Secondary Logo

Journal Logo

Original Articles

Technology Solutions for Nurse Leaders

Clancy, Thomas R. PhD, MBA, RN, FAAN

Author Information
doi: 10.1097/NAQ.0000000000000439
  • Free
  • Continuing the Conversation

Abstract

TODAY'S CHIEF NURSE EXECUTIVES (CNEs) face daunting challenges as they strive to meet their health system's strategic objectives including improving the quality of clinical and administrative outcomes, improving access to health care services, reducing expenses, and improving the productivity of staff and the efficiency of system processes. As a result, CNEs are under intense pressure to meet these objectives. In 6 updated surveys published by the Commonwealth Fund since 2004, the United States was ranked last in 9 industrialized countries (Australia, Canada, France, Germany, the Netherlands, New Zealand, Norway, Sweden, Switzerland, and the United Kingdom) for quality of care, access, efficiency, equity, healthy lives, and health expenditures.1 Yet, the United States now spends 17.7% of its gross domestic product on health care, twice as much or more than any other country in the survey.2 Clearly, CNEs can expect pressure by government and commercial insurance payers to reduce reimbursement for health care services to continue into the future in an effort to contain further costs.

Adding to CNEs' list of challenges is an anticipated global shortage of 18 000 000 health care workers by the year 2030.3 In the United States alone, the Bureau of Labor Statistics estimates that there will be a shortfall of 203 700 registered nurses (RNs) per year through 2026.4 Ironically, this shortage will be seen in the context of an estimated increase in total RNs from 2.9 million to 3.4 million in the same time period. Much of this is the result of shifting demographics where the proportion of adults older than 65 years is outpacing the proportion of children younger than 18 years.5 This shift in demographics will have a significant impact on nursing, as today's nurses are retiring at a much faster rate than nursing schools can graduate new nurses. In 2018, for example, more than 50% of nurses were older than 50 years and it is estimated that more than 1 million nurses will retire in the next 15 years.4 Thus, CNEs will continue to face a growing shortage of nurses at the same time the demand for nursing services is increasing in response to the increased expectations on quality of care as well as from an older, more resource-intensive senior population.

Complicating these challenges further is the performance of the health care industry when compared with other non–health care economic sectors such as manufacturing and agriculture. Non–health care sectors, through advances in technology, have demonstrated significant gains in labor productivity over the last 100 years. For example, at the turn of the 19th century, more than 50% of the US labor force was involved in farming. Since then, automation has driven this percentage down to less than 2%.6 Manufacturing has followed the same path as farming, with the exception that it took only 40 years.7 Today, the global economy is in the midst of a digital revolution that is increasing productivity through automation in professions such as law, finance, and other services.8 In contrast, annual labor productivity gains in the health care sector have been approximately 50% of non–health care sectors. In fact, in recent years, labor productivity in health care has actually declined approximately 1% per year.3 Given the success of technology in other economic sectors, CNEs may be under increasing pressure to achieve the same level of productivity improvements seen in non–health care industries.

In summary, CNEs face formidable future challenges in their desire to meet performance criteria that are aligned with health systems' goals and objectives. Staffing shortages will likely be the greatest barrier to CNEs meeting their strategic goals for quality, productivity, access, and cost. Past efforts to narrow the staffing gap through strategies that increase the number of admissions to nursing school, such as increasing faculty members, clinical sites, preceptors, and scholarships, will likely not be enough to offset the number of nurses retiring. One solution, in part, may be to make today's nurses more productive through the use of technology. By doing so, significant crossover benefits can also be realized for quality, access, and cost measures. The focus of this article is to assist CNEs understand how technology can be leveraged to improve health systems' quality, labor productivity, cost, and access objectives.

THE IMPACT OF WASTE ON QUALITY, PRODUCTIVITY, COST, AND ACCESS

In health care, productivity, quality, cost, and access measures are defined in many ways. Productivity can be measured by dividing the average output per period by the total costs incurred or resources (capital, energy, material, and personnel) consumed in that period. A common and often criticized hospital productivity measure for nursing is calculated by dividing the total number of RN-hours during a 24-hour period (outputs) by the number of patients (inputs) at the midnight census.9 This measure, known as “hours per patient day,” or HPPD, is used by most hospitals to benchmark productivity. Quality is defined as “the degree to which health care services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge.”10(p707) The Agency for Healthcare Research and Quality (AHRQ) has identified 6 domains that represent quality: safety, effectiveness, patient-centered, timeliness, efficiency, and equity.11 Cost in health care is defined by the customer: to providers cost is the expense incurred to deliver health care services to patients; to payers, it is the amount they pay to providers for services rendered; and to patients, cost is the amount they pay out-of-pocket for health care services.12 Finally, access in health care is defined as having “the timely use of personal health services to achieve the best health outcomes.”13 The AHRQ identifies 4 components of adequate access: coverage or the efficient entry into the health system; services or having a usual source of care; timeliness or the ability to provide health care when the need is recognized; and workforce or having capable, qualified, culturally competent providers.14

Clearly, multiple factors can influence a health system's quality, labor productivity, cost, and access measures, but the factors that do not add value to a patient's care can be categorized as a form of waste. An often-used framework to identify waste is Toyota's Lean Production Process. The Lean framework defines waste as “any expense or effort that is expended but which does not transform raw materials into an item the customer is willing to pay for.”15 Today, Toyota recognizes several types of waste that include defects, excess processing, overproduction, waiting, transportation, and motion. By adapting the Lean framework to health care, CNEs can prioritize opportunities to reduce waste and improve quality, productivity, cost, and access. Table 1 identifies some examples of waste in health care.

Table 1. - Examples of Waste in Health Care
Type of Waste Health Care Examples
Defects Errors, failings, or deficiencies that result in patient harm and additional cost to the system. Examples include medical errors, hospital-acquired infections, misdiagnosis, 30-d hospital readmissions, and unnecessary care through failure of illness prevention and wellness programs.
Excess processing Doing more work, making it more complex or more expensive than is necessary. Examples include patients accessing the wrong level of care and treatment such as the ECU when care could be provided in a clinic or at home. Overly cumbersome processes for clinical documentation, medication reconciliation, result reporting, orders, and other through poorly designed workflow in an EHR.
Overproduction Excess utilization of resources. Examples include unnecessary hospitalization, diagnostic tests, uneaten meals, ordering medications that the patient does not need, overutilization of staff (supply) in proportion to the need for nursing services (demand).
Waiting Unnecessary delays. Examples include delays in access to medical providers, admission and discharge from hospitals, clinic visits, and other. Delays in procedures, medications, test results, and supplies (laboratory, radiology, pharmacy, and other) because of poor system processes. Delays in answering call lights and other.
Transportation Unnecessary movement of patients, providers, equipment, and supplies within and throughout facilities in the health system (hospitals, clinics, subacute care, and patient homes). Poorly designed facilities that require patients and staff to walk to multiple different areas to access various services.
Motion Unnecessary movement of people within a health system. Examples include an office or hospital layout not being consistent with workflow. Supplies not being stored at the point of care, causing nurses to search for supplies. Equipment is not conveniently located. Documentation in an EHR is cumbersome and time-consuming, causing providers to search for information (policies, evidence-based guidelines, and other)
Abbreviations: ECU, emergency care unit; EHR, electronic health record.

TECHNOLOGY AS A SOLUTION

When considering technology as a means to improve quality, productivity, cost, and access, it is important for CNEs to identify how technology may reduce waste within the system. This can be accomplished by first determining which of your system's objectives are not being met and then clearly defining the problem in terms of the various types of waste. CNEs will be familiar with the scenario where despite numerous programs and efforts to recruit and retain nurses in your hospitals, the number of traveling nurses hired continues to grow, resulting in significant cost overruns. Technology solutions generally work best when the cause of the problem can be clearly articulated with measurable outcomes. In this case, the supply of nurses required to implement the current care model is unable to accommodate patient demand, leading to increased use of contracted labor (travel nurses). Solutions that focus on redesigning the care model in a way that allows existing core staff to care for the same number of patients with less contracted labor will be of interest to CNEs. Technology, in part, can assist in meeting this goal, but given limited funds, it is important for CNEs to know how and what to invest in?

Innovation in computer technology over the last 100 years has followed an accelerating growth curve that has been doubling approximately every 18 months.16 Technology is characterized by the use of digital Data, machine learning Algorithms, Networks, the Cloud, and Exponential growth, which are known by the acronym “DANCE.”17 DANCE technologies fueled by improvements in computer processing speed, storage capabilities, and sensors have driven advances in the field of data science, robotics, computer science, and biomedical engineering. These developments support new applications within the interdisciplinary field of biomedical informatics, which “studies and pursues the effective use of biomedical data, information, and knowledge for scientific inquiry, problem solving, and decision-making to improve human health.”18 New and existing technological applications can be used to assist CNEs in reducing waste and creating value for key stakeholders. The following is a review of types of waste and how various technologies can reduce waste using a Lean framework in a health care context.

Defects

There are approximately 400 000 medical errors per year in US hospitals, with an estimated annual cost of $20 billion.19 Medical errors are now the third leading cause of mortality after cancer and heart disease. Errors can be divided into 2 main categories: errors of omission, such as forgetting to put a bed rail up, and errors of commission, which might be administering the wrong medication to a patient. Both types of errors can result in adverse patient events or near misses. Technology solutions aimed at eliminating medical errors should be designed to identify these types of errors early and then prevent them from occurring. The use of technology, notably electronic health records (EHRs), to prevent medical errors was the primary impetus for implementation of the HITECH ACT (Health Information Technology for Economic and Clinical Health) in 2009, which increased EHR use significantly. Historically, medication errors have contributed the greatest proportion of overall medical errors in hospitals. Although some organizations have seen an increase in medication errors after the implementation of an EHR, a meta-analysis of pooled results from 2833 US hospitals and 9 different studies demonstrated a 48% decrease in the likelihood of an medication error when using computerized order entry in conjunction with a bar-coded medication management system.20 Key applications such as clinical decision support software (CDSS) have been integrated into EHRs, smart pumps, bar-coded medication management systems, and other forms of technology for years. These applications map inputs from provider orders to a patient's ID, diagnosis codes, allergies, medical history, medication administration record, tests, procedures, and treatments through rules-based algorithms programmed in the software. Alerts and reminders are then electronically pushed to providers when a potential error such as a drug interaction, wrong patient, or other is discovered. Given that 95% of hospitals today have an EHR, most vendors have already built sophisticated CDSS capabilities into their EHR platforms.21

There are a number of issues to consider when evaluating EHRs as a solution to reduce medical errors. Inadequate information flow and poor communication were cited as the top reasons for medical errors in a recent study by John Hopkins Hospital.22 A lack of crucial information needed to influence prescribing decisions, lack of appropriate communication of test results, and poor coordination of medication orders for transfer of care all contribute to an increased risk of medical errors. Medical errors created by inadequate information flow and poor communication are, in part, the result of poorly designed EHR workflow and interoperability. Two key strategies have been shown to reduce medical errors through EHRs. The first is to purchase a fully integrated, enterprise-wide EHR from one vendor with a proven record. In a recent survey of 153 physicians, nurses, and pharmacists, HIMSS (Healthcare Information Management Systems Society) found that 82% of hospitals have transitioned to an enterprise-wide, single-vendor EHR and 71% see this strategy as providing the greatest value in medication management and error reduction.23 Although purchasing an enterprise-wide, single-vendor EHR does not guarantee interoperability and efficient workflow, it is clearly better than a patchwork of new and older “legacy” systems. The second strategy is to develop a robust EHR governance team that can strike the right balance between EHR workflow standardization and customization for individual providers. Thoughtful development of a CDSS to optimize the use of alerts and reminders in regard to their frequency and relevance, as well as respect for how different providers practice, is vital to success.

A second common error is misdiagnosis or failure to recognize and respond to patient deterioration.24,25 Examples include failure to identify in-hospital early signs of sepsis, heart failure, and other disease conditions. Advances in predictive models through machine learning systems (MLS) are increasingly being included (or custom-built) as part of an EHR's clinical decision support system. Machine learning is a subfield of artificial intelligence (AI) whose aim is to simulate intelligent human behavior.26 Through the use of algorithms programmed in computer software, the structure or patterns of data collected from an EHR can be learned and then applied later to unknown data. A simple illustration of a pattern would be the “least-squares, best-fit” line running through a scatter plot that represents the association between body mass index (BMI) and diabetes. This is an example of a predictive model because it predicts the risk of diabetes, given a specific BMI. Predictive models, both diagnostic and prognostic, are increasingly being integrated into EHRs today because of advances in MLS algorithms, computer processing speed, and access to large digital data sets (big data). Sepsis prediction models are likely the most widely accepted applications in hospitals today. In a recent meta-analysis of 135 studies evaluating the accuracy of MLS prediction models for sepsis, the overall pooled estimation showed that machine learning performance for early recognition of patients with and without sepsis performed better than the traditional sepsis scoring tools.27,28 Other common clinical uses of predictive models are for the early diagnosis of heart failure,29 venous thromboembolism,30 asthma,31 and others.32

Similar to prediction models are risk stratification models that classify patients into distinct groups of similar complexity and estimate the probability of an event. For example, using EHR data, a recent study evaluated several machine learning risk stratification algorithms to estimate the probability of falls and the effects of a resultant intervention.33 The best model achieved 78% accuracy in identifying patients who would be readmitted to the emergency care unit (ECU) within the next 6 months for a fall. Risk stratification models for other nurse-sensitive indicators such as urinary tract infection can be enabled through the use of in-home sensors to identify changes in patterns of daily living.34 Both predictive and risk stratification models are in a nascent stage of development; however, most commercial EHRs provide tools that will allow the organization to build models based on their own data.35

Excess processing

The use of AI to streamline workflows by automating routine and repetitive tasks has gained significant traction in recent years. Two emerging technologies, automated speech recognition (ASR) and natural language processing (NLP), exemplify how AI has the potential for improving labor productivity. ASR is the ability for a machine or program to identify words and phrases in spoken language, whereas NLP determines what the “words” mean and then converts them to a machine-readable format (voice to text). A good example of both ASR and NLP is conversational “chatbots” or “virtual assistants,” which use AI systems (ASR and NLP) to interact with users via text or spoken language, mimicking human behavior.36 Chatbots have been employed in a variety of health care settings to provide patients access to such activities as self-scheduling, medical queries, medication guidance, symptom checks, nutrition counseling, exercise advising, health screenings, and cognitive-behavioral therapy.37 If successfully deployed, chatbots can eliminate some simple and repetitive tasks often carried out by nurses. A search of the literature found no large-scale randomized clinical trials regarding the effectiveness of chatbots in comparison with human intervention. However, a number of smaller, observational studies have demonstrated that chatbots are effective in performing specific activities.38 For example, a systematic review of 25 peer-reviewed journals and conference proceedings on the use of conversational chatbots with oncology patients found 3 broad areas of effectiveness: cancer screening for learned prostate cancer knowledge and informed decision-making, mental health with Web-based cognitive-behavioral therapy, and lifestyle change.38 Chabot growth in health care will likely follow the path of other economic sectors such as retail, which already has more than a 50% adoption rate for chatbot technology.39

In addition to communicating directly with patients, chatbots have the potential to streamline the nursing workflows in other ways. For example, missing data fields in EHR patient assessment and medication reconciliation screens can be identified through AI algorithms and reported via conversational chatbots not unlike reminders made by Microsoft's Cortana or Amazon's Alexa. Self-triage symptom checker chatbots will also likely transition to being the first line of access for patients seeking health care and will relieve medical providers of unnecessary phone calls, documentation, and other activities.39,40 If necessary, these “triage” chabots can provide immediate and direct access to a medical provider if a patient meets certain emergent criteria. Current acceptance of chabots in health care is mixed and criticism of them falls into 3 broad areas: they can only be used in simple, noncomplex disease conditions; patients and providers are not always trusting of their decisions; and patients are concerned about giving their private health information to a machine. Nevertheless, the health care chabot market continues to grow and is expected to exceed $1.5 billion by 2024.41Table 2 provides examples of the different types of chatbots used in health care today:

Table 2. - Health Care Chabots
Chatbot Description Web Site
Nuance's Florence Provides advanced ASR and NLP to allow hands-free, voice-activated medical orders, clinical documentation, and test results viewing, messaging, and other. Hands-free documentation has been shown to be 3 times faster than keyboard documentation.42 https://www.youtube.com/watch?v=SbTevkYpoSs
Sensily's Molly An avatar-based virtual nurse assistant—connects patients with clinical advice to assess their condition and suggest appropriate follow-up. https://www.youtube.com/watch?v=AU1nGpOmZpQ
Babylon Health Utilizes deep learning AI methods to provide a symptom checker, health tracker for monitoring mental health, fitness, lifestyle metrics, and an assessment of a user's overall health. It is currently utilized in >60 countries. https://www.babylonhealth.com/product/healthcheck
SafedrugBot Utilizes a messaging service that offers assistance to providers via “Telegram,” its messaging app. The chatbot helps providers access the right information about drug dosage on the go. https://www.safeinbreastfeeding.com/safedrugbot-chatbot-medical-assistant/
Tess Provides a conversational chabot to deliver cognitive-behavioral therapy to patients suffering from depression and anxiety. https://mental.jmir.org/2018/4/e64/
Abbreviations: AI, artificial intelligence; ASR, automated speech recognition; NPR, natural language processing.

Transportation and motion

It is estimated that nurses spend 16% to 20% of their time conducting non–value-added activities caused by human error, delay, equipment malfunction, miscommunication, shortcomings in technology design, implementation, workflow integration, and a lack of standardization.43 The unnecessary and excess movement of providers, patients, equipment, medications, test results, and supplies is a form of waste classified as transportation and motion. Technology has the capacity to reduce this burden on patients and providers through the integration of robotics and AI.

There is no general consensus on what defines a robot. However, there is agreement among experts that robots are machines that can be programed by a computer, controlled either internally or externally, and have the following functionality: data processing, sensing (machine vision ASR, NLP, touch), mobility, multilinked manipulation (ie, grasping arm), intelligent-like behavior, which often mimics humans, and the capacity to carry out a complex series of actions autonomously or semiautonomously.44,45 Using this broad definition, health care devices that could be considered a robot range anywhere from software robots (or bots) used in Internet searches for journal articles to bar-coded medication management systems to mobile, humanoid-like machines that co-assist nurses with delivering supplies and equipment.

The use of robots in nursing has been limited because much of patient care is unpredictable and requires multitasking and emotional intelligence, something robots are not good at. As a result, the use of robots in nursing has primarily focused on specific, unifunctional, and repetitive tasks within a larger process. For example, within the entire medication administration process, verification of the 7 rights of medication administration is, in part, semiautomated through a bar-coded medication management system. However, the actual task of administering the drug and evaluating its effect is performed in person by a nurse. The narrow use of robotics in nursing also bears out in the literature where, in a systematic review of 25 publications investigating the utilization of robots by nurses, Kangasniemi et al46 found all but 3 studies focused on just bar code–assisted medication administration or physiologic monitoring systems (vital signs and urine output). The remaining 3 studies focused on the impact of robots on cost, quality, and productivity when used to assist nurses with treatments such as retrieving supplies and equipment using radio-frequency identification devices, passing instrument to surgeons in the operating room, smart pumps, and turning and bathing patients. However, the results of these studies are mixed, given the small number of subjects.

In recent years, there has been significant growth of service robots in health care and it is estimated that the market will exceed $2.8 billion by the year 2021.47 Improvements in microprocessors, actuator and sensor technology, computer vision and location mapping, Cloud-based AI, the Internet of Things (IoT), force sensing (touch), and ASR have all contributed to the development of todays “smart” robots. Health care robots currently are used in a variety of areas including surgical procedures, intravenous smart pumps, lifting and rehabilitation (exoskeletons), bionics (smart prosthesis), co-assist service robots (supply distribution), mobile telepresence, emotional support (social companion robots), cleaning, pharmacy drug distribution, and other. Table 3 provides more recent examples of different types of service robots and the tasks they can assist nurses with.

Table 3. - Examples of Robots Used to Assist Nurses With Different Tasks
Robot Description Web Site
Diligent System's Moxi A co-assistant robot that helps clinical staff with non–patient-facing tasks such as gathering supplies and bringing them to patient rooms, delivering laboratory samples, fetching items from central supply, and removing soiled linen bags. https://diligentrobots.com/moxi
Panasonic's Atoun Model Y Power Assist Suit A lightweight exoskeleton that provides support for individuals whose job requires long periods of walking, standing, and lifting (ie, nursing). The exoskeleton provides up to 15 kg of assisting force, weighs 6 kg, and has a battery life of 8 h. https://newatlas.com/robotics/panasonic-exoskeleton-world-para-powerlifting-events/
Aethon's Tugs A supply distribution robot that transports medical equipment and supplies, linen, laboratory specimens, meals, and waste removal throughout the hospital. https://aethon.com/mobile-robots-for-healthcare/
Softbank's Pepper A social, companion robot able to recognize faces and basic human emotions. Pepper is optimized for human interaction and is able to engage with seniors in hospitals and nursing homes through conversation and a touch screen. https://www.softbankrobotics.com/emea/en/pepper
RIBA (Robot for Interactive Body Assistance), also known as “Robobear” A service robot that can lift up or set down a real human from or to a bed or wheelchair. RIBA utilizes human-like arms and tactile sensors as a guidance system. http://rtc.nagoya.riken.jp/RIBA/index-e.html

A number of early adopter health systems are piloting the use of service robots similar to those described in Table 3 to understand how best to integrate them into the workflow of nurses. There is little question that as the pace of advances in AI and robotics technology increases, robots, as demonstrated in other economic sectors, will eventually perform many tasks once only completed by nurses. As nurse administrators, it is important to not wait for robots to move into the mainstream of adoption because the learning curve and cultural acceptance by staff and patients take considerable time. It is far better to introduce robots to your organization at this early stage and then iteratively investigate how to use them as best practices and technology evolve.

Overproduction and waiting

Appropriate placement of patients in the right level of care can reduce waste caused by the overutilization of labor, supplies, tests, and treatments. The most obvious example would be referrals to an ECU for a problem that could be treated in an outpatient clinic or at home. Because hospitals and their ECUs must be prepared to handle any health-related event at any time of day, their space needs and inventory of staff, equipment, and supplies are disproportionately higher than lower levels of care. Overuse of ECUs and hospitals for unnecessary care can quickly overwhelm a health system and drive up overhead costs and delays in treatment. Coordination of services to ensure patients receive the right level of care, at the right time and place, can reduce this form of waste. Technology can offer a solution by providing a digital platform that facilitates the transition of care and treatment that was once exclusively provided in hospitals and clinics to patients' homes. This platform, known as the IoT, provides access to communication networks so that mobile devices (eg, a smartphone) can access software hosted remotely.

The IoT has the capacity to deliver care in 3 broad based settings: acute care (hospitals), community care (clinic and home), and long-term care (independent, assisted, and skilled nursing homes). Within each of these settings, the IoT can provide 3 classes of use cases: tracking humans, tracking things, and tracking humans and things in combination.48 Applications for the IoT within each of these use cases include monitoring and control, big data and business analytics, and information sharing and collaboration.49 Combined, the integration of DANCE technologies has driven growth in the IoHT (Internet of Healthcare Things) and its build-out should provide both appropriate access to health care and enable the “information value loop” (IVL). The IVL is the stepwise process of creating data from sensors, transferring them via networks to the Cloud, creating new information from the data through analytics, transmitting the information to health care providers to augment their intelligence, and finally providing it to the patient through technology to augment their behavior.49Table 4 provides some current examples of the IoHT.

Table 4. - Examples of the Internet of Healthcare Things
IoHT Technology Description Examples
Monitoring and Control Big Data and Analytics Information Sharing and Outcomes
Tracking Things
Radio-frequency identification The use of a small sensor (computer chip) with an antenna that is either passive (receives signals only) or active (receives and sends signals) that can be placed on equipment, supplies, and even documents. Supply and inventory control Supply chain management optimization Provides data to supply management and budget reports
Capacity command centers A center that receives real-time data regarding the flow of patient admissions, transfers, discharges, and delays from the operating room, radiology, and other departments across a health system's hospital and clinics. Identification and resolution of system delays AI-enabled workflow optimization Improves readmissions, LOS, wait, and turnaround times, and other metrics
Robotics Cloud-controlled robots such as supply distribution robots and co-assist service robots for nursing staff. Cloud-based location mapping and movement AI-driven decision-making for service robots Improves HPPD and patient and staff satisfaction
Home safety monitors Sensors and cameras in the home for home security (temperature, lighting, door and window locks), fall prevention carpets and cameras, nutrition monitors on refrigerators for seniors or homebound patients. AI-controlled thermostats, lighting, locks, and other Identification of activity patterns that could lead to problems Improves readmissions, population health, ECU visits, falls, nutrition
Tracking Humans
Mechanical sensors Sensors that track the movement activities of a patient that include wandering, geo location, falls, sleep patterns, and other. They allow patients to remain in the home and decrease unnecessary admissions Captures patient-specific movement activity Provides data for MLS algorithms and prediction models for fall risk and other Improves fall rates, activity levels, elopements, and nutrition
Physiological sensors Wearable devices that monitor physiological measures such as vital signs, heart rate and rhythm, weight, impedance (fluid retention), stress, and other for acute care (tele-ICU) or remote in-home care. Allows remote care for patients with stable disease conditions Sensor data can be used for MLS as well as performance dashboards Alerts providers of historical and real-time clinical events requiring follow-up
Biochemical sensors Embedded sensors that monitor clinical laboratory measures such as glucose, alcohol, electrolytes, oxygenation, and pH and external factors such as air pollution. Pill sensors can alert providers of medication noncompliance Allows remote point of testing and monitoring for care for patients with stable disease conditions Sensor data can be used to develop MLS predictive models and performance dashboards Tracks population health for chronic disease conditions (diabetes, asthma, heart failure, and other)
Tracking Humans and Things
Telehealth The broad use of electronic information and telecommunications technologies to support long-distance clinical health care, patient and professional health-related education, public health, and health administration. Technologies include videoconferencing, the Internet, store-and-forward imaging, streaming media, and terrestrial and wireless communications.50 This includes clinical services and nonclinical services, such as provider training, administrative meetings, and continuing medical education, in addition to clinical services. Telehealth in combination with sensor technology allows remote real-time monitoring and communication between patients, providers, and staff AI-enabled CDSS augments providers critical decision-making ability and allows telehealth services to a broader population Decreases the dependence on physical structures such as hospitals and clinics for on-site care and improves access for patients
Telemedicine The use of electronic information and telecommunication technologies to support long-distant clinical care. Telemedicine differentiates itself from telehealth by its narrower focus on the diagnosis, treatment, and prevention of disease and injuries.49 Used primarily for remote specialist consultation with other medical providers Telemedicine can be used in conjunction with population management strategies to provide better management of chronic disease Allows small community hospitals to provide specialist consultation on-site and reduces patient transfers
Telepresence robots The use of real-time, interactive telemedicine through telepresence robots (in a hospital or home) to provide remote access to a medical provider for consultation on a patient. Can monitor patients through daily telepresence rounds Telepresence robots can interface with EHRs that provide specialists clinical analytics on their patients Telemedicine robots improve access to specialists, primarily in rural hospitals, and allow patients to stay in their communities
Virtual visit A form of telemedicine where the traditional on-site medical office visit is replaced by an online visit. The visit can be real time with a provider (synchronous) or store and forward (asynchronous) where a provider can respond to an online assessment form at a later time. Telehealth assessment tools are available to allow nurses to partner with remote providers and conduct a full virtual patient evaluation MLS algorithms can be used to stratify high-risk patients who need to be seen face to face by a medical provider Allows patients convenient remote access to primary care providers and reduces unnecessary on-site clinic visits
Virtual nurse assistant An interactive, avatar-based virtual assistant (chatbot) that utilizes AI to monitor patients' health and to answer specific health questions regarding any disease conditions they may have. Acts as a triage nurse by asking and answering routine health questions on a regular basis via avatar-based software Speech recognition software and NLP provide baseline information to patients and collect data for providers Provides routine information to patients and reduces unnecessary phone calls and follow-up by medical providers
Abbreviations: AI, artificial intelligence; CDSS, clinical decision support software; ECU, emergency care unit; HPPD, hours per patient day; ICU, intensive care unit; IoHT, Internet of Healthcare Things; LOS, length of stay; MLS, machine learning system; NLP, natural language processing.

SUMMARY

The anticipated shortage of future health care workers, an aging demographic, and advancements in technology make it imperative that nurse leaders seek solutions that increase not only the number of care providers but also their productivity. Technology, as demonstrated in other economic sectors, has the capability to increase productivity while improving quality, access, and cost. However, choosing the right technology requires CNEs to carefully research which technologies will actually reduce waste in their organizations and in what sequence they should be implemented. It also requires CNEs to examine the significant social and cultural changes that various forms of technology can impart on staff and patients. The fear of job loss is often associated with the introduction of robots. Patients are frequently mistrustful regarding the accuracy, security, and privacy of IoHT devices such as wearable monitors and biosensors. Overreliance on AI for clinical decision-making can result in poor judgment and medical errors. This article has provided a glimpse of the future of health care technology, how it may offer solutions to tomorrow's challenges, and how it will take thoughtful leadership on how best to apply it.

REFERENCES

1. Schneider EC, Sarnak DO, Squires D, Shah A, Doty MM. System has widest gap between people with higher and lower incomes. The Commonwealth Web site. https://www.commonwealthfund.org/press-release/2017/new-11-country-study-us-health-caresystem-has-widest-gap-between-people-higher. Published July 17, 2017. Accessed January 14, 2020.
2. Historical. Centers for Medicare & Medicaid Services Web site. https://www.cms.gov/Research-Statistics-Dataand-Systems/Statistics-Trends-andReports/NationalHealthExpendData/NationalHealthAccountsHistorical. Published December 17, 2019. Accessed January 14, 2020.
3. Britnell M. Human: Solving the Global Workforce Crisis in Healthcare. Oxford, England: Oxford University Press; 2019.
4. The nursing shortage. American Association of Colleges of Nursing Web site. https://www.aacnnursing.org/news-information/fact-sheets/nursing-shortage. Published April 1, 2019. Accessed January 14, 2020.
5. Vespa J. The U.S. joins other countries with large aging populations. US Census Bureau Web site. https://www.census.gov/library/stories/2018/03/graying-america.html. Published October 8, 2019. Accessed January 14, 2020.
6. Latest U.S. agricultural trade data. USDA Economic Research Service Web site. https://www.ers.usda.gov/topics/farm-economy/farm-labor. Published October 23, 2019. Accessed January 14, 2020.
7. Percent of employment in manufacturing in the United States. US Bureau of Labor Statistics, (DISCONTINUED) [USAPEFANA], retrieved from FRED, Federal Reserve Bank of St. Louis. https://fred.stlouisfed.org/series/USAPEFANA. Published January 14, 2020. Accessed January 17, 2020.
8. Pugliano J. The Robots Are Coming: A Human's Survival Guide to Profiting in the Age of Automation. New York, NY. Ulysses Press; 2017.
9. Beswick S, Hill PD, Anderson MA. Comparison of nurse workload approaches. J Nurs Manag. 2010;18(5):592–598. doi:10.1111/j.1365-2834.2010.01124.x.
10. Harris-Wehling J. Medicare: defining quality. In: Lohr KN; Institute of Medicine Committee to Design a Strategy for Quality Review and Assurance in Medicare, eds. Medicare: A Strategy for Quality Assurance: Volume II Sources and Methods. Washington, DC: National Academies Press; 1990.
11. Six domains of health care quality. Agency for Healthcare Research and Quality Web site. https://www.ahrq.gov/talkingquality/measures/six-domains.html. Published November 2018. Accessed January 14, 2020.
12. Arora V, Moriates C, Shah N. The challenge of understanding health care costs and charges. AMA J Ethics. 2015;17(11):1046–1052.
13. Institute of Medicine, Committee on Monitoring Access to Personal Health Care Services. Access to Healthcare in America. Washington, DC: National Academies Press; 1993. https://www.ncbi.nlm.nih.gov/books/NBK235882. Accessed January 15, 2020.
14. Chartbook on access to health care: elements of access to health care. Agency for Healthcare Research and Quality Web site. https://www.ahrq.gov/research/findings/nhqrdr/chartbooks/access/elements.html. Published June 2018. Accessed January 15, 2020.
15. 7 types of waste. Everlean Web site. https://www.everlean.de/en/7-types-of-waste. Accessed May 28, 2020.
16. Moore's Law Web site. https://www.umsl.edu/∼siegelj/information_theory/projects/Bajramovic/www.umsl.edu/_abdcf/Cs4890/link1.html. Accessed May 28, 2020.
17. McAfee A, Brynjolfsson E. Machine, Platform, Crowd: Harnessing Our Digital Future. New York, NY: WW Norton & Co; 2017.
18. What is biomedical and health informatics? American Medical Informatics Association Web site. https://www.amia.org/fact-sheets/what-informatics. Published June 2011. Accessed January 15, 2020.
19. Rodziewicz TL, Hipskind JE. Medical error prevention. StatPearls Web site. https://www.ncbi.nlm.nih.gov/books/NBK499956. Published May 5, 2019. Accessed January 15, 2020.
20. Radley DC, Wasserman MR, Olsho LEW, Shoemaker SJ, Spranca MD, Bradshaw B. Reduction in medication errors in hospitals due to adoption of computerized provider order entry systems. J Am Med Inform Assoc. 2013;20(3):470–476.
21. AHA reports: electronic access to health information continues to climb. American Hospital Association Web site. https://www.aha.org/news/headline/2018-03-01-aha-reports-electronic-accesshealth-information-continues-climb. Published March 1, 2018. Accessed January 15, 2020.
22. The 8 most common root causes of medical errors. Always Culture Web site https://alwaysculture.com/hcahps/communication-medications/8-most-common-causes-of-medicalerrors/). Accessed January 15, 2020.
23. Reducing medication errors through better technology integration. Health IT News Web site. https://www.healthcareitnews.com/news/reducing-medication-errors-through-better-technologyintegration). Published May 4, 2018. Accessed January 15, 2020.
24. Massey D, Chaboyer W, Anderson V. What factors influence ward nurses' recognition of and response to patient deterioration? An integrative review of the literature. Nurse Open. 2016;4(1):6–23. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5221430. Accessed January 15, 2020.
25. Robert N. How artificial intelligence is changing nursing. Nurs Manag. 2019;50(9):30–39. doi:10.1097/01.NUMA.0000578988.56622.21.
26. Artificial intelligence. Oxford Dictionary: Lexicon Web site. https://www.lexico.com/en/definition/artificial_intelligence. Published August 13, 2020. Accessed January 15, 2020.
27. Islam MM, Nasrin T, Walther BA, Wu CC, Yang HC, Li YC. Prediction of sepsis patients using machine learning approach: a meta-analysis. Comput Methods Programs Biomed. 2019;170:1–9.
28. Kent J. Predictive analytics, EHR big data reduce sepsis mortality by 18%. Health IT Analytics Web site. https://healthitanalytics.com/news/predictive-analytics-ehr-big-data-reduce-sepsismortality-by-18. Published November 9, 2018. Accessed January 16, 2020.
29. Frizzell JD, Liang L, Schulte PJ, et al. Prediction of 30-day all-cause readmissions in patients hospitalized for heart failure: comparison of machine learning and other statistical approaches. JAMA Cardiol. 2017;2(2):204–209. doi:10.1001/jamacardio.2016.3956.
30. Hendriksen JMT, Geersing GJ, Moons KGM, de Groot JAH. Diagnostic and prognostic prediction models. J Thromb Haemost. 2013;11(suppl 1):129–141. doi:10.1111/jth.12262.
31. Daines L, McLean S, Buelo A, Lewis S, Sheikh A, Pinnock H. Systematic review of clinical prediction models to support the diagnosis of asthma in primary care. NPJ Prim Care Respir Med. 2019;29:19. doi:10.1038/s41533 019-0132-z.
32. Cowley LE, Farewell DM, Maguire S, Kemp AM. Methodological standards for the development and evaluation of clinical prediction rules: a review of the literature. Diagn Progn Res. 2019;3:16. doi:10.1186/s41512-019-0060-y.
33. Patterson BW, Engstrom CJ, Sah V, et al. Training and interpreting machine learning algorithms to evaluate fall risk after emergency department visits. Med Care. 2019;57(7):560–566. doi:10.1097/MLR.0000000000001140.
34. Enshaeifar S, Zoha A, Skillman S, et al. Machine learning methods for detecting urinary tract infection and analyzing daily living activities in people with dementia. PLoS One. 2019;14(1):e0209909. doi:10.1371/journal.pone.0209909 PLOS/One.
35. Pennic J. Geisinger, IBM develop predictive model using EHR data to detect sepsis. HIT Consultant Web site. https://hitconsultant.net/2019/09/13/geisinger-ibm-develops-predictive-modelusing-ehr-data-to-detect-sepsis/#.XfKWimRTk2w. Published September 13, 2019. Accessed January 16, 2020.
36. Health care chatbots: overview. Qolty Web site. https://qolty.com/chatbot. Accessed January 16, 2020.
37. Papp J. 3 use cases for chatbots in healthcare. Circulation Web site. https://www.circulation.com/blog/3-use-cases-for-chatbots-in-healthcare. Published January 9, 2019. Accessed January 16, 2020.
38. Bibault JE, Chaix B, Nectoux P, Pienkowsky A, Guillemasse A, Brouard B. Healthcare ex Machina: are conversational agents ready for prime time in oncology?. Clin Transl Radiat Oncol. 2019;16:55–59. doi:10.1016/j.ctro.2019.04.002.
39. Smart Talk: how organizations and consumers are embracing voice and chat assistants. Capgemini Web site. https://www.capgemini.com/us-en/news/smart-talk-how-organizations-and-consumers-areembracing-voice-and-chat-assistants. Published September 5, 2019. Accessed January 16, 2020.
40. Lovett L. AI triage chatbots trekking toward a standard of care despite criticism. Mobilehealthnews Web site. https://www.mobihealthnews.com/content/ai-triage-chatbots-trekking-toward-standardcare-despite-criticism. Published November 2, 2018. Accessed January 16, 2020.
41. Bryant M. Hospitals turn to chatbots, AI for care. HealthcareDive Web site. https://www.healthcaredive.com/news/chatbots-ai-healthcare/516047. Published February 8, 2019. Accessed January 17, 2020.
42. Clipper B, Batcheller J, Thomaz AL, Rozga A. Artificial intelligence and robotics: a nurse leader's primer. Nurse Leader. 2018;16(6):379–384. doi:10.1016/j.mnl.2018.07.015.
    43. Upenieks VV, Akhavan J, Kotlerman J. Value-added care: a paradigm shift in patient care delivery. Nurs Econ. 2008;26(5):294–300; quiz 301.
    44. Robot. Oxford English Dictionary Web site. https://www.oxfordlearnersdictionaries.com/us/definition/american_english/robot. Accessed January 16, 2020.
    45. Service robots. The International Federation of Robotics Web site. https://www.ifr.org/servicerobots. Accessed January 16, 2020.
    46. Kangasniemi M, Karki S, Colley N, Voutilainen A. The use of robots and other automated devices in nurses' work: an integrative review. Int J Nurs Pract. 2019;25(4):e12739. doi:10.1111/ijn.12739.
    47. Arya A. The era of the robot nurse. MedTechIntelligence Web site. https://www.medtechintelligence.com/feature_article/the-era-of-the-robot-nurse. Published April 16, 2019. Accessed January 16, 2020.
    48. Lee I, Lee K. The Internet of Things (IoT): applications, investments, and challenges for enterprises. Bus Horizons. 2015;58(4):431–440. https://www.academia.edu/20706011/The_Internet_of_Things_IoT_Applications_investments_and_challenges_for_enterprises. Accessed June 1, 2020.
    49. Raylor M, Cotteleer M. The more things change: value creation, value capture, and the Internet of Things. Deloitte Web site. https://www2.deloitte.com/insights/us/en/deloitte-review/issue-17/valuecreation-value-capture-internet-of-things.html. Accessed January 17, 2020.
    50. What is telehealth? How is telehealth different from telemedicine? HealthIT.gov Web site. https://www.healthit.gov/faq/what-telehealth-how-telehealth-different-telemedicine. Published October 17, 2019. Accessed January 16, 2020.
      Keywords:

      nursing and artificial intelligence; data analytics; Internet of Things; natural language processing; predictive models; robotics; speech recognition

      © 2020 Wolters Kluwer Health, Inc. All rights reserved.