Detection of malignancy in body fluids: a comparison of the hematology and cytology laboratories.

CONTEXT Body fluids submitted to the hematology laboratory for cell counts may also be examined for the presence of malignancy. Previous studies evaluating the hematology laboratory's performance at detecting malignancy in body fluids have reached conflicting conclusions. OBJECTIVE To investigate the hematology laboratory's ability to detect malignancy in body fluids by comparison with cytology. DESIGN Retrospective analysis of 414 body fluid samples during an 18-month period, with introduction of new quality assurance measures after the first 210 cases. If no concurrent cytology was ordered, results were compared with recent previous and/or subsequent cytologic, histologic, or flow cytometric diagnoses. RESULTS Of the initial 210 cases, the hematology laboratory detected 3 of 13 malignancies diagnosed by concurrent cytology (23% sensitivity), with no false-positives (100% specificity). Malignancy was not identified on retrospective review of the hematology slides in the 10 discrepant cases. After the initial study, educational sessions on morphology for the medical technologists and a more thorough hematology-cytology correlation policy were implemented. The subsequent 204 hematology laboratory cases had increased sensitivity for the detection of malignancy (60%; 6 of 10). Definitive features of malignancy were seen in only one discrepant hematology laboratory slide on retrospective review. This case had not been flagged for hematopathologist review. None of the discrepancies before or after implementation of the additional quality assurance measures impacted patient care. CONCLUSIONS Body fluid processing by the hematology laboratory is not optimized for the detection of malignancy. Concurrent cytologic examination is critical for the detection of malignancy, and needs to be considered as cost-saving measures are increasingly implemented.