key: cord-0901801-h7nfgco5
authors: Samueli, Benzion; Sror, Neta; Jotkowitz, Alan; Taragin, Benjamin
title: Remote pathology education during the COVID-19 era: Crisis converted to opportunity
date: 2020-08-29
journal: Ann Diagn Pathol
DOI: 10.1016/j.anndiagpath.2020.151612
sha: 2700e71ae453f4243a2e29a3369672d64f4f064c
doc_id: 901801
cord_uid: h7nfgco5

CONTEXT: The COVID-19 pandemic led to shutting of education faculties, including clinical clerkships for medical students. OBJECTIVE: To review a selective for a course in diagnostic pathology geared toward undergraduate medical students, including its design, technical implementation, instructor and student evaluations, and suggestions for options for further adjusting and optimizing the selective. DESIGN: Whole slide images (WSI) were anonymized and students were given remote access to university computers, which were prepared with two freely available WSI viewers. Each topic was taught in a four-part module: Self-assigned reading, lecture via Zoom, quiz based on digital slide sets, and a frontal review of the slides via Zoom. Fifty-nine students participated in the selective. Following the course, students completed an anonymous questionnaire. RESULTS: Of the 59 participants, 42% (n = 25) responded. None of the respondents had any previous instruction in diagnostic pathology. Overall, the course was rated very favorably: 68% (n = 17) gave at least 3 points on a 4-point scale on questions relating to course interest, improvement in understanding of the covered diseases, and how strongly they would recommend a student take this course if given an option. The most significant disadvantage of the class, as reported by 80% (n = 20) were technical challenges in accessing the slides. CONCLUSION: We believe the course was a success and can be a model for future virtual pathology electives. Great effort should be done to provide technical support to the students. The selective demonstrated value for students and provided much-needed exposure to diagnostic pathology in clinical practice.

On 11 March 2020, the World Health Organization publicly "made the assessment that COVID-19 can be characterized as a pandemic" [1] . Four days later, the State of Israel announced a shutting of all educational facilities [2] . This decision precluded medical students both from in-person lectures and being present in the hospital wards. The Medical School of International Health (MSIH) of Ben Gurion University of the Negev (BGU) in Israel is a four-year international MD program modeled on the United States educational curriculum. As such, students in their first two years (termed M1 and M2) are primarily focused on lecture-based preclinical studies, whereas upperclassmen (M3 and M4) normally participate in one-month electives and two-week selectives in the clinical setting, such as hospital wards and outpatient clinics.

Guidelines issued by the State required MSIH to implement a variety of alternate educational solutions in order to prepare the students to practice medicine and to maintain their academic timeline for graduation and advancement to fourth-year. While various distance learning solutions have been adapted for pre-clinical instruction, for which it is more suitable, the same format is not as well tailored to M3 and M4 clinical experiences. Nonetheless, distance learning was used for various courses in these academic years, including radiology and psychiatry.

One of the implemented solutions at MSIH was to create a new Diagnostic Pathology Selective, in which distance learning was used to provide upperclassmen with a unique educational opportunity. Herein, we describe the course, including our technical implementation, benefits and limitations, and student satisfaction. Before this course was designed, students at MSIH were given a dedicated introduction to pathology course in their first year which covered general pathology, followed by integration of systems pathology within the later courses. However, with the exception of the very occasional tumor boards attended by some of the students (see Results, Section 3), we feel there is traditionally inadequate exposure to surgical pathology in clinical practice and decision making.

A number of previously published papers provide background and useful insights for developing a education. In the 8 years that have passed since its publication, additional software has been developed for this purpose, which we review here, and importantly incorporate it as a part of a curriculum designed for distance learning. A more recent article [6] provides a brief overview and instructor evaluation of a breadth of these newer resources, including social media; however, it similarly lacks a helpful curriculum outline and retrospective evaluation of resource implementation.

Innumerable web-based slide sharing applications have been developed, too many to be mentioned.

The publication [7] of the American Association for Anatomy's Virtual Microscope Database (VMD) opens with the many advantages of using digital microscopy in education from the student, instructor, and institutional perspective. However, in our experience, "centralized" digital slide archives, including the VMD, have many significant limitations. These include difficulties in student registration, slow response time, limitations in slide availability (both in terms of diagnosis and in ancillary tests that could have also been scanned), friendliness of the user interface, compatibility with various devices and operating systems, and ability to save notations (students and instructors). For all of these reasons, at the current time, we recommend that an educational institution with the resources to host its own slide set should do so. In the future, a fast and well-designed centralized slide archive that offers the ability for instructors to upload their own collection could alleviate the value of a local slide set, and could distribute the costs of maintenance via institutional subscriptions, but after extensive exploration and testing, it seems that nothing of the sort yet exists.

We believe that the following paper will be an invaluable guide to pathology educators in the undergraduate setting for two principal reasons: First, the emphasis on distance learning will likely be greater post-COVID-19 than in the past. Described herein are various options for the specific implementation of distance learning in pathology, including our experiences and evaluations of these options. Second, undergraduate medical students frequently are underexposed to diagnostic pathology, and the curriculum discussed in this paper successfully piqued student interest in the field, at least in the short-term.

In order to prepare for the course, the Course Coordinator (BS, corresponding author) prepared a curriculum with accompanying slide sets from surgical biopsies in the Department of Pathology at J o u r n a l P r e -p r o o f Journal Pre-proof Soroka University Medical Center. In this section, we first describe with technical implementations of the course following by the academic aspects.

Whole slide images (WSI) were prepared using a Pannoramic MIDI automated digital slide scanner (3DHistech, Budapest). The WSI's were anonymized by removing all identifiable information, including the coded biopsy number assigned for internal department tracking purposes. The students were given access to a virtual private network service (VPN), which then enabled remote desktop access to BGU computers. The BGU computers were prepared with two freely available WSI viewers: CaseViewer 2.3 (3DHistech, Budapest) and Aperio ImageScope 12.3.3 (Leica, Illinois). Alternatively, using the VPN, students could map a particular BGU server as a network drive and use any compatible WSI viewer on their computer; this solution was useful for students who had difficulties with remote desktop. These options were selected because they can be implemented on a minimal budget and did not require extensive investment of human resources from the university's computer department. The students are made aware of all three of these options and were asked to provide feedback on their preferred slide viewer (see Supplement 1) . The evaluations of CaseViewer and ImageScope are part of an ongoing process at our institution among the faculty involved in the Histology and Pathology curricula, as we were planning on incorporating digital slides into both of these courses just before the pandemic began.

Other than the two options provided by BGU (VPN-restricted remote desktop or mapping the server as a network drive), other options exist for students to access slides. Paid options that are significantly more user-friendly include software suites from various companies (eg Philips Educational Suite with Preparing anonymized WSI can technically be accomplished in one of two ways. Within the 3DHistech software environment, which is the one employed by BGU, the slide label is stored separately from the biopsy itself and can easily be deleted using Slide Converter 2.3 (3DHistech, Budapest). Further detailed information about the two methods we have used for anonymizing slides, as well as the benefits and limitations of different WSI formats (Mirax, iSyntax, and OME-TIFF) specifically within the context of undergraduate medical education are discussed in Supplement 2. For each module, a number of items on a differential diagnosis were developed ( Table 3 ). The students would first be required to independently prepare their understanding of the differential diagnosis, which variably including diagnostic criteria, histological findings, immunostains, molecular pathways, treatment, and prognostic factors. Required and optional reading selections were freely accessible to

The "frontal" lecture was given over Zoom and was PowerPoint based, lasting 45 minutes. However, attempts were made to keep this interactive. Each disease was introduced with a skeleton slide with the most important headlines listed (clinical aspects, histological findings, molecular pathology, see Figure   1 ), and students were encouraged to tell what information should be completed under the headline, with some success (see Results, Section 3). The Zoom "survey" feature was also used. For students who could not attend the live session, a recording of the lecture was available via the Zoom cloud.

After the lecture, students were given a quiz, which usually consisted of 5 questions based on digital slides. In several modules, the questions were a simply multiple choice of "what is the most likely diagnosis?" In the breast module, students were also asked to interpret receptor stains.

In the final part of each module, each slide was reviewed over Zoom, with focus on the quiz questions but also exploring other important features that would be mentioned in a pathology report. This session was likewise recorded and available for later playback.

During one of the last modules, neuropathology, as part of the course evaluation, pre-test was administered in which students were shown a tumor and asked to classify it as either astrocytic or oligodendroglial. The pre-test was given before the frontal Zoom lecture and was composed of 15 questions, each with a single image taken from one of three neuropathology atlases published by Elsevier/Saunders (Ellison's Neuropathology, 3rd ed; Prayson's Neuropathology, 2nd ed; or Yachnis' High Yield Pathology: Neuropathology). As this module was one of the last modules, the students had already reviewed a significant number of slides under instruction and could appreciate concepts of monomorphism vs pleomorphism and hyperchromasia.

The overall module consisted of 8 sessions spanning 2 weeks. The beginning of each live session represented the fourth part of the 4-part module, and then we transitioned into the first part of the next module. The first session was dedicated to developing a general appreciation for important general concepts in inflammatory and neoplastic diseases, and an additional 2 sessions (one at the half-way point and the final session) were dedicated to more advanced topics of interest, using the main curriculum as a starting point (eg NIFTP was discussed in a session after the thyroid neoplasia module).

The course was graded pass-fail based on evidence of serious student participation, which included the quizzes and the live lectures. Following the course, students were asked to complete an anonymous questionnaire (see Table 4 ).

Informed consent was obtained in accordance with study approval by the Ethics Review Board for the Faculty of Health Sciences, Ben Gurion University of the Negev (No. 26-2020).

A full listing of the survey questions is provided in Table 4 ; select results are presented here. The raw data may be downloaded as Supplement 1. Of the Fifty-nine students who were enrolled in the course, Overall, the course was rated very favorably by the students: 68% (n=17) gave at least 3 points on a 4point scale to questions relating to how interesting the course was, how much the course improved their understanding of the covered diseases, and how strongly they would recommend a student take this course if given an option. Further, 80% (n=20) indicated each that the course reinforced material covered in Year 2 and agreed or strongly agreed that the selective was a positive learning experience.

The format of the course (specifically the 4-step cycle explained in the Methods and Materials, Section 2.2) was similarly rated favorably (see the comments in Supplement 1).

The online format seems to be preferable to students, with 67% (n=16) stating that they are more encouraged to attend the live session if it is online. The top two advantages indicated by the students were the ability to attend from anywhere (68%, n=17) or having a recording available for later review (32%, n=8).

The single worst disadvantage of the class, as reported by 80% of respondents (n=20) were technical challenges in accessing the slides. In fact, due to this problem, 40% of respondents (n=10) used screenshots taken by other students for part or all of the selective. Indeed, when planning out the selective, creating a user-friendly environment within budget was one of our primary concerns. The seven students who used CaseViewer gave more favorable comments than the five who used J o u r n a l P r e -p r o o f ImageScope or three who used QuPath (see Supplement 1) . Negative reviews mainly focused on the technically difficult setup and slow connection speeds, especially related to the VPN and remote desktop setup, although the ImageScope interface was judged unfavorably by one student. One student recommended the instructors offer a tutorial on using the slide readers. The authors feel that a future offering along this format should either be accompanied by significant technical support from the school before the course begins in order to assure that everyone can access the slides, or otherwise instituting one of the other slide access methods discussed in the Materials and Methods section and Supplement 2.

Although M1 and M2 students are exposed to pathology in their education as part of either a dedicated introductory course (M1) or systems courses (M2), these experiences are limited to knowledge transfer and acquisition as occurs in many preclinical settings. Traditionally, very few students in their final two years have any exposure to diagnostic pathology as practiced by anatomic pathologists (Survey Results, Section 3). As such, graduates have minimal familiarity with the relevance of surgical pathology in clinical practice, as well as the significance of the material that they learned in their preclinical years.

The pandemic crisis that forced exclusion of medical students from the in-person patient healthcare setting provided an opportunity to expose the students to material they largely would have remained ignorant of, despite its great use in treating patients.

Overall, we believe that this selective was an excellent addition to the students' education for both technical and educational reasons. Our new slide viewing interface enabled students to truly interact with the slides as a pathologist would, including exploring the slide and zooming into interesting areas.

An additional significant benefit was the ability to add IHC to the course, which had not been done until now in our regular student pathology lab exercises. The addition of IHC added a significant dimension to the students' experience, reinforcing the mechanistic understanding of neoplasia (eg c-myc

The format of the present course may enrich the students' education on numerous fronts. Teaching the students with actual diagnostic pathology slides (H&E and IHC) imparted the difference between telling them about depth of invasion, nuclear grade, and the expression of clinically relevant proteins, versus showing them. At least for some students, visual perceptual learning is a valuable tool for developing expertise in general. [10] As such, certainly the skill of understanding a pathology report should be nurtured in all undergraduate students. Beyond this, basic exposure to some of the profession's specific skills -which is generally lacking in undergraduate medical education -may encourage a student to give a moment's thought to perusing a career in pathology, in the same manner that these students are rightfully taught to grade autonomic reflexes, evaluate the optic fundus, and identify acetowhite epithelium, even though most will not employ such specialized skills in the future. In the neuropathology pre-test described in Materials & Methods, students were shown a tumor and asked to classify it as either astrocytic or oligodendroglial. On this pre-test, which comprised of 15 H&E images, 28 students (70%) scored at least 13/15 correct (86%), and 7 students (17%) got 15/15 correct.

Although we have no control group, we believe that these scores demonstrate some success of the course and its contribution to students' appreciation of pathological findings.

Even within the distance learning construct, there were several limitations to the development of the course. Due to the need for fast curriculum development, case selection was based in largely on availability. The course coordinator and chief teaching assistant for pathology (BS and NS, respectively) were developing some digital labs for medical students and already had some cases selected for this purpose. Additional cases were chosen based on the ease of finding high-quality slides with classic features of the disease in question, and that the instructor had a PowerPoint presentation that could be easily adapted to the new format. Within these restrictions, about 1 hour was invested for curriculum planning of each module (meaning, selecting a differential diagnosis and adapting a PowerPoint presentation for this purpose), and variable time was required for preparing digital slides (depending on J o u r n a l P r e -p r o o f if slides had been pre-selected for digital labs and the ease of findings archival slides that met our quality criteria). Collectively, the course sampled from the most common diagnosis in the core of general pathology (surface epithelial neoplasia in dermatopathology, breast neoplasia) in addition to more specialized topics (neuropathology, thyroid pathology). The otherwise peculiar combination was appropriate to maintain a high level of interest in order to teach key concepts at an undergraduate level rather than the most important skills for an entry-level resident, in which case more emphasis may be placed on the common "general pathology" fields.

Another time-related restriction concerned how much time students were asked to dedicate to the course. On the one hand, this was obviously a serious educational framework which demanded appropriate commitment and investment from the students, both for the "refined" sake of their medical education and to ensure that the course met regulatory requirements for academic credit. On the other hand, we understood that many new uncertainties and expectations suddenly arose regarding fourth

year rotations (for current M3 students) and residency (for current M4 students). In light of this, we made a conscientious decision to limit the time within the module so that students could secure their academic and professional frameworks for the coming year. Very unfortunately, this, in turn, limited the investment we could put into interaction with the students, and can hopefully be rectified in future offerings.

The coronavirus pandemic was international disrupter of education at all levels. It also threatened to disrupt the academic timeline of medical students, specifically those in the third and fourth years. We viewed this crisis as an opportunity to improve our online clinical rotations and specifically to attempt to enhance our students' knowledge of material that they heretofore were not exposed. Using freely available public software, professor led problem based learning and standard texts we were able to create a virtual curriculum simulating a two-week diagnostic pathology selective. Based on subjective responses by students as well as objective assessments of knowledge not previously taught, we feel that the course was a success and can be a model for similar virtual pathology electives in the future, both in times of crises such as this and even during routine medical school education. The most significant improvement to the course that should be implemented in future iterations is providing greater technical support to the students before the beginning and/or use of a friendlier interface. The course 

Format Before this course, did you have any exposure to diagnostic pathology from the pathologist perspective? (eg rotations or electives in pathology) (If you answered yes please elaborate.)

Before this course, did you have any exposure to diagnostic pathology from the clinician perspective? (eg tumor board at which a pathologist presented slides, department pathology meetings) (If you answered yes please elaborate.)

On a scale of 1 (boring) to 4 (fascinating), how did you find the selective? Scale 1-4 On a scale of 1 (not at all) to 4 (very much), how much did the selective improve your understanding of the diseases covered? 

WHO Director-General's opening remarks at the media briefing on COVID-19 -11

No more daycare, restaurants, gyms or shopping malls: The new virus regulations

Pathology education in a multisite urban/rural distributed curriculum

Regional solutions to the physician workforce shortage: The WWAMI experience

Virtual microscopy and digital pathology in training and education

Leveraging technology for remote learning in the era of COVID-19 and social distancing: tips and resources for pathology educators and trainees

The virtual microscopy database-sharing digital microscope images for research and education

Philips Pathology SDK now available to software developers and research scientists

Visual Perceptual Learning and Models