To the Editor.—It has previously been shown that the use of a checklist such as the one used in synoptic reporting can lead to a more complete surgical pathology report,1–3 and that different methods of generating a synoptic report can also affect the completeness of that report.4–6 The presence or absence of clinically significant data elements may impact the clinical management of a patient. Formatting of the synoptic report can also impact pathologists' ability to create a synoptic report and a clinician's ability to interpret that report.4,5,7–15 To date, the impact of the choices that are presented to the pathologists as responses to any data element on the data in the report or the management of the patient has not been studied. In the Word-based synoptic reports offered by the College of American Pathologists (CAP) and its electronic Cancer Checklist, a discrete set of options is presented with limited use of qualifying terms.8 However, previous studies suggest that many pathologists prefer to include qualifying terms (ie, “suspicious for,” “extensive,” etc) and they may use free-text entry to achieve this result when the choice is not offered to them.8 To assess whether presenting qualified responses in a synoptic report has an impact on the report content or patient care, we compared the response for the data element “lymphovascular invasion” (LVI) in our laboratory from 240 consecutive cases (60 each from endometrium, colon, lung, and breast) using the CAP Word-based synoptic reports system (“no qualified choices,” between January 1 and June 30, 2015) with a Web-based method that did include qualifying terms among the selection choices (between August 1, 2016, and December 31, 2019).14
Results are presented in the Table. Free-text options were used less than 2% of the time for both groups. Both “suspicious for LVI” and “extensive LVI” were entered as free-text options (once each) in the reports with no qualified options presented. Although there was no significant difference in the rate of LVI when cases in which LVI was either present or extensive (22% versus 26%; P = .23), there were significantly more cases identified as having LVI when cases diagnosed as suspicious for LVI were included and presented as an option (23% versus 34%; P = .01). There were also significantly fewer cases diagnosed as indeterminate when qualified responses were offered (17% versus 6%; P = .001, 2-tailed χ2 test).
During this time period, whenever an endometrial carcinoma case was presented at Gynecologic Tumor Board in which the use of adjuvant therapy was not clear, the presence or absence of LVI was always discussed. Patients whose LVI status was reported as indeterminate were considered to have the same risk as patients whose LVI status was reported as not identified. However, patients who were labeled as suspicious for LVI were identified as at higher risk of recurrence than patients without LVI or those who were classified as indeterminate, and patients with extensive LVI were identified as having a higher risk of recurrence than patients who were labeled as LVI without any qualifier. In one case, the clinicians decided that they would treat the patient only if the LVI were extensive, and it was asked that the case be rereviewed to assess this. The findings were similar in the tumor boards for other sites, but these results were not specifically tracked.
There are limitations to this study. We compared consecutive series of cases and did not match them for grade, tumor size, or stage. However, there was no difference in the overall rate of cases with LVI when cases diagnosed as suspicious for LVI were excluded, suggesting that the cases were similar, and that the differences we observed are related to pathologists simply choosing “indeterminate” when “suspicious for LVI” was not offered as a choice. In addition, we compared the performance of the same pathologists at different times, and their performance may have changed over time.
In summary, we have shown that not only does the inclusion of qualified responses for LVI in synoptic reports result in increased reporting of LVI and decreased reporting of indeterminate cases, but reporting of LVI directly impacts patient care in a subset of patients with synoptic reports. Clinicians routinely include their own assessment of the significance of these qualifying terms in their recommendations for patient management. Further study to ensure that the response options that are presented in the CAP synoptic report protocols are evidence based seems warranted.
The authors have no relevant financial interest in the products or companies described in this article.