Background The overall profitability of Gram staining in synovial fluid has been estimated at around 50% (Ryan MJ, et al Br J Rheumatol 1997 Mar, 36[3]:370 3). The practice of articular or periarticular assisted procedures guided by ultrasound have led to a reduction in the rate of complications and increased therapeutic effectiveness of infiltration

Objectives The aim of this study was to determine the changes in terms of sensitivity (S), specificity (Sp), positive predictive value ( VVP), negative predictive value (NPV) of the Gram stain in the cases of infectious bursitis in our hospital depending on the technique of bursocentesis used: Conventional or real time ultrasound guided (GETR).

Methods Definitive diagnoses (synovial fluid culture) of infectious bursal disease and clinical diagnoses but not microbiological made between 2011 and 2013 were indentified. Only cases in which there has been a sufficiently documented bursocentesis in the emergency report (electronic registration system Excalibur®) and has been practiced a Gram stain were included. VPP and S were calculated according to the technique of bursocentesis.

Results In the period of time indicated were identified 66 definitive diagnoses of infectious bursitis and 21 synovial fluid cultures from bursae that were reported as sterile. From the final diagnoses, 57 corresponded to patients whose electronic report was sufficiently documented and synovial fluids had undergone a Gram stain. The total of 21 cases of sterile liquids from bursae were sufficiently documented. In GETR procedures performed and whose cultures were positive (17), 10 had a diagnostic Gram stain and in 7 no microorganisms were identified. In procedure performed conventionally and whose cultures were positive (40), 19 had a Gram stain diagnosis, while in 21 no microorganisms were identified. The values of S, E, PPV and NPV of the Gram stain in GETR bursocentesis were 58.8%, 100%, 100% and 41.6% respectively while in the conventional bursocentesis were 47.5%, 93.3%, 95% and 40%, respectively.

Conclusions Despite the restrictions inherent to the sample size of our study, it was observed an increase in the sensitivity and positive predictive value of the Gram stain in bursocentesis performed with ultrasound assistance. This can be explained by the possibility of samplers on specific areas avoiding puncture of tissues or vessels and over the obstacles that represent partitions intrabursales fibrin, commonly present in these types of bursitis. Although infectious bursitis is not a diagnosis of urgency compared with septic arthritis, knowledge and application of techniques to increase the diagnostic profitability of tests that can be performed at point of care positive impact on the management of these patients in terms of greater diagnostic affiliation (definitive diagnosis) and will condition better empirical therapy directed by the results of the Gram stain.

Disclosure of Interest : None declared

DOI 10.1136/annrheumdis-2014-eular.5898