SESSION 34. ORAL ABSTRACTS IN DIAGNOSTIC MICROBIOLOGY. NO. 132. DETECTION OF BLOODSTREAM INFECTIONS IN ADULTS: HOW MANY BLOOD CULTURES ARE NEEDED?1
Lee, Weinstein, Mirrett, and Reller reviewed all positive blood cultures (BC) obtained from adult inpatients from November 1, 2004 through December 31, 2005 at Robert Wood Johnson University Hospital and Duke University Medical Center. Medical records of those inpatients were analyzed, and BC, which were felt to represent true infection, were included in the study. Of 248 unimicrobial episodes with at least 4 BC obtained in 24 hours, 73.4% of bloodstream infections (BSI) were detected with the first BC, 89.9% were detected with the first 2 BC, 96.4% were detected with the first 3 BC, and 99.6% were detected with the first 4 BC. Of 181 unimicrobial episodes with 3 BC obtained over 24 hours, 77.3% were detected with 1 BC, 93.9% were detected with 2 BC, and 100% were detected with 3 BC. Of 37 polymicrobial episodes, 94% were detected with 2 BC and 100% were detected with 3 BC. The authors conclude that 4 BC in 24 hours seem to be necessary to detect greater than 99% of BSI.
Over the years, several studies have been performed to answer the question of how many BC are necessary to detect BSI. Washington2 in a 1975 study found that 3 BC were necessary to detect 99% of bacteremias. Weinstein et al3 in their 1983 study found that only 2 BC were necessary to detect 99% of BSI. Cockerill et al4 in a study published in 2004 found that 3 BC were necessary to detect 95.6% of bacteremias, and 4 BC were needed to detect 99% of bacteremias. In the present study, 4 BC yielded a higher result than 2 or 3.
Although the studies done in the 1970s and 1980s used manual systems, the 2 later studies were performed using automated BACTEC 9240 systems. Thus, one might have expected fewer BC necessary to yield a pathogen, yet this was not the case. Perhaps BSI with lower concentrations of microorganisms are being detected by the new automated systems. Another factor might be that more patients are already on effective antibiotics at the time the BC are being drawn.
In the discussion after the presentation of the article, the other factor brought up was the volume of blood necessary for satisfactory yield. The volume of blood per BC set in the aforementioned studies was as follows: 20 mL in the Washington study, 15 mL in the 1983 Weinstein study, 20 mL in the Cockerill study, and 20 mL in the present study. Thus, the lower yields in the latter two studies do not appear to be due to suboptimal volumes of blood.
Other factors that have changed over the years include an increasing number of indwelling venous catheters and prosthetic devices, such as permanent pacemakers, and a wider range of immunocompromised patients, such as bone marrow and solid organ transplant patients. Thus, the etiology and nature of bacteremias have shifted somewhat over the years, along with host response in some cases. A potential confounding factor in comparing studies is the interpretation of blood culture results as real or not real, which has some element of subjectivity, even when evaluated by an infectious disease physician.
The present study, along with the previous Cockerill data, supports the ordering of 4 BC in 24 hours for maximum yield.
1. Lee A, Weinstein MP, Mirrett MS, et al. Session 34. No. 132 IDSA; 2006.
2. Washington JA. Blood cultures: principles and techniques. Mayo Clin Proc
3. Weinstein MP, Reller LB, Murphy JR, et al. The clinical significance of positive blood cultures: a comprehensive analysis of 500 episodes of bacteremia and fungemia in adults. Rev Infect Dis
4. Cockerill FR, Wilson JW, Vetter EA, et al. Optimal testing parameters for blood cultures. Clin Infect Dis
SESSION 67. LATE BREAKER SYMPOSIUM
Session 67. Late Breaker Symposium. No. LB-2. Frequency of Vaginal Colonization with Community-Associated Methicillin-Resistant Staphylococcus aureus (CA-MRSA) in Pregnant Women1
The present study was undertaken in Nashville, Tenn. Two hundred fifty women with high-risk pregnancies were screened for group B streptococcus infection with vaginal cultures. The cultures were also tested for the presence of Staphylococcus aureus. If S. aureus was detected, further workup of the strain was undertaken.
Of the 250 women, 10.8% had community acquired MRSA (CA-MRSA) isolated from vaginal culture. Panton-Valentine Leukocidin gene was found in 22% of the MRSA cultures. One woman with MRSA vaginal colonization was noted to develop postcesarean section wound infection.
There have been well-described outbreaks of CA-MRSA infection in nurseries and among neonates.2 The source of these infections is often unclear and presumably originate from either health-care workers or mothers or other family members. The present study, along with a study done on urban populations in New York City and Newark, NJ, and presented in poster 365,3 provides evidence that some women are vaginally colonized with MRSA (in the second study, the rate of vaginal colonization with CA-MRSA was 3%). This represents a potential source of infection for neonates and others, although the risk of maternal or neonatal disease is unknown. Other questions that arise include the following: should women with vaginal colonization of CA-MRSA be decontaminated or receive antibiotics around the time of delivery? The optimal approach to the patient with vaginal colonization of CA-MRSA remains undefined.
1. Creech CB, Litzner B, Talbot TR, et al. Session 67. No. LB-2 IDSA; 2006.
2. Centers for Disease Control and Prevention. Community associated methicillin resistant Staphylococcus aureus
infections among healthy newborns: Chicago and Los Angeles County, 2004. MMWR
3. Azar E, D'Amico R, Doymaz M, et al. Session 45. No. 365 IDSA: 2006.
SESSION 86. ORAL ABSTRACT SESSION, EPIDEMIOLOGY
Session 86. Oral Abstract Session, Epidemiology. No.773. First Dengue Hemorrhagic Fever Outbreak in the Continental United States, 20051
In 2005, there was an outbreak of dengue involving more than 6,000 people in Brownsville, Tex, and Matamoros, Tamaulipas in Mexico. The Centers for Disease Control and Prevention investigated 129 persons seen in hospitals or emergency rooms with laboratory-confirmed dengue. These included 25 cases in the United States. Of the cases in the United States, 64% had dengue hemorrhagic fever (DHF). The authors comment that with large populations of Aedes aegypti in the area carrying multiple strains of dengue serotypes, along with a high prevalence of diabetes in southern Texas, there will likely lead to more cases of dengue and DHF in the region.
Since 1989, there have been many outbreaks of DHF in South America, Central America, and Mexico.2,3 There have been sporadic cases of dengue in Texas towns bordering on Mexico over the years.4 The present study details the first DHF outbreak in the continental United States. Clinicians must be mindful of this when taking a travel history in their febrile patients.
1. Kapella BK, Carillo C, Moya C, et al. Session 86. No. 773 IDSA: 2006.
2. Thomas SJ, Strickman D, Vaughn DW, et al. Dengue epidemiology: virus epidemiology, ecology, and emergence. Adv Virus Res
. 2003;61: 235-289.
3. Guzman MG, Kouri G. Dengue and dengue hemorrhagic fever in the Americas: lessons and challenges. J Clin Virol
4. Rawlings JA, Hendricks KA, Burgess CR, et al. Dengue surveillance in Texas, 1995. Am J Trop Med Hyg