https://orcid.org/0000-0002-9623-7533
Abstract
To review a single institution’s second opinion breast imaging process, data tracking, and metrics before and after implementing quality improvement changes and the effect on report turnaround time.
This Institutional Review Board approved retrospective quality improvement project was performed at a tertiary-care academic medical center and included patients 18 years or older who submitted their outside facility imaging for reinterpretation (any combination of mammography, breast ultrasonography, and/or magnetic resonance imaging performed within the last six months) with finalized second opinion reports between June 1, 2016, and July 17, 2017. Significant intradepartmental changes were implemented March 2017 with the goal to improve second opinion report turnaround time. Key metrics from 399 studies were analyzed before and after implemented changes. Two-sided Fisher’s exact test was used to assess the significance of results.
After department interventions, the percentage of outside reports available at the time of surgical consultation improved from 82% (213/259) to 91% (127/140), an 11% improvement (P < 0.05). The average number of days from initial second opinion consultation to the availability of final report improved from 10.2 days to 9 days, a 12% improvement. Prior to the changes, the number of days it took a radiologist to complete a report varied from 1 to 4 days, but afterwards was consistently 1 day or less.
Implementation of second opinion intradepartmental changes demonstrated a significant improvement in report turnaround time and the number of finalized reports available at the time of surgical consultation. An efficient second opinion process is crucial to a breast imaging center, as it ultimately expedites patient surgical and oncological care.
The process of second opinion review of breast imaging studies involves multiple steps, relies on collaboration with the patient and an outside hospital to obtain all necessary components, and, therefore, requires meticulous management.
Components of the quality improvement process which contributed to the overall success of the program included a full-time designated second opinion coordinator, a daily designated clinical consult radiologist to review cases for completeness, and a regular review of case status and failed cases.
Implementation of second opinion intradepartmental changes demonstrated a significant improvement in report turnaround time and the number of finalized reports available at the time of surgical consultation.
Introduction
Breast cancer is the most common noncutaneous cancer in women in the United States (1). For a variety of reasons, patients with abnormal breast imaging findings may seek further consultation with breast surgeons and oncologists at different institutions. Reinterpretation of outside breast imaging is a critical part of the new consultation review process. Review can reveal new or additional sites of malignancy, reduce unnecessary interventions that may contribute to patient psychologic distress, and add value to patient care (2). In addition, many patients choose to complete the remainder of their treatment at the site of second review (3).
Research demonstrates that a review of imaging performed at a different facility plays a pivotal role prior to the initiation of any breast treatment, whether it be surgical or oncologic (2, 3). Coffey et al reviewed 200 breast imaging cases at a major cancer center and found second opinion recommendations led to a change in interpretation in 28% of cases (2). Additional biopsies recommended beyond the initial interpretation yielded 10 additional malignancies and 4 high-risk lesions, while 8 biopsies were averted based on benign interpretation of imaging (2). Surgical management was changed to mastectomy for 6 of 10 patients with new sites of biopsy-proven cancer (2). Spivey et al reviewed 380 patients who submitted outside facility (OF) imaging at a major academic center and found 53.5% of patients had a change in recommended management plan, resulting in an overall change in surgical management in 27.1% of cases (3).
Given the significance of second opinion review (SOR), it is critical for a radiology department to produce a timely report to aid in surgical consultation and management. The process of SOR is surprisingly complex, time consuming, and labor intensive, with multiple moving parts. There are many ways the process can fail or become prolonged, with a delay in interpretation bottlenecking the initiation of treatment. Thus, a streamlined and efficient process is ideal.
Many studies have evaluated the differences in imaging interpretations between institutions, as well as the impact on clinical and surgical management. However, few papers describe their breast imaging department’s process of SOR and/or techniques for improvement. This paper aims to describe our process of second interpretation with implemented changes, methods of data tracking, and the effects department interventions had on key metrics and report turnaround time.
Methods
This retrospective, quality improvement (QI) project was approved by our Institutional Review Board and performed at a tertiary-care academic medical center. A waiver of consent and a waiver of Health Insurance Portability and Accountability Act (HIPAA) authorization was obtained.
Inclusion criteria consisted of (1) patients aged 18 or older who submitted their OF imaging for reinterpretation (modalities included mammography, breast ultrasonography, and magnetic resonance imaging), (2) imaging that was performed within the past six months, and (3) a finalized second opinion report that was signed by the radiologist at our institution between June 1, 2016, and July 17, 2017. Exclusion criteria consisted of (1) the absence of a formal radiology report from the OF, and (2) the absence of a formal SOR report at our institution.
Prior to the implementation of changes for this project, the SOR process would begin with a patient requesting a second opinion consultation appointment with a breast surgeon. Upon doing so, a formal intake questionnaire would be initiated in their electronic medical record, notifying our imaging department to begin the process of image review. The patient would then be required to sign a release form from their original institution to allow transfer of their imaging. Once the release was signed, an available technologist assistant (TA) would request the appropriate images and reports from the past five years. A TA is a full-time staff position whose primary responsibility is to facilitate the radiologist’s diagnostic schedule by helping organize paperwork and orders, uploading prior imaging, and performing other tasks to help expedite patient care in the diagnostic setting. This results in a decrease in workload for the technologist who would otherwise be performing these tasks. The TA would also help with the outside read process when not performing other duties that required immediate attention. Once the prior images were requested by the TA, the SOR would then be set aside until images were received. Upon complete receipt of images and reports, the next available TA (often times a different TA than the one who initially requested the images) would assign the appropriate accession numbers and submit the case to one of 14 radiologists to review, reinterpret, and finalize a report. Without a streamlined process, delays in image reinterpretation often arose due to the lack of continuity between TA staffing, assignment of cases, and the absence of a formal process to check that all images were received and uploaded. This often resulted in a patient undergoing surgical consultation at our institution without a formalized SOR by our radiologists, frequently delaying patient care.
As of March 2017, we implemented multiple department changes with the goal to improve report turnaround time. The new process is outlined in detail in Figure 1. One new full-time staff position was created—a dedicated second opinion coordinator, funded by the Department of Radiology, who was tasked with overseeing all components of the OF imaging review process. The workflow of this coordinator began with (1) ensuring the intake questionnaire was complete, (2) contacting the patient to complete a form to release images from the OF, (3) subsequently mailing, faxing, or emailing the release form to the OF, (4) coordinating and following-up with the OF to receive all appropriate and requested imaging in a timely fashion, (5) confirming that all imaging was uploaded into the Picture Archiving and Communication System (PACS), (6) evenly distributing cases among the 14 breast radiologists, and (7) verifying a final radiologist report was available prior to surgical consultation. Weekly meetings were instituted between the coordinator and a department manager with the goal to identify patients scheduled for consultation in the upcoming week without a completed review in order to prioritize cases given to radiologists.
Second opinion consultation workflow after implemented changes.
Few regional hospitals near our center can electronically transmit images; therefore, we have historically relied on standard mail to transport images, which contributed to significant delays. Lack of routine mail delivery on holidays and Sundays further delayed the receipt of imaging. In an effort to expedite the delivery of images, our department set up a dedicated mail account covering outside facility costs to send imaging via a 1 to 2 day mail service.
Additionally, one faculty radiologist was given the title “clinical consult radiologist” to assist the second opinion coordinator. Previously, an outside review was not given to the radiologist until all imaging from the past five years was available. Although it is optimal to have the entire folder available for review, occasionally components of the patient’s record either were not available or delayed in arrival. In the new process, in cases of missing images or looming surgical consultation, the consult radiologist would determine if the missing image was critical for interpretation. For example, missing stereotactic biopsy procedure images, but the specimen radiograph and post-procedure mammogram were available. If a complete and accurate report could be rendered without the missing information, the coordinator could assign the case to a radiologist, thereby avoiding an interpretation delay. No extra time was provided to this radiologist to perform as consultant; cases were reviewed in between daily casework.
Next, the intake questionnaire, which is required to be completed prior to an OF imaging review, was edited and optimized to improve communication flow between the radiology and surgery departments (Table 1). To help streamline the process of OF image requests, scheduled consultation date, reason for appointment, and a list of previous imaging based on patient memory was added to the existing questionnaire.
Patient Intake Questionnaire
|
|
Patient Intake Questionnaire
|
|
Furthermore, we utilized the Enterprise Data Warehouse (EDW) (Syncsort, Pearl River, NY) a HIPAA-compliant database stored on secure shared university servers and accessible on single log-on password-protected computers, to generate reports summarizing consultation deadlines. These reports were reviewed and monitored periodically to examine cases considered a “miss,” as defined as a radiology report that was not available in time for a surgical consultation. Finally, a monthly “miss meeting” was held to monitor and review outcomes. Cases not meeting goals were isolated and discussed to determine if a miss was preventable. A control plan process map was utilized and followed (Figure 2).
Control plan process map after department quality improvement process.
To measure how our departmental changes influenced report turnaround, we retrospectively pulled data from our EDW system, from June 1, 2016, to July 17, 2017. The EDW provided a shared worksheet for our department managers and, as of March 2017, to the newly appointed second opinion coordinator to keep track of patient imaging requests. As a standard ongoing policy, we routinely kept track of key metrics influencing turnaround time. This included dates for when (1) the patient contacted the breast imaging department to make a surgical consultation (initiating the process of outside imaging review), (2) imaging was requested from the OF, (3) imaging was received, (4) imaging was assigned to the radiologist, and (5) reinterpreted reports were available to the consulting service. Also included was the overall total number of days from initial patient contact to finalized radiologist report. The percentages of SOR reports available at the time of surgical consultation before and after implementation of changes were documented and analyzed using a two-sided Fisher’s exact test.
Results
A total of 490 patients requested OF SORs between June 1, 2016, and July 17, 2017. There were 91 cancelations resulting in a total of 399 outside imaging studies included in our analysis (Figure 3). The SOR was considered canceled if the patient withdrew their surgical consultation request before outside imaging was fully acquired and reported, or if the relevant imaging was found to be greater than six months old.
Case volume per month. *Only data through July 17 was collected.
Our study spanned a total of 412 days and, therefore, 287 business days from which weekends and hospital holidays were excluded. This resulted in an average of 1.4 fully produced reinterpretations per business day. As data collection ended July 17, 2017, instead of at the end of the month, the following numbers exclude July 2017. The average number of cases requested per full month was 36.7 (range, 22–50) (Figure 3). The average number of cancelations per full month was 6.6 (range, 0–12). The average number of cases performed for review per full month was 30.1 (range, 19–38) and the median was 31.
Prior to the implementation of department interventions, an average of 82% (213/259) of outside diagnostic reports were available in the electronic medical record at the time of surgical consultation. After the implementation of department interventions, on average, 91% (127/140) of reports were available at the time of surgical consultation (Figure 4), an 11% improvement (P < 0.05).
Percentage of patients with outside facility imaging interpretations reported in the electronic medical record by the time of scheduled surgical consultation. *Only data through July 17 was collected.
The average number of total days required to produce a second interpretation report (days from initial consult to availability of final report) gradually decreased month by month. This decreased from 10.2 days preintervention to 9.0 days postintervention, an improvement of 12% (Figure 5). The number of days a breast radiologist took to complete their report substantially decreased after interventions. This originally varied from one day to as many as four days, but thereafter remained consistently at one day or less.
Average number of days from consult initiation to time report was finalized in the electronic medical record.
We had complete data for the number of days from (1) patient contact to production of our final imaging report, and (2) from radiologist assignment to completion. However, due to a database error, of the 399 total outside review studies, we only had complete data regarding the number of days it took to obtain outside hospital images for 262 cases. Of these 262 cases, the minimum number of days required to receive imaging was zero (images were received same day) and the maximum number of days was 33. The average number of days required to receive OF images was 4.5 days, with a median of 3 days. The total number of cases in which imaging was received in 2 days or less, 5 days or less, 7 days or less was 116 (44%), 189 (72%), and 220 (84%), respectively. Only 32 (12%) of the cases were received in 10 or more business days (Figure 6).
Time required to receive requested outside facility imaging.
Discussion
Implementing multiple changes to our second opinion process streamlined our review of outside facility images. We found a significant improvement in the number of available second opinion reports available at the time of the surgical consultation and an overall decrease in the length of time it took to generate a SOR.
When we embarked on this project, the challenges of obtaining OF images were immediately obvious. Standard mail contributed significantly to the delay in OF image interpretation. Therefore, it seemed intuitive to have images overnighted to our facility to speed up the process. However, despite having a dedicated coordinator organizing imaging requests, and the availability of a dedicated mail account where we incur the cost of shipping images, less than half of patients’ images were received in a 1 to 2 day timeframe. This may be related to the outside imaging facility’s inability to make imaging available in a timely fashion or other factors not under our control. Nevertheless, it is likely that far fewer patients would have had available imaging if this shipping option was not implemented.
Another obstacle we frequently faced was incomplete imaging, with or without an OF report. By designating a radiologist who preliminarily reviewed the incomplete imaging to determine if it was sufficient for interpretation and expediting some cases that would have otherwise been held by the coordinator, the average number of days required to produce a reinterpreted report decreased from 10.2 days preintervention to 9.0 days postintervention. While this may not seem substantial, what is not readily apparent in this statistic is the overall workflow and organization. After implementing our changes, the imaging that arrived at a radiologist’s desk was complete and ready for interpretation. The necessary and available images were present and uploaded into PACS with corresponding prior reports. Previously, most of the work fell on the radiologists who had to notify staff about missing components or lack of images uploaded into PACS. In our new process, the radiologist reviews the case at their convenience rather than rushing to produce a report within minutes of a looming surgical consultation. Although we did not quantify this in our study, the overall improved process significantly reduced the radiologists’ frustration at the time of interpretation and resulted in a less frantic review for the surgeon.
The number of days a radiologist took to complete their report varied prior to interventions, but thereafter stayed consistent at only one day. While this suggests that our implemented department changes had an impact, the degree to which it did is somewhat uncertain, as there may be confounding variables. For example, improved turnaround could be attributed to having the coordinator keep track of which radiologists were in the office each day, thus preventing the assignment of a case to a radiologist out of the office. It could also have been due to improved image organization and completeness at the time of SOR presentation to the radiologist. However, a confounding variable contributing to radiologist turnaround time may have been the new department changes themselves, placing a new spotlight on report production metrics, and thereby creating a perception of new expectations that turnaround times should be faster. Radiologists may have subconsciously or intentionally self-imposed deadlines, feeling pressured into finishing their reinterpretations the same day rather than waiting until the next day when they might have more time.
Our study has several limitations. One limitation of our study was that OF imaging only included patients with a surgical appointment and did not include patients who were transferring care for other reasons. We chose this patient subset because increasing the percentage of patients with a final report prior to surgical consultation was the primary goal. Therefore, we may not have improved the process for patients seeking a biopsy of a suspicious lesion at our institution or a radiological second opinion. Additionally, in response to increasing volumes and second opinion requests, we implemented multiple departmental changes in a short period of time. Thus, it is not possible to measure whether one intervention was more or less contributory to a successful reduction in turnaround time. For example, the impact of having a dedicated coordinator could not be accurately measured against the benefit of weekly and monthly quality review meetings of missed cases.
Finally, our results and department changes may not be transferrable or feasible for all radiology centers. It may not be possible or necessary for smaller radiology centers or nonsubspecialty groups to hire a dedicated second opinion coordinator. Similarly, the benefit of the EDW, or the labor needed to track all moving parts for every patient, may not be sustainable by every department. Funding for a dedicated mail account to help expedite time required to obtain films from another city or state may not be practical. However, coordination with surrounding hospitals to electronically exchange all images and reports would significantly improve this process and would be universally helpful, regardless of practice type.
Conclusion
The process of SOR of breast imaging studies involves multiple steps, relies on collaboration with the patient and outside facilities to obtain all necessary components, and, therefore, requires meticulous management. Our process improvement plan for SOR demonstrated a significant improvement in report turnaround time and the availability of reports at the time of surgical consultation. An efficient second opinion process is crucial to a breast imaging center, as it ultimately expedites patient surgical and oncological care.
Funding
None declared.
Conflict of interest statement
None declared.
Acknowledgments
We would like to acknowledge Barbara Buckley, RN, MS, and Janine Dorsey for their significant contributions in providing data for the publication of this manuscript. Additionally, we would like to acknowledge the support for the statistical analysis from Julie Anne Blaisdell, MS, Department of Radiology, Northwestern University.
