Insufficient data to answer the study question

It is possible to distinguish between pediatric oncology patients who are at high or low risk for serious infection during periods of fever and treatment related neutropenia based on clinical parameters. Patients with low risk can be safely treated as outpatients primarily using oral antibiotics. It is possible to improve methods of risk stratification through the addition of genomic and proteomic factors.

Outpatient management of patients considered to be at low risk for serious bacterial infection has been explored using risk stratification schema based on clinical parameters. First, patients will be stratified based on a clinical risk stratification schema. Patients stratified to the low risk group will be randomized between treatment using standard inpatient intravenous antibiotic therapy or outpatient antibiotic therapy using primarily an oral regimen. Second, an evaluation of proteins important to the innate immune system will be performed to provide a molecular characterization of episodes based on etiology. Third, single nucleotide polymorphisms in genes important for innate immunity will be evaluated to determine effect of each on infection risk during treatment induced neutropenia. Finally, we will develop a bank of both plasma and DNA specimens correlated with clinical outcomes for future use.

Inclusion Criteria:

1. Any pediatric patient age <21 years with an oncology diagnosis who is undergoing therapy and is expected to have treatment related neutropenia.

Exclusion Criteria:

1. Any patient who has previously undergone autologous or allogeneic bone marrow transplant will be excluded from study enrollment. If a patient is expected to undergo autologous or allogeneic bone marrow transplant as part of therapy at some point after enrollment in the study he/she will be removed from the study at the start of their bone marrow transplant.
2. Any patient with a documented allergy to Levofloxacin or any other fluoroquinolone will be excluded.
3. Patients with a known pregnancy will be excluded.
4. Any patient with an underlying chronic musculoskeletal condition (ie Juvenile rheumatoid arthritis, Systemic lupus erythematosis etc) which may make evaluation for joint toxicity related to quinolone treatment difficult.

Because inpatient management and outpatient management use different treatment methods, blinding is not possible.