencies are inaudible, so higher frequencies should be avoided for alarm signals. Visual alarm signals can be used to address the needs of users with hearing loss, as well as users in noisy environments. Visual alarm signals are also more position-specific than auditory alarm signals. Motivated in part by AAMI/ANSI HE-75: Alarm Design-13.2.1.1; 13.3.8.

1.5.20 For auditory alarm signals, a fundamental frequency as well as at least 3 additional frequencies should be used.

Discussion: If more than one frequency is included in an alarm signal, if users have a frequency-specific hearing loss then they will still be able to hear the auditory signal. Motivated in part by AAMI/ANSI HE-75: Alarm Design-13.3.7.1.1.1.

1.5.21 Designers should minimize the number of false alarms generated by an alarm system.

Discussion: There are many medical devices that emit a large number of false alarms, and it is especially important to minimize false alarms when designing for users with disabilities because of various cognitive and physical limitations that may make users physically or cognitively unable to deal with false alarm conditions. Ideally designed alarm systems have both perfect sensitivity and specificity. A top goal for alarm designs is to minimize the likelihood of false alarm signals. Motivated in part by AAMI/ANSI HE-75: Alarm Design-13.2.2.

1.5.22 Alarm signals should be adjustable (within reason) with the option to store user settings. The device should clearly indicate when the limits have been adjusted from the factory settings.

Discussion: Many people with disabilities rely heavily upon adjustability options, and this is especially important for critical alarm signals. Volume and alarm signal alarm modalities are two adjustments that should be considered when designing alarm systems. However, they must still be capable of fulfilling their basic function of being an effective alarm. The control(s) used to adjust the alarm system should follow the accessibility principles outlined in this section. Motivated in part by AAMI/ANSI HE-75: Alarm Design-13.3.3.1.

[ Old Home | Home | Overview | Partners | Advisors | Projects | Training | Links ]
[ R1: Needs | R2: Usability | R3: Metrics | D1: Tools | D2: Design | D3: Emerging ]
[ All Content © 2005 RERC-AMI | Credits ]
Valid XHTML 1.0! Level Triple-A conformance icon, W3C-WAI Web Content Accessibility Guidelines 1.0