Context.—

Pleural effusion cytology has been widely used in the investigation of pathologic fluid accumulation in pleural spaces. However, up to one-tenth of the cases were not given a definitive diagnosis. These cases have largely been neglected in the bulk of the literature.

Objective.—

To provide real-world data on indefinite diagnoses including “atypia of uncertain significance” (AUS) and “suspicious for malignancy” (SFM) in pleural effusion cytology and to investigate pathologists’ practice patterns on using these diagnostic categories.

Design.—

We reported the diagnoses of 51 675 cases. Descriptive statistics and correlation coefficients were used to analyze the relationships between different diagnostic categories and pathologists’ practice patterns and possible explanatory variables.

Results.—

The diagnoses AUS and SFM were reported in 4060 cases (7.86%) and 1554 cases (3.01%) in the cohort, respectively. The mean rates for these indefinite diagnoses varied up to 3-fold between pathologists. Correlations were found between AUS and SFM, as well as between indefinite diagnoses and negative for malignancy (NFM). No correlations were found between pathologists’ years of experience or case volume and the rates of indefinite diagnosis or diagnostic certainty.

Conclusions.—

A real-world baseline for the rates of indefinite diagnoses in pleural effusion cytology is provided in this large retrospective study. Pathologists show significant variation in their use of indefinite diagnostic categories, and the tendency to use these ambiguous terms was not correlated with individuals’ experience or case volume. How to untangle the intertwined relationship between the uncertainty of indefinite diagnoses and that of NFM requires future prospective studies.

Pleural effusion cytology has been widely used in the analysis of pathologic fluid accumulation in pleural spaces. Malignancy is one of the many underlying causes of pleural effusion, and its identification is crucial for patient triage and management. However, the accuracy of pleural effusion cytology varies widely, with sensitivities ranging from 43% to 71% in larger studies performed in the era without contemporary ancillary studies1  and from 23.5% to 100% in a more recent comprehensive meta-analysis.2  This significant variation can be attributed to several possible causes including preanalytic factors, morphologic overlapping between benign/reactive and malignant diagnoses, and competency of the pathologists who sign out the cases. Additionally, a definitive diagnosis cannot always be made, as 1.8% to 10.1% of the cases were given an indefinite diagnosis.1 

In pleural effusion cytology, there is a fraction of cases, even with contemporary ancillary studies, that cannot be readily classified as negative or positive for malignancy. “Atypia of uncertain significance” (AUS) or “suspicious for malignancy” (SFM) have long been used to report such indefinite cases. However, information for these cases is scarce, as most of the studies focus on negative or positive diagnostic categories and rarely report indefinite ones.3  The lack of clear diagnostic criteria and uniformity in reporting for the indefinite diagnostic categories may also be a source of confusion among both pathologists and clinicians.2 

Cases with indefinite diagnoses are hard to compare interinstitutionally because of different patient populations and different clinical practice patterns. Intra-institutional performance variation for these cases, which is not subject to the above factors, has also never been carefully investigated. Despite the recommendations from the workforce of the International System for Reporting Serous Fluid Cytopathology (TIS),3  intra-institutional monitoring of indefinite diagnosis rates has not yet been reported in the literature. This may be due to a variety of reasons. Some institutions might only have 1 expert cytopathologist responsible for all case sign-outs; therefore, interobserver agreement is not an issue. Some other institutions might have triaged the specimens before assigning them to expert cytopathologists, specialist pathologists, or general pathologists, and therefore cannot guarantee the homogeneity of cases handled by different personnel, thus making comparisons impossible.

At our institution, without access to expert cytopathologists, pleural effusion cytology cases are randomly assigned to practicing pathologists to sign out. This provides a unique opportunity to observe intra-institutional, interobserver performance variation. In this study, we reviewed the reports of pleural effusion cytology during a 20-year period in a tertiary medical center, focusing on cases with indefinite diagnoses including AUS or SFM to investigate the different practice patterns of pathologists.

Patient Cohort

The study was approved by the institutional review board of Chang Gung Medical Foundation, Taoyuan, Taiwan (IRB No. 202201814B0). Pleural effusion cytology reports during a 20-year period from 2002 to 2021 were retrieved from the archive of the Department of Anatomic Pathology at Linkou Chang Gung Memorial Hospital, Taoyuan, Taiwan. The hospital is the largest tertiary medical center and academic hospital in Taiwan, with 4000 beds, and the Department of Anatomic Pathology has annual case volumes approaching 113 000 in surgical pathology and 18 000 in nongynecologic cytology, with 16 full-time practicing pathologists specializing in a wide range of subspecialties from medical renal pathology to hematopathology. There are no expert cytopathologists in our hospital, and the pleural effusion cytology cases are randomly assigned to practicing pathologists without triage. To further ensure that the data included in the cohort were representative, we checked the number of cases issued by each pathologist in each year. Only pathologists who had at least 10 years of practice during the 20-year study period and who issued no fewer than 100 reports each year (n = 11) were included for the analyses of characteristics of the cohort, characteristics of indefinite diagnoses, and relationships between diagnostic categories. For the correlations between indefinite diagnoses and pathologists’ experience, all data during the study period were included for comprehensiveness (n = 16). During the study period, cytologic diagnoses were reported in a 4-category reporting system: negative for malignancy (NFM), AUS, SFM, and positive for malignancy (MAL). Our institution did not have explicitly defined criteria for a diagnosis of AUS. In general, this diagnosis is reserved for cases showing cytomorphologic changes that cannot be readily interpreted as benign/reactive/reparative but are not severe enough to be assigned a diagnosis of suspicious or MAL. A diagnosis of SFM was made when the changes were indicative of malignancy, but available evidence fell short of rendering a definite diagnosis. Selected cases with indefinite diagnoses from January 2022 to November 2022 were blindly reviewed by a pathologist who was not one of the 11 pathologists included in the cohort.

Statistical Analysis

The χ2 homogeneity test was used to test reporting homogeneity in each diagnostic category. One-way analysis of variance (ANOVA) and unpaired 2-sample t test were used to test whether there were differences between group means among pathologists in reporting each diagnostic category. The Grubb test and 1.5× interquartile range (1.5× IQR) method were used to identify and remove outliers. The Kolmogorov-Smirnov test confirms that the data for Grubb tests do not differ significantly from data that are normally distributed. Fisher exact test was used to compare 2 proportions. Pearson correlation coefficient was used to examine correlations between continuous variables. A prespecified significance level of .05 was used in all statistical hypothesis tests.

Basic Characteristics of the Cohort

During the 20-year period from 2002 to 2021, a total of 68 141 pleural effusion cytology reports were issued. The final cohort included 51 675 reports issued by 11 pathologists (Supplemental Table 1; see the supplemental digital content containing 8 tables at https://meridian.allenpress.com/aplm in the August 2024 table of contents). Within this period, the years of practice attributed to the pathologists included in the cohort ranged from 11 to 20 (mean, 15.9 years; median, 16.0 years). The annual number of reports issued by a single pathologist ranged from 115 to 824 (mean ± SD, 295 ± 105 reports per person-year), and the total numbers of reports issued by a single pathologist during the 20-year period ranged from 2326 to 7301 (mean ± SD, 4698 ± 1323 reports). The number of cytology diagnoses for the 51 675 reports included in the cohort was as follows: 39 477 (76.39%), 4060 (7.86%), 1554 (3.01%), and 6584 (12.74%) for NFM, AUS, SFM, and MAL, respectively. The mean prevalence of MAL each year was 13.03% ± 3.06% (mean ± SD) per year.

Characteristics of Indefinite Diagnoses (AUS, SFM, and Combined AUS+SFM)

Table 1 summarizes the percentages and variation in performance in pleural effusion cytology among the 11 pathologists, focusing on indefinite diagnoses. Supplemental Tables 2 through 4 provide more detailed data. The χ2 homogeneity test showed that the proportions of AUS, SFM, and combined AUS+SFM diagnoses reported by pathologists were inhomogeneous (all P < .001). One-way ANOVA confirmed differences in group means for the rates of diagnoses signed out as AUS, SFM, and combined AUS+SFM by pathologists (all P < .001). The mean rate (±SD) of AUS in pleural effusion cytology was 8.14% ± 2.72% and showed a near 3-fold variation (4.96%–14.41%). The mean rate (±SD) of SFM was 3.20% ± 1.16% and showed a 3.5-fold variation (1.54%–5.39%). The mean rate (±SD) of combined AUS+SFM was 11.34% ± 3.71% and showed a near 3-fold variation (7.00%–19.80%) (Figure 1, A and B; and Figure 2). No outliers were found by Grubb test and 1.5× IQR method for all the diagnostic categories. However, the group means significantly differed between the lowest reported rate, the middle rate, and the highest rate in AUS, SFM, and combined AUS+SFM. Upon review of the cases with indefinite diagnoses signed out by the pathologist who issued the largest number of these cases, more cases could have been reclassified to a definite diagnosis of NFM or MAL as compared to those of a pathologist who issued a medium number of indefinite diagnoses (31 of 48 versus 13 of 32, P = .04 by Fisher exact test). The mean rates for indefinite diagnoses among pathologists (interpathologist) had similar statistical dispersions (coefficient of variation = 33.41% for AUS, 36.40% for SFM, and 32.68 for combined AUS+SFM). In contrast, MAL category had a much lower coefficient of variation of 23.97%.

Figure 1.

Variation in performance in pleural effusion cytology. A, Variation in atypia of uncertain significance. B, Variation in suspicious for malignancy. Data are expressed as mean ± standard deviation. The lower limit of 1.5× interquartile range (IQR) is set to zero if the calculated value is less than zero because mean rates cannot be negative numbers.

Figure 1.

Variation in performance in pleural effusion cytology. A, Variation in atypia of uncertain significance. B, Variation in suspicious for malignancy. Data are expressed as mean ± standard deviation. The lower limit of 1.5× interquartile range (IQR) is set to zero if the calculated value is less than zero because mean rates cannot be negative numbers.

Close modal
Figure 2.

Variation in performance in pleural effusion cytology for combined atypia of uncertain significance and suspicious for malignancy. Data are expressed as mean ± standard deviation.

Figure 2.

Variation in performance in pleural effusion cytology for combined atypia of uncertain significance and suspicious for malignancy. Data are expressed as mean ± standard deviation.

Close modal
Table 1.

Variation in Performance in Pleural Effusion Cytology Among 11 Pathologists

Variation in Performance in Pleural Effusion Cytology Among 11 Pathologists
Variation in Performance in Pleural Effusion Cytology Among 11 Pathologists

Relationships Between Diagnostic Categories

There was a significant fair positive correlation between the rates of AUS and SFM (n = 175, R = 0.404, P < .001). Pearson correlation coefficient further confirmed that there were significant fair to moderate negative correlations between the rates of AUS (R = −0.532, P < .001), SFM (R = −0.487, P < .001), and AUS+SFM (R = −0.606, P < .001) and the rates of NFM. The relationships between the rates of AUS, SFM, and AUS+SFM and the rates of MAL were either poor or insignificant (Table 2). A greater number of critically reviewed cases, as mentioned in the previous paragraph, were reclassified into the NFM than the MAL category (67.7% [21 of 31] versus 32.3% [10 of 31], P = .01 by Fisher exact test). There were no correlations between the dispersions of indefinite diagnoses and the dispersions of MAL among individual pathologists. On the other hand, there were significant moderate positive correlations between the dispersions of indefinite diagnoses and the dispersions of NFM among individual pathologists (Table 3).

Table 2.

Correlations Between the Rates of Diagnoses

Correlations Between the Rates of Diagnoses
Correlations Between the Rates of Diagnoses
Table 3.

Correlations Between the Dispersions of Diagnoses

Correlations Between the Dispersions of Diagnoses
Correlations Between the Dispersions of Diagnoses

Indefinite Diagnoses Were Not Correlated With Pathologists’ Experience

Figure 3, A through C, and Supplemental Table 5 summarize the correlations between all (n = 16) pathologists’ average experience in the 20-year study period and their performance. There were no correlations between the years of experience and the rates of AUS, SFM, or combined AUS+SFM diagnoses signed out by the pathologists (P = .45, .99, and .58, respectively). There were no correlations between the total or annual case number signed out and the rates of indefinite diagnoses signed out by the pathologists (P = .67 to .84 and .57 to .70, respectively). There were no correlations between increased personal experience and the rates of indefinite diagnoses signed out among 12 of the 16 pathologists (Figure 4 and Supplemental Tables 6 through 8). Four pathologists whose performance showed a significant correlation with increased experience signed out even more indefinite diagnoses with increased experience (R = 0.480–0.792, P = .006 to .04). Aligning pathologists’ performance according to their “exact” years of experience further confirmed that the rates of indefinite diagnoses did not change during the first 5 years (R = 0.091, P = .89), the first 10 years (R = 0.623, P = .054), or the whole range of the pathologists’ career (R = 0.207, P = .24). Similarly, sorting each pathologist’s performance in each year according to the number of cases they previously signed out revealed a weak, nonsignificant correlation (R = 0.070–0.087, P = .19 to .30). In general, there were no significant correlations between the years of experience, total case number, and annual case number and the variations in the rates of indefinite diagnoses signed out by the pathologists. Restricting analyses using only 11 pathologists yielded the same results (data not shown).

Figure 3.

Correlations between the years of experience of 16 pathologists and the rates and CVs for AUS (A), SFM (B), and combined AUS+SFM (C). Data for the rates are expressed as mean ± standard deviation. Abbreviations: AUS, atypia of uncertain significance; C.V., coefficient of variation; SFM, suspicious for malignancy.

Figure 3.

Correlations between the years of experience of 16 pathologists and the rates and CVs for AUS (A), SFM (B), and combined AUS+SFM (C). Data for the rates are expressed as mean ± standard deviation. Abbreviations: AUS, atypia of uncertain significance; C.V., coefficient of variation; SFM, suspicious for malignancy.

Close modal
Figure 4.

Correlations between increased personal experience and the rates of indefinite diagnoses among 16 pathologists (combined AUS and SFM). *D, *J, *L, and *C represent 4 individual pathologists with significant correlation. Abbreviations: AUS, atypia of uncertain significance; R, Pearson correlation coefficient; SFM, suspicious for malignancy.

Figure 4.

Correlations between increased personal experience and the rates of indefinite diagnoses among 16 pathologists (combined AUS and SFM). *D, *J, *L, and *C represent 4 individual pathologists with significant correlation. Abbreviations: AUS, atypia of uncertain significance; R, Pearson correlation coefficient; SFM, suspicious for malignancy.

Close modal

In this first large retrospective study of indefinite diagnoses in pleural effusion cytology, we made several intriguing findings. First, the rates of indefinite diagnoses signed out by individual pathologists vary widely, up to 3-fold, even within a single institution. Second, many cases with an indefinite diagnosis made by pathologists who tend to sign out more indefinite diagnoses can be reclassified to a definite diagnosis on review. Third, both between pathologists and among individual pathologists, there is more variation in the rates of indefinite diagnoses than for definite diagnoses. Fourth, an intermingled relationship exists between the rates of indefinite diagnoses and that of NFM, but not between that of indefinite diagnoses and a definite diagnosis of malignancy (MAL). Fifth, there were no significant correlations between pathologists’ years of experience, total case number, and annual case number and either the rates of indefinite diagnoses they signed out or the variations in the rates of indefinite diagnoses.

During the 20-year study period, more than 50 000 cases of pleural effusion cytology were signed out by 11 pathologists in our institution. Given this large number, the unique situation with no expert cytopathologists on staff, and the randomly assigned cases, we had the very rare opportunity to investigate pathologists’ practice patterns in a nonexperimental, nonreclassifying, real-world setting. Literature evaluating indefinite diagnoses in pleural effusion cytology is sparse, precluding any prior formal analyses. All literature after the introduction of TIS focuses on reclassification, and therefore cannot give us information about how pathologists perform in their daily practice. We made no attempt to review all the cases to get “reclassified” diagnoses and believe this approach has the strength to reveal the true practice patterns of pathologists in the real world.

In a recent meta-analysis of 73 studies, only 0.6% and 3.5% of pleural effusions were reported as AUS and SFM, respectively. After the implementation of TIS, 4.3% (0.6%–7.8%) and 6.2% (1.2%–18.7%) were reported as AUS and SFM in retrospective reclassification studies (Table 4).4–15  Our study shows a slightly higher rate of AUS of 7.86%, just above the upper limit but still reasonably within the current working benchmark (5%–10%),3  and we found a rate of SFM (3.0%) that is well within the reported range. Most published large series, however, do not emphasize cases with indefinite diagnoses.3  Therefore, it is difficult to determine the true rates (either before or after reclassifying) of atypical or suspicious cells in pleural effusions before the introduction of TIS. Unfortunately, all studies after the introduction of TIS were retrospective and reclassifying in nature. They can be considered only “optimized” performance rather than representative of what pathologists actually do in routine sign-out. This study, for the first time, provides valuable real-world evidence of intra-institutional performance variation. Our laboratory is planning to introduce TIS, and it will be interesting to observe the changes in diagnosis rates after the transition.

Table 4.

Studies After Implementation of the International System for Reporting Serous Fluid Cytopathology and Their Reported Rates of Diagnoses

Studies After Implementation of the International System for Reporting Serous Fluid Cytopathology and Their Reported Rates of Diagnoses
Studies After Implementation of the International System for Reporting Serous Fluid Cytopathology and Their Reported Rates of Diagnoses

We found large variation in the rates of indefinite diagnoses between pathologists, with a smaller rate of variation with MAL. We speculate this implies that the diagnostic criteria applying to MAL are stricter and more agreed upon among pathologists. Pathologists are exacting in making a definite diagnosis of malignancy. However, this rigorous attitude seems somewhat relaxed when making a distinction between indefinite diagnoses and NFM. It is interesting to observe the intermingled relationship between indefinite diagnoses and NFM. It is not surprising to see that the variation in the rates of indefinite diagnostic categories correlate well with each other among individual pathologists, but it is somewhat unexpected that this phenomenon would extend to the variation in the rates of NFM, but not to the MAL category. This again implies that pathologists draw a clear line between indefinite diagnoses and a definite diagnosis of malignancy, but the line between indefinite diagnoses and NFM is much blurrier. We believe one of the major sources of variability is the ambiguity and lack of strict definition for AUS and SFM. Without a clear definition, AUS and SFM serve more like wastebaskets rather than specific diagnostic categories in daily practice. However, owing to the lack of direct comparisons between the original and the reclassified diagnoses in most studies, whether implementation of TIS can play a role in avoiding overuse of indefinite diagnoses in a daily, real-world setting is currently unknown and requires further prospective studies.

Another interesting but surprising finding is that we were unable to detect any clear correlation between pathologists’ experience and their confidence level in rendering a diagnosis. Neither years of experience nor total volume and annual volume of cases during the study period significantly correlated with the pathologists’ indefinite diagnosis rates and diagnostic certainty within indefinite diagnostic categories. In addition, during the 20-year study period, the indefinite diagnosis rates were constant for 12 of 16 pathologists, and for those whose performance changed with increased experience, they signed out even more, rather than fewer, indefinite diagnoses. According to the above findings, plus the positive correlation between the rates of AUS and SFM diagnoses signed out by individual pathologists, it seems that a pathologist’s willingness to give a definite or indefinite diagnosis depends more on personal characteristics than on their experience. This kind of “nonrandom response” has been documented in other fields of pathology.16  The finding that experience is likely noncontributory is consistent with previous studies in pancreatic lesion fine-needle aspiration,17  thyroid fine-needle aspiration,18  and uterine cervical cytology.19 

The strength of this study is inclusion of a large number of cases and a large number of board-certified pathologists working at the same institution. All pathologists’ cases were randomly assigned. Therefore, we were able to minimize any confounding effect from preanalytic factors and can more confidently attribute the differences of performance to individual pathologists’ ability and practice pattern. This strength of the study is also its limitation. We would be very interested in how expert cytopathologists perform in the real-world setting. Nevertheless, access to expert cytopathologists is not ubiquitous, and we consider that our findings represent a large part of the real-world situation. Another limitation of the study is that we were unable to investigate the follow-up outcome of all cases. Therefore, the true incidence of malignancy in pleural effusion specimens during the 20-year study period is unknown.

Although we found no significant relationship between pathologists’ experience and their confidence level in rendering a diagnosis, we did not investigate the potential effects of continuous monitoring, feedback, and education. A Dutch study showed a major discrepancy between the initial diagnoses rendered by general pathologists and those made after review by an expert cytopathologist in up to 19.4% of cases with uncertain diagnosis in pleural effusion cytology.20  Whether monitoring and feedback would help to avoid uncertain diagnoses requires further studies. We found that case review by a general pathologist can reclassify more indefinite cases into definite cases from pathologists with less confidence in their diagnoses, supporting the view that focused review can contribute to better quality of patient care. Double reading, even by general pathologists, is likely better than nothing. Our institution, as a part of continuing education, regularly participates in the College of American Pathologists Interlaboratory Comparison Program in Nongynecologic Cytopathology, which provides participants with 5 cases on a quarterly basis. However, the program is not specifically designed for pleural effusion cytology and assesses only interinstitutional performance. To the authors’ best knowledge, data on intra-institutional monitoring of the diagnosis of pleural effusion cytology, as recommended by TIS, have never been reported in the literature.3 

Our study reveals the heterogeneity in the practice pattern of pathologists in a diagnostic gray area and suggests poor interobserver agreement, but our study did not formally analyze this particular aspect. To the best of the authors’ knowledge, there is only one small study regarding interobserver agreement on pleural effusion cytology, and the agreement was indeed poor between AUS and SFM even with TIS applied.21  Clearly, more studies are required to investigate the agreement of pathologists on indefinite diagnoses.

In conclusion, this study provided a baseline of the rates of indefinite diagnoses in pleural effusion cytology signed out in the real world. Future studies should help to refine the acceptable rates. We found that pathologists show significant variation in their use of atypical and suspicious diagnostic categories in daily practice, and both the frequency and variation in the use of indefinite diagnostic categories were not correlated with pathologists’ experience or case volume. Whether the blurred line between NFM and indefinite diagnoses could be made more distinct by implementation of TIS criteria requires future prospective observational studies. We should emphasize that while literature regarding reclassification does exist, it is lacking in studies that report pathologists’ true daily practice patterns.

1.
Motherby
H,
Nadjari
B,
Friegel
P,
Kohaus
J,
Ramp
U,
Böcking
A.
Diagnostic accuracy of effusion cytology
.
Diagn Cytopathol
.
1999
;
20
(
6
):
350
357
.
2.
Farahani
SJ,
Baloch
Z.
Are we ready to develop a tiered scheme for the effusion cytology: a comprehensive review and analysis of the literature
.
Diagn Cytopathol
.
2019
;
47
(
11
):
1145
1159
.
3.
Chandra
A,
Crothers
B,
Kurtycz
D,
Schmitt
F.
The International System for Serous Fluid Cytopathology
. 1st ed.
Cham, Switzerland: Springer Nature
;
2020
.
4.
Jha
S,
Sethy
M,
Adhya
AK.
Application of the International System for Reporting Serous Fluid Cytopathology in routine reporting of pleural effusion and assessment of the risk of malignancy
.
Diagn Cytopathol
.
2021
;
49
(
10
):
1089
1098
.
5.
Lobo
C,
Costa
J,
Petronilho
S,
Monteiro
P,
Leça
L,
Schmitt
F.
Cytohistological correlation in serous effusions using the newly proposed International System for Reporting Serous Fluid Cytopathology: experience of an oncological center
.
Diagn Cytopathol
.
2021
;
49
(
5
):
596
605
.
6.
Pergaris
A,
Stefanou
D,
Keramari
P,
et al.
Application of the International System for Reporting Serous Fluid Cytopathology with cytohistological correlation and risk of malignancy assessment
.
Diagnostics (Basel)
.
2021
;
11
(
12
):
2223
.
7.
Pinto
D,
Cruz
E,
Branco
D,
et al.
Cytohistological correlation in pleural effusions based on the International System for Reporting Serous Fluid Cytopathology
.
Diagnostics (Basel)
.
2021
;
11
(
6
):
1126
.
8.
Xu
Y,
Hu
AY,
Wang
SM,
Wang
Q,
Pan
YC,
Zhang
SH.
A retrospective analysis of pleural effusion specimens based on the newly proposed International System for Reporting Serous Fluid Cytopathology
.
Diagn Cytopathol
.
2021
;
49
(
9
):
997
1007
.
9.
Ahuja
S,
Malviya
A.
Categorisation of serous effusions using the International System for Reporting Serous Fluid Cytopathology and assessment of risk of malignancy with diagnostic accuracy
.
Cytopathology
.
2022
;
33
(
2
):
176
184
.
10.
Bharti
S,
Nalwa
A,
Elhence
PA,
et al.
Risk stratification of pleural fluid cytology based on the International System for Reporting Serous Fluid Cytology in a tertiary care centre
.
Acta Cytol
.
2022
;
66
(
5
):
449
456
.
11.
Kolte
S,
Zaheer
S,
Aden
D,
Ranga
S.
Application of the international system for reporting serous fluid cytopathology on reporting various body fluids; experience of a tertiary care hospital
.
Cytojournal
.
2022
;
19
:
52
.
12.
Lim
KH,
Ahmed
SS,
Cheng
XM,
et al.
A single tertiary institution review of the international system for serous fluid cytopathology and the impact of cell block and ancillary studies on its performance
.
J Am Soc Cytopathol
.
2023
;
12
(
1
):
48
57
.
13.
Straccia
P,
Chiappetta
M,
Magnini
D,
Cancellieri
A.
Application of the International System for Reporting Serous Fluid Cytopathology (TIS): a retrospective institutional study
.
Cytopathology
.
2022
;
33
(
3
):
305
311
.
14.
Sun
T,
Wang
M,
Wang
H.
Risk of malignancy assessment of the International System for Reporting Serous Fluid Cytopathology: experience in a community hospital setting and comparison with other studies
.
Cancer Cytopathol
.
2022
;
130
(
12
):
964
973
.
15.
Zhu
YL,
Ren
WH,
Wang
Q,
Jin
HZ,
Guo
YY,
Lin
DM.
A retrospective analysis of serous effusions based on the newly proposed international system for reporting serous fluid cytopathology: a report of 3633 cases in an oncological center
.
Diagn Pathol
.
2022
;
17
(
1
):
56
.
16.
Usubutun
A,
Mutter
GL,
Saglam
A,
et al.
Reproducibility of endometrial intraepithelial neoplasia diagnosis is good, but influenced by the diagnostic style of pathologists
.
Mod Pathol
.
2012
;
25
(
6
):
877
884
.
17.
Virk
RK,
Gamez
R,
Mehrotra
S,
et al.
Variation of cytopathologists' use of the indeterminate diagnostic categories “atypical” and “suspicious for malignancy” in the cytologic diagnosis of solid pancreatic lesions on endoscopic ultrasound-guided fine-needle aspirates
.
Diagn Cytopathol
.
2017
;
45
(
1
):
3
13
.
18.
Vanderlaan
PA,
Krane
JF,
Cibas
ES.
The frequency of ‘atypia of undetermined significance’ interpretations for thyroid fine-needle aspirations is negatively correlated with histologically proven malignant outcomes
.
Acta Cytol
.
2011
;
55
(
6
):
512
517
.
19.
Juskevicius
R,
Zou
KH,
Cibas
ES.
An analysis of factors that influence the ASCUS/SIL ratio of pathologists
.
Am J Clin Pathol
.
2001
;
116
(
3
):
331
335
.
20.
Kuijpers
CC,
Visser
M,
Sie-Go
DM,
et al.
Improved cytodiagnostics and quality of patient care through double reading of selected cases by an expert cytopathologist
.
Virchows Arch
.
2015
;
466
(
6
):
617
624
.
21.
Layfield
LJ,
Yang
Z,
Vazmitsel
M,
Zhang
T,
Esebua
M,
Schmidt
R.
The international system for serous fluid cytopathology: interobserver agreement
.
Diagn Cytopathol
.
2022
;
50
(
1
):
3
7
.

Author notes

Supplemental digital content is available for this article at https://meridian.allenpress.com/aplm in the August 2024 table of contents.

The authors have no relevant financial interest in the products or companies described in this article.

Supplementary data