- Email: [email protected]
Disruption of Radiologist Workflow
Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
Current Problems in Diagnostic Radiology journal homepage: www.cpdrjournal.com
Disruption of Radiologist Workflow Akash P. Kansagra, MD, MSa,n, Kevin Liu, MDa, John-Paul J. Yu, MD, PhDb a b
Mallinckrodt Institute of Radiology, Washington University School of Medicine, Saint Louis, MO Department of Radiology, University of California, San Francisco, San Francisco, CA
The effect of disruptions has been studied extensively in surgery and emergency medicine, and a number of solutions—such as preoperative checklists— have been implemented to enforce the integrity of critical safety-related workflows. Disruptions of the highly complex and cognitively demanding workflow of modern clinical radiology have only recently attracted attention as a potential safety hazard. In this article, we describe the variety of disruptions that arise in the reading room environment, review approaches that other specialties have taken to mitigate workflow disruption, and suggest possible solutions for workflow improvement in radiology. & 2015 Mosby, Inc. All rights reserved.
Introduction Society has recognized the role of disruptions in creating accidents and mishaps. Wherever the ensuing mishaps have the potential to cause harm or loss of life, society has enacted restrictions to minimize disruptions of normal workflow. For this reason, drivers in many states are prohibited from sending text messages or using handheld devices while driving, and airline pilots are mandated to maintain a “sterile cockpit” during critical phases of flight where only mission-related tasks are discussed. Medicine—where disruptions can easily cause harm and loss of life—has also started to respond to these challenges. The most publicized and mature examples to date involve the use of preprocedural checklists before surgery or central line insertion to enforce the integrity of critical safety-related workflows, resulting in dramatic improvements in patient safety and clinical outcomes.1-3 However, the potential for significant workflow disruption extends far beyond periprocedural care. In a busy radiology reading room, for example, radiologists must contend with a complex and fast-paced workflow characterized by frequent disruptions, disruptions that may be particularly problematic given the high cognitive demand of image interpretation.4-6 Unfortunately, the nonstandard nature of most radiology workflows reduces the potential effectiveness of basic interventions such as checklists and may require more sophisticated solutions.7 In this review, we describe the workflow disruptions with which radiologists must contend in daily practice, highlight steps that other specialties have taken to respond to workflow disruptions, and suggest measures that can be taken to mitigate similar disruptions in radiology. It is our hope that this review will draw
n Reprint requests: Akash P. Kansagra, MD, MS, Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S. Kingshighway Blvd, Campus Box 8131, Saint Louis, MO 63110. E-mail address: [email protected] (A.P. Kansagra).
http://dx.doi.org/10.1067/j.cpradiol.2015.05.006 0363-0188/& 2015 Mosby, Inc. All rights reserved.
attention to the urgent need for improved clinical workflow in the reading room and provide a blueprint for safer and more effective radiological care.
Workflow Disruptions in Radiology The working environment in diagnostic radiology has undergone a tremendous change over the past 2 decades because of the widespread adoption of filmless imaging, introduction of speech recognition systems for report dictation, and the incorporation of electronic medical records (EMR) into an increasingly informationrich interpretive workflow.8 Unfortunately, the promised efficiency gains of these systems have been partially offset by a paradoxical increase in the complexity of radiologists’ workflow. This complexity reflects a number of converging trends, including the central and growing role of medical imaging in patient evaluation and management, as well as increasing fragmentation and disruption of interpretive workflows. In addition to their primary task of image interpretation and reporting, radiologists in modern practice must shoulder added responsibilities that can include frequent telephone communication, in-person physician consultations, technologist supervision, patient consent, ultrasound scanning, and management of contrast agent injections and adverse reactions.4,5,8 Although these additional tasks are important, they distract and detract from the primary workflow of image interpretation, create barriers to productivity, and likely contribute to errors in the knowledge-intensive service environment of clinical radiology.9 Of these many potential sources of disruption, telephone communication is particularly problematic, in part because many different sources of disruption are funneled through this common communication channel. As an example, incoming telephone calls may come from clinical providers inquiring about imaging findings or selection of appropriate imaging tests, or from technologists requesting “scan checks” to assess study adequacy or seeking
2
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
guidance for image acquisition (eg, protocol selection or solutions for specific technical challenges). Similarly, outgoing phone calls are often directed to providers to communicate time-sensitive imaging findings or solicit additional patient information. Recent work by Yu et al5 describes their experience with telephone-based disruption of on-call radiologists’ workflow. Their work reveals telephone interruptions of radiologists on a surprisingly large scale, with more than 10,000 after-hours calls directed to a single resident over a 3-month period. Incoming telephone calls occurred as frequently as once every 4 minutes during peak hours, which translated to roughly 2-3 expected interruptions during interpretation of a single CT of the abdomen and pelvis. Related work by Balint et al4 has suggested that the frequency of telephone disruptions in the hour preceding resident interpretation of a study was positively correlated with the likelihood of an incorrect interpretation. Importantly, disruptions to radiologists are not confined to the reading room; dysfunctional or inefficient workflows in other areas of the radiology department can secondarily increase burdens on radiologists. For example, inadequate systems to identify and triage patients for imaging may create a need for frequent radiologist involvement in study prioritization.8 Similarly, poorly designed systems for information transmission between different members of a radiology department (eg, radiologists, technologists, and patient transporters) can hinder effective care of patients, thereby requiring increased radiologist involvement to maintain appropriate and timely care.10 If left unchecked, the scale of workflow disruption is likely to increase. As the information economy of medicine continues to grow in scale and complexity, there is likely to be increased reliance on specialties such as radiology that can create and share objective patient information. Against this backdrop, inefficient or ill-defined clinical workflows are likely to produce ever-increasing disruptions to radiologists. Thus, implementing solutions to dysfunctional workflows is a key component in building and maintaining an efficient information economy.
Managing Workflow Disruption in Nonradiological Settings Disruptions in workflow are not unique to radiology but are also experienced by other hospital-based specialties such as emergency medicine,11-19 critical care,20-25 and surgery.26-32 The solutions to these disruptions vary based on the specific workflow patterns in each patient care setting, but they can generally be grouped into several themes. Filtering Interruptions by Activity A basic strategy to improve workflow is to create physical or temporal barriers to interruption during activities that are of critical importance or particularly susceptible to disruption. In some cases, this strategy may amount simply to having individuals address potential sources of interruption at a convenient time, such as refilling intravenous fluids before nursing handoffs to prevent unnecessary alarming during transfer of care.23 In other cases, physical barriers—including possibilities such as signs or colored vests for individuals seeking to avoid interruption, or colored floor tiles or shields for specially designated areas—may be of value.23,33,34 These basic interventions can have a profound effect. One study found that implementing a visible “No Interruption Zone” around a medication dispensing station resulted in a 41% decrease in interruptions,33 while another found that erecting a wall around the medication dispensing station decreased interruptions by 81%.34 Alternatively, there may be value in gentler approach that allows for interruptions even during critical
activities, provided the interruption conveys important patientrelated information. In this context, interruptions may be mitigated by increasing the transparency of task importance so that potential interrupters can determine if interruption is appropriate.22 Filtering Interruptions by Acuity A large percentage of interruptions—even during critical tasks such as transfer of care discussions (“sign out”) and clinical rounds —are nonessential, with only 11% of interruptions during morning sign out and 27% during morning rounds being essential to patient care.21 As such, filtering nonessential interruptions may streamline workflow. Young et al35 describe a system in which nurse requests to send after-hours pages to resident physicians are reviewed by a charge nurse and categorized by acuity, with emergent pages transmitted immediately, urgent pages batched, and nonurgent pages deferred until the morning. Following implementation of this system, the total number of pages and number of nonurgent pages sent after-hours to house staff decreased. Asynchronous Communication Synchronous channels of communication require simultaneous participation of both parties, preventing the recipient of an interruption from managing the timing of that interruption.11 In contrast, asynchronous channels of communication provide the recipient of a message with control over the timing of disruptions, and this may therefore represent a practical method for acuitybased filtering, task prioritization, and reduced communication burden.36 Voicemail capability may be an effective means to reduce disruption. In a study of emergency department (ED) providers equipped with mobile phones, the lack of voicemail capability contributed significantly to workflow interruption, as the providers were forced to immediately answer any incoming call.13 Alternatively, landlines with a clerical receptionist may serve a similar role and help to reduce unnecessary interruptions.11 Alphanumeric pagers may also permit filtering of nonurgent interruptions,11,37 provided that the recipient of a message is provided with sufficient information to judge the urgency of the page. Unfortunately, a large percentage of alphanumeric pages contain only basic callback information, thereby preventing the receiver from performing effective task prioritization and mandating an immediate callback to determine the acuity of the page.37 A proposed explanation for this behavior is that synchronous communication provides receipt confirmation for the interrupter; asynchronous communication may benefit from a confirmation mechanism to encourage broader adherence.36 Technology-Assisted Workflow Electronic and nonelectronic whiteboards have been used extensively in a variety of care settings to organize and facilitate communication and workflow.36,38-42 For example, when used in the operating room as a basic information display system, electronic whiteboards can facilitate integration of safety checklists into preoperative workflow and aid intraoperative communication between multiple team members.43,44 Chaotic and disruptive workflows can be further streamlined with electronic systems that go beyond basic information display to serve as an integrated information technology (IT) solution.45 Aronsky et al46 have described the implementation of such a system in an ED, which allowed for easy information access, information sharing, and decision support using data from multiple hospital information systems, with resulting dramatic
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
increases in revenue, enhanced hospital-wide triage, and improved staffing. Such systems can facilitate high-fidelity communication, improve multitasking ability, and promote physician efficiency through automated results notification and performance feedback.47-49 Indeed, IT tools of this sort may help to reduce both the number and the adverse effects of interruptions.50,51 Eliminating Unnecessary Tasks Some time-consuming tasks can be entirely obviated through process redesign or IT solutions. For instance, automated systems to identify and physically locate patients, providers, and equipment can eliminate the need for multiple phone calls and pages.36,52 Similarly, normal triage pathways and full patient registration can be bypassed in some ED settings,40,42,48 while the considerable complexity of patient transfer between wards can be avoided with acuity-adjustable beds.49 Workload Redistribution Workload is not evenly distributed between individuals on a health care team. For instance, junior physicians in the intensive care unit (ICU) setting must contend with competing demands on their attention far more often than senior physicians do.53 Accordingly, sharing some aspects of workload may help to improve overall workflow. Simulations have shown that the time efficiency of rounding in the ICU can be dramatically improved by having all team members share responsibility for handling interruptions, thereby unburdening the trainee primarily responsible for a patient under active discussion.50 Similarly, workflow in the ED can be improved by sharing triage responsibilities among a larger number of providers or having a pool of on-demand nurses and physicians who can be recruited to help manage excessive workloads.40,54 There may also be workflow benefit from transferring some responsibilities to other appropriate individuals. For instance, designating a receptionist or nurse to answer phone calls and pages can help to reduce interruption to physicians.11,26 Similarly, assigning nonphysician staff to handle basic but time-consuming aspects of workflow such as medication reconciliation or intrahospital patient transfers can also promote workflow efficiency.55,56 Indeed, even direct patient care tasks can be assigned to nonphysician providers if there are clearly established protocols of care.40,42
3
has recommended that radiologists use communication channels other than the radiology report to convey this information to the treating physician in a timely fashion.58 To minimize the effect of these nonroutine communications on workflow, many practices have developed efficient channels of asynchronous communication that eliminate time-consuming tasks, redistribute workload away from radiologists, or incorporate technology-assisted workflows. A common approach is to automate communication of important but nonemergent findings, such as lung nodules and solid organ lesions. Johnson et al59 described the implementation of a clickable button within their dictation software that conveyed such findings to referring physicians via facsimile. Other groups have reported systems to generate e-mail alerts to referring physicians when specific, predefined statements were included in the final report text.60,61 In an alternative approach, Eisenberg et al62 redistributed rather than eliminated this workload by having the interpreting radiologist submit a case to a web-based portal and assigning a “communications facilitator” to the task of contacting the referring physician by e-mail, telephone, or pager. Most of these tools offered some form of receipt confirmation, and each was effective for communicating nonemergent findings with minimal workflow disruption and a high level of satisfaction from both radiologists and referring clinicians. Lacson et al63 developed an automated notification system that could be used even for emergent findings. In their system, radiologists were able to indicate the level of urgency of a finding within notification software that was integrated with their picture archiving and communication system (PACS) environment; the system then automatically contacted the appropriate provider using e-mail for nonurgent findings or pager for urgent findings and documented this notification within the EMR. Triage Assistants
Strategies to Reduce Workflow Disruption in Radiology
As part of larger efforts to more meaningfully integrate medical students into the clinical practice of radiology,64 several medical schools have experimented with using medical students to assist with triage in radiology reading rooms. Authors from at least 3 academic radiology departments have separately reported on the benefits of paying medical students to answer telephone calls and pages, and protocol radiology studies with the help of the on-call radiology resident.65-67 This arrangement provides educational and financial benefits to medical students while reducing disruptions to radiology residents. In particular, medical students can enforce acuity- and task-based filtering of interruptive communications (eg, holding nonurgent communications until the resident is between studies). Furthermore, some of the workload of soliciting history, performing medical record review, or contacting referring physicians could be redistributed from the resident to the medical student. Mamlouk et al68 have taken the idea of workload redistribution further and designated a radiology fellow as a “quality control” radiologist on a rotating basis during daytime hours. This radiologist is largely excused from interpretive tasks but handles most incoming phone calls from referring providers, protocols and prioritizes imaging studies, performs real-time scan checks, and customizes imaging for complex cases for the entire medical center, thereby preserving an efficient and undisrupted workflow for the remaining staff and trainees of the section. Notably, this solution simultaneously preserves efficiency of workflow while increasing the availability of consultative services and other valuebased practices.
Efficient Asynchronous Communication
Computerized Order Entry, Decision Support, and Protocoling
When presented with emergent or unexpected findings or interpretative discrepancies, the American College of Radiology
Computerized order entry for imaging services may help to avoid issues related to insufficient, misleading, or illegible clinical
Physical Layout Physical layout has been implicated as a source of greater-thannecessary interruption. In the operating room, 33% of workflow disruptions were attributable to suboptimal layout.57 Similarly, intensive care units can be designed in a way that discourages disruption of important tasks such as medication administration.52 Some authors have hypothesized that the high rate of interruptions in the ICU relates to a large number of collaborating providers being concentrated into a relatively compact space, and that a linear arrangement of rooms along a busy thoroughfare may lead to more interruptions than a more open arrangement that separates patient care zones from high traffic areas.20
4
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
information that often accompanies handwritten or otherwise nonstandardized requests for imaging.69-71 Indeed, these systems can aggregate relevant clinical data—referring physician contact information and patient location, among others—from multiple hospital systems and display this information in a standardized format.69 Thus, implementation of a technology-assisted workflow for clinicians may improve radiologists’ work efficiency by reducing the need for direct communication with clinicians or excessively detailed exploration of patients’ medical records before study protocoling or interpretation. Coupling computerized decision support tools with computerized order entry systems may amplify these benefits further. By encouraging appropriate imaging utilization at the time of order entry, these tools likely reduce the need for radiologists to interact directly with clinical providers to modify or eliminate unhelpful aspects of imaging evaluation. As an example, Rosenthal and colleagues have developed and implemented a decision support system that provides clinicians with a “utility score” based on clinical indication in the form of an ICD-9 code.72 Following implementation of this system, several major insurance carriers agreed to eliminate required preauthorization for high-cost imaging studies. Similarly, a system to alert referring providers to the presence of a contrast agent allergy and suggest prophylactic premedication dramatically increased the number of patients who were appropriately premedicated without radiologist intervention.73 Computerized order entry for radiologists in the form of electronic study protocoling can also improve radiologists’ workflow. As with computerized order entry for clinicians, these systems can effortlessly aggregate important patient-related data such as serum creatinine, allergies, and cardiac pacemakers from the EMR and implement automated safety checks related to specific study types (eg, magnetic resonance imaging and cardiac pacemakers) or pharmacologic agents (eg, iodinated contrast agent and renal function). Moreover, recallable electronic protocols may greatly expedite protocoling in cases where serial imaging is performed, such as multiple sclerosis or lung nodule surveillance. Integrated Ecosystem of Imaging Applications Large efficiency gains can also be realized with a tightly integrated ecosystem of clinical applications, including PACS, EMR, radiology information system (RIS), voice dictation software, electronic protocoling applications, teaching files, and image postprocessing software.74 As an example, context integration between PACS, EMR, RIS, and dictation software can allow an entirely PACS-driven workflow that automatically opens the patient’s information patient’s chart in the EMR and initiates reporting of the correct case in the dictation software and RIS. Similarly, an electronic protocoling system with single sign-on login and automatic patient lookup in the EMR may allow dramatically more efficient study protocoling than other, less tightly integrated systems.75,76 This same approach may be useful for radiologists who are asked to review studies performed at other facilities and stored on physical media such as compact discs. When no defined workflow exists, referring physicians may bring these CDs directly to the radiologist, who must divert his or her attention from other interpretive tasks to perform the time-consuming task of loading images from the CD. Finally, these images must actually be reviewed, often using one of the many cumbersome and often limited software image-viewing applications provided on those CDs. Instead, radiologists’ workflow efficiency may be substantially improved using software designed to facilitate image importation, such as LILA (LifeIMAGE, Newton, MA), or direct upload to PACS.77 These workflows unburden radiologists and allow images to be
reviewed in an application that is more suited to interpretative tasks. Scheduled Rounds The frequency of in-person consultation between radiologists and clinical practitioners has dramatically decreased with the widespread adoption of distributed PACS. Today, most in-person consultations occur sporadically, though some services—most commonly neonatology and trauma surgery—advocate for or continue to practice formalized radiology rounds as a means to improve patient care.78 In this setting, radiologists’ workflow may be improved by identifying a specific, mutually convenient time at which radiology rounds will occur. This practice would allow radiologists to address other clinical demands at a different time so that they may meet their interpretative responsibilities while also increasing their availability for regular consultation. Workload Optimization and Balancing Improved optimization of radiology workload can be achieved by identifying high-priority cases to be interpreted before less urgent cases, thereby reducing the number of incoming interruptive communications from clinicians in need of results to guide time-sensitive management conditions. Halsted and Froehle devised an automated work queuing system to prioritize cases based on patient medical acuity, wait time, psychological state, and other factors to present a prioritized queue to radiologists and demonstrated a nearly 28% increase in the mean time between workflow interruptions.8 Improved balancing of radiology workload can be achieved through more diversified scheduling of nonurgent examinations to more fairly utilize radiologists scattered across different reading rooms. If a disproportionate number of imaging studies on a given day are directed to a single reading room (eg, a large number of brain MRIs on a day that coincides with a scheduled neurooncology clinic), it may lead to excessive workflow for a small group of radiologists that can produce a backlog of studies awaiting interpretation and corresponding increase in disruptive communications from clinicians requesting results. In contrast, scheduling nonurgent studies to minimize day-to-day variations in imaging volume between separate reading rooms may permit more predictable and easily managed workloads. Embedded Radiologists Recent efforts to “embed” radiologists near their clinical counterparts may also affect radiologists’ workflow. A recent study demonstrated wide variability in the preferred modes of communication between radiologists and clinical providers depending on their physical proximity.79 Not surprisingly, embedded radiologists were more likely to communicate in person and less likely to use automated result notification systems. Nevertheless, it remains unclear if this alteration in physical layout improves or hinders a radiologist’s efficiency, as the concept of embedding radiologists has been promoted because of the promise of adding value to clinical care rather than increasing interpretative efficiency.
Conclusion Disruptions of radiologists’ workflow occur commonly and come from a wide range of sources, including phone calls, pages, in-person consultation, coordination of care, and burdensome IT tools. Other hospital-based specialties use different workflows than that of radiology but have developed solutions to mitigate
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
workflow disruption that can inspire improvements in radiology. Through process redesign and improved IT tools, radiologists can streamline workflow to reduce the number of disruptions and competing demands during the cognitively intense task of image interpretation and can thereby facilitate the delivery of more efficient and higher-quality radiological care.
References 1. de Vries EN, Prins HA, Crolla RM, et al. Effect of a comprehensive surgical safety system on patient outcomes. N Engl J Med 2010;363(20):1928–37. 2. Haynes AB, Weiser TG, Berry WR, et al. A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med 2009;360(5): 491–9. 3. Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med 2006;355 (26):2725–32. 4. Balint BJ, Steenburg SD, Lin H, et al. Do telephone call interruptions have an impact on radiology resident diagnostic accuracy? Acad Radiol 2014;21 (12):1623–8. 5. Yu JP, Kansagra AP, Mongan J. The radiologist’s workflow environment: Evaluation of disruptors and potential implications. J Am Coll Radiol 2014;11 (6):589–93. 6. Reiner BI, Krupinski E. The insidious problem of fatigue in medical imaging practice. J Digit Imaging 2012;25(1):3–6. 7. Berbaum K, Franken EA Jr, Caldwell RT, et al. Can a checklist reduce SOS errors in chest radiography? Acad Radiol 2006;13(3):296–304. 8. Halsted MJ, Froehle CM. Design, implementation, and assessment of a radiology workflow management system. AJR Am J Roentgenol 2008;191(2):321–7. 9. Froehle CM, White DL. Interruption and forgetting in knowledge-intensive service environments. Prod Oper Manag 2014;23(4):704–22. 10. Larson DB, Froehle CM, Johnson ND, et al. Communication in diagnostic radiology: Meeting the challenges of complexity. AJR Am J Roentgenol 2014;203(5): 957–64. 11. Brixey JJ, Robinson DJ, Turley JP, et al. The roles of MDs and RNs as initiators and recipients of interruptions in workflow. Int J Med Inform 2010;79(6):e109–115. 12. Woloshynowych M, Davis R, Brown R, et al. Communication patterns in a UK emergency department. Ann Emerg Med 2007;50(4):407–13. 13. Brixey JJ, Tang Z, Robinson DJ, et al. Interruptions in a level one trauma center: A case study. Int J Med Inform 2008;77(4):235–41. 14. Brixey JJ, Robinson DJ, Johnson CW, et al. Towards a hybrid method to categorize interruptions and activities in healthcare. Int J Med Inform 2007;76(11-12):812–20. 15. Chisholm CD, Dornfeld AM, Nelson DR, et al. Work interrupted: A comparison of workplace interruptions in emergency departments and primary care offices. Ann Emerg Med 2001;38(2):146–51. 16. France DJ, Levin S, Hemphill R, et al. Emergency physicians’ behaviors and workload in the presence of an electronic whiteboard. Int J Med Inform 2005;74(10):827–37. 17. Spencer R, Coiera E, Logan P. Variation in communication loads on clinical staff in the emergency department. Ann Emerg Med 2004;44(3):268–73. 18. Fairbanks RJ, Bisantz AM, Sunm M. Emergency department communication links and patterns. Ann Emerg Med 2007;50(4):396–406. 19. Chisholm CD, Collison EK, Nelson DR, et al. Emergency department workplace interruptions are emergency physicians “interrupt-driven” and “multitasking”? Acad Emerg Med 2000;7(11):1239–43. 20. Kalisch BJ, Aebersold M. Interruptions and multitasking in nursing care. Jt Comm J Qual Patient Saf 2010;36(3):126–32. 21. Nimmo GR, Mitchell C. A preliminary audit of interruptions in intensive care: Implications for patient safety. J Intensive Care Soc 2008;9(3):240–2. 22. Sasangohar F, Donmez B, Easty A, et al. Interruptions experienced by cardiovascular intensive care unit nurses: An observational study. J Crit Care 2014;29 (5):848–53. 23. Spooner AJ, Corley A, Chaboyer W, et al. Measurement of the frequency and source of interruptions occurring during bedside nursing handover in the intensive care unit: An observational study. Aust Crit Care 2015;28(1):19–23. 24. Weigl M, Muller A, Zupanc A, et al. Hospital doctors’ workflow interruptions and activities: An observation study. BMJ Qual Saf 2011;20(6):491–7. 25. Drews FA. The frequency and impact of task interruptions in the ICU. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2007;51 (11):683-686. 26. Healey AN, Sevdalis N, Vincent CA. Measuring intra-operative interference from distraction and interruption observed in the operating theatre. Ergonomics 2006;49(5-6):589–604. 27. Antoniadis S, Passauer-Baierl S, Baschnegger H, et al. Identification and interference of intraoperative distractions and interruptions in operating rooms. J Surg Res 2014;188(1):21–9. 28. Christian CK, Gustafson ML, Roth EM, et al. A prospective study of patient safety in the operating room. Surgery 2006;139(2):159–73. 29. Wiegmann DA, ElBardissi AW, Dearani JA, et al. Disruptions in surgical flow and their relationship to surgical errors: An exploratory investigation. Surgery 2007;142(5):658–65.
5
30. Weigl M, Antoniadis S, Chiapponi C, et al. The impact of intra-operative interruptions on surgeons’ perceived workload: An observational study in elective general and orthopedic surgery. Surg Endosc 2015;29(1):145–53. 31. Sevdalis N, Undre S, McDermott J, et al. Impact of intraoperative distractions on patient safety: A prospective descriptive study using validated instruments. World J Surg 2014;38(4):751–8. 32. Zheng B, Martinec DV, Cassera MA, et al. A quantitative study of disruption in the operating room during laparoscopic antireflux surgery. Surg Endosc 2008;22(10):2171–7. 33. Anthony K, Wiencek C, Bauer C, et al. No interruptions please: Impact of a No Interruption Zone on medication safety in intensive care units. Crit Care Nurse 2010;30(3):21–9. 34. Colligan L, Guerlain S, Steck SE, et al. Designing for distractions: A human factors approach to decreasing interruptions at a centralised medication station. BMJ Qual Saf 2012;21(11):939–47. 35. Young JB, Baker AC, Boehmer JK, et al. Using NNAPPS (Nighttime Nurse and Physician Paging System) to Maximize Resident Call Efficiency within 2011 Accreditation Council for Graduate Medical Education (ACGME) work hour restrictions. J Surg Educ 2012;69(6):819–25. 36. Xiao Y, Hu P, Moss J, et al. Opportunities and challenges in improving surgical work flow. Cogn Technol Work 2008;10(4):313–21. 37. Espino S, Cox D, Kaplan B. Alphanumeric paging: A potential source of problems in patient care and communication. J Surg Educ 2011;68(6):447–51. 38. Fairbanks RJ, Bisantz AM, Sunm M. Emergency department communication links and patterns. Ann Emerg Med 2007;50(4):396–406. 39. France DJ, Levin S, Hemphill R, et al. Emergency physicians’ behaviors and workload in the presence of an electronic whiteboard. Int J Med Inform 2005;74(10):827–37. 40. Yen K, Gorelick MH. Strategies to improve flow in the pediatric emergency department. Pediatr Emerg Care 2007;23(10):745–9 [quiz 750-741]. 41. Xiao Y, Schenkel S, Faraj S, et al. What whiteboards in a trauma center operating suite can teach us about emergency department communication. Ann Emerg Med 2007;50(4):387–95. 42. Wiler JL, Gentle C, Halfpenny JM, et al. Optimizing emergency department front-end operations. Ann Emerg Med 2010;55(2):142–160 e141. 43. Mainthia R, Lockney T, Zotov A, et al. Novel use of electronic whiteboard in the operating room increases surgical team compliance with pre-incision safety practices. Surgery 2012;151(5):660–6. 44. Parush A, Kramer C, Foster-Hunt T, et al. Communication and team situation awareness in the OR: Implications for augmentative information display. J Biomed Inform 2011;44(3):477–85. 45. Lemke HU, Ratib OM, Horii SC. Workflow in the operating room: A summary review of the Arrowhead 2004 Seminar on Imaging and Informatics. Int Congress Ser 2005;1281(0):862–7. 46. Aronsky D, Jones I, Lanaghan K, et al. Supporting patient care in the emergency department with a computerized whiteboard system. J Am Med Inform Assoc 2008;15(2):184–94. 47. Laxmisan A, Hakimzada F, Sayan OR, et al. The multitasking clinician: Decisionmaking and cognitive demand during and after team handoffs in emergency care. Int J Med Inform 2007;76(11-12):801–11. 48. Baumlin KM, Shapiro JS, Weiner C, et al. Clinical information system and process redesign improves emergency department efficiency. Jt Comm J Qual Patient Saf 2010;36(4):179–85. 49. Alameddine M, Dainty KN, Deber R, et al. The intensive care unit work environment: Current challenges and recommendations for the future. J Crit Care 2009;24(2):243–8. 50. Gupta A, Sharda R, Dong Y, et al. Improving rounding in critical care environments through management of interruptions. Decis Supp Syst 2013;55 (2):516–27. 51. Grundgeiger T, Sanderson P. Interruptions in healthcare: Theoretical views. Int J Med Inform 2009;78(5):293–307. 52. Thompson DR, Hamilton DK, Cadenhead CD, et al. Guidelines for intensive care unit design. Crit Care Med 2012;40(5):1586–600. 53. Malhotra S, Jordan D, Shortliffe E, et al. Workflow modeling in critical care: Piecing together your own puzzle. J Biomed Inform 2007;40(2):81–92. 54. Lu W. An agent-based simulation for workflow in Emergency Department. Paper Presented at Systems and Information Engineering Design Symposium, SIEDS’09; 24-24 April 2009. 55. Hummel J, Evans PC, Lee H. Medication reconciliation in the emergency department: Opportunities for workflow redesign. Qual Saf Health Care 2010;19(6):531–5. 56. Abraham J, Reddy MC. Challenges to inter-departmental coordination of patient transfers: A workflow perspective. Int J Med Inform 2010;79(2):112–22. 57. Palmer G 2nd, Abernathy JH 3rd, Swinton G, et al. Realizing improved patient care through human-centered operating room design: A human factors methodology for observing flow disruptions in the cardiothoracic operating room. Anesthesiology 2013;119(5):1066–77. 58. (ACR) ACoR. ACR practice guideline for communication of diagnostic imaging findings. American College of Radiology (ACR), 2010. p. 6. 59. Johnson E, Sanger J, Rosenkrantz AB. Important nonurgent imaging findings: Use of a hybrid digital and administrative support tool for facilitating clinician communication. Clin Imaging 2015;39(3):493–6. 60. Hayes SA, Breen M, McLaughlin PD, et al. Communication of unexpected and significant findings on chest radiographs with an automated PACS alert system. J Am Coll Radiol 2014;11(8):791–5.
6
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
61. Abujudeh HH, Kaewlai R, Choy G, et al. Important imaging finding e-mail alert system: Experience after 3 years of implementation. Radiology 2009;252 (3):747–53. 62. Eisenberg RL, Yamada K, Yam CS, et al. Electronic messaging system for communicating important, but nonemergent, abnormal imaging results. Radiology 2010;257(3):724–31. 63. Lacson R, O’Connor SD, Andriole KP, et al. Automated critical test result notification system: Architecture, design, and assessment of provider satisfaction. AJR Am J Roentgenol 2014;203(5):W491–496. 64. Naeger DM, Phelps A, Kohi M, et al. Reading room electives: Say goodbye to the “radi-holiday”. J Am Coll Radiol 2013;10(6):442–8. 65. Chen JY, Lewis PJ. The value of a medical student radiology triage program in enhancing clinical education and skills. Acad Radiol 2014;21(7):829–33. 66. Davis DJ, Moon M, Kennedy S, et al. Introducing medical students to radiology as paid emergency department triage assistants. J Am Coll Radiol 2011;8 (10):710–5. 67. Dobre MC, Maley J. Medical student radiology externs: Increasing exposure to radiology, improving education, and influencing career choices. J Am Coll Radiol 2012;9(7):506–9 e505. 68. Mamlouk MD, Saket RR, Hess CP, et al. Adding value in radiology: Establishing a designated quality control radiologist in daily workflow. J Am Coll Radiol 2015 [Epub ahead of print]. 69. Troude P, Dozol A, Soyer P, et al. Improvement of radiology requisition. Diagn Interv Imaging 2014;95(1):69–75. 70. Khorasani R. Optimizing communication of critical test results. J Am Coll Radiol 2009;6(10):721–3.
71. Juluru K, Eng J. Internet-based radiology order-entry, reporting, and workflow management system for coordinating urgent study requests during off-hours. AJR Am J Roentgenol 2005;184(3):1017–20. 72. Rosenthal DI, Weilburg JB, Schultz T, et al. Radiology order entry with decision support: Initial clinical experience. J Am Coll Radiol 2006;3(10):799–806. 73. Lee J, Baughman C, Ferguson R. Results of implementation of integrated electronic health record contrast allergy decision support. Paper Presented at 2014 IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI); 1-4 June 2014. 74. Branstetter BF, editor. Practical Imaging Informatics: Foundations and Applications for PACS Professionals. New York, NY: Springer; 2009. 75. Davenport MS, Gardner CS, Jaffe TA. Mandated radiologist-performed electronic order entry: Effect on CT oral contrast administration. AJR Am J Roentgenol 2012;198(3):616–20. 76. Wasser EJ, Galante NJ, Andriole KP, et al. Optimizing radiologist e-prescribing of CT oral contrast agent using a protocoling portal. AJR Am J Roentgenol 2013;201 (6):1298–302. 77. van Ooijen PM, Roosjen R, de Blecourt MJ, et al. Evaluation of the use of CDROM upload into the PACS or institutional web server. J Digit Imaging 2006;19 (suppl 1):72–7. 78. Hoff WS, Sicoutris CP, Lee SY, et al. Formalized radiology rounds: The final component of the tertiary survey. J Trauma 2004;56(2):291–5. 79. Tillack AA, Borgstede JP. An evaluation of the impact of clinically embedded reading rooms on radiologist-referring clinician communication. J Am Coll Radiol 2013;10(5):368–72.
Current Problems in Diagnostic Radiology journal homepage: www.cpdrjournal.com
Disruption of Radiologist Workflow Akash P. Kansagra, MD, MSa,n, Kevin Liu, MDa, John-Paul J. Yu, MD, PhDb a b
Mallinckrodt Institute of Radiology, Washington University School of Medicine, Saint Louis, MO Department of Radiology, University of California, San Francisco, San Francisco, CA
The effect of disruptions has been studied extensively in surgery and emergency medicine, and a number of solutions—such as preoperative checklists— have been implemented to enforce the integrity of critical safety-related workflows. Disruptions of the highly complex and cognitively demanding workflow of modern clinical radiology have only recently attracted attention as a potential safety hazard. In this article, we describe the variety of disruptions that arise in the reading room environment, review approaches that other specialties have taken to mitigate workflow disruption, and suggest possible solutions for workflow improvement in radiology. & 2015 Mosby, Inc. All rights reserved.
Introduction Society has recognized the role of disruptions in creating accidents and mishaps. Wherever the ensuing mishaps have the potential to cause harm or loss of life, society has enacted restrictions to minimize disruptions of normal workflow. For this reason, drivers in many states are prohibited from sending text messages or using handheld devices while driving, and airline pilots are mandated to maintain a “sterile cockpit” during critical phases of flight where only mission-related tasks are discussed. Medicine—where disruptions can easily cause harm and loss of life—has also started to respond to these challenges. The most publicized and mature examples to date involve the use of preprocedural checklists before surgery or central line insertion to enforce the integrity of critical safety-related workflows, resulting in dramatic improvements in patient safety and clinical outcomes.1-3 However, the potential for significant workflow disruption extends far beyond periprocedural care. In a busy radiology reading room, for example, radiologists must contend with a complex and fast-paced workflow characterized by frequent disruptions, disruptions that may be particularly problematic given the high cognitive demand of image interpretation.4-6 Unfortunately, the nonstandard nature of most radiology workflows reduces the potential effectiveness of basic interventions such as checklists and may require more sophisticated solutions.7 In this review, we describe the workflow disruptions with which radiologists must contend in daily practice, highlight steps that other specialties have taken to respond to workflow disruptions, and suggest measures that can be taken to mitigate similar disruptions in radiology. It is our hope that this review will draw
n Reprint requests: Akash P. Kansagra, MD, MS, Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S. Kingshighway Blvd, Campus Box 8131, Saint Louis, MO 63110. E-mail address: [email protected] (A.P. Kansagra).
http://dx.doi.org/10.1067/j.cpradiol.2015.05.006 0363-0188/& 2015 Mosby, Inc. All rights reserved.
attention to the urgent need for improved clinical workflow in the reading room and provide a blueprint for safer and more effective radiological care.
Workflow Disruptions in Radiology The working environment in diagnostic radiology has undergone a tremendous change over the past 2 decades because of the widespread adoption of filmless imaging, introduction of speech recognition systems for report dictation, and the incorporation of electronic medical records (EMR) into an increasingly informationrich interpretive workflow.8 Unfortunately, the promised efficiency gains of these systems have been partially offset by a paradoxical increase in the complexity of radiologists’ workflow. This complexity reflects a number of converging trends, including the central and growing role of medical imaging in patient evaluation and management, as well as increasing fragmentation and disruption of interpretive workflows. In addition to their primary task of image interpretation and reporting, radiologists in modern practice must shoulder added responsibilities that can include frequent telephone communication, in-person physician consultations, technologist supervision, patient consent, ultrasound scanning, and management of contrast agent injections and adverse reactions.4,5,8 Although these additional tasks are important, they distract and detract from the primary workflow of image interpretation, create barriers to productivity, and likely contribute to errors in the knowledge-intensive service environment of clinical radiology.9 Of these many potential sources of disruption, telephone communication is particularly problematic, in part because many different sources of disruption are funneled through this common communication channel. As an example, incoming telephone calls may come from clinical providers inquiring about imaging findings or selection of appropriate imaging tests, or from technologists requesting “scan checks” to assess study adequacy or seeking
2
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
guidance for image acquisition (eg, protocol selection or solutions for specific technical challenges). Similarly, outgoing phone calls are often directed to providers to communicate time-sensitive imaging findings or solicit additional patient information. Recent work by Yu et al5 describes their experience with telephone-based disruption of on-call radiologists’ workflow. Their work reveals telephone interruptions of radiologists on a surprisingly large scale, with more than 10,000 after-hours calls directed to a single resident over a 3-month period. Incoming telephone calls occurred as frequently as once every 4 minutes during peak hours, which translated to roughly 2-3 expected interruptions during interpretation of a single CT of the abdomen and pelvis. Related work by Balint et al4 has suggested that the frequency of telephone disruptions in the hour preceding resident interpretation of a study was positively correlated with the likelihood of an incorrect interpretation. Importantly, disruptions to radiologists are not confined to the reading room; dysfunctional or inefficient workflows in other areas of the radiology department can secondarily increase burdens on radiologists. For example, inadequate systems to identify and triage patients for imaging may create a need for frequent radiologist involvement in study prioritization.8 Similarly, poorly designed systems for information transmission between different members of a radiology department (eg, radiologists, technologists, and patient transporters) can hinder effective care of patients, thereby requiring increased radiologist involvement to maintain appropriate and timely care.10 If left unchecked, the scale of workflow disruption is likely to increase. As the information economy of medicine continues to grow in scale and complexity, there is likely to be increased reliance on specialties such as radiology that can create and share objective patient information. Against this backdrop, inefficient or ill-defined clinical workflows are likely to produce ever-increasing disruptions to radiologists. Thus, implementing solutions to dysfunctional workflows is a key component in building and maintaining an efficient information economy.
Managing Workflow Disruption in Nonradiological Settings Disruptions in workflow are not unique to radiology but are also experienced by other hospital-based specialties such as emergency medicine,11-19 critical care,20-25 and surgery.26-32 The solutions to these disruptions vary based on the specific workflow patterns in each patient care setting, but they can generally be grouped into several themes. Filtering Interruptions by Activity A basic strategy to improve workflow is to create physical or temporal barriers to interruption during activities that are of critical importance or particularly susceptible to disruption. In some cases, this strategy may amount simply to having individuals address potential sources of interruption at a convenient time, such as refilling intravenous fluids before nursing handoffs to prevent unnecessary alarming during transfer of care.23 In other cases, physical barriers—including possibilities such as signs or colored vests for individuals seeking to avoid interruption, or colored floor tiles or shields for specially designated areas—may be of value.23,33,34 These basic interventions can have a profound effect. One study found that implementing a visible “No Interruption Zone” around a medication dispensing station resulted in a 41% decrease in interruptions,33 while another found that erecting a wall around the medication dispensing station decreased interruptions by 81%.34 Alternatively, there may be value in gentler approach that allows for interruptions even during critical
activities, provided the interruption conveys important patientrelated information. In this context, interruptions may be mitigated by increasing the transparency of task importance so that potential interrupters can determine if interruption is appropriate.22 Filtering Interruptions by Acuity A large percentage of interruptions—even during critical tasks such as transfer of care discussions (“sign out”) and clinical rounds —are nonessential, with only 11% of interruptions during morning sign out and 27% during morning rounds being essential to patient care.21 As such, filtering nonessential interruptions may streamline workflow. Young et al35 describe a system in which nurse requests to send after-hours pages to resident physicians are reviewed by a charge nurse and categorized by acuity, with emergent pages transmitted immediately, urgent pages batched, and nonurgent pages deferred until the morning. Following implementation of this system, the total number of pages and number of nonurgent pages sent after-hours to house staff decreased. Asynchronous Communication Synchronous channels of communication require simultaneous participation of both parties, preventing the recipient of an interruption from managing the timing of that interruption.11 In contrast, asynchronous channels of communication provide the recipient of a message with control over the timing of disruptions, and this may therefore represent a practical method for acuitybased filtering, task prioritization, and reduced communication burden.36 Voicemail capability may be an effective means to reduce disruption. In a study of emergency department (ED) providers equipped with mobile phones, the lack of voicemail capability contributed significantly to workflow interruption, as the providers were forced to immediately answer any incoming call.13 Alternatively, landlines with a clerical receptionist may serve a similar role and help to reduce unnecessary interruptions.11 Alphanumeric pagers may also permit filtering of nonurgent interruptions,11,37 provided that the recipient of a message is provided with sufficient information to judge the urgency of the page. Unfortunately, a large percentage of alphanumeric pages contain only basic callback information, thereby preventing the receiver from performing effective task prioritization and mandating an immediate callback to determine the acuity of the page.37 A proposed explanation for this behavior is that synchronous communication provides receipt confirmation for the interrupter; asynchronous communication may benefit from a confirmation mechanism to encourage broader adherence.36 Technology-Assisted Workflow Electronic and nonelectronic whiteboards have been used extensively in a variety of care settings to organize and facilitate communication and workflow.36,38-42 For example, when used in the operating room as a basic information display system, electronic whiteboards can facilitate integration of safety checklists into preoperative workflow and aid intraoperative communication between multiple team members.43,44 Chaotic and disruptive workflows can be further streamlined with electronic systems that go beyond basic information display to serve as an integrated information technology (IT) solution.45 Aronsky et al46 have described the implementation of such a system in an ED, which allowed for easy information access, information sharing, and decision support using data from multiple hospital information systems, with resulting dramatic
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
increases in revenue, enhanced hospital-wide triage, and improved staffing. Such systems can facilitate high-fidelity communication, improve multitasking ability, and promote physician efficiency through automated results notification and performance feedback.47-49 Indeed, IT tools of this sort may help to reduce both the number and the adverse effects of interruptions.50,51 Eliminating Unnecessary Tasks Some time-consuming tasks can be entirely obviated through process redesign or IT solutions. For instance, automated systems to identify and physically locate patients, providers, and equipment can eliminate the need for multiple phone calls and pages.36,52 Similarly, normal triage pathways and full patient registration can be bypassed in some ED settings,40,42,48 while the considerable complexity of patient transfer between wards can be avoided with acuity-adjustable beds.49 Workload Redistribution Workload is not evenly distributed between individuals on a health care team. For instance, junior physicians in the intensive care unit (ICU) setting must contend with competing demands on their attention far more often than senior physicians do.53 Accordingly, sharing some aspects of workload may help to improve overall workflow. Simulations have shown that the time efficiency of rounding in the ICU can be dramatically improved by having all team members share responsibility for handling interruptions, thereby unburdening the trainee primarily responsible for a patient under active discussion.50 Similarly, workflow in the ED can be improved by sharing triage responsibilities among a larger number of providers or having a pool of on-demand nurses and physicians who can be recruited to help manage excessive workloads.40,54 There may also be workflow benefit from transferring some responsibilities to other appropriate individuals. For instance, designating a receptionist or nurse to answer phone calls and pages can help to reduce interruption to physicians.11,26 Similarly, assigning nonphysician staff to handle basic but time-consuming aspects of workflow such as medication reconciliation or intrahospital patient transfers can also promote workflow efficiency.55,56 Indeed, even direct patient care tasks can be assigned to nonphysician providers if there are clearly established protocols of care.40,42
3
has recommended that radiologists use communication channels other than the radiology report to convey this information to the treating physician in a timely fashion.58 To minimize the effect of these nonroutine communications on workflow, many practices have developed efficient channels of asynchronous communication that eliminate time-consuming tasks, redistribute workload away from radiologists, or incorporate technology-assisted workflows. A common approach is to automate communication of important but nonemergent findings, such as lung nodules and solid organ lesions. Johnson et al59 described the implementation of a clickable button within their dictation software that conveyed such findings to referring physicians via facsimile. Other groups have reported systems to generate e-mail alerts to referring physicians when specific, predefined statements were included in the final report text.60,61 In an alternative approach, Eisenberg et al62 redistributed rather than eliminated this workload by having the interpreting radiologist submit a case to a web-based portal and assigning a “communications facilitator” to the task of contacting the referring physician by e-mail, telephone, or pager. Most of these tools offered some form of receipt confirmation, and each was effective for communicating nonemergent findings with minimal workflow disruption and a high level of satisfaction from both radiologists and referring clinicians. Lacson et al63 developed an automated notification system that could be used even for emergent findings. In their system, radiologists were able to indicate the level of urgency of a finding within notification software that was integrated with their picture archiving and communication system (PACS) environment; the system then automatically contacted the appropriate provider using e-mail for nonurgent findings or pager for urgent findings and documented this notification within the EMR. Triage Assistants
Strategies to Reduce Workflow Disruption in Radiology
As part of larger efforts to more meaningfully integrate medical students into the clinical practice of radiology,64 several medical schools have experimented with using medical students to assist with triage in radiology reading rooms. Authors from at least 3 academic radiology departments have separately reported on the benefits of paying medical students to answer telephone calls and pages, and protocol radiology studies with the help of the on-call radiology resident.65-67 This arrangement provides educational and financial benefits to medical students while reducing disruptions to radiology residents. In particular, medical students can enforce acuity- and task-based filtering of interruptive communications (eg, holding nonurgent communications until the resident is between studies). Furthermore, some of the workload of soliciting history, performing medical record review, or contacting referring physicians could be redistributed from the resident to the medical student. Mamlouk et al68 have taken the idea of workload redistribution further and designated a radiology fellow as a “quality control” radiologist on a rotating basis during daytime hours. This radiologist is largely excused from interpretive tasks but handles most incoming phone calls from referring providers, protocols and prioritizes imaging studies, performs real-time scan checks, and customizes imaging for complex cases for the entire medical center, thereby preserving an efficient and undisrupted workflow for the remaining staff and trainees of the section. Notably, this solution simultaneously preserves efficiency of workflow while increasing the availability of consultative services and other valuebased practices.
Efficient Asynchronous Communication
Computerized Order Entry, Decision Support, and Protocoling
When presented with emergent or unexpected findings or interpretative discrepancies, the American College of Radiology
Computerized order entry for imaging services may help to avoid issues related to insufficient, misleading, or illegible clinical
Physical Layout Physical layout has been implicated as a source of greater-thannecessary interruption. In the operating room, 33% of workflow disruptions were attributable to suboptimal layout.57 Similarly, intensive care units can be designed in a way that discourages disruption of important tasks such as medication administration.52 Some authors have hypothesized that the high rate of interruptions in the ICU relates to a large number of collaborating providers being concentrated into a relatively compact space, and that a linear arrangement of rooms along a busy thoroughfare may lead to more interruptions than a more open arrangement that separates patient care zones from high traffic areas.20
4
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
information that often accompanies handwritten or otherwise nonstandardized requests for imaging.69-71 Indeed, these systems can aggregate relevant clinical data—referring physician contact information and patient location, among others—from multiple hospital systems and display this information in a standardized format.69 Thus, implementation of a technology-assisted workflow for clinicians may improve radiologists’ work efficiency by reducing the need for direct communication with clinicians or excessively detailed exploration of patients’ medical records before study protocoling or interpretation. Coupling computerized decision support tools with computerized order entry systems may amplify these benefits further. By encouraging appropriate imaging utilization at the time of order entry, these tools likely reduce the need for radiologists to interact directly with clinical providers to modify or eliminate unhelpful aspects of imaging evaluation. As an example, Rosenthal and colleagues have developed and implemented a decision support system that provides clinicians with a “utility score” based on clinical indication in the form of an ICD-9 code.72 Following implementation of this system, several major insurance carriers agreed to eliminate required preauthorization for high-cost imaging studies. Similarly, a system to alert referring providers to the presence of a contrast agent allergy and suggest prophylactic premedication dramatically increased the number of patients who were appropriately premedicated without radiologist intervention.73 Computerized order entry for radiologists in the form of electronic study protocoling can also improve radiologists’ workflow. As with computerized order entry for clinicians, these systems can effortlessly aggregate important patient-related data such as serum creatinine, allergies, and cardiac pacemakers from the EMR and implement automated safety checks related to specific study types (eg, magnetic resonance imaging and cardiac pacemakers) or pharmacologic agents (eg, iodinated contrast agent and renal function). Moreover, recallable electronic protocols may greatly expedite protocoling in cases where serial imaging is performed, such as multiple sclerosis or lung nodule surveillance. Integrated Ecosystem of Imaging Applications Large efficiency gains can also be realized with a tightly integrated ecosystem of clinical applications, including PACS, EMR, radiology information system (RIS), voice dictation software, electronic protocoling applications, teaching files, and image postprocessing software.74 As an example, context integration between PACS, EMR, RIS, and dictation software can allow an entirely PACS-driven workflow that automatically opens the patient’s information patient’s chart in the EMR and initiates reporting of the correct case in the dictation software and RIS. Similarly, an electronic protocoling system with single sign-on login and automatic patient lookup in the EMR may allow dramatically more efficient study protocoling than other, less tightly integrated systems.75,76 This same approach may be useful for radiologists who are asked to review studies performed at other facilities and stored on physical media such as compact discs. When no defined workflow exists, referring physicians may bring these CDs directly to the radiologist, who must divert his or her attention from other interpretive tasks to perform the time-consuming task of loading images from the CD. Finally, these images must actually be reviewed, often using one of the many cumbersome and often limited software image-viewing applications provided on those CDs. Instead, radiologists’ workflow efficiency may be substantially improved using software designed to facilitate image importation, such as LILA (LifeIMAGE, Newton, MA), or direct upload to PACS.77 These workflows unburden radiologists and allow images to be
reviewed in an application that is more suited to interpretative tasks. Scheduled Rounds The frequency of in-person consultation between radiologists and clinical practitioners has dramatically decreased with the widespread adoption of distributed PACS. Today, most in-person consultations occur sporadically, though some services—most commonly neonatology and trauma surgery—advocate for or continue to practice formalized radiology rounds as a means to improve patient care.78 In this setting, radiologists’ workflow may be improved by identifying a specific, mutually convenient time at which radiology rounds will occur. This practice would allow radiologists to address other clinical demands at a different time so that they may meet their interpretative responsibilities while also increasing their availability for regular consultation. Workload Optimization and Balancing Improved optimization of radiology workload can be achieved by identifying high-priority cases to be interpreted before less urgent cases, thereby reducing the number of incoming interruptive communications from clinicians in need of results to guide time-sensitive management conditions. Halsted and Froehle devised an automated work queuing system to prioritize cases based on patient medical acuity, wait time, psychological state, and other factors to present a prioritized queue to radiologists and demonstrated a nearly 28% increase in the mean time between workflow interruptions.8 Improved balancing of radiology workload can be achieved through more diversified scheduling of nonurgent examinations to more fairly utilize radiologists scattered across different reading rooms. If a disproportionate number of imaging studies on a given day are directed to a single reading room (eg, a large number of brain MRIs on a day that coincides with a scheduled neurooncology clinic), it may lead to excessive workflow for a small group of radiologists that can produce a backlog of studies awaiting interpretation and corresponding increase in disruptive communications from clinicians requesting results. In contrast, scheduling nonurgent studies to minimize day-to-day variations in imaging volume between separate reading rooms may permit more predictable and easily managed workloads. Embedded Radiologists Recent efforts to “embed” radiologists near their clinical counterparts may also affect radiologists’ workflow. A recent study demonstrated wide variability in the preferred modes of communication between radiologists and clinical providers depending on their physical proximity.79 Not surprisingly, embedded radiologists were more likely to communicate in person and less likely to use automated result notification systems. Nevertheless, it remains unclear if this alteration in physical layout improves or hinders a radiologist’s efficiency, as the concept of embedding radiologists has been promoted because of the promise of adding value to clinical care rather than increasing interpretative efficiency.
Conclusion Disruptions of radiologists’ workflow occur commonly and come from a wide range of sources, including phone calls, pages, in-person consultation, coordination of care, and burdensome IT tools. Other hospital-based specialties use different workflows than that of radiology but have developed solutions to mitigate
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
workflow disruption that can inspire improvements in radiology. Through process redesign and improved IT tools, radiologists can streamline workflow to reduce the number of disruptions and competing demands during the cognitively intense task of image interpretation and can thereby facilitate the delivery of more efficient and higher-quality radiological care.
References 1. de Vries EN, Prins HA, Crolla RM, et al. Effect of a comprehensive surgical safety system on patient outcomes. N Engl J Med 2010;363(20):1928–37. 2. Haynes AB, Weiser TG, Berry WR, et al. A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med 2009;360(5): 491–9. 3. Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med 2006;355 (26):2725–32. 4. Balint BJ, Steenburg SD, Lin H, et al. Do telephone call interruptions have an impact on radiology resident diagnostic accuracy? Acad Radiol 2014;21 (12):1623–8. 5. Yu JP, Kansagra AP, Mongan J. The radiologist’s workflow environment: Evaluation of disruptors and potential implications. J Am Coll Radiol 2014;11 (6):589–93. 6. Reiner BI, Krupinski E. The insidious problem of fatigue in medical imaging practice. J Digit Imaging 2012;25(1):3–6. 7. Berbaum K, Franken EA Jr, Caldwell RT, et al. Can a checklist reduce SOS errors in chest radiography? Acad Radiol 2006;13(3):296–304. 8. Halsted MJ, Froehle CM. Design, implementation, and assessment of a radiology workflow management system. AJR Am J Roentgenol 2008;191(2):321–7. 9. Froehle CM, White DL. Interruption and forgetting in knowledge-intensive service environments. Prod Oper Manag 2014;23(4):704–22. 10. Larson DB, Froehle CM, Johnson ND, et al. Communication in diagnostic radiology: Meeting the challenges of complexity. AJR Am J Roentgenol 2014;203(5): 957–64. 11. Brixey JJ, Robinson DJ, Turley JP, et al. The roles of MDs and RNs as initiators and recipients of interruptions in workflow. Int J Med Inform 2010;79(6):e109–115. 12. Woloshynowych M, Davis R, Brown R, et al. Communication patterns in a UK emergency department. Ann Emerg Med 2007;50(4):407–13. 13. Brixey JJ, Tang Z, Robinson DJ, et al. Interruptions in a level one trauma center: A case study. Int J Med Inform 2008;77(4):235–41. 14. Brixey JJ, Robinson DJ, Johnson CW, et al. Towards a hybrid method to categorize interruptions and activities in healthcare. Int J Med Inform 2007;76(11-12):812–20. 15. Chisholm CD, Dornfeld AM, Nelson DR, et al. Work interrupted: A comparison of workplace interruptions in emergency departments and primary care offices. Ann Emerg Med 2001;38(2):146–51. 16. France DJ, Levin S, Hemphill R, et al. Emergency physicians’ behaviors and workload in the presence of an electronic whiteboard. Int J Med Inform 2005;74(10):827–37. 17. Spencer R, Coiera E, Logan P. Variation in communication loads on clinical staff in the emergency department. Ann Emerg Med 2004;44(3):268–73. 18. Fairbanks RJ, Bisantz AM, Sunm M. Emergency department communication links and patterns. Ann Emerg Med 2007;50(4):396–406. 19. Chisholm CD, Collison EK, Nelson DR, et al. Emergency department workplace interruptions are emergency physicians “interrupt-driven” and “multitasking”? Acad Emerg Med 2000;7(11):1239–43. 20. Kalisch BJ, Aebersold M. Interruptions and multitasking in nursing care. Jt Comm J Qual Patient Saf 2010;36(3):126–32. 21. Nimmo GR, Mitchell C. A preliminary audit of interruptions in intensive care: Implications for patient safety. J Intensive Care Soc 2008;9(3):240–2. 22. Sasangohar F, Donmez B, Easty A, et al. Interruptions experienced by cardiovascular intensive care unit nurses: An observational study. J Crit Care 2014;29 (5):848–53. 23. Spooner AJ, Corley A, Chaboyer W, et al. Measurement of the frequency and source of interruptions occurring during bedside nursing handover in the intensive care unit: An observational study. Aust Crit Care 2015;28(1):19–23. 24. Weigl M, Muller A, Zupanc A, et al. Hospital doctors’ workflow interruptions and activities: An observation study. BMJ Qual Saf 2011;20(6):491–7. 25. Drews FA. The frequency and impact of task interruptions in the ICU. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2007;51 (11):683-686. 26. Healey AN, Sevdalis N, Vincent CA. Measuring intra-operative interference from distraction and interruption observed in the operating theatre. Ergonomics 2006;49(5-6):589–604. 27. Antoniadis S, Passauer-Baierl S, Baschnegger H, et al. Identification and interference of intraoperative distractions and interruptions in operating rooms. J Surg Res 2014;188(1):21–9. 28. Christian CK, Gustafson ML, Roth EM, et al. A prospective study of patient safety in the operating room. Surgery 2006;139(2):159–73. 29. Wiegmann DA, ElBardissi AW, Dearani JA, et al. Disruptions in surgical flow and their relationship to surgical errors: An exploratory investigation. Surgery 2007;142(5):658–65.
5
30. Weigl M, Antoniadis S, Chiapponi C, et al. The impact of intra-operative interruptions on surgeons’ perceived workload: An observational study in elective general and orthopedic surgery. Surg Endosc 2015;29(1):145–53. 31. Sevdalis N, Undre S, McDermott J, et al. Impact of intraoperative distractions on patient safety: A prospective descriptive study using validated instruments. World J Surg 2014;38(4):751–8. 32. Zheng B, Martinec DV, Cassera MA, et al. A quantitative study of disruption in the operating room during laparoscopic antireflux surgery. Surg Endosc 2008;22(10):2171–7. 33. Anthony K, Wiencek C, Bauer C, et al. No interruptions please: Impact of a No Interruption Zone on medication safety in intensive care units. Crit Care Nurse 2010;30(3):21–9. 34. Colligan L, Guerlain S, Steck SE, et al. Designing for distractions: A human factors approach to decreasing interruptions at a centralised medication station. BMJ Qual Saf 2012;21(11):939–47. 35. Young JB, Baker AC, Boehmer JK, et al. Using NNAPPS (Nighttime Nurse and Physician Paging System) to Maximize Resident Call Efficiency within 2011 Accreditation Council for Graduate Medical Education (ACGME) work hour restrictions. J Surg Educ 2012;69(6):819–25. 36. Xiao Y, Hu P, Moss J, et al. Opportunities and challenges in improving surgical work flow. Cogn Technol Work 2008;10(4):313–21. 37. Espino S, Cox D, Kaplan B. Alphanumeric paging: A potential source of problems in patient care and communication. J Surg Educ 2011;68(6):447–51. 38. Fairbanks RJ, Bisantz AM, Sunm M. Emergency department communication links and patterns. Ann Emerg Med 2007;50(4):396–406. 39. France DJ, Levin S, Hemphill R, et al. Emergency physicians’ behaviors and workload in the presence of an electronic whiteboard. Int J Med Inform 2005;74(10):827–37. 40. Yen K, Gorelick MH. Strategies to improve flow in the pediatric emergency department. Pediatr Emerg Care 2007;23(10):745–9 [quiz 750-741]. 41. Xiao Y, Schenkel S, Faraj S, et al. What whiteboards in a trauma center operating suite can teach us about emergency department communication. Ann Emerg Med 2007;50(4):387–95. 42. Wiler JL, Gentle C, Halfpenny JM, et al. Optimizing emergency department front-end operations. Ann Emerg Med 2010;55(2):142–160 e141. 43. Mainthia R, Lockney T, Zotov A, et al. Novel use of electronic whiteboard in the operating room increases surgical team compliance with pre-incision safety practices. Surgery 2012;151(5):660–6. 44. Parush A, Kramer C, Foster-Hunt T, et al. Communication and team situation awareness in the OR: Implications for augmentative information display. J Biomed Inform 2011;44(3):477–85. 45. Lemke HU, Ratib OM, Horii SC. Workflow in the operating room: A summary review of the Arrowhead 2004 Seminar on Imaging and Informatics. Int Congress Ser 2005;1281(0):862–7. 46. Aronsky D, Jones I, Lanaghan K, et al. Supporting patient care in the emergency department with a computerized whiteboard system. J Am Med Inform Assoc 2008;15(2):184–94. 47. Laxmisan A, Hakimzada F, Sayan OR, et al. The multitasking clinician: Decisionmaking and cognitive demand during and after team handoffs in emergency care. Int J Med Inform 2007;76(11-12):801–11. 48. Baumlin KM, Shapiro JS, Weiner C, et al. Clinical information system and process redesign improves emergency department efficiency. Jt Comm J Qual Patient Saf 2010;36(4):179–85. 49. Alameddine M, Dainty KN, Deber R, et al. The intensive care unit work environment: Current challenges and recommendations for the future. J Crit Care 2009;24(2):243–8. 50. Gupta A, Sharda R, Dong Y, et al. Improving rounding in critical care environments through management of interruptions. Decis Supp Syst 2013;55 (2):516–27. 51. Grundgeiger T, Sanderson P. Interruptions in healthcare: Theoretical views. Int J Med Inform 2009;78(5):293–307. 52. Thompson DR, Hamilton DK, Cadenhead CD, et al. Guidelines for intensive care unit design. Crit Care Med 2012;40(5):1586–600. 53. Malhotra S, Jordan D, Shortliffe E, et al. Workflow modeling in critical care: Piecing together your own puzzle. J Biomed Inform 2007;40(2):81–92. 54. Lu W. An agent-based simulation for workflow in Emergency Department. Paper Presented at Systems and Information Engineering Design Symposium, SIEDS’09; 24-24 April 2009. 55. Hummel J, Evans PC, Lee H. Medication reconciliation in the emergency department: Opportunities for workflow redesign. Qual Saf Health Care 2010;19(6):531–5. 56. Abraham J, Reddy MC. Challenges to inter-departmental coordination of patient transfers: A workflow perspective. Int J Med Inform 2010;79(2):112–22. 57. Palmer G 2nd, Abernathy JH 3rd, Swinton G, et al. Realizing improved patient care through human-centered operating room design: A human factors methodology for observing flow disruptions in the cardiothoracic operating room. Anesthesiology 2013;119(5):1066–77. 58. (ACR) ACoR. ACR practice guideline for communication of diagnostic imaging findings. American College of Radiology (ACR), 2010. p. 6. 59. Johnson E, Sanger J, Rosenkrantz AB. Important nonurgent imaging findings: Use of a hybrid digital and administrative support tool for facilitating clinician communication. Clin Imaging 2015;39(3):493–6. 60. Hayes SA, Breen M, McLaughlin PD, et al. Communication of unexpected and significant findings on chest radiographs with an automated PACS alert system. J Am Coll Radiol 2014;11(8):791–5.
6
A.P. Kansagra et al. / Current Problems in Diagnostic Radiology ] (2015) ]]]–]]]
61. Abujudeh HH, Kaewlai R, Choy G, et al. Important imaging finding e-mail alert system: Experience after 3 years of implementation. Radiology 2009;252 (3):747–53. 62. Eisenberg RL, Yamada K, Yam CS, et al. Electronic messaging system for communicating important, but nonemergent, abnormal imaging results. Radiology 2010;257(3):724–31. 63. Lacson R, O’Connor SD, Andriole KP, et al. Automated critical test result notification system: Architecture, design, and assessment of provider satisfaction. AJR Am J Roentgenol 2014;203(5):W491–496. 64. Naeger DM, Phelps A, Kohi M, et al. Reading room electives: Say goodbye to the “radi-holiday”. J Am Coll Radiol 2013;10(6):442–8. 65. Chen JY, Lewis PJ. The value of a medical student radiology triage program in enhancing clinical education and skills. Acad Radiol 2014;21(7):829–33. 66. Davis DJ, Moon M, Kennedy S, et al. Introducing medical students to radiology as paid emergency department triage assistants. J Am Coll Radiol 2011;8 (10):710–5. 67. Dobre MC, Maley J. Medical student radiology externs: Increasing exposure to radiology, improving education, and influencing career choices. J Am Coll Radiol 2012;9(7):506–9 e505. 68. Mamlouk MD, Saket RR, Hess CP, et al. Adding value in radiology: Establishing a designated quality control radiologist in daily workflow. J Am Coll Radiol 2015 [Epub ahead of print]. 69. Troude P, Dozol A, Soyer P, et al. Improvement of radiology requisition. Diagn Interv Imaging 2014;95(1):69–75. 70. Khorasani R. Optimizing communication of critical test results. J Am Coll Radiol 2009;6(10):721–3.
71. Juluru K, Eng J. Internet-based radiology order-entry, reporting, and workflow management system for coordinating urgent study requests during off-hours. AJR Am J Roentgenol 2005;184(3):1017–20. 72. Rosenthal DI, Weilburg JB, Schultz T, et al. Radiology order entry with decision support: Initial clinical experience. J Am Coll Radiol 2006;3(10):799–806. 73. Lee J, Baughman C, Ferguson R. Results of implementation of integrated electronic health record contrast allergy decision support. Paper Presented at 2014 IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI); 1-4 June 2014. 74. Branstetter BF, editor. Practical Imaging Informatics: Foundations and Applications for PACS Professionals. New York, NY: Springer; 2009. 75. Davenport MS, Gardner CS, Jaffe TA. Mandated radiologist-performed electronic order entry: Effect on CT oral contrast administration. AJR Am J Roentgenol 2012;198(3):616–20. 76. Wasser EJ, Galante NJ, Andriole KP, et al. Optimizing radiologist e-prescribing of CT oral contrast agent using a protocoling portal. AJR Am J Roentgenol 2013;201 (6):1298–302. 77. van Ooijen PM, Roosjen R, de Blecourt MJ, et al. Evaluation of the use of CDROM upload into the PACS or institutional web server. J Digit Imaging 2006;19 (suppl 1):72–7. 78. Hoff WS, Sicoutris CP, Lee SY, et al. Formalized radiology rounds: The final component of the tertiary survey. J Trauma 2004;56(2):291–5. 79. Tillack AA, Borgstede JP. An evaluation of the impact of clinically embedded reading rooms on radiologist-referring clinician communication. J Am Coll Radiol 2013;10(5):368–72.