Performing Gram stain directly on catheter tips: assessment of the quality of the observation process

Abstract

A previous study performed in our institution showed that catheter tip (CT) staining by combining acridine orange and Gram stain (GS) before culture anticipated catheter colonization with exhaustive and careful observation by a highly trained technician. Our objective was to assess the validity values of GS without acridine orange on an external smear of CT for predicting catheter colonization and catheter-related bloodstream infection (C-RBSI). We compared different periods of observation and the results of two technicians with different levels of professional experience. Over a 5-month period, the roll-plate technique was preceded by direct GS of all CTs sent to the microbiology laboratory. The reading was taken at ×100 by two observers with different skill levels. Each observer performed a routine examination (3 min along three longitudinal lines) and an exhaustive examination (5 min along five longitudinal lines). The presence of at least one cell was considered positive. All slides were read before culture results were known. We included a total of 271 CTs from 209 patients. The prevalence of catheter colonization and C-RBSI was 16.2 % and 5.1 %, respectively. Routine and exhaustive examinations revealed only 29.5 % and 40.9 % of colonized catheters, respectively (p < 0.001). In contrast, they revealed high negative predictive values for C-RBSI (96.5 % and 96.3 %, respectively). Our study shows that the yield of GS performed directly on CTs is greater when staining is performed exhaustively. However, the decision to implement this approach in daily routine will depend on the prevalence rate of catheter colonization at each institution.