Improving the quality of cytologic diagnoses: Root cause analysis of indeterminate cytologic diagnoses in body cavity fluid specimens
Shah A, Schultz D, Stone C, Noyes L, Cole G, and Zhang Z. Improving the quality of cytologic diagnoses: Root cause analysis of indeterminate cytologic diagnoses in body cavity fluid specimens. Lab Invest 2018; 98:174.
Background: Compared to a definitive positive or negative diagnosis of malignancy, indeterminate (abnormal/suspicious) cytologic diagnoses have less value in clinical decision making; in addition, they may cause avoidable delays in treatment or psychologic stress. This study aims to develop a strategy to minimize indeterminate cytologic diagnoses as part of continuous quality improvement of diagnostic services at the Henry Ford Hospital Cytopathology Laboratory. Design: We conducted a retrospective study and selected all pleural and peritoneal fluid cases with a primary diagnosis of abnormal/ atypical or suspicious for malignancy from July 2016 through June 2017. All cases were second reviewed for morphologic confirmation. The following data were collected: quantity submitted to cytology; actual amount collected; availability of cell block and immunostains and follow up data, if available. We categorized identified cases into <50mL and >50mL groups. Results: 1506 body cavity fluids were retrieved; 238 with positive diagnosis, 50 with indeterminate diagnosis (38 abnormal/atypical, 12 suspicious). 40/50 cases had <50mL (average 10.7mL submitted to the lab, but the average amount collected was 1833mL); slide review of 40 cases showed scarcity of abnormal cells morphologically undistinguished between reactive mesothelial cells vs metastatic adenocarcinoma. Scant quantity precludes cell blocks and subsequent immunostains; follow up data showed that 22/40 patients had repeat effusions, from which the diagnosis changed from atypical/suspicious to malignant in 19/22 (86%) with greater volume submitted (average 532mL on repeat samples). 10/50 cases have >50mL submitted to the lab (range 100-1600mL, average 780mL), 7/10 cases had available cell blocks but inconclusive immunostains; follow up data showed two diffuse large B cell lymphomas and one solitary fibrous tumor. Conclusions: Low body cavity fluid volume submitted to the lab is a major cause for indeterminate diagnosis. Of note, 86% of these cases prove to be positive for malignancy when repeat sample with larger volume is examined. When an “indeterminate” diagnosis is rendered in low volume specimens it is essential to add a diagnostic comment requesting a larger volume if the effusion re-accumulates. Prior publications have suggested at least 50-75 mL volumes for optimal analysis. Simultaneous training of clinicians to submit a large volume, or >75mL, of effusion fluid at time of first submission is likely to lead to more conclusive diagnoses.