Search

BACKGROUND Rapid onsite evaluation (ROSE) has several potential benefits but also can prolong procedures if one must wait for a cytopathologist, and it can involve a considerable time commitment on the part of the cytopathologist. At the University of Arkansas for Medical Sciences, interventional pulmonologists have routinely reviewed cytology specimens. This study was performed to determine prospectively how accurately pulmonologists could perform ROSE and whether they could contribute to the efficiency of the process. METHODS For sequential cases, the procedural pulmonologist documented a ROSE reading before the reading by the cytopathologist. Readings were compared between the two for agreement and for accuracy. The time commitment for the cytopathologist was also recorded. RESULTS One hundred sixty-four sites were biopsied in 102 patients. With respect to onsite adequacy, there was a high level of concordance between pulmonology and cytopathology as evidenced by the κ score ( ± standard error) of 0.72 ± 0.15 and by disagreement in only 3 cases (2%). For the diagnostic category, there was once again a high level of concordance; there was agreement in 141 of the 164 cases (86%), and the weighted κ score was 0.89 ± 0.02. The cytopathologist's time in the endoscopy suite averaged 4.02 ± 6.9 minutes per procedure. CONCLUSIONS Procedural pulmonologists can effectively learn enough cytology to be able to make ROSE a collaborative process and to greatly increase the efficiency of the cytopathologist. Cancer (Cancer Cytopathol) 2015. © 2015 American Cancer Society.

BACKGROUND There is no widely accepted rapid on-site evaluation (ROSE) reporting system for endobronchial ultrasound-guided transbronchial needle aspiration. At the University of Arkansas for Medical Sciences, ROSE reporting was unstructured. The goal was to evaluate, compare, and improve upon 2 structured approaches proposed in the literature. METHODS One hundred eighteen consecutive nodal aspirates were retrospectively reviewed by a pathology resident and a staff cytopathologist, both of whom were blinded to the original unstructured readings. Each reviewer interpreted every specimen with 2 different structured criteria proposed in the literature: criteria from the University of Minnesota (the Minnesota [MN] criteria) and criteria from the North Shore Long Island Jewish Health System (the New York [NY] criteria). The data allowed a comparison of the original unstructured ROSE system with the MN and NY scoring schemes and the final diagnosis. RESULTS Original on-site adequacy (OSA) had been assessed at 96%. Three cases were false-adequate according to the original unstructured approach; these had been called adequate on site, but a subsequent slide review including cell blocks did not show definite nodal tissue. OSA dropped to 86% with the MN criteria and to 85% with the NY criteria. No false-adequate on-site diagnoses would have been rendered with the application of either structured criteria. There were no significant differences between the MN and NY criteria with respect to the determination of OSA. An assessment of ease of application favored the NY criteria. With respect to diagnostic categories, each of the systems (MN and NY) was felt to have a category of value not used by the other system. CONCLUSIONS A standardized intra- and inter-institutional system for ROSE reporting is needed. On the basis of comparative analyses and consensus, modifications to prior criteria have been proposed in the hope of approaching this goal. Cancer (Cancer Cytopathol) 2015;123:362–72. © 2015 American Cancer Society

Rapid onsite evaluation (ROSE) has several potential benefits but also can prolong procedures if one must wait for a cytopathologist, and it can involve a considerable time commitment on the part of the cytopathologist. At the University of Arkansas for Medical Sciences, interventional pulmonologists have routinely reviewed cytology specimens. This study was performed to determine prospectively how accurately pulmonologists could perform ROSE and whether they could contribute to the efficiency of the process.For sequential cases, the procedural pulmonologist documented a ROSE reading before the reading by the cytopathologist. Readings were compared between the two for agreement and for accuracy. The time commitment for the cytopathologist was also recorded.One hundred sixty-four sites were biopsied in 102 patients. With respect to onsite adequacy, there was a high level of concordance between pulmonology and cytopathology as evidenced by the κ score ( ± standard error) of 0.72 ± 0.15 and by disagreement in only 3 cases (2%). For the diagnostic category, there was once again a high level of concordance; there was agreement in 141 of the 164 cases (86%), and the weighted κ score was 0.89 ± 0.02. The cytopathologist's time in the endoscopy suite averaged 4.02 ± 6.9 minutes per procedure.Procedural pulmonologists can effectively learn enough cytology to be able to make ROSE a collaborative process and to greatly increase the efficiency of the cytopathologist.

There is no widely accepted rapid on-site evaluation (ROSE) reporting system for endobronchial ultrasound-guided transbronchial needle aspiration. At the University of Arkansas for Medical Sciences, ROSE reporting was unstructured. The goal was to evaluate, compare, and improve upon 2 structured approaches proposed in the literature.One hundred eighteen consecutive nodal aspirates were retrospectively reviewed by a pathology resident and a staff cytopathologist, both of whom were blinded to the original unstructured readings. Each reviewer interpreted every specimen with 2 different structured criteria proposed in the literature: criteria from the University of Minnesota (the Minnesota [MN] criteria) and criteria from the North Shore Long Island Jewish Health System (the New York [NY] criteria). The data allowed a comparison of the original unstructured ROSE system with the MN and NY scoring schemes and the final diagnosis.Original on-site adequacy (OSA) had been assessed at 96%. Three cases were false-adequate according to the original unstructured approach; these had been called adequate on site, but a subsequent slide review including cell blocks did not show definite nodal tissue. OSA dropped to 86% with the MN criteria and to 85% with the NY criteria. No false-adequate on-site diagnoses would have been rendered with the application of either structured criteria. There were no significant differences between the MN and NY criteria with respect to the determination of OSA. An assessment of ease of application favored the NY criteria. With respect to diagnostic categories, each of the systems (MN and NY) was felt to have a category of value not used by the other system.A standardized intra- and inter-institutional system for ROSE reporting is needed. On the basis of comparative analyses and consensus, modifications to prior criteria have been proposed in the hope of approaching this goal.