Pseudomeningitis again. Association with cytocentrifuge funnel and Gram-stain reagent contamination.

OBJECTIVE: To report an "epidemic" of pseudomeningitis related to cytocentrifuge funnel and Gram-stain reagent contamination, and our evaluation and responses.
DESIGN: Investigation was stimulated by the recognition of Gram-stained, smear-positive, culture-negative cerebrospinal fluid (CSF) specimens. Cytofunnels, glass slides, Gram-staining reagents, and an automated Gram-staining apparatus were subjected to repeated staining and culture. Control stains and cultures using fetal bovine serum (simulated CSF) were performed for comparison.
SETTING: The clinical microbiology laboratory of Parkland Memorial Hospital, a large acute-care teaching hospital. SPECIMENS: Cerebrospinal fluid specimens were submitted to the clinical microbiology laboratory in the course of routine patient care. MAIN OUTCOME MEASURES: Gram's stains and cultures of test and control preparations.
RESULTS: Most of the smear-positive, culture-negative, original CSF specimens contained Gram-positive bacilli or Gram-negative bacilli. Smears of cytofunnels revealed similar organisms, and cultures revealed Bacillus species. Cytofunnels from several lots were culture-positive. Glass slides were not contaminated. Of 25 CSF specimens stained during the initial week of investigation, 23 were negative by culture and two grew Cryptococcus neoformans (from acquired immunodeficiency syndrome patients). Control stains and cultures of simulated CSF were negative. Gram-stain reagents were frequently smear-positive, and cultures repeatedly yielded Flavimonas oryzihabitans from the crystal violet well of an automated Gram-staining apparatus. These latter contaminants could not be eliminated consistently.
INTERVENTIONS: No alternative sources of cytofunnels were found. The Gram-staining apparatus was cleaned and reagents changed frequently. Cytocentrifugation and use of automated Gram staining was discontinued for CSF and other normally sterile fluids. The laboratory staff was repeatedly educated about the problem.
CONCLUSIONS: Contamination of cytocentrifuge funnels and an automated Gram-staining apparatus contributed to an "epidemic" of pseudomeningitis. The problem was corrected by education of the laboratory staff and by altered management of CSF and other sterile body fluid specimens.