Time to Detection of Bacterial Cultures in Infants Aged 0 to 90 Days

RESEARCH ARTICLE Time to Detection of Bacterial Cultures in Infants Aged 0 to 90 Days AUTHORS Rianna C. Evans, MD, Bryan R. Fine, MD, MPH Division of Pediatric Hospital Medicine, Department of Pediatrics, Children s Hospital of The King s Daughters and Eastern Virginia Medical School, Norfolk, Virginia KEY WORDS bacterial culture, febrile infants, serious bacterial infection ABBREVIATIONS CSF: cerebrospinal fluid CONS: coagulase-negative Staphylococcus ED: emergency department SBI: serious bacterial infection TTD: time to detection UTI: urinary tract infection www.hospitalpediatrics.org doi:10.1542/hpeds.2012-0025 Address correspondence to Rianna C. Evans, MD, Children s Hospital of The King s Daughters, Division of Pediatric Hospital Medicine, 601 Children s Lane, Norfolk, VA 23507. E-mail: rianna.evans@chkd.org HOSPITAL PEDIATRICS (ISSN Numbers: Print, 2154-1663; Online, 2154-1671). Copyright 2013 by the American Academy of Pediatrics FINANCIAL DISCLOSURE: The authors have indicated they have no fi nancial relationships relevant to this article to disclose. FUNDING: No external funding. abstract OBJECTIVE: To determine the time to detection (TTD) of positive results on blood, urine, and cerebrospinal fluid (CSF) cultures taken during the evaluation for serious bacterial infection (SBI) in otherwise healthy infants aged 0 to 90 days. METHODS: This study was a retrospective chart review of infants aged 0 to 90 days with positive blood, urine, or CSF cultures drawn during evaluation for SBI in the emergency department or inpatient setting. The TTD of positive culture results, reason for testing, and age of the infant were recorded. RESULTS: A total of 283 charts were reviewed related to 307 positive culture results. Of the 101 positive results on blood culture, 38% were true pathogens with a mean TTD of 13.3 hours; 97% were identified in 36 hours. Blood cultures with TTD 36 hours were 7.8 times more likely to be contaminants compared with those with TTD <36 hours. Of 192 positive results on urine culture, 58% were true pathogens with a mean TTD of 21 hours; 95% were identifi ed in 36 hours. Fifty percent of 14 positive CSF culture results were true pathogens with a mean TTD of 28.9 hours; 86% were identified in 36 hours. When data for infants 28 days of age were analyzed separately, TTD followed the same patterns for positive blood and urine culture results as seen in all infants aged 0 to 90 days. CONCLUSIONS: In certain clinical situations, the inpatient observation period for infants under evaluation for SBI may be decreased to 36 hours. INTRODUCTION The management of young infants with fever is evolving with the advent of improved bacterial culture systems and various protocols that aid in predicting the risk for serious bacterial infections (SBIs). Febrile infants (temperature 38.0 C) are initially seen in an outpatient setting such as the emergency department (ED) or physician s office and may be either managed in an outpatient or inpatient setting while awaiting bacterial culture results. 1 Infants with fever or concern for SBI who are admitted to the hospital often receive empiric antibiotic coverage and are clinically monitored for 48 hours while awaiting blood, urine, and cerebrospinal fluid (CSF) culture results. This time period is based on multiple previous studies reporting that nearly all clinically significant organisms will show growth on bacterial culture media within 48 hours 2 4 ; however, many of these studies were performed before the institution of continuous blood culture monitoring systems, potentially inflating the reported 97

time to detection (TTD). Several more recent studies have shown that the TTD of the majority of pathogenic organisms on blood cultures in children is <24 hours, 5 8 but these studies were performed on infants in the ICU, nursery, or those >1 month of age, potentially excluding a large proportion of infants with fever managed by general pediatricians. An automated blood culture monitoring system has been in place in our hospital system since 1991. Many infants with fever have urine and CSF cultures drawn as well, which are read manually once per day in our laboratory. There are limited published data on the TTD of urine and CSF cultures of febrile infants and no identifiable study including the TTD of bacterial cultures in otherwise healthy infants aged 0 to 28 days. The current study was undertaken to determine if bacterial cultures taken in young infants would show growth of pathogenic bacteria in <36 hours, with the goal of potentially discharging otherwise healthy infants earlier than the standard 48-hour observation period. METHODS Study Population and Design This study was a retrospective chart review of positive results on blood, urine, and CSF cultures drawn in the ED or inpatient setting during the evaluation for SBI in previously healthy infants aged 0 to 90 days. Culture data were obtained by using the PathNet Microbiology Database (North Kansas City, MO) within the Cerner electronic medical record system. Inclusion and exclusion criteria are listed in Table 1. All blood, urine, and CSF cultures drawn in the specified population were identified, and charts from all infants TABLE 1 Inclusion and Exclusion Criteria Inclusion Criteria Age 0 90 d at time culture was drawn Report of blood, urine, and CSF culture present in microbiology database between October 1, 2007, and February 28, 2011 Cultures drawn in ED or inpatient setting with positive culture results were manually reviewed by the authors. The time period for charts selected for review was from October 2007 (earliest available data in the database) through February 2011. Cultures were deemed true pathogens if the patient was treated for that particular infection as determined by the attending physician. Cultures were deemed contaminant if the patient was not treated for infection with that organism. Data gathered from chart review included the cultured organism, TTD (in hours), infant s age (in days) at the time the culture was drawn, chief complaint on presentation, and true pathogen versus contaminant in culture specimen based on clinical course. The institutional review board at Eastern Virginia Medical School re - viewed and approved this study. Laboratory Protocols The standard procedure in our institution is to obtain a 3- to 5-mL blood specimen for culture, with a minimum of 1 ml. Specimens are inoculated into BACTEC Peds Plus/F culture vials and placed on the BACTEC 9000 Fluorescent Series instrument for Exclusion Criteria Blood cultures taken from central line Urine cultures obtained while Foley catheter in place CSF cultures taken from a ventricular shunt Urine cultures taken after urologic surgery (ie, ureterostomy) Any bacterial cultures taken while in the NICU, PICU, or transitional care unit Significant underlying medical or surgical conditions Repeat bacterial cultures taken from the same patient during the same hospitalization incubation and reading (BD, Franklin Lakes, NJ). The bottles are scanned every 10 minutes, and if flagged as positive, the sample is removed from the bottle and Gram stained. The specimen is then plated and read daily for further identification of the organism and susceptibilities. The minimum fluid collection requirement for urine and CSF culture is 0.5 ml. Urine and CSF samples are immediately plated and placed in an incubator. They are then removed and read manually once daily for growth. Statistical Analysis The data were analyzed by using SAS version 9.2 (SAS Institute, Cary, NC) and SPSS version 19 (SPSS Inc, Chicago, IL) software. The average TTD for blood, urine, and CSF samples were determined by using data from all positive samples. To determine the relationship between average TTD and percent contaminant, linear regression analysis was performed. To examine the relationship between outcome (treatment or no treatment) and TTD ( 36 hours or <36 hours), Pearson χ 2 tests were performed. If significant (P <.05), this analysis was followed by odds ratio analysis to further quantify the relationship s significance. Similarly, χ 2 analyses 98 VOLUME 3 ISSUE 2 www.hospitalpediatrics.org

were conducted to determine the relationship between TTD ( 36 hours or <36 hours) and age of patient ( 28 days or <28 days). RESULTS A total of 283 charts were reviewed related to 307 positive culture results. The mean TTD of the isolated organisms for each culture type is outlined in Table 2. A total of 2092 blood cultures were drawn in the selected population, of which 115 were positive (5.5%); 14 positive culture results met exclusion criteria. Of the remaining 101 positive blood culture results, 38 (38%) were deemed to be true pathogens with a mean TTD of 13.3 hours. The 63 (62%) blood cultures deemed contaminant specimens had a mean TTD of 24.9 hours. Of the blood cultures treated as true pathogens, 97% were identified in 36 hours (Fig 1). Linear regression demonstrated a significant relationship between average TTD and percent contaminant with an R 2 of 0.7876 (Fig 2). Blood cultures with TTD 36 hours were 7.827 times more likely to be treated as a contaminant compared with blood samples identified in <36 hours (P =.028). There was a total of 2283 urine cultures drawn in the selected population; 232 had positive results (10.2%). Forty positive culture results met exclusion criteria. Of the remaining 192 positive results on urine culture, 111 (58%) true pathogens had a mean TTD of 21 hours versus 26.7 hours for the 81 (42%) contaminant urine cultures. Of the urine cultures treated as true pathogens, 95% were identified in 36 hours (Fig 1). Linear regression demonstrated a weak correlation between mean TTD and percent contaminant for positive urine samples, mainly due to the inclusion of coagulasenegative Staphylococcus (CONS), which is commonly considered a urine contaminant and has a relatively rapid TTD. If CONS were excluded, the R 2 would increase to 0.8328 (Fig 3). However, with inclusion of all pathogens, a relationship was still demonstrated, with those positive urine culture results with a mean TTD 36 hours being 4.005 times more likely to be treated as a contaminant compared with samples with a mean TTD <36 hours. A total of 1159 CSF cultures were drawn in the selected population, of which 14 had positive (1.2%) results, and no exclusions were necessary. Of the 14 positive culture results, 7 (50%) were true pathogens with a mean TTD of 28.9 hours, and 7 (50%) were contaminant specimens with a mean TTD of 57.7 hours. Of the CSF cultures treated as true pathogens, 86% were identified in 36 hours (Fig 1). Due to small sample size, linear regression could not be performed for CSF cultures. Analysis of data for infants aged 28 days was limited due to small sample TABLE 2 Mean TTD in Hours of Isolated Species for All Subjects Organism Blood(n) Urine(n) CSF(n) Pathogen Contaminant Pathogen Contaminant Pathogen Contaminant Total 13.3 (38) 24.9 (63) 21.3 (111) 26.7 (81) 28.9 (7) 57.7 (7) α-hemolytic streptococci 0 20.8 (12) 0 31.8 (6) 24 (1) 0 Bacillus sp 0 12.3 (5) 0 0 0 0 Candida albicans 0 0 0 44 (1) 0 0 Citrobacter sp 16.1 (1) 0 22 (2) 0 0 0 CONS 27.2 (1) 28 (35) 0 18.5 (8) 0 52.3 (4) Clostridium subterminale 0 0 0 0 0 119 (1) Escherichia coli 15.8 (14) 10.6 (1) 20.4 (84) 18.7 (15) 19 (1) 0 Enterobacter sp 12.3 (2) 0 27.5 (4) 21.5 (2) 0 0 Enterococcus sp 12.3 (2) 14.2 (1) 31.6 (5) 29.7 (15) 65 (1) 36 (1) Streptococcus agalactiae 7.5 (7) 0 23.3 (4) 28.3 (8) 25 (2) 0 K pneumoniae 12.2 (4) 0 21.4 (11) 0 0 0 Lactobacillus sp 0 0 0 44 (2) 0 0 Methicillin-resistant Staphylococcus aureus 0 0 0 35 (2) 0 0 Methicillin-sensitive S aureus 17.4 (1) 0 0 31.9 (11) 0 40 (1) Micrococcus sp 0 59.8 (1) 0 0 0 0 Mixed respiratory flora 0 11.9 (2) 0 0 0 0 Moraxella sp 0 90.4 (1) 0 0 0 0 Nutritionally variant Streptococcus 0 16.6 (1) 0 0 0 0 S pneumoniae 12.3 (4) 0 0 0 22 (2) 0 Streptococcus salivarius 11.1 (1) 10.7 (1) 0 0 0 0 Three contaminant specimens with growth of multiple bacteria were excluded from the table. 99

FIGURE 1 TTD of true pathogen bacterial cultures. size. Forty percent of true pathogens in blood cultures occurred in infants aged 28 days (n = 16), and all produced positive results in <36 hours. Twenty-seven percent of true pathogens in urine cultures occurred in children aged 28 days (n = 30) with 1 sample detected at >36 hours. There were 2 true positive results in CSF samples in infants 28 days of age, both detected in <36 hours. When looking at the relationship between TTD ( 36 hours or <36 hours) and age of patient ( 28 days or <28 days), there were no significant relationships found; all P values were >.05. This finding demonstrates that age of the infant (in days) likely has no relationship to TTD. DISCUSSION Multiple studies of infants <90 days of age have shown that the incidence of FIGURE 2 Mean TTD for blood cultures according to percent contaminants for predominant organism types. a Klebsiella pneumoniae and S pneumoniae with similar data points. b Other includes all other samples with numbers too small for individual data point inclusion. SBI is great enough to warrant evaluation for the source of fever or illness. Current recommendations for the evaluation and management of the febrile infant involve risk stratification based on the age of the infant and findings on laboratory testing; it is recommended that infants in high-risk categories be monitored on empirical antibiotics in the inpatient setting while awaiting bacterial culture results. 1 Previous studies have evaluated the TTD of bacterial cultures in the newborn population, as well as in infants aged 28 to 90 days 2 8 ; however, there are many infants evaluated for SBI who do not fall into these categories, and culture methods have changed since these studies were published. These data suggest that, in the appropriate clinical and social circumstances, the physician may discontinue antibiotic therapy at 36 hours if bacterial cultures continue to have negative results during evaluation for SBI in otherwise healthy infants. In practice, the extent of the laboratory evaluation and choice to administer antibiotics varies, and children are frequently managed based on provider preference and clinical judgment of the degree of illness. 9,10 With existing variation in practice, many otherwise healthy infants with fever are admitted to the hospital based on provider experience and practice, not necessarily with strict adherence to published guidelines. Although the risk and consequences of SBI in young infants cannot be overlooked, those in - fants who are admitted to the hospital and who are determined to be low risk have, at most, an SBI rate of 1% to 3%. 11 When weighing the risk of SBI in young infants with fever, the downsides of hospitalization must be considered. 100 VOLUME 3 ISSUE 2 www.hospitalpediatrics.org

FIGURE 3 Mean TTD for urine according to percent contaminants for predominant organism types. CONS were excluded (see text). a Other includes all other samples with numbers too small for individual data point inclusion. Within the Children s Hospital As - sociation network, an ED evaluation and a 48-hour hospital stay in an otherwise healthy infant with fever generally incurred a charge of approximately $8000 in 2010. 12 An inpatient admission also exposes the infant to iatrogenic complications such as health care associated infections, intravenous infiltration, and antibiotic adverse effects as well as causing significant parental emotional and financial stress. Two studies have described the iatrogenic complication rate involved with the hospitalization of the young febrile infant to be 20% to 30%. 13,14 These studies describe complications including intravenous infiltrations, diarrhea or bone marrow suppression attributed to antibiotic use, acquired infection, and drug errors. Although the majority of these complications were self-resolving or easily treated, any complication is difficult to face from the child s or parent s perspective and may be less likely to occur with shorter length of hospitalization. Notable to the current study was the determination of true positive and contaminant culture results based on attending physician decision for treatment course. This methodology was chosen due to subtle clinical and social circumstances that are present with each case which cannot reliably be determined in chart review. We have therefore included in the true positive analysis 3 samples with positive blood culture results with species generally considered contaminants. A limitation to evaluating the TTD of bacterial specimens is the method of detection of bacterial growth on culture in our microbiology laboratory. As in many laboratory systems, urine and CSF cultures are read once daily, thereby potentially lengthening the reported TTD. If urine and CSF cultures were read twice daily, there would likely be a shorter reported TTD for pathogenic bacteria. There were 5% of true positive urine cultures detected in >36 hours; the details of each case are outlined in Table 3 and demonstrate increased level of suspicion for UTI, treatment before culture, or physician discretion for treatment. In addition, 2 pathogenic specimens in CSF were detected at 36 hours, which require discussion. The first case was an illappearing infant with a blood culture result at 11 hours positive for Streptococcus pneumoniae in which the CSF demonstrated growth of the same organism at 36 hours; this specimen was read initially at 4 hours and not read again until the 36-hour mark due to our laboratory s process of once-daily reading of cultures. The second was an 8-day-old infant of a 35-week twin gestation without prenatal care presenting with hypothermia. This infant s CSF culture had detection of Enterococcus faecium at 65 hours; despite the lack of TABLE 3 Case Details for Positive Urine Culture Results With TTD >36 Hours Age, d TTD, h Organism Case Details 44 60 E faecalis >100 000 CFU UA negative; history of previous UTI 75 55 Enterobacter amnigenus >100 000 CFU UA negative but not performed until 24 h after culture taken 47 48 E faecalis 40000 CFU UA with 0 2 WBC, few bacteria 75 41 K pneumoniae 10000 CFU UA with 0 2 WBC 10 38 E coli 1000 CFU 10 25 WBC, many bacteria, large LE, and positive nitrate; first urine culture at another facility and this sample after antibiotic given 81 37 E coli 1000 CFU UA with 0 WBC, moderate LE; on antibiotic at time of culture CFU, colony-forming unit; LE, leukocyte esterase ; UA, urinalysis; WBC, white blood cell count. 101

CSF pleocytosis and long TTD of the organism, this infant was treated due to ongoing hypothermia and apnea with bradycardia. This study is limited in the small number of CSF samples available for analysis in the selected population. Of 1159 CSF cultures performed in the population there were 7 positive samples treated as true pathogens (0.6%). Our study found that 86% of true pathogens in CSF culture were identified in 36 hours; therefore, we suggest using these data to support discharge before 48 hours in a well-appearing infant without significant CSF pleocytosis. Future collaborative studies would be helpful to gather more data regarding the TTD of pathogenic bacteria in the CSF of infants being evaluated for SBI. We also suggest that physicians become familiar with the methods used for bacterial culture sampling and monitoring in their institutions. Due to the nature of this retrospective chart review, we were unable to determine several factors that should be taken into consideration in determining likelihood of bacterial growth in culture; these include the bacterial burden in the infant, amount of culture fluid obtained, and previous administration of antibiotics. In addition, when considering discharge of an infant hospitalized during the evaluation for SBI, the infant s medical history and social circumstances should be investigated and close follow-up arranged. CONCLUSIONS In the wake of improved vaccination for several serious bacterial pathogens in young infants and improved bacterial culture technique, the evaluation and management of SBIs are evolving. Our study demonstrates that pathogenic bacteria are frequently identified in <36 hours. These data suggest that, in the right clinical and social circumstances, the physician may discontinue antibiotic therapy if bacterial culture results remain negative at 36 hours in infants admitted to the hospital with fever. Future research in the area of the TTD of CSF cultures would strengthen this conclusion. ACKNOWLEDGMENT We thank Erin McGuire Kren, MS, and Amy Perkins, MS, for their help with statistical analyses. REFERENCES 1. Baraff LJ. Management of fever without source in infants and children. Ann Emerg Med. 2000;36(6):602 614. 2. Pichichero ME, Todd JK. Detection of neonatal bacteremia. J Pediatr. 1979;94(6): 958 960. 3. Hurst MK, Yoder BA. Detection of bacteremia in young infants: is 48 hours adequate? Pediatr Infect Dis J. 1995;14(8):711 713. 4. Friedman J, Matlow A. Time to identification of positive bacterial cultures in infants under three months of age hospitalized to rule out sepsis. Paediatr Child Health (Oxford). 1999;4(5):331 334. 5. Alpern ER, Alessandrini EA, Bell LM, Shaw KN, McGowan KL. Occult bacteremia from a pediatric emergency department: current prevalence, time to detection, and outcome. Pediatrics. 2000;106(3):505 511. 6. McGowan KL, Foster JA, Coffin SE. Outpatient pediatric blood cultures: time to positivity. Pediatrics. 2000;106(2 pt 1):251 255. 7. Kaplan RL, Harper MB, Baskin MN, Macone AB, Mandl KD. Time to detection of positive cultures in 28- to 90-day-old febrile infants. Pediatrics. 2000;106(6). Available at: www. pediatrics.org/cgi/content/full/106/6/e74. 8. Garcia-Prats JA, Cooper TR, Schneider VF, Stager CE, Hansen TN. Rapid detection of microorganisms in blood cultures of newborn infants utilizing an automated blood culture system. Pediatrics. 2000;105(3 pt 1):523 527. 9. Pantell RH, Newman TB, Bernzweig J, et al. Management and outcomes of care of fever in early infancy. JAMA. 2004;291(10):1203 1212. 10. Meehan WP 3rd, Fleegler E, Bachur RG. Adherence to guidelines for managing the well-appearing febrile infant: assessment using a case-based, interactive survey. Pediatr Emerg Care. 2010;26(12):875 880. 11. Huppler AR, Eickhoff JC, Wald ER. Performance of low-risk criteria in the evaluation of young infants with fever: review of the literature. Pediatrics. 2010;125(2): 228 233. 12. Proprietary PHIS Database, administrated through Children s Hospital Association. Accessed on-line February 2011 by independent query. 13. Condra CS, Parbhu BP, Lorenz D, Herr SM. Charges and complications associated with the medical evaluation of febrile young infants. Pediatr Emerg Care. 2010;26(3):186 191. 14. DeAngelis C, Joffe A, Wilson M, Willis E. Iatrogenic risks and financial costs of hospitalizing febrile infants. Am J Dis Child. 1983;137(12):1146 1149. 102 VOLUME 3 ISSUE 2 www.hospitalpediatrics.org