Attention Among Health Care Professionals

Abstract

Background:

The concept of attention can provide insight into the needs of clinicians and how health systems design can impact patient care quality and medical errors.

Purpose:

To conduct a scoping review to 1) identify and characterize literature relevant to clinician attention; 2) compile metrics used to measure attention; and 3) create a framework of key concepts.

Data Sources:

Cumulated Index to Nursing and Allied Health Literature (CINAHL), Medline (PubMed), and Embase (Ovid) from 2001 to 26 February 2024.

Study Selection:

English-language studies addressing health care worker attention in patient care. At least dual review and data abstraction.

Data Extraction:

Article information, health care professional studied, practice environment, study design and intent, factor type related to attention, and metrics of attention used.

Data Synthesis:

Of 6448 screened articles, 585 met inclusion criteria. Most studies were descriptive (n = 469) versus investigational (n = 116). More studies focused on barriers to attention (n = 387; 342 descriptive and 45 investigational) versus facilitators to improving attention (n = 198; 112 descriptive and 86 investigational). We developed a framework, grouping studies into 6 categories: 1) definitions of attention, 2) the clinical environment and its effect on attention, 3) personal factors affecting attention, 4) relationships between interventions or factors that affect attention and patient outcomes, 5) the effect of clinical alarms and alarm fatigue on attention, and 6) health information technology’s effect on attention. Eighty-two metrics were used to measure attention.

Limitations:

Does not synthesize answers to specific questions. Quality of studies was not assessed.

Conclusion:

This overview may be a resource for researchers, quality improvement experts, and health system leaders to improve clinical environments. Future systematic reviews may synthesize evidence on metrics to measure attention and on the effectiveness of barriers or facilitators related to attention.

Primary Funding Source:

None.

A clinician’s typical workday involves careful decision making, technical expertise, and, often, moments of deep human connection. All of this work can be meaningful and satisfying and, to be done well, requires a clinician’s full attention. However, care environments do not always facilitate a state of attention, leading to harm for both patients (medical errors) and clinicians (burnout) (1–4). Patient complexity continues to increase, as measured by various characteristics including multimorbidity, polypharmacy, and risk for readmission (5), while the complexity of the systems and tasks required of the clinician navigating care systems is also increasing (6). Cognitive failures and attentional lapses contribute to medical errors (7, 8), and, sometimes, the well-intentioned efforts to avoid them create environments that are less amenable to the interpersonal and individualized work of care (2).

We propose that the concept of an “ecology of attention” can serve as a useful guiding framework with which to understand the needs of clinicians and the impact of health systems design on patient care (9–11). We define attention as a state of presence, focus, and selective incorporation of information within clinical environments (12). In biology, “ecology” describes the entirety of interrelationships between an organism and its environment. A comprehensive ecology of clinician attention would describe analogous interrelationships between clinician attention and the clinical environment, suggesting key measurable factors that influence the ability of clinicians to maintain presence and focus. This concept is informed not only by ecology but also by thinkers in narrative theory (13), anthropology (14), philosophy (15), organizational psychology (16), medical education (17), ethics (18), and quality and safety (19) and emphasizes the ways that environments shape practice.

A well-functioning environment enables good care without overtaxing the clinician’s cognitive or emotional resources. But simply creating a “sterile cockpit” (20) for a distraction-free environment is unrealistic in most health care settings. Furthermore, because so many different aspects of the environment influence attention, a siloed approach that focuses on improving only one care aspect at a time will likely fall short. To achieve the goal of well-divided attention, it is necessary to take an ecological approach (21, 22) in which interdependent elements are considered, and measured, together.

From this starting place, we sought to gain a comprehensive understanding of the elements of the clinical ecosystem that contribute to an ecology of attention. We used the scoping review method to cast a wide net to identify key bodies of work on many asyet-unconnected elements and to identify gaps to guide future work. The goal of this work is to describe the breadth of work for clinicians, researchers, and administrators and to enhance the understanding of the ecology of attention and its impact on health care professionals and patient care.

Methods

We conducted a scoping review to examine the broad literature on clinician attention in the context of direct patient care. Consistent with scoping review methods following the Joanna Briggs Institute (JBI) (23), the Arksey and O’Malley scoping review guidelines (24), and our peer-reviewed protocol (25), we systematically mapped the literature and identified key concepts, theories, evidence, and research gaps. The review is reported following the published protocol and the PRISMA-ScR (Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews) checklist (25, 26).

Data Sources and Searches

The search was conducted in iterative stages per JBI guidelines and in consultation with a health sciences research librarian (23). In phase 1, we first developed and refined the search strategy, analyzing the text of selected articles to identify additional terms and optimizing the search terms using key conceptual, population, and exclusion terms. Subsequently, we conducted a complete search of Cumulated Index to Nursing and Allied Health Literature (CINAHL), Medline (PubMed), and Embase (Ovid) using the refined search strategy (Supplement Tables 1a and 1b in Supplement 1, available at Annals.org [Supplement Tables 2 to 5 can also be found in Supplement 1]). In phase 2, we handsearched references of included studies for relevant articles. In phase 3, we updated the search based on our revised terms to include articles published while completing the initial 2 phases of the review. The search strategy was iterative to be as comprehensive as possible. The last search was performed on 26 February 2024.

Study Selection

We included English-language studies published between 2001 and 26 February 2024. Eligible studies needed to describe attention-related concepts that could impact patient care. Patient care settings could be inpatient, outpatient, and simulated care provision settings. Health care professionals included nurses, physician assistants, advanced practice registered nurses, physicians, pharmacists, and any other patient-facing health care worker. Macroeconomics, policy, database architecture, or theoretical and neuroscientific studies of cognitive functioning removed from clinical work were not considered direct patient care.

We excluded studies focused only on non–health care personnel, non–patient care–related activities, patient-specific cognitive and behavioral experiences of attention, and books. Articles that could not be accessed in English and published before 2001 and after 26 February 2024 were also excluded. A complete list of both inclusion and exclusion criteria can be found in Supplement Table 2, available at Annals.org (25).

Titles and abstracts from each article were screened independently by at least 2 researchers (M.J.K., M.K., S.P., M.B., K.K., A.K.) using Rayyan (Qatar Computing Research Institute). The interrater agreement was 93% in phase 1, 89% in phase 2, and 92% in phase 3, all greater than the 75% threshold indicated in our protocol. Conflicts were resolved by 2 researchers (M.J.K. and S.P.). Full texts were again reviewed by at least 2 reviewers (M.J.K., M.K., S.P., K.K., A.K., M.B.) for each potentially relevant article (Figure 1). All articles meeting the full inclusion criteria were included.

Figure 1. PRISMA flow diagram.

Phase 1 represents the initial search through 19 June 2021 using the refined search strategy (Supplement Tables 1a and 1b). Phase 2 is a reference review from phase 1 studies. Phase 3 updates the search through 26 February 2024.

Data Extraction and Quality Assessment

Reviewers used a precalibrated form (Supplement Table 3, available at Annals.org) informed by the matrix method of scientific literature review (27) to capture key concepts of interest. Three reviewers (M.J.K., S.P., and M.K.) independently extracted data on title, authors, journal, publication year, health care professional population (clinicians, including physicians, physician assistants, advance practice registered nurses, and pharmacists, or nurses; multiple; other, such as cardiopulmonary perfusionists, respiratory therapists, anesthesia technicians; or none), practice environment (inpatient, intensive care unit, operating room, emergency department, outpatient, simulated, or other), study methods (quantitative, qualitative, mixed methods, or other), and any metrics of attention used (Supplement 2, available at Annals.org). Papers categorized as “other” environments, methods, or articles not specific to a health care professional population may include perspective or review papers, laboratory studies, or studies of a general clinical environment, such as those measuring noise, or studies in nursing homes or tumor boards.

Two reviewers (M.J.K. and S.P.) also categorized papers as descriptive or interventional and by whether each study was focused on barriers to or facilitators of attention. Descriptive studies had no interventional component but outlined or defined a problem or concept related to clinician attention or quantified interruptions or cognitive load without any intervention. Interventional papers measured the effect of an action in a simulated or actual clinical environment compared with a control group or a preintervention state. Descriptive barriers-to-attention studies might quantify or describe factors that pull attention away from relevant information or from the patient. Interventional barriers-to-attention studies might introduce interruptions and measure the effect on clinician attention. Descriptive facilitators-of-attention studies may quantify or describe factors or approaches that enhance clinicians’ ability to pay attention to relevant information or to the patient. Interventional facilitators-of-attention studies may take action to reduce distractions and measure the effect on clinician attention or patient care.

As this was a scoping review and not a systematic review or meta-analysis, study quality was not assessed.

Data Synthesis and Analysis

Three reviewers (M.J.K., S.P., and M.K.) met throughout the process of abstracting and analyzing the data to discuss emerging findings. Through this iterative process, similarities and themes emerged. Key terms and concepts were identified and mapped using a modified open-coding method (28) by taking major outcomes and topics of each study as themes to code, rather than line-by-line coding as would be done in a qualitative synthesis. This yielded categories that we used to develop an analytic framework.

Metrics of attention were also collected and categorized. Harvest plots and figures were used to organize and summarize study populations and settings, study types, and purpose.

Role of the Funding Source

This work was not funded.

Results

Using an iterative search strategy, we screened a total of 6448 abstracts and 701 full-text articles, ultimately selecting 585 articles for inclusion. We found that most studies were descriptive (469 papers) as opposed to investigational (116 papers). More studies focused on barriers to attention (387 papers; 342 descriptive and 45 investigational) versus facilitators to improving attention (198 papers; 112 descriptive and 86 investigational) (Figure 2).

Figure 2.

Included studies that are descriptive versus interventional and focused on barriers or facilitators of attention.

Square areas are proportional to the number of studies in each quadrant. Tree plots demonstrate the specific barriers to attention and facilitators of attention that were studied. These are also coded in Supplement 1. EHR = electronic health record.

Categories of Studies on Attention

Six major categories emerged in the process of reviewing studies on attention, as summarized in the analytic framework (Figure 3) and Table 1. One category (category A) focused on definitions and key terms related to attention. Four categories (categories B, C, E, F) examined how different factors related to attention. The last category (category D) examined the effect of attention on patient care quality.

Figure 3.

Analytic framework demonstrating factors that influence clinician attention and consequences of enhanced or diminished attention.

Circled letters correspond with categories of literature as outlined in our scoping review. ( + ) means facilitators or facilitation of attention, and (−) means barriers or diminishment of attention. A = Definitions of Attention and Related Concepts. B = The Clinical Environment and Its Effect on Attention. C = Personal Factors Affecting Attention. D = Relationship Between Attention and Patient Outcomes. E = Clinical Equipment, Alarms, and Alerts on Attention, and Alert Fatigue. F = Health IT on Attention and Usability. IT = information technology.

Table 1.

Six Category Groupings of Studies Related to Attention

Category	Description	Studies, n (%)
A: Definitions of Attention and Related Concepts	Papers that focused on foundational concepts, defining individual areas of study that connect to the broader concept of attention.	22 (4)
B: The Clinical Environment and Its Effect on Attention	Descriptive and observational studies characterizing the care environment with respect to workflow, interruptions, noise, and measures of attention and cognitive load.	221 (38)
C: Personal Factors Affecting Attention	Studies addressing the effect of sleep, burnout, rapid shift turnover, and similar factors on neurocognitive outcomes.	57 (10)
D: Relationship Between Attention and Patient Outcomes	QI and research that took additional steps toward characterizing patient-facing outcomes (such as medication errors) associated with personal and environmental factors affecting attention (categories B and C).	116 (20)
E: Clinical Equipment, Alarms, and Alerts on Attention, and Alert Fatigue	Distinct body of literature on the development, testing, and implementation of clinical alarms and alerts.	50 (8)
F: Health IT on Attention and Usability	Studies on the design and implementation of technology in the clinical encounter, including EHR design, data visualization interfaces, and clinical decision support, and the way these technologies influence clinician cognitive load and attention.	119 (20)

EHR= electronic health record; IT= information technology; QI= quality improvement.

Harvest plots summarize the study design methods, health care settings, and health care professionals of individual studies for each category (Figure 4).

Figure 4.

Harvest plots: Synthesis of study design across all 6 categories.

A supermatrix covering all 6 identified categories consisting of 7 rows (1 for each study setting) and 5 columns (1 for each study population). Study design method is indicated by bar shade; number of studies is indicated by height of the bar. ED = emergency department; ICU = intensive care unit; IT = information technology.

Category A: Definitions of Attention and Related Concepts

We identified 22 articles (4%) focused on foundational concepts of attention. Per our protocol, we extracted key terms and their definitions (Supplement Table 4, available at Annals.org). Most studies provided some definition of attention or concepts related to attention, such as cognitive load theory and vigilance. Other related concepts came from disciplines such as psychology (29), anthropology (30), and communication (19, 31) and included less-familiar terms such as socio-institutional enclosure, interaction design, and manageable cockpits. A subset of articles in this category addressed the phenomena of clinician presence, attentiveness, and philosophies of care as an element in interpersonal dimensions of care (14, 30–34).

From these definitions, we found that attention allows clinicians to be receptive to new information and open to the needs of others (35) through “attentiveness” (14, 30, 35) or “presence” (34). Such attentiveness is often in itself meaningful to clinicians (14, 33). Outside of the interpersonal realm, attention also helps clinicians provide safe, high-quality care as they remain vigilant for the deterioration of high-risk patients (36) or interact with technology that informs care decisions through clinical decision support (19, 37). But the human capacity for attention is limited, and a person’s mental resources can be overwhelmed. One framework for understanding this is cognitive load theory (38). Cognitive resources can be depleted by personal factors, such as lack of sleep, exercise, or burnout, and by unnecessary distractions.

Category B: The Clinical Environment and Its Effect on Attention

The largest percentage of articles (n = 221; 38%) involved direct observations of the clinical environment to assess which aspects interfere with or promote focus and presence such as communication patterns, workflow, interruptions, task switching, and multitasking. Methodologically, this category was descriptive and observational, and differentiated from category D in that this category only looked at the relationship between the clinical environment and its effect on clinician attention and did not establish correlations between the clinical environment and clinical outcomes. Most studies were in the inpatient (n = 77) or emergency setting (n = 44) with relatively fewer in the outpatient setting (n = 29). Most studied clinicians (n = 111) as opposed to nurses (n = 52). Methods were mostly quantitative (n = 147) and included direct observation time-motion studies, retrospective surveys, gaze and eye-tracking measures, and ambient noise measurements. Interruptions and multitasking were largely categorized as barriers to attention, though several studies (39–43) suggested that some interruptions can have positive effects on patient care, and that studying interruptions with a higher level of specificity (such as by characterizing the types, motivations, and outcomes of interruptions) is an important part of workflow assessment. Several standardized tools were identified to measure workflow (for example, Work Observation Method by Activity Timing [WOMBAT]) (44) or cognitive load (for example, National Aeronautics and Space Administration Task Load Index [NASA-TLX]) (45–47).

Few studied new communication technologies, such as electronic messaging, or evaluated ways to organize culture, workload, and technology to affect attention. Future research could address how to differentiate and triage beneficial from detrimental interruptions.

Category C: Personal Factors Affecting Attention

Fifty-seven studies (10%) addressed factors that might influence the cognition of working clinicians, including sleep, fatigue, shift timing, and burnout. Most of these studies were in the inpatient (n = 21) or emergency (n = 8) setting. Most studied clinicians (n = 36) versus nurses (n = 12). Most of these studies were quantitative (n = 50) and many used psychometric tests to measure traits such as vigilance, object discrimination, reaction time, and situational awareness in both controlled and uncontrolled settings, whereas other studies used more neurologically based metrics such as electroencephalogram measurements (48). This category had the largest absolute number of neuropsychological metrics of all categories, including visual, cognitive, memory, and motor measurements. However, most of these studies were related to sleep deprivation rather than other elements of the cognitive environment.

Future research might include tests of individualized work preferences (like a preference for multitasking, known as “polychronicity”) or of the aspects of attentional capacity that may change situationally (such as lack of sleep or high levels of workload). A validated set of these metrics, combined with metrics of cognitive load, would allow for comprehensive evaluation and monitoring of individual clinicians’ attentional capacity.

Category D: Relationship Between Interventions or Factors That Affect Attention and Patient Outcomes

One hundred sixteen studies (20%) analyzed the effect of interruption, distraction, multitasking, and structured interventions to increase or decrease attention on clinical processes or outcome measures. A large proportion of these studies (n = 37) involved medication administration by nurses in the inpatient environment. Several interventions were proposed in these studies, including interruption-free zones (49–54), changes to the physical design of medication rooms (52), and signals, such as tabards, worn by nurses during medication administration (53–55). The remainder of the studies were heterogenous and included varied topics such as impact of increasing cognitive load on simulated surgical performance (for example, via multi-tasking), quality improvement projects aimed at noise reduction, and structured rounds in the critical care environment. Several studies noted an increased risk for errors and procedural failures with a higher incidence of interruption, distraction, or multitasking (56, 57). There was a relative lack of studies that had robust links between these factors and quantifiable patient harm.

Future research can build on quality improvement techniques to change the clinical environment to impact attention, monitor subsequent improvements in patient outcomes, and disseminate positive results and best practices. In particular, high-risk tasks such as bedside procedures and cardiopulmonary resuscitation; high-stakes interpersonal tasks such as breaking bad news, articulating a complex care plan, or negotiating a discharge against medical advice; and the complexity of patient assignments could benefit from quality improvement efforts.

Category E: Clinical Equipment, Alarms, and Alerts Affecting Attention

Fifty studies (8%) specifically analyzed the impact of clinical alarms and alerts on attention and alert fatigue, mostly in the inpatient (n = 22) or intensive care unit (n = 10) setting. This category was distinct from the human–computer interface category (category F) in that the focus of these studies was on interruptive alerts and approaches to reduce alert fatigue. Although studies in this category are reminiscent of category B, the unique concept of alert fatigue has its own extensive body of literature that goes beyond a mere description of alarms in the environment. Studies in this category characterize how alarm qualities impact their usability in the clinical setting and investigate the balance between alarm sensitivity and specificity, with an acknowledgment of the potential for errors and inefficiencies associated with overalarming or underalarming (58, 59). Some articles proposed various measures for alarm fatigue, including measurement of alert volume, number of alarms accepted or overridden, time to process an order in the presence of alerts, and false-positive rates (39).

Although studies examined the appropriate thresholds for alarm sensitivity (60–63), fewer articles examined how clinicians actually interact with alarms and how alarms change their work day or affect their cognitive fatigue or situational awareness. Future research may study how to integrate alarms and alerts with new forms of data visualization (discussed more in category F) to consolidate the various sources of information in the clinical environment.

Category F: Health Information Technology Affecting Attention

One hundred nineteen studies addressed technology (20%), including electronic health records (EHRs), clinical decision support systems, and data visualization outputs, and their relationship to cognitive load, error rates, and system usability. About half of these studies involved clinicians alone (n = 58), either in the outpatient (n = 18) or inpatient (n = 11) setting, and many in simulation settings (n = 14). Several articles specifically addressed human-centered design principles in creating and implementing information technologies (37, 64, 65). During the 23 years covered by this review, computers and health information technology have increasingly been integrated into communication technologies within the health care environment. A 2016 scoping review (66) summarized the ways that computers and health information technology affected the clinical encounter, “impacting eye contact and gaze, information sharing, building relationships, and pauses in the conversation.” General themes encountered in this category included the presentation of clinical information to the clinician in a timely and context-specific manner, user-centric data synthesis, and ways to appropriately guide the user through tasks while minimizing extraneous action. This literature spanned informatics publications, clinical usability and accuracy studies, and simulations.

Metrics of Attention

From 585 studies, 84 different metrics were used to define different aspects of attention (Supplement Table 5, available at Annals.org). We identified 36 different aspects of attention to organize metrics (Table 2). Over half of the metrics (n = 43) were related to just 6 aspects of attention: cognitive load (n = 13), user-centered design (a set of principles by which input about improving a process or product is iteratively taken from those who will be using it [n = 4]), user experience design (the iterative process of creating an experience using input from a user [n = 6]), alarm fatigue (n = 6), classification of interruptions (n = 5), and EHR activity (such as time spent doing various EHR tasks [n = 9]).

Table 2.

Summary of Metrics for Particular Aspects of Attention Identified in Included Studies

Aspectof Attention Measured	Metrics, n	Metrics
Active/passive listening	1	Handoff Evaluation Assessing Receivers (HEAR) Checklist
Alarm fatigue	6	Video monitoring and alarm categorization
		Inappropriate overrides/alert quantity
		Nurse perception of alarm positive predictive value
		Duration of alarm/alarm response time
		Alert value
		Number of overridden alarms
Alertness	1	AlertMeter
Attention, concentration	1	Pauli Test
Attention, multidomain	2	Test for Attentional Performance
		MindStreams Global Assessment Battery
Attention, working memory, visual processing	1	Wisconsin Card Sorting Test
Attentional control	1	Attentional Network Test
Attentional failure	1	Continuous electrooculography
Cognitive functioning	1	Digit Symbol Substitution Test
Cognitive load	13	Dual-Task Test/Secondary Task Analysis
		Blink rate
		Paas Scale
		Pupillary diameter
		Cognitive Load Component Questionnaire (CLC)
		NASA-TLX + Nielsen’s Heuristics
		Cognitive task analysis
		Convex hull area
		Nearest Neighbor Index
		Mean saccade length
		Mean fixation duration
		Heart rate variability (HRV)
		Mean arterial pressure (MAP)
Decision making	1	Judgment policy change
Dispositional mindfulness	1	Mindful Attention Awareness Scale
Distraction burden	2	Disruptions in Surgery Index (DiSI)
		Medication Administration Distraction Observation Sheet (MADOS)
EHR activity	9	Medication Administration Distraction Observation Sheet (MADOS)
		Total EHR time
		Work outside of work
		Time on encounter note documentation
		Time on prescriptions
		Time on inbox
		Teamwork for orders
		Undivided attention
		Physician Documentation Quality Instrument (PDQI-9)
Interruption classification	5	Interruption measurement and classification
		Interruption taxonomy-purpose of interruption
		Task Tracker
		Interruption taxonomy-disturbing the work process
		Dual perspectives method for measuring interruptions
Memory performance	1	Rey Auditory Verbal Learning
Memory, multidomain	1	Wechsler Memory Scale
Multitasking	1	Multi-Tasking Assessment Tool (MTAT)
Neuropsychological attention	1	D2 test
Polychronicity	1	Inventory of Polychronic Values
Processing speed	2	Stroop task
		Symbol Searching Test
Reactive inhibition	1	Stop signal task
Recall (immediate and long)	1	California Verbal Learning Test
Situational awareness	2	Situational Awareness Rating Technique (SART)
		Situational Awareness Global Assessment Technique (SAGAT)
Sustained attention	1	Psychomotor vigilance task
Sustained attention, selective attention	1	Continuous performance task
Task switching	1	Trail Making Test B
Time and motion interruption tracking	1	TimeCat
Time and motion	1	Work Observation Method by Activity Timing (WOMBAT)
User design	4	Instrument for Evaluating Human Factors Principles in Medication-Related Decision Support Alerts (I-MeDeSa)
		Video observation
		Technology acceptance model (TAM)
		Clinical, Human and organizational, Educational, Administrative, Technical, and Social (CHEATS) Evaluation of User Design
User experience	6	Think Aloud Protocol
		Cognitive walkthrough
		System Usability Scale
		Time per task/time on task
		Clicks per task
		TURF
Variableaspects	1	Visual Analog Scale (VAS)
Visual attention	3	Trail Making Test A
		Reaction time testing
		Eye tracking
Working memory capacity	4	Eye tracking
		Digit span
		Span spatial
		Operation span task
Working memory	2	Running memory span
		N-back test
Workload	2	NASA-TLX
		Subjective Workload Assessment Technique (SWAT)

EHR= electronic health record; NASA-TLX= The National Aeronautics and Space Administration Task Load Index; TURF= task, user, representation, and function.

We did not assess whether metrics have been validated, particularly in a clinical environment.

Discussion

This review organizes a large and heterogeneous body of work around the role of attention in the clinical environment and how it is distributed between patient-centered concerns, cognitive effort, and the salient environment. Studies are categorized by the type of factors studied and how they relate to different aspects of attention or related patient-centered outcomes with diverse goals and methods. We sought a deeper understanding of attention and its importance in medical care to motivate work that is not siloed but instead considers all related factors to optimize clinician attention. In a previously published perspective piece (9), we proposed that the metaphor of an “ecology of attention” was useful for understanding the interdependence of these factors, their impact on clinicians, and ways to improve them. Drawing from others outside of medicine (21, 22), we believe working toward this kind of ecological framework can help guide efforts to improve patient safety, patient experience, quality, and health care worker wellness.

We identified 6 categories of literature related to attention. These were the clinical environment and its effect on attention, personal factors affecting attention, relationships between interventions or factors that affect attention and patient outcomes, the effect of clinical alarms and alarm fatigue on attention, and health information technology’s effect on attention. Clinicians must be attentive to both their patients and to clinical data to provide good care. As we show in the analytic framework (Figure 3), both the clinical and personal environment, and the flow of information can positively or negatively affect a clinician’s attention to patients and information. If clinicians’ attention is too taxed by the environment or the flow of information, it might contribute to deterioration in patient–physician relationships or even lead to adverse events due to missed diagnoses or inappropriate triage. However, relatively fewer studies investigated the effect of interventions on outcomes, and this is an area for future research. Across these categories, the bulk of studies were primarily descriptive of barriers and facilitators of attention, but there were sufficient studies on how different factors may improve (176 studies) or detract from attention (409 studies). A subsequent systematic review could examine the effectiveness of different factors to improve the clinical environment.

An important concept that was present to some degree in each category was cognitive load theory, or the theory of how humans filter, process, store, and retrieve information needed for decision making (38). Cognitive load is a helpful concept for understanding how clinician attention is directed or misdirected to various pieces of clinical information or data, but it may fall short in understanding how clinician attention is directed toward patients to create meaningful relationships and absorb nonquantitative data. Other concepts like clinician presence, or “a purposeful practice of awareness, focus, and attention with the intent to understand and connect with individuals/patients” (34), will need to be further developed in future work with regard to how to measure it and how to design environments that support it to reach parity with cognitive load. Future research into how to optimize attention will need to understand both minimizing extraneous cognitive load and facilitating presence.

The metrics of attention collected from the studies included in this review may be considered by researchers and quality improvement experts as candidates for use in existing and future studies. These metrics are organized by the aspect of attention that they claim to measure according to their definitions, and citations for the studies are provided in Supplement 1. However, future systematic reviews could evaluate the discrimination, calibration, reclassification, validity, and utility of such measures in the clinical environment.

Emerging technologies are continuing to change the ways that we perform clinical work (67) and measuring factors related to attention in the workplace is more achievable than ever before. The EHRs can record work patterns during clinicians’ routine work by capturing clinician actions through audit log data (68). These technologies offer the ability to unobtrusively measure some types of distractions, such as instant messages, pop-ups, and scattered data visualization, and link to cognitive errors and even burnout (69). This emerging science will soon allow for real-time assessments of work patterns, particularly when work processes are changing, allowing organizations to assess distractions in ways not previously feasible. Researchers and those involved in quality improvement should pay particular attention to ensuring that measurement is not invasive (for example, that the measurement does not add to distractions and interruptions by trying to measure them). Just as importantly, health care innovators and leaders must grapple with the potential implications of these measurements on worker autonomy and privacy, and that we are “measuring what matters” (2) with regard to patient safety, quality, and human-centered design.

We believe the findings from our scoping review and the analytic framework will help researchers or quality improvement experts studying or looking to improve the “ecology of attention.” They should familiarize themselves with the fields of alarm fatigue, human-centered design, and cognitive load theory and should be conversant in theories of attentiveness developed outside of medical practice (including cognitive science, psychology, anthropology, and the humanities).

A strength of our work is that it comprehensively collates and categorizes the various measures that might be used to study different domains of attention within the clinical environment. We cast a wide net to include and organize all studies on attention and related concepts with widely different methods.

Our study also has several limitations. Consistent with scoping review methods, we did not assess the quality or methodological differences between articles, nor did we perform data synthesis. We aimed to identify the breadth of papers related to attention and thus included a range of literature, including editorials, perspectives, and other gray literature beyond the traditional research study. We provide an overview of the field but did not focus on specific questions related to what improves or distracts attention and affects outcomes, and this may be an area for future research. Despite using broad inclusion criteria and an iterative approach to article inclusion, we may have missed key areas of study due to the inherent breadth of this topic. Finally, developing categories and summarizing trends all require interpretive steps, possibly allowing unconscious biases to contribute to the narrative descriptions. We attempted to account for this by using several reviewers.

In conclusion, to be effective, health care quality improvement efforts need to transition from siloed quality improvement interventions to multidisciplinary efforts that understand that health care is composed of complex and overlapping environments. Understanding the importance of clinician attention is a key element in this transition. We performed a broad scoping review to understand the current state of attentional research, used this research to identify a suite of metrics that can be used to study attentional domains, and organized our findings in 6 key categories of studies on attention and related concepts with the overarching goal of helping researchers, quality improvement experts, and health system leaders think more cohesively about designing systems meant to optimize clinician attention. Future systematic reviews may synthesize evidence on metrics to measure attention and on the effectiveness of barriers or facilitators related to attention.