Clinical relevance of viridans and nonhemolytic streptococci isolated from blood and cerebrospinal fluid in a pediatric population.

Laboratory records were reviewed to assess the clinical relevance of isolating viridans (VS) and nonhemolytic (NHS) streptococci from blood and cerebrospinal fluid (CSF) specimens in a pediatric setting. During a nine-month period, 722 of 6,569 blood cultures and 113 of 2,023 CSF cultures were positive for one or more organisms. There were 26 VS and 10 NHS blood isolates from 30 patients and five NHS isolates from the CSF of five additional patients. The patients ranged in age from five weeks to 16 years. The charts of 34 patients were reviewed for evidence of sepsis or meningitis and the physician's response to the positive cultures. Three patients had subacute bacterial endocarditis (SBE) with multiple positive blood cultures. All other patients, including six oncology patients, failed to show a positive correlation between the isolation of VS or NHS and the disease process. Speciation and MIC testing were performed on 13 isolates, including those from all SBE and four oncology patients. Because of the lack of significance of VS and NHS from blood and CSF specimens in patients other than those with SBE, the authors conclude that extensive microbiologic workup of VS and NHS is not necessary without appropriate clinical indications such as SBE or immunosuppression.

Do regional variations in prevalence of cryptosporidiosis occur? The central Ohio experience.

We screened 2,780 consecutive stool specimens submitted for routine ova and parasite examination to assess the prevalence of cryptosporidiosis in a pediatric patient population in central Ohio. The stools were prepared by formalin-ethyl acetate concentration followed by cold Kinyoun acid-fast stain of the sediment. In addition, 912 consecutive intestinal biopsies were monitored for the presence of the parasite. Cryptosporidium oocysts were found in only 0.3 per cent of stool specimens (seven specimens from three patients) and in none of the intestinal biopsies. Due to this low prevalence of cryptosporidiosis, we conclude that routine screening of stool specimens for Cryptosporidium sp. is unnecessary in our patient population. Screening should be targeted to immune compromised patients and patients with persistent diarrhea and no apparent etiology. Our study also supports the concept that there are geographic variations in the prevalence of cryptosporidiosis.

Comparison of throat and nasopharyngeal swab specimens for culture diagnosis of Bordetella pertussis infection.

During a 9-month period, we evaluated the relative sensitivity of throat and nasopharyngeal swab cultures for isolation of Bordetella pertussis. Of 38 pertussis cases, 36 (95%) had positive nasopharyngeal cultures, while only 16 of 36 (44%) had positive throat cultures. There were no cases of nasopharyngeal-negative, throat-positive cultures. The sensitivity of the direct fluorescent-antibody test was 70% when compared with culture.

Onset of Hemophilus influenzae type-b meningitis during cefaclor therapy for preseptal cellulitis.

Second generation cephalosporins are frequently used for the treatment of bacteremic Hemophilus influenzae type b infections. "Breakthrough" meningitis during cefamandole therapy has documented the need for adequate cerebrospinal fluid penetration by these antibiotics if they are to be used in the therapy of Hemophilus infections. A child with H. influenzae type b preseptal cellulitis is reported who initially responded to treatment with intravenous cefuroxime and oral cefaclor. However, while still receiving cefaclor, the child was readmitted with H. influenzae meningitis. Microtiter broth dilution susceptibility testing performed during the second admission showed the isolate to be relatively resistant to cefuroxime (minimum bactericidal concentration [MBC] = 4 micrograms/ml) and resistant to cefaclor (MBC greater than 16 micrograms/ml). This experience documents the need to monitor the clinical response closely during therapy of H. influenzae bacteremic infections with these second generation cephalosporin treatment regimens. In addition, attention should be paid to minimum inhibitory concentrations of these cephalosporins, since variations in H. influenzae type b susceptibility to these agents may limit their efficacy.

Rapid screening and microbiologic processing of pediatric urine specimens.

Urinary nitrite and leukocyte esterase dipstick tests were evaluated as rapid screening procedures to select probable culture-positive urines for direct identification (AutoMicrobic System urine cards) and modified Kirby-Bauer susceptibility testing. Approximately 73% of significant culture-positive (greater than 10(5) organisms per milliliter, pure culture) urine specimens could be selected by nitrite testing alone with very high specificity (approximately 99%). The leukocyte esterase test detected 85% of culture-positive urines when used alone and approximately 91% when used in combination with nitrite testing (if either test was positive it was considered a positive screening); however, the esterase test was significantly less specific for bacteriuria than the nitrite test. Based on these results, the nitrite test was selected for use as the screening test. Rapid, direct identification and susceptibility tests on screen-positive urines showed 97% correlation with standard testing methods. Significant positive urines processed in this manner could be reported with quantitation, identification, and susceptibility results within 24 hr.

Evaluation of rapid identification of gram-positive cocci in positive blood cultures by use of the AutoMicrobic system Gram-Positive Identification Card.

Because rapid identification of gram-positive organisms from blood cultures may provide valuable information for patient care and because the AutoMicrobic system Gram-Positive Identification (AMS-GPI) Card (Vitek Systems, Inc., Hazelwood, Mo.) is designed for the identification of these organisms in 4 to 13 h, we designed this study to evaluate the performance of the AMS-GPI Card in the direct identification of gram-positive organisms upon detection of growth in blood culture bottles. We compared direct identification by the AMS-GPI Card with the final AMS-GPI Card identification and with our standard identification methods. We evaluated 51 gram-positive organisms from clinical blood cultures as well as 49 simulated blood cultures. The isolates included Streptococcus pneumoniae (17), Streptococcus pyogenes (13), group D enterococci (12), Streptococcus agalactiae (11), viridans streptococci (10), coagulase-negative staphylococci (21), Staphylococcus aureus (15), and Listeria monocytogenes (1). The AMS-GPI Card identified all of the group D enterococci, viridans streptococci, and coagulase-negative staphylococci and all but one each of the Streptococcus pyogenes and Streptococcus agalactiae isolates. L. monocytogenes was also correctly identified. However, the AMS-GPI Card identified only 12 of 17 Streptococcus pneumoniae and 9 of 15 Staphylococcus aureus isolates by direct inoculation. We therefore conclude that the results of direct identification of gram-positive organisms by the AMS-GPI Card may be used cautiously for rapid direct identification of gram-positive organisms from positive blood cultures.

Comparative laboratory evaluation of three antigen detection methods for diagnosis of Haemophilus influenzae type b disease.

Cerebrospinal fluid, urine, serum, and other body fluid specimens from pediatric patients with systemic disease were tested with Bactigen latex agglutination (555 specimens), Phadebact coagglutination (319 specimens), and counterimmunoelectrophoresis (335 specimens) for the presence of Haemophilus influenzae type b antigen. All three methods showed good sensitivity for detecting antigen in the cerebrospinal fluid of patients with culture-positive meningitis (greater than or equal to 86% sensitivity). However, coagglutination and counterimmunoelectrophoresis were much less sensitive (less than or equal to 40%) than latex agglutination (96%) for detecting antigen in other body fluid specimens in culture-positive, nonmeningeal H. influenzae disease. Bactigen latex agglutination was also more sensitive than the other procedures for detecting antigen in specimens from patients with culture-negative, presumed H. influenzae disease. Comparative testing of fluids spiked with known quantities of purified H. influenzae b polyribosephosphate capsular polysaccharide revealed an apparent 100-fold greater sensitivity with Bactigen as compared with the other two methods. Although all three methods showed good specificity (greater than 98%), both agglutination methods gave a few false-positive results. In a clinical setting where both meningeal and nonmeningeal H. influenzae b disease are encountered frequently, Bactigen latex agglutination appears to be superior to coagglutination and counterimmunoelectrophoresis for detecting antigen in body fluids.

Comparison of three major antigen detection methods for the diagnosis of Group B streptococcal sepsis in neonates.

Because of the difficulty encountered in diagnosing early onset Group B streptococcal disease (GBS) in neonates and because of the proliferation of tests to detect the antigen in urine, we made qualitative and quantitative comparisons among the three major, commercially available, antigen detection systems. The methods compared were Wellcogen latex agglutination, Phadebact coagglutination, and counterimmunoelectrophoresis (CIE). We tested urine, with or without serum, and tracheal or gastric aspirates from 176 neonates admitted to Columbus Children's Hospital, with suspected GBS disease. Wellcogen and Phadebact were equally sensitive indicators of neonatal GBS sepsis (100%) as compared to CIE which was only 30% sensitive. CIE, however, did not produce any false-positives (100% specificity) while Phadebact coagglutination and Wellcogen latex agglutination were approximately 98% specific. As a side bar to the main study, we also set out to determine whether tracheal or gastric aspirates would be consistent and convenient sources of antigen as compared to urine. Consequently we determined that neither aspirate is a good source of antigen as performed by our method.