Introduction. All the microbiology texts underline the importance of performing a gram stain on blood culture broth, if positive, with a double purpose: address the microbiologist in the choice of subculture media and provide the clinician with preliminary indication for the choice of the treatment. However, there are few data and literature on the concordance between gram stain and definitive result.The study aims at evaluating, in the reality of our hospital, the reliability of the gram stain performed starting from the blood culture broth. Materials and methods. The results of the gram stain on broth and of the final detection of all positive blood cultures performed from January 1st 2003 to December 31st 2004 by the Microbiology Lab based at Ospedali Riuniti Bergamo have been analysed. Results. In two years 52909 bottles have been processed; 6328 (11.96%) resulted positive, of which 5921 (11.19%) for monomicrobial flora and 407 (0.77%) for polymicrobial flora. Concerning the 6643 definitive identifications, the gram stain resulted fully correct in 6140 cases (92.43%); errors have been interpreted as “minor” in 285 cases (4.29%) for a partial or absent definition of the morphologic-dyeing features, “major” in 14 cases (0.21%) for notification of micro-organisms non grown later, “serious” in 204 (3.07%) for wrong reading of Gram (36 cases, equal to 0.54%) or no interpretation, in case of mixed cultures, of micro-organisms grown later (168 cases equal to 2.53%). Conclusions. Our study confirms the reliability of Gram stain and its role in providing in advance the definitive results of blood culture; however, it highlights the risk that Gram stain cannot detect polymicrobial aetiologies.
PlumX Metrics provide insights into the ways people interact with individual pieces of research output (articles, conference proceedings, book chapters, and many more) in the online environment. Examples include, when research is mentioned in the news or is tweeted about. Collectively known as PlumX Metrics, these metrics are divided into five categories to help make sense of the huge amounts of data involved and to enable analysis by comparing like with like.