Challenges and Opportunities For Improving Patient Safety Through Human Factors and Systems Engineering

Abstract

Despite progress on patient safety since the Institute of Medicine 1999 report “To Err is Human”, significant problems remain. Human factors and systems engineering (HF/SE) has been increasingly recognized and advocated for its value in understanding, improving, and redesigning processes for safer care, especially for complex interacting sociotechnical systems. Broad awareness and adoption of HF/SE into safety improvement work, however, has been frustratingly slow. We provide an overview of HF/SE, its demonstrated value to a wide range of patient safety problems, in particular medication safety, and challenges to its broader implementation across health care. We propose seven recommendations and policy implications to maximize the spread of HF/SE, including formal and informal education programs, greater adoption of HF/SE by healthcare organizations, expanded funding to foster greater clinician-engineer partnerships, and coordinated national efforts to design and operationalize a system for spreading HF/SE into health care nationally.

Introduction

With reports by the National Academies (1, 2) and the President’s Council of Advisors on Science and Technology,(3) human factors and systems engineering (HF/SE) has gained recognition as an innovative approach for improving patient safety. A 2005 joint report by the National Academy of Engineering and the Institute of Medicine (IOM) provided a framework for partnerships between systems engineers and healthcare practitioners to address a range of operational and strategic problems, including patient safety.(2) Subsequent reports by the IOM, Agency for Healthcare Research and Quality (AHRQ), and others have called for greater involvement of HF/SE. For example, a 2015 report by the National Academies highlights diagnostic errors as a major patient safety issue, with most people likely to experience at least one diagnostic error with possible negative consequences in their lifetime.(4) The conceptual model of that report describes the diagnostic process as a series of activities that engage the patient with health care over time and are embedded in a work system composed of several interacting elements – people (healthcare professionals, diagnostic team members, patients, caregivers), tasks, technologies and tools, physical environment, organization, and external environment – adapted from the Systems Engineering Initiative for Patient Safety (SEIPS) model of work system and patient safety.(5, 6) Beyond HF itself, numerous patient safety issues also manifest from system design and operational issues that systems engineering methods can address,(7) such as optimizing staffing levels for safe staff-to-patient ratios or smoothing bed flow to prevent ED boarding.(8)

While the value of HF/SE to improve patient safety has been demonstrated, the field remains significantly underused and not well-understood.(9) We therefore describe HF/SE, its different facets and approaches, and applications illustrating the role of HF/SE in patient safety. Because adoption of systems approaches to patient safety remains challenging,(10) we conclude with a discussion of barriers, policy implications, and recommendations to more broadly integrate HF/SE into patient safety improvement locally and nationally.

What is human factors and systems engineering (HF/SE)?

HF/SE encompasses a range of methods and principles to help understand, model, improve, optimize, and integrate complex sociotechnical systems (and systems of systems), often with multiple goals and stakeholders, to yield the best overall system performance, including safety. The HF/SE field brings a systems perspective to patient safety that emphasizes multi-level, spatio-temporal analyses of care processes (e.g., work system analysis, error propagation and recovery, patient journey modeling). These approaches are anchored in the broad discipline of industrial and systems engineering (11) and its human factors engineering component (also called ergonomics).(12)

An important principle of HF/SE is to go beyond improving single system elements, such as technology or tasks, and rather to analyze and improve the entire system, i.e. elements of the system and their interactions.(13) For example, when developing and implementing a patient safety practice, such as preoperative checklists, the entire system needs to be considered where the checklist is viewed as a tool that positively or negatively affects other system elements such as team communication and workflow. As described in Appendix Exhibit A1ⁱ, the core of HF/SE is a systems viewpoint, including system design principles (e.g., usability, situation awareness, system level alignment), implemented through user participation and use of multiple analytical methods in continuous improvement cycles with learning and feedback loops.

Applications of HF/SE to Patient Safety

HF/SE has been applied to many domains of patient safety (14) and made significant contributions to the design of processes, technologies, devices, physical environment, and other aspects of work systems. A majority of this work has focused on hospital-based care, most notably on medication safety and health information technology,(15) HAIs,(16) patient falls,(17, 18) and patient identification.(19) Another important contribution of HF/SE has been the systematic analysis of safety events, including retrospective and prospective methods such as the Human Factors Analysis Classification System (HFACS),(20, 21) Failure Mode and Effect Analysis (FMEA),(22) and others.(23) Exhibit 1 summarizes common examples of HF/SE approaches used to address important patient safety issues.

Exhibit 1.

– Examples of Applications of Human Factors and Systems Engineering to Improve Patient Safety

Safety Issues	HF/SE Approaches	Examples
Patient safety events and near misses	HF classification frameworks and methods for analyzing system factors that contribute to near misses and safety events	Human Factors Analysis Classification System (HFACS) (20)
Medication safety	Human-centered design of medication processes, e.g. prescription and administration	HF design principles and HF methods for safer design of order prescribing interfaces (15), and code cart medication drawer (24)
Healthcare-associated infections	Analysis of system factors that contribute to HAIs	Identification of work system barriers and facilitators to adherence to contact isolation for patients with suspected or confirmed C-difficile infection (16)
Patient falls	HF design of work systems for reducing inpatient falls	Human-centered design of fall prevention toolkit (18)
Patient identification	Human-centered design of armband for preventing patient misidentification	HF design of armband for improving patient identification by reducing number of visual scans required (19)
Patient safety in primary care	Work system analysis for patient safety	Information chaos experienced by primary care physicians that can lead to patient safety events (29)
Patient safety in home care	HF/SE analysis of medical devices and information technologies used in the home	Analysis of usability and system integration of haemodialysis technology (31) and infusion pump (32) HF design of consumer health information technologies for home use (30)
Patient safety in care transitions	Process analysis of transitions between hospital and home (36)	Description of transition process and safety vulnerabilities over multiple phases of care, especially for older adults (35)

SOURCE: Authors’ Analysis

NOTES: HF is Human Factors. SE is Systems Engineering.

For example, a code cart medication drawer used during emergencies was redesigned using HF design principles, e.g. grouping, visibility, and organization, resulting in improvements in completion time, non-wasteful action (e.g. turning an incorrect vial to find the label), perceived visibility, usability and organization, and clinician code cart satisfaction as compared to the usual medication drawer.(24) An analysis of inpatient nursing work, including time studies and the SEIPS model,(5, 6) helped optimize medication retrieval and preparation, improve supply management, and reduce interruptions and distractions, resulting in 59% fewer requests for missing medications, 30% fewer medication room visits, and fewer medication errors reported by ICU nurses.(25) Using a human-centered design process, members of the Surgical Patient Safety Systems Collaborative Group developed the SURPASS perioperative checklist,(26) resulting in a 39% reduction in surgical patient complications and 47% reduction of in-hospital mortality.(27)

Less patient safety research has been conducted outside of acute care settings, and in parallel less use of HF/SE. In primary care, most HF/SE work focuses on understanding work system factors that negatively affect safety and the work of primary care professionals.(28, 29) In the home environment, some HF/SE safety work has focused on health information technologies,(30) haemodialysis technology (31) and infusion pumps,(32) typically addressing usability, the broader system of care,(33) and the patient’s ‘work system’.(34) HF/SE researchers also have examined safe transitions between hospital and home, highlighting the need to examine safety over longer time periods, especially for older adults.(35, 36)

Challenges to Greater Adoption of HF/SE in Patient Safety

Despite many examples of the value of HF/SE,(37) greater adoption by health care remains challenging.(1–3) Five challenges limit broader spread: (1) cultural differences between engineers and healthcare professionals, (2) resource and expertise limitations, (3) organizational environment, (4) care process fragmentation, and (5) policy and market issues.

First, cultural differences between HF/SE and health care are profoundly important but often unrecognized or under-appreciated. Carayon and Xie (38) identified four HF/SE core values (people at the center, systems thinking, continuous improvement, balancing multiple objectives) and four core healthcare culture values (scientific inquiry, individual responsibility, autonomy, excellence), some of which directly conflict. For example, healthcare cultures often emphasize the work, knowledge, and skills of individuals, which can produce a tendency to blame individuals for patient safety incidents, whereas HF/SE seeks to proactively build systems and processes to prevent errors or mitigate their impact. Despite many calls for systems approaches to patient safety, efforts to hold individuals accountable for errors remain,(10) demonstrating how deeply engrained this mindset is in health care. HF/SE approaches also tend to follow a different and more time-consuming work style, viewpoint, and pace that can be at odds with rapid healthcare improvement projects.

A second barrier is resource limitations in technical expertise, time availability, and data infrastructure. Healthcare professionals often have little time away from clinical work (39) and limited knowledge to apply HF/SE approaches.(1) Although this can be offset through clinician-engineer partnerships, few HF/SE professionals are equipped or trained to work in health care.(2) The participatory methods of HF/SE, moreover, require focused time and involvement from healthcare professionals,(40) significantly more than usually available. Additional work therefore could adapt and streamline HF/SE methods for the time-constrained environment of health care. For instance, the time consuming nature of FMEA and other proactive risk assessment methods (41) in one case led to development of a faster hybrid risk assessment method to identify computerized physician order entry (CPOE) vulnerabilities.(42) Current data infrastructures also tend to not capture the type of patient safety data needed by HF/SE professionals for system redesign and improvement.(1–3)

A third barrier to greater HF/SE use is organizational environments not open to innovation and new ideas such as HF/SE. Learning organizations with sufficient resources and decentralized decision-making structures tend to facilitate innovative practices,(40, 43) whereas organizations that are hierarchical and respond to failure punitively create obstacles to using HF/SE.(1) Leaders in healthcare organizations can remove such barriers by committing resources to improvement efforts, raising visibility of improvements, setting priorities, and managing expectations.(3)

A fourth barrier is fragmented and siloed care processes, often designed and managed separately, making the holistic systems approach of HF/SE challenging.(3) Examples include communication and coordination problems across boundaries, separate scheduling systems resulting in unsafe staffing levels or patient rooming practices, and clinicians-in-training being educated primarily within their individual disciplines rather than inter-professionally.

Lastly, policy and market considerations can limit HF/SE adoption, especially when given little incentive to improve healthcare processes under fee-for-service reimbursement models.(1–3) This situation has improved with the renewed focus on population care accountability under the Affordable Care Act (ACA) and the movement of the Centers for Medicare and Medicaid Services (CMS) towards value-based care, which might foster greater adoption of HF/SE. Nonetheless, further work is needed and hiring HF/SE experts, dedicating clinician time, and improving data infrastructures all will require adequate resources. Smaller, rural care settings especially may struggle for such resources,(3) suggesting broader HF/SE deployment might not rest entirely on individual organizations to resource.

Recommendations for Accelerating Integration of HF/SE in Patient Safety

The benefit of HF/SE to patient safety can be more broadly realized through the following mechanisms: more widespread adoption and use of HF/SE tools and methods, greater awareness and training of HF/SE knowledge among healthcare professionals, and hiring of human factors and systems engineers into health systems.(40) We propose seven recommendations and policy implications that address these mechanisms to foster and accelerate greater spread of HF/SE for improving patient safety.

Educational programs for clinicians

Healthcare professionals, health system leaders, and clinicians-in-training would benefit from greater opportunities to learn and apply basic HF/SE methods. Educational programs for physicians, nurses, pharmacists, and others at a minimum should provide an introduction to HF/SE and its role in patient safety. Fellowships should be created to provide clinicians-in-training with deeper HF/SE understanding and skills. As examples, federal funding such as through AHRQ, HRSA, NIH, and NSF training mechanisms could support HF/SE healthcare traineeship programs.

Foundation-managed programs also could be supported, similar to the Kellogg Foundation’s national fellowship program in the 1970s, to embed HF/SE engineers in health systems and provide systems with education in HF/SE. Individual health systems could support HF/SE fellows or expand existing internal quality and safety programs to include modules in HF/SE topics. Closer relationships between HF/SE academic programs and health sciences schools could be fostered and incentivized.

These recommendations are aligned with those by the AMA Education Consortium that describes “health systems science” (HSS) as the third pillar of medical education, in addition to basic science and clinical science.(44) HSS includes HF/SE methods for improving patient safety, e.g. systems approach, PDSA, usability of health information technologies, and process analysis. We recommend a partnership between AMA and HF/SE professionals and experts to further refine HSS and implement it in medical education.

Educational programs for human factors and systems engineers

For more advanced HF/SE work, many more human factors and systems engineers need to be trained in healthcare applications. This ideally would follow a two-pronged approach to engage both students enrolled in HF/SE programs and faculty teaching and advising these students. Beyond classroom education, experiential curricula should provide HF/SE students with opportunities to work on applied projects or internships embedded in healthcare organizations, critical to help understand the organization, environment, and culture of healthcare delivery.

Similar opportunities are needed for HF/SE faculty to be exposed to healthcare’s unique challenges, constraints, and opportunities, such as through embedded summer or sabbatical fellowships. Such immersion experiences could result in new HF/SE courses in patient safety, improved training of HF/SE students, and identification of HF/SE research problems for faculty research, grants, and publications, further enhancing the value of HF/SE to patient safety. Ironically little systems thinking has been conducted to design bi-directional value-added approaches to partnering healthcare systems and engineering academic programs, although several approaches easily could be identified and developed.

Healthcare leadership and boards

Leadership and boards of healthcare organizations should play important roles in realizing the potential of HF/SE. Ultimately, health systems should hire and engage human factors and systems engineers to help improve patient safety. However, leadership awareness of and commitment to the value of HF/SE needs to be fostered, such as through large-scale projects to demonstrate the value of HF/SE, continued advocacy by national organizations such as the National Academies, and other visibility efforts that develop and disseminate convincing evidence of the value of HF/SE.

Healthcare organizations also will need to create or identify mechanisms for retaining and promoting human factors and systems engineers, including work opportunities beyond basic applications and in some cases joint academic appointments. To maximize the value of HF/SE, more robust data infrastructures, beyond those for providing care, also need to be available for analyzing work systems and care processes, which boards of health systems should support.

Notably, a small number of healthcare organizations have long legacies of supporting, engaging, and partnering with human factors and systems engineers. Appendix Exhibit A2ⁱⁱ describes two examples of such healthcare organizations, types of projects typically conducted, and lessons learned. However, hiring or partnering with human factors and systems engineers by healthcare organizations still remain more the exception than the norm. Many hospitals downsized internal SE groups in the cost-cutting era of the late 1980s and 1990s, and the Society for Health Systems, a professional organization consisting of roughly 1,200 hospital-employed industrial engineers, has not grown much since its formation in 1980.

Technology designers and vendors

Although significant research has shown the value of HF/SE in improving the design of technologies, such as smart infusion pump and subsequent medication safety,(22, 45, 46) patient safety problems continue to occur related to poor HF design and technology implementation.(47, 48) Significant work remains to more systematically integrate HF/SE (e.g. usability) in designing health information technologies,(49) in particular among vendors,(50) including training of designers and implementers in human-centered design.(51) Designers and vendors of health information technologies should give more serious consideration to human-centered design and routine use of multiple HF/SE methods, including understanding of the actual work process, incorporation of usability techniques (e.g. heuristic evaluation) early in the design process, user testing in simulated environments, and monitoring of safety problems of technology-in-use.(15, 52)

Regulatory and reimbursement agencies

Agencies and organizations involved in healthcare regulation and financing have important roles to accelerate HF/SE adoption into patient safety. As three examples, they can support large-scale demonstration projects needed to create evidence and broader visibility for HF/SE value, provide incentives for healthcare organizations to integrate HF/SE into patient safety improvement, and create infrastructures to support HF/SE application to small and rural healthcare organizations that are unlikely to have sufficient internal resources to do so themselves.

A large majority of healthcare organizations that are investing in HF/SE tend to be either academic medical centers, large non-academic health systems, or healthcare systems conveniently co-located with HF/SE university programs, all representing only a subset of those that would benefit from greater HF/SE use. Federal and state agencies thus should play important roles in developing mechanisms to assist rural, smaller, and under-resourced healthcare organizations leverage HF/SE to improve patient safety, reduce disparities, and control costs.

Funding organizations

Significant funding increases are needed for HF/SE research in patient safety, in particular in the areas of diagnostic safety, behavioral health, home care, and care transitions. Home-based patients interact with diverse actors and entities during their care journeys, all of which impact safety through a system of interacting systems that HF/SE can help improve through analysis, modeling, and design of detection, failure and vulnerability mitigation, and recovery processes.

Funding agencies such as AHRQ, NIH, and NSF recently have created programs to stimulate integration of HF/SE and systems science with health services research, including patient safety research. These include AHRQ’s investment in 13 Patient Safety Learning Laboratories, several NIH funding announcements for systems engineering modeling in disparities and behavioral health, and joint NSF-NIH funding program (e.g. Smart and Connected Health) to partner healthcare information technology, engineering modeling, and cognitive/behavioral scientists. While relatively new programs, results-to-date provide further evidence of the value of partnering healthcare and HF/SE researchers.(18, 53) However, such opportunities are a minority and greater funding is needed, especially given the size and scope of the healthcare industry and patient safety issues.

Future research should demonstrate the value of HF/SE to patient safety in primary care and home care. In particular, opportunities exist to demonstrate the value of HF/SE so that devices and health information technologies better support information access, communication, patient self-management, and other essential aspects of home care critical to medication safety, early identification of health problems and prompt resolution, and support of safe self-care by patients.

National coordinated efforts

While individual activities along the above lines should continue and expand, more coordinated and widespread efforts will be required to fully realize the potential of HF/SE. As an example, in 2010 the Veterans Health Administration funded the creation of four Veterans Engineering Resource Centers to deploy systems engineering at scale throughout its system. The CMS Innovation Center (CMMI) funded a similar effort to test a healthcare systems engineering regional extension center model, and the University of Texas health system funded a demonstration project to partner clinicians and engineers across its medical center campus system.

While each initiative achieved partial success, they all faced organizational resource, deployment, and scale challenges. We therefore recommend a more coordinated larger national program charged with developing and implementing a long-term (10-year) vision for spreading HF/SE throughout health care, similar in spirit to the national effort of the Office for the National Coordinator and associated investments in health information technology. A national HF/SE initiative could be led by either the federal government, a coalition of foundations, or a coordinated network of large health systems and might be responsible for operationalizing some of the above workforce development, HF/SE value visibility initiatives, fellowship and training programs, large-scale national impact projects, academic partnership mechanisms, curricula unification, and deployment and assistance programs.

Conclusion

The value of HF/SE for improving patient safety is clear yet under-realized. Important and demonstrated application areas include hospital-acquired infections, medication safety, medical device design and usability, reliability design and fault tolerance, cognition and cognitive burden, workload analysis, clinician burnout, and complexity management. Several challenges exist, however, to broaden the scale, application, and benefit from these methods. Appreciation of systems thinking and systems-of-systems awareness has been slow in health care and should be cultivated more broadly. The above recommendations would foster significantly greater and impactful healthcare HF/SE application, education, and research initiatives to help address the many patient safety challenges facing healthcare.