Portal de Periódicos da CAPES

Increasing Median Time between Interruptions in a Busy Reading Room

2021; Radiological Society of North America; Volume: 41; Issue: 2 Linguagem: Inglês

10.1148/rg.2021200094

1527-1323

Ethan A. Smith, Andrew H. Schapiro, Rachel J. Smith, Sarah E. O’Brien, Sara Smith, Amy L. Eckerle, Alexander J. Towbin,

Interpreting and Communication in Healthcare

HomeRadioGraphicsVol. 41, No. 2 PreviousNext Practice Policy and Quality InitiativesFree AccessIncreasing Median Time between Interruptions in a Busy Reading RoomEthan A. Smith , Andrew H. Schapiro, Rachel Smith, Sarah E. O’Brien, Sara N. Smith, Amy L. Eckerle, Alexander J. TowbinEthan A. Smith , Andrew H. Schapiro, Rachel Smith, Sarah E. O’Brien, Sara N. Smith, Amy L. Eckerle, Alexander J. TowbinAuthor AffiliationsFrom the Department of Radiology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave, MLC 5031, Cincinnati, OH 45229-3026 (E.A.S., A.H.S., R.S., S.E.O., S.N.S., A.L.E., A.J.T.); and Department of Radiology, University of Cincinnati College of Medicine, Cincinnati, Ohio (E.A.S., A.H.S., A.J.T.).Address correspondence to E.A.S. (e-mail: [email protected]).Ethan A. Smith Andrew H. SchapiroRachel SmithSarah E. O’BrienSara N. SmithAmy L. EckerleAlexander J. TowbinPublished Online:Mar 1 2021https://doi.org/10.1148/rg.2021200094MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In AbstractIntroductionRadiologists are required to perform complex cognitive tasks during the course of their daily work, often interpreting studies that include thousands of images. Prior research has shown that repeated interruptions can have a negative effect on humans performing complex tasks, resulting in errors and decreased efficiency. Interruptions can increase the amount of time required to perform a specific task, resulting in decreased clinical output (1–7). Finally, repeated interruptions cause increased stress and frustration among radiologists and other staff, potentially leading to higher rates of fatigue and burnout.However, it must be acknowledged that some interruptions are necessary for radiologists to provide safe, high-quality clinical care. For example, if a technologist calls to ask the radiologist a specific question about a patient currently undergoing an examination, the direct and timely communication between the technologist and radiologist may be important for the patient’s safety or for obtaining high-quality diagnostic images. During the course of providing care for patients, these types of interruptions are appropriate. Therefore, while eliminating all interruptions may seem to be the optimal goal, in reality it is not feasible. Instead, the goal should be to limit the number of unnecessary interruptions while optimizing those interruptions deemed necessary for patient safety, high-quality diagnostic imaging, and appropriate consultation with clinical colleagues.We undertook a project whereby the goal was to use quality improvement methodology to increase the median time between interruptions that occur daily in our busiest reading room. The SMART (specific, measurable, achievable, relevant, timely) aim was to increase the median time between interruptions by 50% between July 2018 and January 2019.MethodsInstitutional review board approval was not required for this quality improvement project, for which no patient data or protected health information were collected. This quality improvement project was performed in the radiology department of a large freestanding children’s hospital. The radiology department has 35 faculty radiologists, eight pediatric radiology fellows, more than 200 technologists, and up to 12 reading room assistants. The specific setting for this quality improvement project was the busiest of four reading rooms in our radiology reading area. The reading area consists of an l-shaped hallway with four smaller reading rooms, or pods, off of the main hallway (Fig 1). Each smaller reading room represents a subspecialty area (neuroradiology, thoracoabdominal radiology, musculoskeletal radiology, and cardiac/fetal radiology). In the back of the hallway, there is a front-facing desk where the main reading room assistant works, as well as up to two additional work areas for reading room assistants along the sides of the main hallway. The targeted reading room for this study was the thoracoabdominal reading room, which also serves as the work area for the on-call radiologists after hours.Figure 1. Diagram of the radiology reading room. The different reading room areas are designated by subspecialty type (eg, neuroradiology). **This quality improvement project was focused on the thoracoabdominal reading room. RRA = reading room assistant.Figure 1.Download as PowerPointOpen in Image Viewer The project team included two attending radiologists (E.A.S. and A.H.S.); the departmental director of operations (R.S.), who also manages a quality improvement team; a reading room assistant (S.E.O.); a radiography technologist (A.L.E.); and a CT technologist (S.N.S.). It was believed that a single representative technologist could represent both CT and MRI services, as the interactions of these services (such as confirming protocols, checking images, and requesting signatures for contrast-enhanced imaging orders) with the radiologists in the reading room are similar. US services were not included because US images are not routinely read in this specific reading room during regular business hours. The team also enlisted the ad hoc counsel of a third radiologist (A.J.T), who has quality improvement experience and also serves as the departmental chief for clinical operations.Data were collected by means of in-person observation and by using a sampling method. Observers sat in the reading room for 1 hour per day during the weekdays (Monday to Friday) between the hours of 10:00 am and 3:00 pm, with the hour between 12:00 pm and 1:00 pm excluded because an educational conference typically occurs during that time. Observers were tasked with recording the time between interruptions, rounded to the nearest second. This was done by using the stopwatch function on the observers’ personal smartphones. Additional data recorded included the start-stop times for the observation period and the number of interruptions, divided into one of three categories according to how the interruption occurred (ie, by telephone, pager, or in-person communication).Data were manually collected and recorded on a worksheet and then transferred to a computer spreadsheet by the team leader. For the first few weeks, data collection was performed by members of the quality improvement team. Starting at approximately week 5, data collection was performed by one of two administrative assistants who had been trained by the team leader.Once baseline data had been collected, the team met to perform a simplified failure mode effects analysis. A simplified process map (Fig 2) was developed by the team leader and approved by the team members. The team members were then asked to list potential failures for each step of the process map; this was done individually by each team member. The potential failures were then shared with the team for discussion. Next, the team was asked to list possible solutions to the potential failures at each step of the process map. The potential failures and solutions were recorded by the team leader. The potential failure modes recorded during the team meeting were categorized by the team leader and the quality improvement specialist.Figure 2. Simplified process map shows the general steps involved in an interruption and the resumption of work by the radiologist. This basic process map was the basis for our simplified failure mode effects analysis.Figure 2.Download as PowerPointOpen in Image Viewer A Pareto chart (Fig 3) was created to visualize the frequency of failures and identify processes that would be the highest-yield targets for intervention. From the Pareto chart, a key driver diagram (Fig 4) was developed and approved by the team members. This diagram was used to drive the different interventions.Figure 3. Pareto chart shows the different categories of interruptions based on the frequency of potential failures. Availability was defined as an interruption that occurred in the wrong place (eg, a question about neuroradiology that was directed to the thoracoabdominal reading room) or that involved looking for a specific radiologist. Technical was defined as automated interruptions that occurred, such as automated pages informing the radiologist that he or she had a contrast medium order to enter in the electronic medical record. Medication/line refers to interruptions that occur to confirm or clarify a medication order or to ask questions about contrast medium administration routes.Figure 3.Download as PowerPointOpen in Image Viewer Figure 4. Key driver diagram based on the Pareto chart and failure mode effects analysis. This diagram was used to drive the different interventions. The interventions highlighted in green are the different key drivers and specific interventions that the team worked on during the project period. LOR = level of reliability, QC = quality check, RR = reading room, RRA = reading room assistant.Figure 4.Download as PowerPointOpen in Image Viewer Interventions Based on Key DriversAppropriate Image ChecksCT and MRI.—Our department had already limited the number of mandatory image checks, or situations in which the technologist is required to call the radiologist before the patient leaves the imaging suite to ensure that the images are of sufficient quality and no additional imaging is needed. The primary goal of mandatory image checks is to prevent callbacks, in which a patient is required to return to the department for additional imaging or even undergo repeat anesthesia for additional imaging. Both of these situations are considered serious safety-compromising events. Several years before this quality improvement project started, we had eliminated routine checks for outpatient CT and MRI studies that were performed without anesthesia. Although technologists were encouraged to call the radiologist if there were any questions, they were authorized to allow the patient to leave if the technologist believed that the study was of good quality. We believed that our MRI protocols were robust enough such that even for patients undergoing general anesthesia for MRI, the technologist could be allowed to make the decision to release the patient as long as there were no questions. However, we still required image checks for CT performed with general anesthesia, inspiratory-expiratory dual-phase chest CT, patients being treated for trauma, and patients from the intensive care unit.For patients who underwent imaging with general anesthesia, image checks were performed to ensure that the amount of atelectasis that often occurs during mechanical ventilation did not limit the diagnostic quality of the images, and to make sure the anatomic coverage was adequate. For dual-phase inspiratory-expiratory CT, image checks were required to ensure that true inspiratory and expiratory phases were performed. For trauma patients, images were checked to look for evidence of urinary tract injury or other indications to perform delayed contrast-enhanced imaging. For patients in the intensive care unit, image checks were mandated to avoid having to bring back patients who were critically ill and/or difficult to transport for additional or repeat CT.For trauma patients, it was decided that mandatory CT image checks by the radiologist were important for patient care, and this practice was continued. Similarly, as evaluating the adequacy of dual-phase inspiratory and expiratory images is somewhat subjective, it was believed that the radiologist should still be involved in checking all of these studies. However, for patients undergoing imaging with general anesthesia and patients in the intensive care unit, we decided that authorizing the technologists to use their knowledge and judgment rather than requiring mandatory checks of these studies was worth investigating.To start testing this intervention, the team leader talked informally to several of the CT technologists to gauge their understanding and confidence in three areas: (a) judging whether atelectasis was present, (b) assessing whether intravascular contrast material was present (ie, the contrast medium bolus did not extravasate), and (c) ensuring adequate anatomic coverage. The technologists uniformly agreed that they felt confident performing assessments in these three areas. The one area of concern was that of assessing the amount of atelectasis: They feared that differences in different radiologists’ tolerance for small amounts of atelectasis would lead to confusion. However, when the question used to assess atelectasis was reframed as a dichotomy (ie, atelectasis is present or not present) as opposed to a subjective judgment, their confidence increased.Next, we performed a 2-day trial of allowing the CT technologists to assess their images in these three areas without mandated radiologist checking of images for patients undergoing anesthesia and patients from the intensive care unit. The technologists were still encouraged to call the radiologist if they had any questions. Over the 2-day trial, there were no callback studies and the technologists reported no issues. We decided to continue this intervention, monitoring the callback rates and technologists’ comfort with the new process. This intervention was adopted.Radiography.—The technologists were required to have the radiologist check two types of radiographic studies before allowing a patient to leave the department: skeletal surveys for suspected nonaccidental trauma, and two-view airway radiographs. Skeletal survey images were checked to ensure that the positioning and image quality were adequate and that additional problem-solving views (eg, lateral view of an extremity) were not needed. The quality improvement team agreed that this was important for patient care and thus required the expertise of the radiologist, so mandatory checks of skeletal surveys were continued.The quality improvement team also decided to focus on checks of airway radiographs. Obtaining high-quality diagnostic airway radiographs in young children can be challenging because these patients often have difficulty holding still and are not able to follow the technologist’s instructions. Small differences in the head positioning and phase of respiration can lead to nondiagnostic images or even false-positive findings owing to underdistention of the hypopharynx. Therefore, we decided to perform a test of change to determine whether our technologists were able to evaluate the diagnostic adequacy of airway radiographs, with hopes of limiting mandatory checks of these studies.First, the team leader discussed the idea with several of the more experienced radiography technologists to get their opinion. There were mixed responses. While most of these technologists believed that they knew what a “good” airway radiograph looked like in general, the majority of them were not confident that they knew exactly what the radiologists needed and were not comfortable with the relevant anatomy. At this point, the quality improvement team had to decide whether to undertake a laborious educational effort to increase the technologists’ knowledge and confidence or attempt to perform a small-scale test of change in which they would rely on the technologists’ general gestalt in terms of what a “good” airway radiograph looks like. We decided to do the latter and allowed the technologists to judge the quality of these studies on the basis of experience. To mitigate any potential harm, we decided to do this only with patients from the emergency department or inpatient units who were undergoing imaging, as they could be easily called back to the department for additional imaging if needed.As a balancing measure, we monitored the total number of radiographs obtained for each two-view airway study to make sure that unnecessary repeat images were not being obtained in patients. For example, if we found that the number of patients in whom three or four images were acquired per study was increasing, this might mean that the technologists were not confident in their assessments and were obtaining repeat images “just in case.”Ultimately, this intervention was not successful and had to be abandoned. Some of the technologists did not feel comfortable assessing their images and still called the reading room to have images checked. There was confusion as to which airway studies needed to be checked and which ones did not, leading to frustration among both the radiologists and the technologists. Several radiologists reported that they did not feel comfortable calling a patient back to obtain more images if the study was of borderline diagnostic quality and thus felt forced to interpret suboptimal images. The quality improvement team tried to clarify this process at faculty meetings and through e-mails, but this was not successful. Additional creative ideas such as developing an artificial intelligence algorithm that could automatically assess image quality were proposed, but they were outside the scope of this project.Appropriate Telephone RoutingTelephone calls are by far the most common source of interruptions in our reading rooms. Several of these calls are necessary for quality patient care. However, in the course of collecting baseline data, we noted that there are also times when the phone rings inappropriately in the reading room; these include times when the call is not for a radiologist or should have been triaged by the reading room assistant first. Several different phone numbers all were connected to phones in the reading room, and multiple phones would ring simultaneously as a result of a single call. This created a chaotic environment in which there was confusion as to who was responsible for answering the phone.Frequently we witnessed the following sequence: The phone would ring in the reading room, but the radiologists would not answer it because they expected the reading room assistants to answer it. Once the reading room assistant answered the phone, he or she would then have to either call back into the reading room or physically come to the reading room to tell the radiologist that the call was for him or her. This caused two interruptions in quick succession: the original phone ringing and then the reading room assistant interrupting.The quality improvement team decided that a more coordinated approach to the phone system was necessary. The team members met with representatives from the hospital information technology department to develop a comprehensive telephone tree based on a similar phone tree that had been developed for the laboratory services department in the hospital. One of the quality improvement team members had participated in the development of that phone tree system and was able to use that experience to help model our new phone system.The key components of the new phone system were as follows: First, the actual direct phone number to the reading room was restricted, so that people calling into the department would be forced to use the new phone tree. Any phone call coming into the reading room would first ring at only the reading room assistant’s workstation. This enabled all calls to be appropriately triaged. If the reading room assistant was unavailable, the call would be saved in a queue until the reading room assistant was available. If a reading room assistant was still not available after 60 seconds, the phone would then ring at the radiologist’s workstation, to be answered by the radiologist. With this new system, the radiologists were now aware that if the phone in the reading room rang, either the call had already been triaged to them or the reading room assistant was unavailable. In either case, the radiologists now knew that they must answer the phone. Use of this system also decreased the number of double interruptions that had been occurring in quick succession.The radiologists and technologists had concerns about the new phone system. Some of them were uncomfortable with not knowing the direct number to the reading room. The quality improvement team appreciated this concern and made the radiologists and technologists aware that this was the only way to ensure that callers used the phone tree and that radiologists would realize the benefits of the new system.In addition, some individuals were concerned that callers might be placed on hold for long periods and this would led to frustration and dissatisfaction among clinicians and others calling the reading room to speak to a radiologist. To mitigate this concern, we kept the potential on-hold time to a maximum of 60 seconds. If the call was not answered by the reading room assistant by 60 seconds, it was automatically transferred to the reading room. In addition, we monitored the number of callers who dropped out of the on-hold queue, with these dropped calls considered an indirect indication that the on-hold time was unacceptable to these callers.Owing to the complexity of building the new phone system, this intervention was not implemented until after the primary study period had ended. However, we have adopted this intervention and received positive feedback from both radiologists and reading room assistants, with both groups reporting that the new system is an improvement over the chaotic and disorganized system that was previously in place.Clear Role ExpectationsIn our department, one radiologist was designated as the “resource” person each day. This radiologist was available to answer technologists’ questions, perform image checks, and answer general questions from referring clinicians. This radiologist was responsible for addressing questions regarding all subspecialty areas except neuroradiology. The role of the resource person was based on the radiologist’s clinical assignment for the day, and the physical location of this radiologist varied daily. However, there were unclear expectations as to who would assume this role when the reading room was understaffed—for instance, during educational conferences held in the morning and during the noon hour. This led to confusion among the radiologists and technologists, with the technologists often asking a different radiologist to perform the tasks of the resource radiologist according to the radiologist’s proximity to their work area, the ease of working with a specific radiologist, and other factors.We hypothesized that having more clearly defined and consistent role expectations might have positive benefits, including a reduced number of times that a radiologist who was not assigned the resource role was interrupted to perform a task that should have been the responsibility of the resource radiologist. To accomplish this, the resource role was delegated to a specific clinical assignment so that it would be consistent across days of the week. We decided that the primary thoracoabdominal radiologist would assume this role each day. In addition, we clearly identified the radiologist who would be responsible for covering the reading room during understaffed times (such as during the noon educational conference) and required that person to be physically present in the reading room at that time. Although these interventions did not lead to an appreciable improvement in terms of the median time between interruptions, the clarity of roles and responsibilities was appreciated by both the technologists and the radiologists, and this intervention was adopted.Clear Visitor ExpectationsThe physical design of our radiology reading area is such that any clinicians who come to consult with a radiologist must walk past several reading rooms before reaching the reading room assistant. The reading room targeted in this project is one of those. This reading room is also where most of the general on-call work is done, so a radiologist or radiology fellow is present 24 hours a day. Because of the ready availability of radiologists, clinicians are naturally inclined to visit this specific reading room if they have questions.At the onset of this project, we decided that maintaining the radiologists’ availability to perform clinical consults and interact with referring clinicians was important. Therefore, we did not wish to limit the number of clinical visitors to the reading room. However, we did want to make these interactions as efficient and nondisruptive as possible. For example, a clinical team might enter the reading area where musculoskeletal studies are read and ask about a head CT case. This requires the musculoskeletal radiologist to stop what he or she is doing to redirect the team to the neuroradiology area.To address this issue, signs were created and placed near the entrance to the reading area to direct visitors to the reading room assistant, who could then assist them in finding the appropriate radiologist. This intervention was not successful, as most clinical visitors ignored the signage and continued to go directly to the radiologists. However, because this intervention required only a one-time effort and no effort to maintain, it was adopted. In the future, additional interventions may include changing the physical layout of the reading area and stationing the reading room assistants toward the front. However, changing the physical space of the reading area was beyond the scope of this project.Data AnalysisData were collected manually and transferred to a computer spreadsheet. The times between interruptions were recorded in seconds. A run chart was used to display the data. The total number of interruptions per sampling period also was recorded and was displayed on a separate run chart. The median of the data points was used to evaluate for common cause variation and special cause variation. Significance was determined by using standard process control measures. The rules used to determine special cause variation that indicated a significant improvement were as follows: eight or more consecutive values above the median and six consecutive increasing values. When the special cause variation rules were met, the median time between interruptions was recalculated.ResultsObservational data were collected for the initial study period of August 21, 2018, to January 10, 2019. A total of 92 hours of sampling data were collected during the initial study period. To determine whether our improvement had been sustained, a second set of observational data were obtained by using the same method and by the same observers between July 22 and August 2, 2019. A total of 9 hours of sampling data were collected during this second study period.Baseline data were collected between August 21 and September 18, 2018. During this time, there were 19 baseline 1-hour sampling periods. The median time between interruptions was recorded in seconds. At baseline, the median time between interruptions was 187 seconds (Figs 5, 6). The total number of interruptions per hour also was recorded. The mean number of interruptions per hour at baseline was 10.5 (Fig 7).Figure 5. Baseline run chart shows the median time between interruptions before any interventions were attempted. The median time between interruptions at baseline (red line) was 187 seconds. The dashed green line indicates the time between interruptions goal. The green arrow indicates the desired effect of the interventions—in this case, to increase the median time between interruptions.Figure 5.Download as PowerPointOpen in Image Viewer Figure 6. Run chart shows the median time between interruptions (red line) during the primary study period (August 21, 2018, to January 10, 2019) and shorter second data collection period (July 22 to August 2, 2019). The median time between interruptions at baseline was 187 seconds and had increased to 336 seconds by the end of the project. The dashed green line indicates the time between interruptions goal. The green arrow indicates the desired effect of the interventions—in this case, to increase the median time between interruptions.Figure 6.Download as PowerPointOpen in Image Viewer Figure 7. Run chart shows the total number of interruptions per hour of data collection during the primary study period and the shorter second data collection period. The number of interruptions decreased from a baseline mean of 10.5 interruptions per hour to a mean of six interruptions per hour (red line).Figure 7.Download as PowerPointOpen in Image Viewer There was wide variation in both the median time between interruptions and the number of interruptions per hour during the primary (ie, initial) study period. The median time between interruptions ranged from 35 to 1019 seconds (16 minutes, 58 seconds) (Fig 6). The total number of interruptions ranged from two per hour to 27 per hour (Fig 7).Approximately 7 weeks into the primary study period, there was a shift in the median time between interruptions. The median time between interruptions increased from 187 seconds to 438.5 seconds, an increase of 134%. Another shift in the median time between interruptions occurred at approximately week 12, with the median time between interruptions decreasing to 336 seconds (23% decrease), but this still represented a significant increase in time between interruptions above the baseline.At the end of the primary study period, the median time between interruptions remained at 336 seconds, representing an 80% increase above baseline. The median time between interruptions of 336 seconds was maintained at the second data collection period performed approximately 7 months after the primary study period. The mean number of interruptions per hour also decreased during the study period, from a baseline of 10.5 interruptions per hour to six interruptions per hour, an approximately 43% decrease (Fig 7).In terms of counterbalancing measures, we monitored the number of callbacks for CT. Before this project began, our department averaged less than one callback per year for CT. There were no callbacks for CT during the study period. As a counterbalance for the new phone system, we also monitored the amount of time callers spent on hold and how many calls were abandoned (indicating that the caller was frustrated and hung up before completing the call). The average on-hold time was 6 seconds. The mean number of callers dropping out of the on-hold queue was 3.6% of the total call volume per month. To our knowledge, we received no negative feedback from callers regarding the new system.DiscussionThe radiology reading room is the center of diagnostic function of the radiology department. Radiologists in the reading room must be able to concentrate on the imaging studies with minimal distractions, as the studies that radiologists are required to interpret are becoming increasingly complex and include large numbers of individual images. Interruptions in the reading room can disrupt a radiologist’s concentration and may lead to errors or an increased time to interpret an imaging study (3–7). Although most interruptions are brief, each one still disrupts the radiologist’s primary task of image interpretation. Frequent interruptions may also increase a radiologist’s stress and anxiety levels, leading to higher risks of job dissatisfaction, fatigue, and burnout.We used quality improvement methodology to address the problem of frequent reading room interruptions. By using this process, we were able to better understand the source of interruptions; this enabled us to then develop interventions to eliminate unnecessary interruptions while still allowing appropriate interruptions that were important for patient care. For this project, we chose the median time between interruptions as our primary variable, as the time between interruptions represents the time when the radiologist is able to perform the primary function of image interpretation. In this quality improvement project, we were able to increase the median time between interruptions by 80%.One important concept that we tried to emphasize during this project is that not all interruptions are bad. Interruptions that are directly related to patient safety or patient care occur, and these interruptions should be considered appropriate and necessary. With this in mind, our goal was to optimize the number of appropriate interruptions while minimizing the number of inappropriate or unnecessary interruptions. We also decided as a group that direct collaboration with clinical colleagues, whether in person or over the phone, is an important part of the radiologist’s role and a source of job satisfaction for many radiologists. While it might be tempting to decrease interruptions by “locking” the door to the reading room and discouraging phone calls and visits from clinical colleagues, we decided that for our clinical practice, encouraging clinical consultations, rather than limiting them, is more appropriate. With this in mind, we focused our interventions on looking for unnecessary interruptions that originated in the radiology department and those that did not add value to clinical care.There were limitations to our project and data. First, owing to the complexity of the data that we were trying to record, we believed that in-person observation was the only feasible method for data collection (8). In addition, because of the labor-intensive nature of data collection, we were able to collect sample data for only 1 hour per day during the regular workday. It is possible that the sample data were skewed by minor variations in the time of day that data were obtained or that outside of the sample times, there was really no change in the median time between interruptions (eg, evening or on-call shifts). For some of the interventions, such as decreasing the number of CT image checks, we had no reliable way to determine whether the technologists were uniformly and consistently following the new process. Finally, we were unable to account for daily and seasonal variations in patient volume, which may have affected the number of interruptions in the reading room.Despite these limitations, the general feeling in the department was that the project was a success. The new processes have for the most part been well received by both the radiologists and the technologists. Several radiologists have commented that just the act of our department leadership and quality improvement team acknowledging and addressing the issue of reading room interruptions has improved the general morale of the reading room staff. Although the primary phase of this project is now complete, we plan on revisiting this project in the near future and hope to make continued improvements to optimize interruptions in our reading rooms.ConclusionBy using quality improvement methodology and a team approach, we were able to increase the time between interruptions in our busiest reading room by 80%. Although we have met our initial project goal, we plan to continue our work to optimize interruptions and further increase the time between interruptions and thus decrease radiologist stress levels and increase efficiency.Disclosures of Conflicts of Interest.—A.J.T. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: serves on advisory boards of KLAS Research and IBM Watson Health and as a consultant for Applied Radiology; institution received grants from Guerbet and the Cystic Fibrosis Foundation; received royalties from Elsevier. Other activities: disclosed no relevant relationships.A.J.T. has provided disclosures; all other authors have disclosed no relevant relationships.References1. Eyrolle H, Cellier JM. The effects of interruptions in work activity: field and laboratory results. Appl Ergon 2000;31(5):537–543. Crossref, Medline, Google Scholar2. Bowman LL, Levine LE, Waite BM, Gendron M. Can students really multitask? an experimental study of instant messaging while reading. Comput Educ 2010;54(4):927–931. Crossref, Google Scholar3. Balint BJ, Steenburg SD, Lin H, Shen C, Steele JL, Gunderman RB. Do telephone call interruptions have an impact on radiology resident diagnostic accuracy? Acad Radiol 2014;21(12):1623–1628. Crossref, Medline, Google Scholar4. Yu JPJ, Kansagra AP, Mongan J. The radiologist’s workflow environment: evaluation of disruptors and potential implications. J Am Coll Radiol 2014;11(6):589–593. Crossref, Medline, Google Scholar5. Kansagra AP, Liu K, Yu JPJ. Disruption of radiologist workflow. Curr Probl Diagn Radiol 2016;45(2):101–106. Crossref, Medline, Google Scholar6. Ratwani RM, Wang E, Fong A, Cooper CJ. A human factors approach to understanding the types and sources of interruptions in radiology reading rooms. J Am Coll Radiol 2016;13(9):1102–1105. Crossref, Medline, Google Scholar7. Williams LH, Drew T. Distraction in diagnostic radiology: how is search through volumetric medical images affected by interruptions? Cogn Res Princ Implic 2017;2(1):12. Crossref, Medline, Google Scholar8. Towbin AJ. Collecting data to facilitate change. J Am Coll Radiol 2019;16(9 pt B):1248–1253. Crossref, Medline, Google ScholarArticle HistoryReceived: Apr 26 2020Revision requested: June 12 2020Revision received: July 2 2020Accepted: July 9 2020Published online: Mar 01 2021Published in print: Mar 2021 FiguresReferencesRelatedDetailsCited ByThe Virtual Radiology Reading Room: Initial Perceptions of Referring Providers and RadiologistsJoseph H.Yacoub, Matthew D.Bourne, PranayKrishnan2023 | Journal of Digital ImagingBurnout Is the Symptom, Moral Injury Is the Cause: The Need for Systems-Level Changes in RadiologyRama S.Ayyala, GraysonBaird2023 | American Journal of Roentgenology, Vol. 220, No. 2Communication — a lost art?Janet R.Reid2022 | Pediatric Radiology, Vol. 52, No. 4Direct Access and Skill Mix Can Reduce Telephone Interruptions and Imaging Wait Times: Improving Radiology Service Effectiveness, Safety and SustainabilityChristopherWatura, CharlotteKendall, PaulSookur2022 | Current Problems in Diagnostic Radiology, Vol. 51, No. 1Workflow Interruptions and Effect on Study Interpretation EfficiencySummit H.Shah, Lamya A.Atweh, Corey A.Thompson, StephanieCarzoo, RajeshKrishnamurthy, Nicholas A.Zumberge2022 | Current Problems in Diagnostic Radiology, Vol. 51, No. 6Recommended Articles Increasing the Utilization of Moderate Sedation Services for Pediatric ImagingRadioGraphics2021Volume: 41Issue: 7pp. 2127-2135Quality Improvement Report: Improving Pre- and Postprocedure Care Area Workflows at a Busy Urban Academic Hospital Using Lean Management PrinciplesRadioGraphics2022Volume: 43Issue: 2Customer Service in Radiology: Satisfying Your Patients and ReferrersRadioGraphics2018Volume: 38Issue: 6pp. 1872-1887Magician’s Corner: 6. TensorFlow and TensorBoardRadiology: Artificial Intelligence2020Volume: 2Issue: 3Increasing Patient Access to MRI Examinations in an Integrated Multispecialty PracticeRadioGraphics2021Volume: 41Issue: 1pp. E1-E8See More RSNA Education Exhibits Ergonomics In The Radiology WorkplaceDigital Posters2021Anesthesiology Concepts For The Interventional RadiologistDigital Posters2021Transition From Peer Review To Peer Learning: Lessons Learned.Digital Posters2021 RSNA Case Collection Allergic Bronchopulmonary AspergillosisRSNA Case Collection2021Pulmonary SequestrationRSNA Case Collection2022Sacral Insufficiency FracturesRSNA Case Collection2021 Vol. 41, No. 2 Metrics Altmetric Score PDF download