Can sputum gram stain be used to predict lower respiratory tract infection and guide empiric antimicrobial treatment: Experience from a tertiary care hospital

Abstract

BackgroundThe mortality associated with lower respiratory tract infection is high. Indiscriminate use of antimicrobials leads to alteration of respiratory tract flora and development of multi-drug resistance. Rapid diagnostic tests to confirm infection can guide the clinicians about antimicrobial treatment. So the present study was planned to evaluate the role of direct gram stain examination as a rapid and simple test to help clinicians for appropriate patient management. MethodsThe present study was conducted on 1000 respiratory specimens which were processed using conventional microbiological techniques. Gram stain smear and culture results were compared statistically to assess the sensitivity, specificity, positive and negative predictive value. The agreement between gram stain smear examination and culture was calculated using kappa statistics. ResultsPotential pathogens were obtained from 28 of 209 deeply coughed out sputum samples (13.3%) and from 19 of 315 saliva mixed sputum samples (6%). Out of 473 tracheal aspirates, 115 (24.3%) had potential pathogens. The sensitivity for predicting infection was higher for good quality sputum samples (54%) as compared to poor quality sputum samples (37%). The gram stain and culture of tracheal samples had a good agreement for predicting infection whereas there was only moderate agreement for sputum sample. ConclusionGram stain smear examination from respiratory samples can be used to guide empiric antibiotic therapy pending final culture sensitivity results if the attending physicians ensure appropriate sample collection and transport. In absence of these supportive measures smear examination should not be relied upon for empiric treatment.