Analysis using HCD framework pinpointed three specific areas for improvement of current surgical pathology reporting: 1) final margin status, 2) ambiguous anatomic relationships, and 3) lack of relevance to downstream clinicians. Our multidisciplinary approach enabled the development of an improved pathology report that enhances clarity of final margin status and reconciles the extent and location of supplemental margins.
HCD methodology is not standardized, and the three phases (Understand, Explore, and Materialize) may use slightly different terminology across institutions. During the Understand phase, researchers gather as much information as possible about the target population and problem at-hand. During the Explore phase, researchers work with end-users to brainstorm effective solutions to solve identified problems. Finally, during the Materialize phase, the prototype is placed into practice, followed by end-user feedback, improved prototypes based on feedback, and retesting of those prototypes.
Several medical institutions have invoked HCD in their research programs and medical education curricula [14, 16]. Additionally, HCD methodology has been increasingly used in medicine to provide creative solutions—such as a triage tool for the pediatric emergency departments [17]—that improve patient care. We invoked these methods to co-design an effective surgical pathology report that addresses gaps identified by the downstream providers who regularly rely on pathology reports. This project is uniquely positioned for HCD methodology because of the collaboration that the permanent pathology report invokes between surgeons, pathologists, oncologists, and other clinicians involved in the postoperative care of cancer patients.
It is not surprising that there was a divide between surgeons/pathologists and medical/radiation oncologists with respect to satisfaction with current pathology reporting. Surgeons and pathologists are intimately involved in the activities in the operating room and frozen section lab. However, the reimagined pathology report allows even pathologists to understand how well supplemental margins actually address margins at-risk by identifying the exact location and the breadth of supplemental margins harvested. This allows the pathologist to confidently report on the final margin status. Previously, it was not possible to bridge this essential information gap relying solely on the labeling of supplemental margins.
The surgical pathology report is integral to the complex multidisciplinary management of postoperative oncology patients. Ideally, it should present findings in a manner that is easily understood by all. A systematic review by Mossanen et al. (2014) assessed 25 articles on the impact of pathology report formatting on clarity of communication. Their review highlighted the need for better organization within the report, permitting succinct understanding of results and effective interactions between pathologists and clinicians [18]. The pathology report also plays a pivotal role in holding surgeons and pathologists accountable, and significantly impacts surgical liability [1]. A well-defined, comprehensive documentation system can reduce human error and ultimately improve patient care.
We have previously described our workflow for 3D specimen scanning [5, 12, 13]. The complexity of large head and neck resections render them difficult to document, either by drawing or descriptions. A surgeon needing to understand the location of margins at-risk might have to walk to the frozen laboratory, only to find a dissected specimen bearing no resemblance to the one that was resected. Videoconferencing the margin mapping results using the specimen representations streamlines this process, avoids surgeon down-time, and decreases frustrations for both surgeon and pathologist. The optical scans are produced in the frozen laboratory prior to specimen dissection. After specimen scanning, the pathologist works on the actual resection specimen, identifying margin planes, inking them, and dissecting in the usual manner. Perpendicular scalpel cuts are produced through all margin planes, to visually identify where histological margin sampling is needed. As the tissue blocks are being cut and frozen slides are produced, the pathologist annotates the location of sampled margins onto the virtual specimen representation. Following videoconferencing to describe the margin results, the surgeon annotates 3D defect scans to document location and size of harvested supplemental margins. We have previously demonstrated that 3D defect scans are superior to standard conventional defect photography, as the defect scans are rotated in space to optimize documentation of supplemental margins [12].
The current literature has yet to address the relationship between pathologists, surgeons and medical and radiation oncologists, and their collaborative quest to deliver vital surgical / pathology information via surgical pathology reporting. We call attention to the pressing need for a more integrated approach among providers, reducing informational silos and increasing understanding of interdisciplinary perspectives. Designing a surgical pathology report that can effectively convey visual anatomic information helps to bridge the communication gap between specialties and promotes a more collaborative approach to patient care. Embedding screen shots selected from 3D specimen and defect models enhance the reader’s ability to appreciate visuospatial information about the locations of margins at-risk, and the harvesting of supplemental margins. This enables downstream clinicians to better participate in shared decision-making processes.
With respect to the final margin status, the established final pathology reporting is flawed and fraught with vagaries regarding the labeling process for harvested supplemental margins [19]. The specimen label may bear little resemblance to the anatomic location of a known margin at-risk. The pathologist may be unable to confirm that a specific supplemental margin actually addresses a particular margin at-risk. Indeed, the difficulty in returning to exact regions at-risk is discussed by Maxwell and colleagues with respect to low-stage oral cancer patients requiring supplemental margins [20]. They compared outcomes for 3 groups of patients; Group 1: patients with en-bloc resections and negative specimen-driven margins; Group 2: patients with inadequate preliminary specimen-driven margins requiring supplemental margins; Group 3: patients with surgeon sampled margins taken only from the surgical defect. Not surprisingly, the margin status for Group 3 patients did not correlate with outcome, and these patients had the worst local control. These data support resection specimen-driven margin assessment over defect only margin sampling. The Kaplan-Meier curves also demonstrated a trend (p = 0.06) for worst local recurrence for Group 2 compared with Group 1. One possible explanation is the lack of accuracy when returning to the surgical defect to harvest additional supplemental margins. Another explanation may be with the distribution of worst pattern of invasion, not assessed in these groups. Group 1 may have been skewed towards nonaggressive pattern of invasion, which is more amenable to resection with negative margins, as compared to Group 2, which might have contained more patients with aggressive pattern of invasion, which is more likely to be associated with incomplete resection, and inherently poorer prognosis [21]. Our approach provides precise anatomical context for margin locations and offers a comprehensive analysis of their status, description, and visualization, which further improves comprehensibility of surgical interventions.
Radiation oncologists play an important role in a patient’s postoperative care, which includes planning target volume, dosing boosts, and limiting toxicity to adjacent structures [22]. We believe that our redesigned pathology report offers downstream providers a better understanding of important anatomic details. Embedding preoperative radiographs into final pathology reports, with the surgeon’s annotations for overall resection contours, and at-risk regions, which required supplemental margins, will help stakeholders to correlate intraoperative findings with postoperative radiographs. Our preliminary survey highlights the difficulties for downstream physicians to understand the precise locations of supplemental margins, especially if several rounds of supplemental margin harvesting are required. Visuospatial annotations of at-risk margins onto the 3D specimen models and harvested supplemental margins onto the 3D defect models provide an excellent means for transmitting information. The importance of embedded annotated radiographs becomes very clear when considering the radiation oncologist’s perspective. The radiation oncologist must embark on an electronic health record treasure hunt, synthesizing data from pathology reports, operative notes, and imaging, to understand where difficulties in tumor clearance were encountered, and what additional tissues were harvested to attain negative margins. These anatomic regions at risk may receive radiation dose boosts. Usually, radiation oncologists might discuss margin status with surgeons several weeks postoperatively, when a surgeon’s memory is murkier. We hope that our redesigned reports will obviate the need for those postoperative discussions. The embedded radiographic details convey 3D landmarks and becomes a vital link from the specimen and defect scans to the overall postoperative reconstruction. This is extremely important as complex surgical resections often require complex reconstructions that include bone and soft tissue taken from other parts of the body. Annotation of defect scans and preoperative radiographs provides a direct link to areas at-risk in postoperative / pre-adjuvant therapy scans.
Limitations
Response and selection bias were study limitations. Survey responses were limited to only US-based medical institutions and a small sample of physicians. Future work should expand the sample size of included stakeholders from more varied institutional types to further guide development. In this study, we present the best-case scenario of a workflow between head and neck surgeons, pathologists, designated space, equipment, and technical research staff located in the frozen section laboratory and operating room. We recognize that the additional time, effort, and personnel may seem daunting to surgeons and pathologists in other institutions attempting to emulate this process. However, one cannot contest the additional value and inherent beauty of this approach in delivering more detailed information on pathology findings and the actions taken, which sets the stage for more precise postoperative management and surveillance. In our effort to establish a new gold standard of information transfer, we anticipate that future streamlining of this process will reduce the time required to achieve the desired end.
Another limitation lies in the inherent challenges of using intraoral 3D scanning to accurately capture complex anatomical areas in smaller defect regions such as the base of the tongue, tonsils, hard and soft palate, oropharynx, and hypopharynx [5]. We are currently developing a library of “stock” 3D defect models to use when necessary.