gms | German Medical Science

Objective: Gram staining yields a very low sensibility in the predilection of positive microbiological cultures in septic arthritis. The role of Gram and acridine-orange staining in non-arthritic hand phlegmona remains unknown.

Methods: Retrospective study at Geneva University Hospitals. Paediatric patients, atypical pathogens and arthritis cases were excluded. Stains were performed in the hospital-own laboratory by collaborators with long experience in microbiology.

Results: Among 126 episodes of pyogenic hand phlegmona, only 42 episodes (33%) revealed a positive Gram-staining and 16 episodes (13%) revealed a positive acridine-orange staining. When there was an abscess, the Gram-stain was positive in 50% of cases. The Spearman-correlation coefficient between Gram and acridine -staining was 0.49.

In a multivariate logistic regression analysis, the presence of an abscess/collection was positively correlated with a positive Gram stain (odds ratio 11.6, 95% confidence interval 3.0-44.3), while the duration of empirical antibiotic therapy (continuous variable) before the first surgical intervention was inversely related (OR 0.2, 95%CI 0.1-0.8). Age, immune suppression, tumefaction, number of fingers involved, fever or the serum C-reactive protein level on admission did not improve the low performance of Gram-staining. Local costs for Gram and acridine-orange staining were 23 and 29 US$, respectively.

Conclusion: Gram-staining is relatively costly and not helpful in the immediate confirmation of clinically suspected infection in hand phlegmona, especially if the patients are pre-treated with antibiotics. Acridine-orange yielded even worse results.