Antifungal prophylaxis associated with decreased induction mortality rates and resources utilized in children with new-onset acute myeloid leukemia.

BACKGROUND: Invasive fungal infections cause significant morbidity and mortality for children with acute myeloid leukemia (AML). Data on the comparative effectiveness of antifungal prophylaxis in this population are limited. METHODS: A pediatric AML cohort was assembled from the Pediatric Health Information System database using ICD-9 codes and pharmacy data. Antifungal prophylaxis status was determined by pharmaceutical data review within 21 days of starting induction chemotherapy. Patients were followed until end of induction, death, or loss to follow-up. Cox regression analyses compared induction mortality and resources utilized between patients receiving and not receiving antifungal prophylaxis. A propensity score accounted for variation in demographic factors, location of care, and severity of illness at presentation. RESULTS: Eight hundred seventy-one AML patients were identified; the induction case fatality rate was 3.7%. In the adjusted Cox regression model, patients receiving antifungal prophylaxis (57%) had a decreased hazard for induction mortality (hazard ratio [HR], 0.42; 95% confidence interval [CI], .19-.90). Children receiving prophylaxis were less frequently exposed to broad-spectrum gram-positive (incidence rate ratio [IRR], 0.87; 95% CI, .79-.97) and antipseudomonal ß-lactam agents (HR, 0.91; 95% CI, .85-.96), had fewer blood cultures (IRR, 0.78; 95% CI, .71-.86), and had fewer chest CT scans (IRR, 0.73; 95% CI, .60-.88). CONCLUSIONS: Antifungal prophylaxis in pediatric AML patients was associated with reduced induction mortality rates and supportive care resources. Further investigation is necessary to determine whether antifungal prophylaxis should include antimold activity.

Outcome of pediatric acute myeloid leukemia patients receiving intensive care in the United States.

OBJECTIVE: Children with acute myeloid leukemia are at risk for sepsis and organ failure. Outcomes associated with intensive care support have not been studied in a large pediatric acute myeloid leukemia population. Our objective was to determine hospital mortality of pediatric acute myeloid leukemia patients requiring intensive care. DESIGN: Retrospective cohort study of children hospitalized between 1999 and 2010. Use of intensive care was defined by utilization of specific procedures and resources. The primary endpoint was hospital mortality. SETTING: Forty-three children's hospitals contributing data to the Pediatric Health Information System database. PATIENTS: Patients who are newly diagnosed with acute myeloid leukemia and who are 28 days through 18 years old (n = 1,673) hospitalized any time from initial diagnosis through 9 months following diagnosis or until stem cell transplant. A reference cohort of all nononcology pediatric admissions using the same intensive care resources in the same time period (n = 242,192 admissions) was also studied. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: One-third of pediatric patients with acute myeloid leukemia (553 of 1,673) required intensive care during a hospitalization within 9 months of diagnosis. Among intensive care admissions, mortality was higher in the acute myeloid leukemia cohort compared with the nononcology cohort (18.6% vs 6.5%; odds ratio, 3.23; 95% CI, 2.64-3.94). However, when sepsis was present, mortality was not significantly different between cohorts (21.9% vs 19.5%; odds ratio, 1.17; 95% CI, 0.89-1.53). Mortality was consistently higher for each type of organ failure in the acute myeloid leukemia cohort versus the nononcology cohort; however, mortality did not exceed 40% unless there were four or more organ failures in the admission. Mortality for admissions requiring intensive care decreased over time for both cohorts (23.7% in 1999-2003 vs 16.4% in 2004-2010 in the acute myeloid leukemia cohort, p = 0.0367; and 7.5% in 1999-2003 vs 6.5% in 2004-2010 in the nononcology cohort, p < 0.0001). CONCLUSIONS: Pediatric patients with acute myeloid leukemia frequently required intensive care resources, with mortality rates substantially lower than previously reported. Mortality also decreased over the time studied. Pediatric acute myeloid leukemia patients with sepsis who required intensive care had a mortality comparable to children without oncologic diagnoses; however, overall mortality and mortality for each category of organ failure studied was higher for the acute myeloid leukemia cohort compared with the nononcology cohort.

Variation in hospital antibiotic prescribing practices for children with acute lymphoblastic leukemia.

Antibiotic variation among pediatric oncology patients has not been well-described. Identification of significant variability in antibiotic use within this population would warrant evaluation of its clinical impact. We conducted a retrospective cohort study of newly diagnosed patients with pediatric acute lymophoblastic leukemia (ALL) hospitalized from 1999 to 2009 in 39 freestanding US children's hospitals within the Pediatric Health Information System. Medication use data were obtained for the first 30 days from each patient's index ALL admission date. Antibiotic exposure rates were reported as antibiotic days/1000 hospital days. Unadjusted composite broad-spectrum antibiotic exposure rates varied from 577 to 1628 antibiotic days/1000 hospital days. This wide range of antibiotic exposure was unaffected by adjustment for age, gender, race and days of severe illness (adjusted range: 532-1635 days of antibiotic therapy/1000 hospital days). Antibiotic use for children with newly diagnosed ALL varies widely across children's hospitals and is not explained by demographics or illness severity.

Use of Administrative Data for the Identificationof Laboratory-Confirmed Influenza Infection: The Validity ofInfluenza-Specific ICD-9 Codes.

We used Pediatric Health Information System data and laboratory records from 3 children's hospitals to determine whether administrative data accurately identify children with laboratory-confirmed influenza. Among 23 282 inpatients, diagnosis codes for influenza detected 73% of laboratory-confirmed influenza cases, whereas <1% of patients without a diagnosis code had laboratory-confirmed influenza.

Results from a prospective, international, epidemiologic study of invasive candidiasis in children and neonates.

BACKGROUND: Candida species are the third most common cause of pediatric health care-associated bloodstream infection in the United States and Europe. To our knowledge, this report from the International Pediatric Fungal Network is the largest prospective, multicenter observational study dedicated to pediatric and neonatal invasive candidiasis. METHODS: From 2007 to 2011, we enrolled 196 pediatric and 25 neonatal patients with invasive candidiasis. RESULTS: Non-albicans Candida species predominated in pediatric (56%) and neonatal (52%) age groups, yet Candida albicans was the most common species in both groups. Successful treatment responses were observed in pediatric (76%) and neonatal patients (92%). Infection with Candida parapsilosis led to successful responses in pediatric (92%) and neonatal (100%) patients, whereas infection with Candida glabrata was associated with a lower successful outcome in pediatric patients (55%). The most commonly used primary antifungal therapies for pediatric invasive candidiasis were fluconazole (21%), liposomal amphotericin B (20%) and micafungin (18%). Outcome of pediatric invasive candidiasis was similar in response to polyenes (73%), triazoles (67%) and echinocandins (73%). The most commonly used primary antifungal therapies for neonatal invasive candidiasis were fluconazole (32%), caspofungin (24%) and liposomal amphotericin B (16%) and micafungin (8%). Outcomes of neonatal candidiasis by antifungal class again revealed similar response rates among the classes. CONCLUSIONS: We found a predominance of non-albicans Candida infection in children and similar outcomes based on antifungal class used. This international collaborative study sets the foundation for large epidemiologic studies focusing on the unique features of neonatal and pediatric candidiasis and comparative studies of therapeutic interventions in these populations.

Oseltamivir shortens hospital stays of critically ill children hospitalized with seasonal influenza: a retrospective cohort study.

BACKGROUND: Antiviral therapy reduces symptom duration and hospitalization risk among previously healthy and chronically ill children infected with seasonal influenza. The effect of oseltamivir on outcomes of hospitalized children is unknown. The primary objective of this study was to determine whether oseltamivir improves outcomes of critically ill children hospitalized with influenza. METHODS: We performed a retrospective cohort study of children with influenza infection admitted to a pediatric intensive care unit during 6 consecutive winter seasons (2001-2007). We used the Pediatric Health Information System database, which contains resource utilization data from 41 children's hospitals. We matched oseltamivir-treated patients with oseltamivir-nontreated patients by the probability of oseltamivir exposure using a propensity score we derived from patient and hospital characteristics. We subsequently compared the outcomes of critically ill children treated with oseltamivir within 24 hours of admission with propensity score matched children who were not treated with oseltamivir. RESULTS: We identified 1257 children with influenza infection, 264 of whom were treated with oseltamivir within 24 hours of hospital admission. Multivariable analysis of 252 oseltamivir-treated patients and 252 propensity score-matched untreated patients demonstrated that patients treated with oseltamivir experienced an 18% reduction in total hospital days (time ratio: 0.82, P = 0.02), whereas intensive care unit stay, in-hospital mortality, and readmission rates did not differ. CONCLUSION: For critically ill children infected with seasonal influenza, treatment with oseltamivir within 24 hours of hospitalization was associated with a shorter duration of hospital stay. Additional study is needed to determine the effect of delayed initiation of oseltamivir on clinical outcomes.

Improving surveillance for pediatric Clostridium difficile infection: derivation and validation of an accurate case-finding tool.

BACKGROUND: The incidence of Clostridium difficile infection (CDI) is increasing. Multicenter studies of CDI have been limited by the lack of valid case-finding tools. To facilitate pediatric studies of CDI, we constructed a case-finding tool using administrative data. METHODS: A cross-sectional study was performed using the Pediatric Health Information System database and microbiologic data from 4 member hospitals. Using patients with laboratory-confirmed CDI as the standard, we determined the sensitivity, specificity, positive (PPV), and negative (NPV) predictive value of an ICD-9-CM code for identifying children with laboratory-confirmed CDI. RESULTS: We identified 109 patients with laboratory-confirmed CDI and 119 patients with CDI ICD-9-CM code. The sensitivity, specificity, PPV, and NPV were 80.73%, 99.89%, 73.95%, and 99.92%, respectively, for this comparison. The addition of a billing charge for both C. difficile laboratory test and treatment medication to the ICD-9-CM code increased the specificity and PPV, but resulted in a slight decrease in the sensitivity and NPV. The use of administrative data for identifying pediatric cases of CDI was also compared with that of chart review, and was found to be a stronger surrogate for identifying cases of CDI when compared with microbiology data alone. CONCLUSIONS: These results demonstrate that the use of administrative data for CDI is a reliable and accurate method for identifying pediatric patients with CDI. The use of administrative data could facilitate the completion of larger studies due to its greater accessibility and reduced costs.

Risk factors for renal failure in pediatric patients with acute myeloid leukemia: a retrospective cohort study.

BACKGROUND: In children receiving treatment for acute myeloid leukemia (AML) there is often concern for the development of acute renal failure (ARF). Despite this, data are limited to define the incidence of ARF in this population. This study aims to evaluate the rate of ARF in AML patients and to delineate the impact of age, race, various co-morbid conditions and antimicrobial agents on the development of ARF. METHODS: A cohort of newly diagnosed AML patients from children's hospitals across the United States was identified using the Pediatric Health Information Systems database. Information regarding demographics, discharge diagnoses, pharmaceutical exposures, and hospital resource utilization were collected for each hospitalization for up to 1 year from AML diagnosis. Cox regression analysis was used to define the hazard ratios for development of ARF by demographic variables, co-morbid conditions, and exposure to various antimicrobial agents. RESULTS: Within 1 year of AML diagnosis, 135 (16.2%) patients were diagnosed with ARF. After adjustment for the presence of co-morbid conditions, the risk for ARF was greater in older patients and in black patients. Vancomycin exposure duration of greater than 48 hr and carbapenem exposure duration greater than 10 days were associated with an increased risk for ARF. CONCLUSION: ARF is a relatively common problem in children with AML. Future studies should address the different risks of ARF by age and race. Empiric therapy with potentially nephrotoxic agents did not increase the risk of nephrotoxicity. Patients on prolonged vancomycin therapy should be monitored closely for development of ARF.

Risk factors for healthcare-associated, laboratory-confirmed influenza in hospitalized pediatric patients: a case-control study.

We conducted a case-control study of 46 hospitalized pediatric patients with healthcare-associated laboratory-confirmed influenza (HA-LCI). We sought to determine the characteristics and outcomes of children with HA-LCI and to identify risk factors for HA-LCI. Although we failed to identify any differences in clinical exposures during the 3 days prior to onset of HA-LCI, multivariate analysis showed that asthma was an independent risk factor for HA-LCI (odds ratio, 3.49 [95% confidence interval, 1.25-9.75]).

Blood culture contamination rates after skin antisepsis with chlorhexidine gluconate versus povidone-iodine in a pediatric emergency department.

OBJECTIVE: To determine blood culture contamination rates after skin antisepsis with chlorhexidine, compared with povidone-iodine. DESIGN: Retrospective, quasi-experimental study. SETTING: Emergency department of a tertiary care children's hospital. PATIENTS: Children aged 2-36 months with peripheral blood culture results from February 2004 to June 2008. Control patients were children younger than 2 months with peripheral blood culture results. METHODS: Blood culture contamination rates were compared using segmented regression analysis of time-series data among 3 patient groups: (1) patients aged 2-36 months during the 26-month preintervention period, in which 10% povidone-iodine was used for skin antisepsis before blood culture; (2) patients aged 2-36 months during the 26-month postintervention period, in which 3% chlorhexidine gluconate was used; and (3) patients younger than 2 months not exposed to the chlorhexidine intervention (ie, the control group). RESULTS: Results from 11,595 eligible blood cultures were reviewed (4,942 from the preintervention group, 4,274 from the postintervention group, and 2,379 from the control group). For children aged 2-36 months, the blood culture contamination rate decreased from 24.81 to 17.19 contaminated cultures per 1,000 cultures (P < .05) after implementation of chlorhexidine. This decrease of 7.62 contaminated cultures per 1,000 cultures (95% confidence interval, -0.781 to -15.16) represented a 30% relative decrease from the preintervention period and was sustained over the entire postintervention period. No change in contamination rate was observed in the control group (P = .337). CONCLUSION: Skin antisepsis with chlorhexidine significantly reduces the blood culture contamination rate among young children, as compared with povidone-iodine.

Community acquired influenza requiring hospitalization: vaccine status is unrelated to morbidity in children with cancer.

BACKGROUND: Community acquired influenza can be severe and there are few data regarding hospitalization for children with cancer and influenza. Association between prior vaccination and infection severity has not been studied, although vaccination is standard practice. PROCEDURE: Patients with malignancy or prior stem cell transplant (SCT) were identified using a database of children with laboratory confirmed influenza (2000-2005). Other data collected included receipt of vaccine, absolute neutrophil count (ANC) and absolute lymphocyte count (ALC). These were compared with intensive care unit (ICU) stay, respiratory complications and hospital days. RESULTS: There were 39 patients with laboratory-confirmed influenza with a median age of 6.9 years. Twenty-four (62%) were on cancer therapy at time of infection and 18 (46%) had received the influenza vaccination that season. Measures of immune status included ANC at time of infection (median 1,530 cells/microl; inter-quartile range, 315, 4347), presence of graft versus host disease 2 (5%) and steroid therapy 4 (10%) patients. All had a low ALC (median 448 cells/microl; IQR 189, 861). Respiratory complications occurred in 8 (20%), ICU admissions in 4 (10%) and death in 2 (5%) patients. Median hospital stay was 2 days. All ICU admissions occurred in unvaccinated patients (P = 0.1). Vaccine status, ANC (<1,000 cells/microl vs. >1,000) and ALC (<500 cells/microl vs. >500) were not associated with length of stay or respiratory complications. CONCLUSIONS: Influenza infection can be severe in children with cancer and complications occur despite vaccination. Prospective evaluation of vaccine response is worthy of future study.

Use of antibiotics in children hospitalized with community-acquired, laboratory-confirmed influenza.

Many children with influenza are treated with antibiotics. In this report, we describe the rate and indications for antibacterial use in children hospitalized with influenza. A total of 333 of 729 (46%) patients received >2 days of treatment with antibacterial medications, of whom 36% did not have an apparent indication for therapy.

Cefepime and mortality in pediatric acute myelogenous leukemia: a retrospective cohort study.

BACKGROUND: Based on 2 meta-analyses, the Food and Drug Administration issued a communication in 2009 regarding the potential risk of death in patients treated with cefepime. Pediatric patients with acute myelogenous leukemia (AML) have frequent episodes of fever necessitating the use of antibiotics such as cefepime. We evaluated the association of cefepime and other beta-lactam antibiotic exposures with all cause in-hospital mortality in pediatric AML patients. METHODS: We performed a retrospective cohort study using the Pediatric Health Information System, an inpatient database. Exposure to cefepime, ceftazidime, antipseudomonal penicillin, and carbapenems was evaluated for each 30-day period within the first year from AML diagnosis. Cox regression analysis was used to compute hazard ratios (HR) for death adjusting for demographics, clinical variables, and clustering by hospital. The final analysis used 2 distinct time periods (0-3 months and >3-12 months) to account for variation in proportional hazards over time. RESULTS: No differences between the HRs for mortality were observed for the time period of 0 to 3 months (cefepime vs. ceftazadime: HR=1.33, 95% CI: 0.70-2.52; cefepime vs. antipseudmonal penicillin: HR=0.86, 95% CI: 0.34-2.13; and cefepime vs. carbapenems: HR=1.08, 95% CI: 0.50-2.35) or the time period of >3 to 12 months after diagnosis (cefepime vs. ceftazadime: HR=1.29, 95% CI: 0.53-3.15; cefepime vs. antipseudomonal penicillin: HR=1.08, 95% CI: 0.44-2.66; and cefepime vs. carbapenems: HR=1.03, 95% CI: 0.45-2.33). CONCLUSIONS: In this cohort of pediatric AML patients, cefepime exposure in the 30 days preceding death did not result in an increased mortality risk when compared with ceftazidime, antipseudomonal penicillins, or carbapenems.

Treatment with oseltamivir in children hospitalized with community-acquired, laboratory-confirmed influenza: review of five seasons and evaluation of an electronic reminder.

BACKGROUND: When initiated within 48 hours of the onset of symptoms, oseltamivir has been shown to reduce severity and length of influenza illness. Few studies have evaluated the use of oseltamivir in patients hospitalized with influenza. OBJECTIVE: To describe the prescribing practices for oseltamivir in children hospitalized with influenza and to evaluate a mechanism to improve the rate of appropriate prescription. DESIGN, SETTING, PATIENTS: Retrospective cohort study of 929 patients aged 21 years or younger hospitalized with community-acquired laboratory-confirmed influenza (CA-LCI) during 5 consecutive seasons (2000-2005). We examined oseltamivir eligibility, which included patients 1 year of age or older with an influenza test result available within 48 hours of symptom onset. During the 2005-2006 season, an observational trial of an electronic reminder was conducted to improve the frequency of oseltamivir prescription. MEASUREMENTS: Oseltamivir prescription. RESULTS: Of 305 patients (32.8%) eligible for treatment with oseltamivir, 49 (16.1% of those eligible) were prescribed oseltamivir during hospitalization. Prescription rates for indications consistent with the US Food and Drug Administration (FDA) approval ("on label") increased from 0% to 37.2% over 5 seasons (P < 0.0001). Prescriptions outside this recommendation ("off label") also increased over 5 seasons (P < 0.0001). Twenty-nine (5%) of 624 patients were treated with oseltamivir off label; 11 were less than 1 year of age. Initiation of a reminder had no impact on prescription (P > 0.05). CONCLUSIONS: Oseltamivir was used infrequently for children hospitalized with influenza. In addition, use inconsistent with the FDA label of oseltamivir occurs. Mechanisms are needed to improve appropriate prescription of oseltamivir.

Prolonged intravenous therapy versus early transition to oral antimicrobial therapy for acute osteomyelitis in children.

OBJECTIVES: Early transition from intravenous to oral antimicrobial therapy for acute osteomyelitis in children has been suggested as a safe and effective alternative to traditional prolonged intravenous therapy via central venous catheter, but no studies have directly compared these 2 treatment modalities. We sought to compare the effectiveness of early transition from intravenous to oral antimicrobial therapy versus prolonged intravenous antimicrobial therapy for the treatment of children with acute osteomyelitis. METHODS: We conducted a retrospective cohort study of children aged 2 months to 17 years diagnosed with acute osteomyelitis between 2000 and 2005 at 29 freestanding children's hospitals in the United States to confirm the extent of variation in the use of early transition to oral therapy. We used propensity scores to adjust for potential differences between children treated with prolonged intravenous therapy and logistic regression to model the association of outcome (treatment failure rates within 6 months of diagnosis) and difference in the mode of therapy within hospitals and across hospitals. RESULTS: Of the 1969 children who met inclusion criteria, 1021 received prolonged intravenous therapy and 948 received oral therapy. The use of prolonged intravenous therapy varied significantly across hospitals (10%-95%). The treatment failure rate was 5% (54 of 1021) in the prolonged intravenous therapy group and 4% (38 of 948) in the oral therapy group. There was no significant association between treatment failure and the mode of antimicrobial therapy. Thirty-five (3.4%) children in the prolonged intravenous therapy group were readmitted for a catheter-associated complication. CONCLUSIONS: Treatment of acute osteomyelitis with early transition to oral therapy is not associated with a higher risk of treatment failures and avoids the risks of prolonged intravenous therapy through central venous catheters.

Epidemiological features of Clostridium difficile-associated disease among inpatients at children's hospitals in the United States, 2001-2006.

OBJECTIVE: Clostridium difficile is the main cause of nosocomial and antibiotic-associated diarrhea in adults. Recently, the incidence and severity of C difficile-associated disease in adults have been increasing. Whether similar phenomena are occurring among children remains unknown. Our study describes the epidemiological features of C difficile-associated disease in hospitalized children. METHODS: We conducted a retrospective cohort study of hospitalized children with C difficile-associated disease at 22 freestanding children's hospitals in the United States, from 2001 to 2006. Cases of C difficile-associated disease were defined as a hospitalized child with a discharge code for C difficile infection, a laboratory billing charge for a C difficile toxin assay, and receipt of antimicrobial therapy for C difficile-associated disease. RESULTS: We identified 4895 patients with C difficile-associated disease. Over the study period, the annual incidence of C difficile-associated disease increased from 2.6 to 4.0 cases per 1000 admissions and from 4.4 to 6.5 cases per 10 000 patient-days. The median age of children with C difficile-associated disease was 4 years. Twenty-six percent of patients were <1 year of age. The majority of patients (67%) had underlying chronic medical conditions. The colectomy and all-cause mortality rates among children with C difficile-associated disease did not increase during the study period. CONCLUSIONS: The annual incidence of C difficile-associated disease in hospitalized children increased significantly from 2001 to 2006. However, the rates of colectomy and in-hospital death have not increased in children with C difficile-associated disease as they have among adults. The risk factors and outcomes for children with C difficile-associated disease remain to be defined in future studies.

Pediatric antifungal utilization: new drugs, new trends.

BACKGROUND: The frequency and severity of invasive fungal infections in immunocompromised patients has increased steadily over the last 2 decades. In response to the increased incidence and high mortality rates, novel antifungal agents have been developed to expand the breadth and effectiveness of treatment options available to clinicians. Despite these therapeutic advances, the impact of the availability of new antifungal agents on pediatric practice is unknown. METHODS: A retrospective cohort study was conducted using the Pediatric Health Information System database to describe the changes in pediatric antifungal therapy at 25 freestanding United States children's hospitals from 2000 to 2006. All pediatric inpatients who received a charge for one or more of the following agents were included in the analysis: conventional amphotericin B (AMB), lipid amphotericin B, fluconazole, itraconazole, voriconazole, flucytosine, caspofungin, and micafungin. Underlying conditions and fungal infection status were ascertained. RESULTS: A total of 62,842 patients received antifungal therapy, with prescriptions significantly increasing during the 7-year study period (P = 0.03). The most commonly prescribed antifungal agent was fluconazole (76%), followed by amphotericin preparations (26%). Prescription of AMB steadily decreased from 2000 to 2006 (P = 0.02). Prescription of voriconazole steadily increased during the study period and replaced AMB for the treatment of aspergillosis. The echinocandins steadily increased in prescription for treatment of fungal infections, particularly in disseminated/systemic candidiasis. CONCLUSIONS: We found that the number of pediatric inpatients requiring antifungal therapy has increased significantly and the choice of treatment has changed dramatically with the introduction of newer antifungal agents.

Methicillin-resistant Staphylococcus aureus colonization in children with atopic dermatitis.

Children with atopic dermatitis are more frequently colonized with Staphylococcus aureus than children without atopic dermatitis. However, little epidemiological data exist regarding the prevalence of methicillin-resistant S. aureus among children with atopic dermatitis. Recent studies have revealed an increasing prevalence of community-associated methicillin-resistant S. aureus among patients presenting to hospitals with serious bacterial infections, particularly those with cutaneous and soft tissue infections. As many atopic dermatitis patients are treated empirically with antibiotics for secondary skin infections, an understanding of the epidemiology of bacterial colonization and superinfection is essential for directing proper treatment in the atopic patient population. This study investigates the prevalence of risk factors for community-associated, methicillin-resistant S. aureus colonization among pediatric atopic dermatitis patients encountered at an academic pediatric dermatology clinic. An observational cross-sectional study was conducted at the Children's Hospital of Philadelphia in which 54 patients previously diagnosed with atopic dermatitis were enrolled. A detailed patient questionnaire, a complete cutaneous examination, and an evaluation of eczema severity according to the Eczema Area and Severity Index were completed at the time of enrollment. Bacterial cultures from the skin and nares were obtained to determine the frequency of colonization with either methicillin-sensitive S. aureus or methicillin-resistant S. aureus. Although most atopic dermatitis patients studied were colonized with S. aureus (43/54 [80%]), methicillin-resistant S. aureus was isolated from only seven atopic dermatitis patients (7/43 [16%]). Patients colonized with S. aureus were more likely to be male, to have been previously hospitalized, to have used a topical calcineurin inhibitor in combination with a topical steroid, and less likely to have used topical antibiotics. Bivariable analysis, however, revealed that only previous hospitalization was independently associated with an increased risk of methicillin-resistant S. aureus colonization. We observed that 80% of atopic dermatitis patients were colonized with S. aureus, and that of these patients, 16% of colonized patients were colonized with a methicillin-resistant strain. Methicillin-resistant S. aureus colonization was found to be significantly associated with previous hospitalization. Evidence also indicates that topical calcineurin inhibitors used in conjunction with topical steroids is associated with increased S. aureus colonization, while topical antibiotic use appears to decrease S. aureus colonization.