Microbiological confirmation of a urinary tract infection (UTI) takes 24-48 h.
In the meantime, patients are usually given empirical antibiotics, sometimes inappropriately. We assessed the feasibility of sequentially performing a Gram stain and MALDI-TOF MS mass spectrometry (MS) on urine samples to anticipate clinically useful information. In May-June 2012, we randomly selected 1000 urine samples from patients with suspected UTI. All were Gram stained and those yielding bacteria of a single morphotype were processed for MALDI-TOF MS. Our sequential algorithm was correlated with the standard semiquantitative urine culture result as follows: Match, the information provided was anticipative of culture result; Minor error, the information provided was partially anticipative of culture result; Major error, the information provided was incorrect, potentially leading to inappropriate changes in antimicrobial therapy. A positive culture was obtained in 242/1000 samples. The Gram stain revealed a single morphotype in 207 samples, which were subjected to MALDI-TOF MS. The diagnostic performance of the Gram stain was: sensitivity (Se) 81.3%, specificity (Sp) 93.2%, positive predictive value (PPV) 81.3%, negative predictive value (NPV) 93.2%, positive likelihood ratio (+LR) 11.91, negative likelihood ratio (-LR) 0.20 and accuracy 90.0% while that of MALDI-TOF MS was: Se 79.2%, Sp 73.5, +LR 2.99, -LR 0.28 and accuracy 78.3%. The use of both techniques provided information anticipative of the culture result in 82.7% of cases, information with minor errors in 13.4% and information with major errors in 3.9%. Results were available within 1 h. Our serial algorithm provided information that was consistent or showed minor errors for 96.1% of urine samples from patients with suspected UTI. The clinical impacts of this rapid UTI diagnosis strategy need to be assessed through indicators of adequacy of treatment such as a reduced time to appropriate empirical treatment or earlier withdrawal of unnecessary antibiotics.
Burillo A, Rodríguez-Sánchez B, Ramiro A, Cercenado E, Rodríguez-Créixems M, Bouza E. Are you the author?
Department of Clinical Microbiology & Infectious Diseases, Hospital General Universitario Gregorio Marañón - Instituto de Investigación Sanitaria Gregorio Marañón, Madrid, Madrid, Spain; Facultad de Medicina, Universidad Complutense de Madrid (UCM), Madrid, Madrid, Spain; Red Española de Investigación en Patología Infecciosa (REIPI RD06/0008/1025), Sevilla, Sevilla, Spain; CIBER de Enfermedades Respiratorias (CIBERES CB06/06/0058), Palma de Mallorca, Islas Baleares, Spain.
Reference: PLoS One. 2014 Jan 22;9(1):e86915.