Vascular imaging plays an important role in baseline assessment and monitoring disease progression in large-vessel vasculitis (LVV).1 However, interpretation and documentation of vessel wall involvement can be challenging due to a lack of standardized radiology reporting frameworks and unfamiliarity with vascular anatomy among rheumatologists.2,3 Currently, the workflow at our institution involves manually copying and pasting vascular imaging reports into the patient chart at each clinical encounter. The treating physician at times may have to compare individual reports, which are often verbose and inconsistent in reporting format, to make an accurate assessment of change in vessel wall involvement over time. Therefore, the main objectives of this project were to (1) outline the current process of interpreting vessel wall imaging reports for clinical use in LVV; (2) understand the barriers to accurately tracking vessel involvement over time; and (3) develop a tool to improve the documentation of vessel wall imaging and tracking changes over time. To our knowledge, there are currently no such tools being used in clinical practice.
An exploratory stakeholder analysis was conducted in the fall of 2023 at Mount Sinai Hospital’s Vasculitis Clinic, the largest tertiary care vasculitis clinic in Canada that assesses over 3000 patients annually. We disseminated a standardized survey to 2 rheumatology attending physicians, 3 vasculitis fellows, and 8 rheumatology residents, rotating or having previously rotated through the vasculitis clinic, to characterize the clinicians’ experience with the current workflow. Specifically, the survey aimed to identify challenges (if any) associated with visualizing vessel involvement over time using the current workflow and solicited potential stakeholder-led solutions to improve documentation of vessel involvement in LVV.
We obtained a 100% response rate in our survey. In total, 9/13 (69%) respondents found it challenging to clearly visualize vessel involvement from reading radiology reports at initial consultation and 10/13 (77%) found it difficult to track vessel involvement over time at subsequent follow-up visits. Moreover, 11/13 (85%) respondents found the current process of comparing vessel involvement between radiology reports time consuming. Free-text responses identified reasons why respondents found this challenging, including lack of standardized reporting frameworks (endorsed by 5 trainees), unfamiliarity with vessel anatomy (endorsed by 4 trainees), inconsistent terminology used by radiologists, and lengthy computed tomography or magnetic resonance angiography reports. Last, 9/13 (69%) respondents indicated that a diagram where clinicians could label vessels would help improve visualization and tracking of vessels over time.
Expanding upon themes identified in the survey, a fishbone cause and effect diagram was developed to highlight components of the current workflow that contributed to difficulty in tracking LVV vessel involvement (Figure 1). Of the various contributors identified, the lack of a standard documentation tool was felt to be the most important to address. This is because such a tool could address and/or improve upon reasons highlighted in the survey for why respondents found it challenging to visualize and track vessel involvement in LVV, through standardizing documentation processes, educating providers on vessel wall anatomy, and succinctly summarizing vessel wall involvement. After discussion with stakeholders at the vasculitis clinic including attending physicians, trainees, and administrative staff, the creation of such a documentation tool was also deemed to be the most feasible solution to implement within our clinic, and one that we had the most control over. This led to our initial plan-do-study-act (PDSA) cycle, which involved the creation of a novel standardized labeled visual tool to monitor LVV disease involvement.4 The tool consisted of the aorta and major large arteries digitally illustrated by a medical student (ELVN) in consultation with a rheumatology attending physician (MS) to ensure clinical accuracy. This tool allowed clinicians to annotate vessel involvement at each clinical encounter, facilitating easier comparison between visits.
Next, we assessed our tool’s uptake and implementation by performing a random chart audit of 11 patients with LVV, 1 month after the tool’s implementation at the vasculitis clinic. Our primary outcome measure was the percentage of patient charts where the visual tool was completed by the physician who assessed the patient at the most recent clinical encounter with accompanying serial imaging following tool implementation; the pertinent process measures were the percentage of patient charts where the tool was included but not completed, and the percentage of charts where the tool was partially completed. Based on our initial chart audit, the tool was found to be included in 9/11 (82%) charts, and fully completed in 7/11 (64%) charts. In discussion with stakeholders, challenges with using the visual tool included small font and inconsistent labeling due to the lack of a standardized labeling scheme. For example, vessel thickening was labeled differently by various clinicians. Additionally, there was an absence of dates corresponding to vessel changes.
As part of a second PDSA cycle, the visual tool was revised to address these issues. The updated visual tool included labels with a larger font, and a legend with standardized symbols to capture active inflammation in the form of vessel wall thickening, damage including stenosis/occlusions and dilatation/aneurysms, and prior surgical interventions including vessel bypass and stenting (Figure 2). A second random chart audit of 5 charts suggested that the modified tool demonstrated clarity, organization, comprehensibility, and accuracy as assessed by comparing those with original radiology reports. We further disseminated a second survey to the same group of rheumatology attending physicians, vasculitis fellows, and rheumatology residents rotating through the vasculitis clinic to characterize the clinicians’ experience of the modified tool. Of the 5 respondents (1 attending physician, 3 vasculitis fellows, 1 rheumatology resident), 4/5 (80%) found the tool to be useful in visualizing vessel involvement for initial consultations, and 5/5 (100%) found it useful for monitoring disease progression for follow-up visits. In terms of balancing measures, 4/5 (80%) of respondents felt that the tool did not affect clinic efficiency; however, our study did not collect objective data (ie, time spent per patient encounter pre- and post intervention) to confirm this. Free-text responses from clinicians who used the modified tool also highlighted that the tool was easy to use and served to improve their own knowledge of vascular anatomy (endorsed by 1 attending physician and 3 trainees). Free-text responses also alluded to the potential use of the visual tool in educating patients on their disease.
Limitations of our approach include not having a radiologist stakeholder involved in tool development. Further, we did not collect data on the time it took clinicians to complete the tool or other associated interruptions to clinic workflow. Future steps include further validating the accuracy of the modified tool with construction of receiver-operating characteristic curves compared to gold standard radiologist interpretations, improving uptake and dissemination with stakeholder engagement through further PDSA cycles, capturing the end user as well as patient experience, and assessing its effect on clinic efficiency.
In conclusion, the accurate documentation of vessel wall involvement is critical in quality LVV care as it allows the treating rheumatologist to monitor disease progression, which may at times be asymptomatic, so that therapy can be appropriately tailored.1,5 We hereby describe the creation of a visual tool to illustrate vessel wall involvement in LVV, allowing for ease of comparison between serial imaging reports (Figure 2), and report preliminary data of its success after 2 iterative PDSA cycles at a large quaternary vasculitis clinic. We intend to further explore the tool’s role as a patient education vehicle by performing future qualitative studies with patients with LVV. Moreover, we aim to better delineate balancing measures such as potential disruptions to clinic flow with the time required to document vessel involvement using this tool. We anticipate that the generalizability of this tool will allow for its adaptation for use in other clinics where patients with LVV are treated, thereby improving the quality of care delivered to this patient population on a larger scale.
ACKNOWLEDGMENT
The authors would like to thank Dr. Christian Pagnoux and the vasculitis fellows at Mount Sinai Hospital for their clinical content input in the development of this tool.
Footnotes
T.S.H. Kwok and A.L. Zhou contributed equally to this manuscript.
The authors declare no conflicts of interest relevant to this article. Research ethics board (REB) approval was obtained (23-0126-E) from the Mount Sinai Hospital REB.
- Copyright © 2024 by the Journal of Rheumatology