Home Health Care Tasks and Tools
HOME CAREGIVING TASKS1
Home caregiving tasks are extremely diverse, including help with activities of daily living, transportation, interaction with medical personnel and a care recipient’s family or social group, use of medical devices, negotiation with insurance carriers, and use of the Internet and other information sources, said Colin Drury. These tasks call on the physical, cognitive, social, and emotional abilities of caregivers. Relating these task-derived demands to actual caregiver capabilities is one aspect of the discipline known as task analysis.
The errors committed during the delivery of home health care can range from the trivial to the deadly, said Drury. Because the task demands made of home caregivers can exceed human capabilities, these tasks need to be carefully analyzed.
Care recipients and care providers are extremely diverse, and all are under more stresses than in the past. Many people, including home care providers, are working harder than they have in the past, Drury said. “There are more people doing a lot of small part-time jobs, and there are people working large hours of overtime on one job. The good old 40-hour week … is disappearing.”
A standard finding from human factors research is that, as task demands exceed human capabilities, error rates increase. Errors may occur occasion-
ally when a caregiver is distracted, sick, or otherwise incapacitated. But as the diversity of tasks and the diversity of people increase, the potential for errors grows.
To bring task demands in line with capabilities, there are five things that can be changed:
the person providing or receiving care,
the technology being used,
the environment surrounding the task, and
the social system surrounding the task.
Making changes in these five areas implies better procedures, better training, better equipment design, better home environment design, and better social interactions. In this context, “better” means fitting tasks to people, Drury said.
Task analysis has two parts. First, assessing the demands of a task requires a task description—a detailed and hierarchical breakdown of every step involved in the task. Second, assessing human capabilities draws on the literature on human factors plus contributions from other disciplines, such as psychology and biomechanics, supplemented with professional judgment.
In addition, there are two methods of understanding tasks, and both are needed for error-proof designs. The first is to analyze errors or system failures in existing systems, as was done in the 2000 report on medical errors by the Institute of Medicine, To Err Is Human: Building a Safer Health System, National Academy Press, Washington, DC. The second is to analyze the functioning of a system, starting with its objectives and then focusing on the task elements for existing and proposed systems.
A common pitfall of task analysis is to assume that everyone looks like you, Drury said. Avoiding this false assumption requires that the people performing the task be involved along with someone who can integrate the various tasks being analyzed. As an example, Drury cited the transport of a care recipient, whether from the bed to a chair, from the home to a hospital, or from the hospital to a care facility. Each of these overall tasks requires planning to tie its constituent tasks together. The job is simple if the constituent tasks are lined up in a logical and linear order. “You just go down the checklist and you do them. But lots of them have branches. If it says this, you do this. It may not be that. This may be blocked. You may have to do something else.” Because of this complexity, task analyses generally involve multiple levels of detail.
Drury also drew an analogy with task analysis in aviation, in which he has done considerable work. The task of inspecting the safety of airplanes has many built-in safeguards, Drury said. It often involves both humans and technologies. It is designed to discover possible errors and identify good practices that lead to error reduction. It leads to advice for the people running the systems—something that “adds to your knowledge, not just your rule base.”
Drury listed some important components of task analysis.
First, task analyses begin by specifying what has to be done rather than focusing on specifically who does what, since different parts of a task can be done by different individuals or even by technology. Once tasks have been identified, the appropriate person or technology to do each part of a task can be identified (task allocation).
Task analyses are the basis for design recommendations in the form of good practices for general use and specific design changes for specific tasks. There is a well-developed methodology for task analysis that can be adapted for home caregiving, and other domains also demonstrate how to format the results for maximum impact and how to use the results in a design or redesign context.
Task analysis needs both human factors practitioners and subject matter experts to be successful. “You need people who have been doing the job. You need the potential users and the real users.”
Finally, for any new equipment or procedure, a task analysis is the start of the design to reduce future user errors. These analyses are best performed by a team with knowledge of both human factors and the subject matter.
Responses to Questions
In response to a question about how to make task analysis a more standard practice in home health care, Drury replied that the practitioners of task analysis need to describe its benefits to the individuals who are in a position to use it. Technology manufacturers and designers of the built environment in particular can be approached, as well as the professional organizations that support home health care. “You need to go to those communities and say that this ought to be done.”
Task analysis in health care is somewhat different than in many other areas because of the communications, social, and emotional issues involved. Greater understanding of the dimensions of human capabilities is needed, such as the human subsystems likely to be overstressed by caregiving.
In response to a question about the kinds of tasks involved in home health care, Drury cited the high-level division of physical/cognitive needs, psychological/emotional needs, and social needs. “That may not be an exhaustive list…. But that struck me, from what I have read, as an appro-
priate starting point.” Drury also observed that a taxonomy of the kinds of errors that occur in home health care could be a useful precursor to a more generic task analysis.
MEDICAL DEVICES AND EQUIPMENT2
According to the Center for Devices and Radiological Health of the Food and Drug Administration (FDA), a medical device is “an instrument, apparatus, implement, machine, contrivance, implant, or in vitro reagent or other similar article that is … intended for use in the diagnosis of disease or other conditions, or in the care, mitigation, treatment, or prevention of disease.” Similarly, the Home Health Committee of the center has defined a home medical device as “a device intended for use in a nonclinical or transitory environment [that] is managed partly or wholly by the user, requires adequate labeling for the user, and may require training for the user by a health care professional in order to be used safely and effectively.”
These definitions lay out the three dimensions that must be considered in applying human factors research to the design of home medical devices: (1) the device itself, (2) the people who use it, and (3) the environment in which it is used, said Molly Story. These dimensions, in turn, become more complex as the complexity of medical devices used in the home increases. Today, such devices as ventilators, infusion pumps, and dialysis machines are frequently being used outside the hospital or clinic, often by lay users, even though many of these devices were not designed for and were not specifically labeled for this use, Story said.
Devices used in the home are not always the same models used in health care facilities. They may be older or lower in quality. Professionals who encounter them in the home or in a clinic may not be familiar with their operation. Consumers are giving these devices to each other and are selling and buying them on the Internet. Such devices are less likely to be appropriate, to be properly operated or maintained, or even to come with complete instructions.
Many different people use medical devices in the home, including physicians, nurses, nurse practitioners, various therapists, workers, home care aides, independent contractors, family members, friends, neighbors, care recipients, or even someone who gets pulled in from the street in an emergency. These users may be of any age, may have various kinds of disabilities, or may be sick themselves. A person’s ability to use a home health device depends on many factors, including
This section is based on the presentation by Molly Story, president of Human Spectrum Design. For more information and for references to the information cited in this presentation, see Chapter 8.
physical capabilities, such as their size, strength, stamina, dexterity, flexibility, and coordination;
sensory capabilities, including not only vision and hearing but also sometimes touch;
cognitive abilities, including their memory, literacy, language skills, knowledge, and experience base;
mental and emotional state;
personal history and experience with home health care and medical care in general; and
ability and willingness to learn how to use new devices and adapt to having new devices in the home.
Many environmental factors also affect a person’s ability to use a medical device. Space issues can be very important, especially if there are obstacles in a home or if the device needs to be moved. Floor surfaces can make a difference, such as wood versus carpeting. Lighting, noise levels, temperature, and humidity can be very high or very low. “All of these can make devices misbehave,” Story said.
The activity level in the environment can be confusing and can conflict with the operation of a device. The environment may not be clean. Animals—pets, service animals, vermin—can affect devices. Electromagnetic interference can come from other devices in the environment, such as computer gear or videogames. “You have heard that beeping that your cell phone makes on the radio when you are in the car? It does the same to your medical devices in the home.” In addition, the electrical power may go out for a variety of reasons and an emergency backup system may be needed, especially if a device is keeping a person alive.
If a device needs to move into and out of a home, other questions arise. How portable is the device? What does it weigh? Does it have wheels? Does it have a handle? Is it discreet? If someone sees it fall out of your pocket, will you be embarrassed? How long is the battery going to last? Device durability and ruggedness are also factors when a device is taken out of a home or clinic.
Taxonomy of Home Medical Devices
Story has developed a 12-category taxonomy of home health care devices:
Medication administration equipment, such as syringes, cups, eye-droppers, sprays, patches, and syringes.
Test kits, from pregnancy and allergy kits to cholesterol and hormone tests.
First aid equipment, such as bandages, traction equipment, ostomy care, and defibrillators.
Assistive technologies, such as glasses, hearing aids, prostheses, orthotics, crutches, wheelchairs, and mobility aids.
Durable medical equipment, including beds, specialized mattresses, specialized chairs, lift equipment that may be either ceiling-mounted or portable, commodes, urinals, and bedpans.
Meters and monitors, such as thermometers, blood glucose meters, electrocardiogram monitors, and fetal monitors.
Treatment and therapy equipment, such as infusion pumps, dialysis equipment, transcutaneous electrical nerve stimulation equipment, and intravenous equipment.
Respiratory equipment, such as ventilators, forced airway devices, oxygen, masks, and suction.
Feeding equipment, such as feeding tubes and food pumps.
Voiding equipment, catheters, and colostomy gear.
Infant care equipment, such as incubators, warmers, bilirubin lights, and apnea monitors.
Telehealth equipment, such as cameras, sensors, and computers.
More technologies will move into the home in the future. Telehealth, in particular, is expected to grow vigorously in the coming years. For example, wireless technologies offer continuous monitoring and a greater range of mobility for care recipients. Remote monitoring allows for long-term monitoring, encourages adherence to treatment regimens, and provides for reminder alerts to perform certain acts, such as taking medication or scheduling an appointment.
Future technological advances will bring new types of medical devices into the home, like improved pacemakers, cochlear implants, corneal implants, and artificial retinas. Nanotechnology will be embedded into devices, allowing for much more sophisticated biosensing. Smart fabrics will detect events happening in the body. Heads-up displays with pattern recognition software will help people with vision impairments or cognitive impairments recognize objects and faces. Skin surface mapping can keep track of things like moles on the skin to see if they are changing. Other types of biosensors will be embedded in all kinds of familiar objects, such as toothbrushes. And many other kinds of devices are on the way, including “things that we can’t yet imagine,” said Story.
Good designers of medical devices understand the needs of both average users and users who have capabilities far from the average. Device designers also need to give attention to the positive or negative aspects of using the device and the potential individualization of the device. “Once people’s needs for safety, functionality, and usability are satisfied, designers should address their needs for pleasure and self-actualization.”
These considerations are factors in the concept known as universal design, which has seven basic principles:
Equitable use, so that everyone can use the same device. “Just as we are not going to have the accessible MRI machine and the regular MRI machine, the same should hold true for all home health care devices as well.”
Flexibility in use, so that the design accommodates the full range of individual preferences and abilities. “We need to accommodate individual operational styles, as well as learning styles, such as using things left-handed.”
Simple and intuitive use, so that the design is easy to understand regardless of the user’s experience, knowledge, language skills, or current concentration level. “Keep it simple. Remember that not everyone reads or understands English.”
Perceptible information, so that the design communicates necessary information effectively to the user, regardless of ambient conditions or the user’s sensory abilities. “Everything that is visible on the device also should be auditory—and vice versa.”
Tolerance for error, so that the design minimizes hazards and the adverse consequences of accidental or unintended actions. “We need to minimize the risk of injury to both the user and the device.”
Low physical effort, so that the device can be used efficiently, comfortably, and with a minimum of fatigue. “It needs not to wear you out just to turn it on.”
Size and space for approach and use, regardless of the user’s body size, posture, or mobility. “There has to be sufficient space available for whatever body parts may be involved, as well as whatever assistive technologies—wheelchairs, crutches, service dogs, or personal assistants—may be present.”
Less tangible factors may also come into play. Users may have powerful emotions knowing that they or their loved ones are seriously ill. They may be overwhelmed by the critical new responsibilities they have had to take
on. They may be acutely aware of the potential for harm to the equipment, to their loved ones, or to themselves. They may be confused by the new terminology that they have to master in a hurry. They may be confused by the care instructions and the device instructions. They may not have the personal or institutional support that they need.
National and international standards play an important role in medical device development. A U.S. human factors engineering process standard, referred to as ANSI/AAMI HE74 and published in 2001, is for use in fulfilling user interface design requirements in the development of medical devices and systems, including hardware, software, and documentation. An international human factors engineering process standard, referred to as ISO/IEC 62366 and published in 2007, specifies a process for a manufacturer to analyze, specify, design, verify, and validate usability as it relates to the safety of a medical device.
In addition, a guidance document published by the FDA in 2000, Medical Device Use-Safety: Incorporating Human Factors Engineering into Risk Management (see http://www.fda.gov/downloads/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm094461.pdf [accessed August 2010]), describes how the agency wants hazards related to medical device use to be addressed during device development, noting that they should be addressed in the context of a thorough understanding of how a device will be used.
Finally, a committee with which Story has been involved is working on the standard ANSI/AAMI HE75, scheduled to be released in 2010, which provides detailed human factors engineering design guidance to those who are responsible for human factors engineering work in medical device companies.
These documents provide information, guidance, and models of best practices to designers and manufacturers. They also enable manufacturers to show that they are aware of the processes in the guidance and that they have followed them. “Standards are helpful,” said Story, “but you still have to know what you are doing with them.”
Instructions and Training
Device labeling and user instructions are important for home health care, and Story said they get too little attention. These resources for users include the packaging, the graphics and text on the box, the printed instructions, user manuals, quick-start guides, user brochures, leaflets, advertise-
ments, and all other forms of information, including video and audio files that may be offered on DVD or on Internet websites. “These things have to be written for lay users,” said Story. “They are too often written for health care professionals—that is, to the education and knowledge levels of people who know about medical technology in general and the subject device in particular.” Written procedures and diagrams need to be user-tested and offered in alternative modes and formats, not just print, because not everybody can read print. “Put it on a disk. Put it on the Internet. Even if the person doesn’t have Internet access themselves, they may know someone who can get it for them.”
Training for home users may have deficiencies, including being presented too quickly, using jargon, not providing enough practice for the user, or not providing enough explanation of the problems that may arise if the required steps are not done correctly. Training has to be designed for lay users and needs to be available in multiple modes. “Hands-on training is far and away the best way to do this. Have people practice using the device—there is no substitute.” A lot of people may use a device just occasionally, so designers need to minimize the need for long-term memory.
Information needs to be provided where and when it is needed. “Stick it to the device itself. Embed it in the user interface. Don’t make me go find the manual. I have no idea where it is.” And some form of user support should be available 24 hours a day, 365 days a year.
Voice output in a device offers many benefits. It reinforces the visual messages. It reduces misinterpretation of visual information. It is especially helpful for infrequent users. It improves user confidence and trust in the device. It reduces the burden of customer support for health care professionals. And it is vitally important for people with vision impairments.
Many different human factors methods can be used to assess device safety, functionality, and usability, including task analysis, risk assessment of potential errors and their consequences, evaluation by a group of testers against a set of heuristics or general principles, expert review, and formative and summative user testing. “It’s really important to identify the people who are at highest risk. Those are the people you need to be testing on the risk-critical tasks that are identified through your task analysis. By doing that, you can identify the sources and the nature of difficulties that they are having and develop design solutions to mitigate the risks.”
Improving the Use of Human Factors Research in Medical Devices and Equipment
Story had a number of ideas for action, research, and development. In the area of action, she suggested the following:
Professional caregivers, lay caregivers, and home care recipients need better mechanisms to provide feedback about a device to designers.
In the area of research, Story highlighted some questions to be addressed:
For users, what factors influence people’s ability and willingness to follow their doctors’ recommendations and adhere to treatment regimens?
For manufacturers, what factors influence their ability and willingness to address the human factors needs of their users and customers?
For purchasers, what factors influence the medical device purchasers and what factors influence prescribers to consider the needs of their end-users when they choose a device?
In the area of development, she suggested the following:
For users, tools are needed to improve people’s ability and willingness to follow doctor’s orders and adhere to treatment regimens.
For health care providers, assessment tools and mechanisms are needed to gauge whether a medical device is appropriate for a specific user.
For manufacturers, higher standards are needed for home health devices in such areas as safety, accuracy, and ease of use for more diverse user populations.
Users need to be more demanding of the devices they use to provide care in the home, Story concluded. “People seem reluctant to blame the devices…. Lay users tend to blame themselves when they have trouble. I think we need to turn that around and blame the device.”
Responses to Questions
Committee member Mary Weick-Brady added that the availability of clean water is an environmental factor that can affect the use of a home medical device. She also urged designers to design the errors out of a device rather than just adding warnings to a flawed one. And she noted that users often are being required to purchase rather than rent some of the devices they use, even though they then become responsible for maintenance and upgrades.
Committee member Jon Pynoos reminded the workshop participants of
one of the most feared phrases in the English language: Assembly required. “If you can’t even get it together, you can’t use it.”
In response to a question about the difficulty of using some devices, Story speculated that some engineers may design devices for themselves. “They think, if I can use it, then anybody can.” User testing is essential to discover the problems in a design. “I have been doing user testing for 16 years, and real people always teach me things I didn’t expect.”
There was some discussion of how considerations of good design can be integrated into the education of students, including the possibility of infusing human factors education into the basic engineering curriculum. Story agreed that such education is critically needed, and not just for engineers. “In medicine it’s critical, [and] you certainly need it in lots of other professions, too…. The question is, where is it, who does it, how do you do it? It is a complicated project.”
Story also noted that many of the technologies used for home health care would not necessarily be defined as medical devices by the FDA. For example, software used in various contexts is not necessarily a device but can make a critical difference in home health care applications.
Several factors have greatly increased interest in the use of information technology in home health care, including the need to reach people in rural and underserved areas, a clinical workforce shortage, and technological advances, such as social networking. In addition, said George Demiris, there is great potential for new technologies to empower care recipients and involve them more actively in health care delivery.
Active Monitoring and Management
Demiris divided the use of information technology into two categories: (1) active monitoring and management and (2) passive monitoring.
Active monitoring implies that the end user is involved with and, in most cases, operates the equipment. Technologies falling into this category include telehealth applications, social networking systems, and personal health record systems.
Telehealth technologies are a diverse set of devices that collect and transmit data over phone lines or other communications media, so that care providers or others can access data remotely. These technologies also
This section is based on the presentation by George Demiris, associate professor of biobehavioral nursing and health systems at the University of Washington. For more information and for references to the information cited in this presentation, see Chapter 9.
include video devices, including low-cost videoconferencing solutions and videophones that are currently available. These can enable home care recipients and their families to communicate with care providers remotely. Video technologies can also link home care recipients with distant caregivers, such as family members, friends, or other parts of a social network.
Some systems integrate video with monitoring devices or have other components to allow for self-report. For example, they might have built-in screens on which people can respond to predetermined questions. Kiosks that are publicly accessible can be used by multiple users, with each user entering a password or swiping a card so that the system knows who the person is.
Research on telehealth applications has focused on care recipients with chronic conditions, including asthma, diabetes, chronic obstructive pulmonary disease, congestive heart failure, stroke rehabilitation, wound care, oncology, and post-transplant care. This research has produced several important findings with human factors implications:
In most cases a significant component of end-user training is involved. The end-user may be the care recipient or, in many cases, the family caregiver, the spouse, or other caregivers who are entering the data or learning how to use the equipment.
The residential infrastructure can be critical. Technologies that rely on phone lines are becoming less usable as people give up their landlines. Technologies that rely on broadband Internet service cannot be used in regions, or individual homes, without such access.
Storing and managing data raises issues associated with security and privacy. These issues also come up in considering how to allow health care providers to process new large data sets gathered from telehealth applications.
Users have to accept and be comfortable with the use of such systems or devices in their homes.
A growing body of literature deals with the effectiveness of those systems on clinical outcomes. However, most of the studies have had small sample sizes and have been focused on feasibility. “We don’t have a very solid evidence base as of yet in terms of these types of telehealth systems reducing rehospitalization rates or improving other specific clinical outcomes,” Demiris said.
Web-based communities, often referred to as virtual communities, are groups of people with a common purpose and common interests who communicate without meeting face to face. They use telecommunications, the Internet, or other technologies to bridge geographic distances. For example, they may use web applications to link care recipients remotely with family
members, with health care providers, or with peer-to-peer communities. In other cases, virtual communities may link health care teams to each other or to groups of domain experts.
Again, not enough evidence exists to demonstrate that these communities improve clinical outcomes. Some individual applications do seem to improve specific clinical outcomes. However, peer-to-peer or web-based communities generally are parts of larger interventions that may include other aspects, such as education or cognitive therapy, making it difficult to attribute positive outcomes solely to the use of peer-to-peer communities.
A relatively new use of web technologies is to create social networks that do not require registering with a website but instead use readily available social networking systems. Early studies have looked at Facebook, for example, in which people link to peers and seek feedback on their progress. Some web-based applications use synchronous communications in which people have to be present virtually at the same time, such as a chat room. Others rely on asynchronous communications, in which people can use discussion boards or other tools to communicate at their discretion. Some of these applications are moderated to control communications and make sure that rules are followed. Other applications do not have a moderator.
Finally, a personal health record is an individual’s electronic medical record that is managed, shared, and controlled by the individual. People own their own data and decide who will have access to them, creating care that is more patient- than institution-centered. Many vendors have shown an interest in personal health records. The Department of Veterans Affairs has implemented an early prototype of a personal health record, called MyHealtheVet, which allows patients to log in and access health-related information, notes and comments about their well-being, and records of health care transactions. Google has been investing in a platform called Google Health that allows users to store health-related data and choose to export data from the application to health care providers or other third parties. In addition, Microsoft has introduced Health Vault as a personal health record platform, with an emphasis on chronic conditions and people who frequently use multiple health care providers.
Although much effort has been invested in the design of personal health records, they have not been tested extensively to see how they affect the quality of home health care, Demiris said.
With passive monitoring, the end-user does not have to operate any equipment and continues with daily activities. Technologies in the environment collect information and direct it to health care providers and other recipients.
“Smart homes” are equipped with an infrastructure allowing passive monitoring of residents to improve their quality of life. For example, the Aware Home developed by the Georgia Institute of Technology provides a display of a resident’s well-being that can be accessed remotely by family members. Another smart home developed in Florida, GatorTech, includes technologies like a smart mirror that provides reminders to residents. TigerPlace, a smart home designed for an independent retirement community, has motion sensors, heat sensors, stove sensors, and bed sensors to track such things as sleep quality, activity in the home, and time outside the home.
Smart homes are a relatively new technology and for the most part have not been systematically evaluated. Existing studies have looked mostly at safety monitoring and assistance, cognitive and sensory assistance, cognitive aids, and overall wellness. As with other technologies, an extensive body of evidence on clinical effectiveness is lacking. Furthermore, clinical trials of smart homes are even more costly than for traditional telehealth or virtual communities.
Privacy and Confidentiality
With any technology involving the collection and movement of information, privacy and confidentiality are concerns.
The Health Insurance Portability and Accountability Act (HIPAA) plays a major role in telehealth applications and web-based applications in which individuals transmit personal health information over the Internet. However, HIPAA cannot address some of the new and emerging trends in health information technology. For example, many of the vendors introducing personal health records are not covered entities according to HIPAA. “There is a debate about whether we need to actually rethink what it means to be a covered entity and how we would deal with a vendor who collects personal health record information for other purposes,” said Demiris.
Interoperability is a major consideration for different information systems that can be employed in the home setting. An infrastructure needs to be in place that will enable data sets to be transmitted among different systems, such as a remote monitoring system and a personal health record. Vocabulary and workflow standards, along with provisions to address security issues, will be needed to make interoperability possible.
Information technologies need to be accessible for people with diverse physical or cognitive limitations. This will require usability testing among users with various limitations. A major problem with information technology systems in the past is that end-users have not been involved in their design and development. Device development often has been driven by
what a technology can do rather than by clinical needs. Human factors research offers a variety of methods, such as prototyping, sketching, or cognitive walkthroughs, to solicit end-user feedback and assess how users interact with a technology.
Several important policy issues will influence the future use of information technologies in home health care.
Technologies can improve access to care, yet they can also be associated with barriers to access. For example, they may be too expensive for some people to afford, the infrastructure may be missing to access a technology in all locations, or some people may lack the education and training to use a technology.
Reimbursement for health information technologies will heavily influence their use. “Who is going to pay for those types of systems, and how are health care providers going to be reimbursed for their time to use the systems or to process data that are resulting from these systems?” asked Demiris.
Data streams may include large quantities of information that are difficult to interpret. “We don’t want to burden health care providers with too much information that may not even be significant, but rather find the right ways to display the data in aggregate form to allow them to identify trends or patterns and detect emergencies in an effective and efficient way.”
Ensuring the safety and efficacy of information technology devices becomes especially complex when additional software or hardware can be added to a system to enhance functionality but is perhaps not subjected to the same scrutiny as the earlier system. In addition, telehealth often will involve delivering care across state borders, raising issues bearing on liability and accreditation.
Finally, introducing information technology in the home environment can have ethical impacts, such as creating dependence on automation, dehumanizing interpersonal relationships, reducing social interaction, generating stigma associated with the use of technology, or being overly intrusive.
Demiris made several suggestions:
Integrate usability, interoperability, and human factors considerations in all phases of the design, implementation, and evaluation of information technology systems.
Explore technical and clinical guidelines proposed by different groups that inform the reimbursement debate.
In the area of research, he suggested the following:
Move away from small pilot studies of technical feasibility toward wide-scale implementation of technologies with clinical studies to assess their effectiveness.
Focus on clinical outcomes and on current gaps in the literature.
Define and assess the empowerment of care recipients and shared decision making.
Enlist the expertise of an interdisciplinary group to conduct translational research that will inform users.
Examine both processes and outcomes.
Responses to Questions
When asked about the human factors issues involved in the use of personal medical records, Demiris observed that commercial vendors claim that their systems will be intuitive to most end-users who are familiar with their other products. Vendors also claim that work focused on human factors has been done on their systems because they are patient-centered systems, not electronic medical records that are focused on clinical encounters. “People are recognizing that human-factors considerations need to inform the design. But it remains to be seen if indeed it will become the case.” An additional challenge will be interactions between personal health records and electronic medical records, especially if multiple entities desire access to those records.
Demiris also said that the category of applications with the best evidence for efficiency and efficacy is telehealth technologies, with clinical trials of web-based virtual communities also showing some effects. Studies tend to be difficult to do because people in a test group cannot be deprived of a standard of care, so they often receive standard care plus additional virtual visits. “The argument there is that you are obviously increasing the attention you are paying to your patients, and it’s not necessarily the technology that is doing great things; it’s just that they get to talk to the health care provider more frequently.” Even if a test group forgoes some in-person visits, the equipment being used may greatly increase their contact with care providers.
In response to a question about the ethical effects of information technologies, Demiris observed that no evidence is yet available showing a detrimental effect on human relationships. However, he pointed to people who sometimes refuse to carry wearable devices when they interact with each other. Making devices less visible, such as making them part of jewelry, might lessen such stigma. In some cases, technologies will have the positive effects of increasing communication and a sense of community. But
designers must also try to anticipate unexpected reactions to new technologies. For example, he cited the case of an elderly resident of a smart home who mistakenly believed that sensors were capturing images of residents.
A discussion began that continued during the session wrap-up about the value of the information being collected by information technologies. Demiris stated that “the verdict is still out in terms of whether it’s really useful to know all the things we are now capturing with sensors.” Case studies have demonstrated successful applications, such as sensors that detect large amounts of wandering, falls, or long periods of inactivity. “The challenge is to have the right infrastructure to respond to [emergencies], because we can detect an emergency, but if there is no plan in terms of how to address it or who would go and check that somebody has indeed fallen, then the system won’t really work.”
Paul Crawford, director of research in the Digital Health Group at the Intel Corporation, led a discussion of some of the most important points and missing elements in the session. In the area of task analysis, risk-based systems engineering approaches have not been widely adopted in health care, he said. Also, a larger research infrastructure than currently exists will allow people to build on each other’s work.
The regulation and surveillance of home-based medical devices need to be reexamined. “We can’t just force-fit what we have been using in the institutional-care setting into the home health care setting.” For example, different standards may be needed for regulatory clearance of home health care devices. Such devices are subject to different demands and expectations and move from person to person in different ways. As a specific example, should medical devices controlled by mobile phones be subject to the same regulatory standards as other devices?
Finally, information technology offers “game-changing possibilities,” but its effects must be better understood and its benefits clinically validated for usage to increase. “How can you identify effectively those characteristics and attitudes that will allow your [technology] solution to flourish as opposed to fail?” Crawford asked. Executives of the companies developing these technologies want to see returns on investments, while regulatory agencies want to see proof of efficacy.
In all three areas, said Crawford, an important step will be to establish priorities. It also is important to determine what legacies of the existing health care system will extend into the future and which can be discarded and reimagined. For example, “Do we need to build a whole separate workforce and education system … as home health care grows at the expense of institutional health care?” A cohesive community and leading
journal—equivalent to the New England Journal of Medicine, but for home health care—“would obviously be productive.”
Committee member Christopher Gibbons said that reimagining the delivery of health care requires asking what recipients want and need, not just doing what others think they want or need. Crawford agreed, saying “there is certainly a lot of interesting feedback out there that is not obvious.” Taking such steps requires a sound infrastructure for regulation, technology development, and reimbursement, said committee member Laura Gitlin. “It’s not developing the same infrastructure we have for [institutional] settings but what a new infrastructure is going to look like.”
In response to a question about incompatibilities caused by technology upgrades, Crawford said that Intel has emphasized backward compatibility, so that capabilities are not lost when a system is upgraded. The issue also arose of incorrect data entry into home health care technologies by users, whether a care recipient or a formal or informal provider. The possibility of erroneous data entry requires both user testing and safeguards built into technologies that could cross-check entries. But full capabilities in this area will require that systems be integrated across different devices and technologies, which will require even greater attention to human factors issues.
Committee member Judith Matthews raised the issue of trust. “Does the system do what it’s intended to do?” Airplanes rely on redundant systems, which increase the level of trust among fliers. “It’s not just a matter of the information being put in or the output at the other end to the recipient of that information. It’s also knowing that the system is working, that it’s calibrated, that it’s good to go.”
Carolyn Humphrey observed that a large number of formal caregivers have had extensive experience with home health care devices and technologies. These individuals could inform discussions about user needs and trust. She also mentioned that current reimbursement structures do not necessarily lend themselves to the widespread adoption of particular medical devices, including telehealth applications. “If we do get telehealth for a patient while they are on formal home care, we most of the time don’t have a way to get it continued after their discharge from home care.” And the removal of equipment can be traumatic for people who have learned to trust it. “We had people crying, literally, when their systems were leaving,” said Crawford.
Several physician participants at the workshop questioned the value of at least some of the data generated by new technologies. Much of this information is simply discarded by physicians too busy to consult or use it. The data need to be filtered and analyzed. Gibbons said, “This is why health care has to change. There are some things we do now that we shouldn’t do, that we don’t need to do. But there are some things that we are not doing that we should do.”