Prevalence of Pneumococcal Bacteremia in Children with Sickle Cell Disease Hospitalized for Management of a Febrile Illness

Children with sickle cell disease are at a high risk of death from pneumococcal bacteremia due to functional asplenia. The common practice for management of any fever in a child with sickle cell disease is in-patient antibiotic therapy for presumed bacteremia, until the blood cultures have remained negative for 48 hours. We hypothesize that interventions such as penicillin prophylaxis, use of polyvalent pneumococcal vaccine at 2 years of age in addition to the primary immunization series with pneumococcal conjugate vaccine 13 have decreased the prevalence of pneumococcal bacteremia in children with sickle cell disease. Outpatient management of an uncomplicated febrile illness may be a safe alternative.

We conducted a retrospective chart review of children with sickle cell disease hospitalized with a fever at Children's and Women's Hospital between January 2006 and June 2012. We collected demographical parameters, temperature, white blood cell count and C-reactive protein level at the time of presentation as possible predictors of true bacteremia. The continuous variables were compared between the two groups (positive and negative blood cultures) using independent t-test. We also recorded the time to positivity of blood cultures, pathogens identified in positive blood cultures, serotypes and sensitivities of pneumococcal bacteria.

A total of 133 patients, aged 2 months to 18 years, met the inclusion criteria with 456 hospital admissions. The prevalence of culture proven bacteremia was 3.7% with only two (0.44%) cases of pneumococcal bacteremia. We have not had any patient with pneumococcal bacteremia reported since the start of pneumococcal conjugate vaccine 13, in May 2010. Both the patients with pneumococcal bacteremia were older than five years and had discontinued taking penicillin prophylaxis. The first patient was fully vaccinated with pneumococcal conjugate vaccine 7 but had not received polyvalent vaccine. She contracted a non-vaccine serotype (23B) that was susceptible to penicillin, ceftriaxone, and vancomycin. The second patient was partially vaccinated and acquired a serotype (23F) that is included in pneumococcal conjugate vaccine 7. The bacterium was resistant to penicillin but sensitive to ceftriaxone and vancomycin. In the analysis of predictors of bacteremia, the average C-reactive protein level was found to be 6.3mg/dL in patients with positive blood cultures and 4.5mg/dL in patients with negative blood cultures (p=0.02). The mean age of patients with positive blood cultures was 11 years compared to 7.4 years in patients with negative blood cultures (p=0.03). The differences in body temperature and white blood cell count between the two groups were not statistically significant. The average time to positivity measured from the time of collection was 25.4 hours.

Penicillin prophylaxis and complete immunization, in accordance with the centers for disease control and prevention guidelines are important for patients with sickle cell disease. Patients with an uncomplicated febrile illness may not necessarily need to be hospitalized to continue antibiotics. We support an outpatient care model proposed by Williams et al, comprising of a single 24-hour dose of ceftriaxone followed by oral cefixime for 5 days and follow-up telephone calls by a nurse. Good patient education may help ensure compliance with outpatient care. Outpatient management of an uncomplicated febrile illness will be more cost-effective and beneficial to the patients and families. With this approach, however, patients may not be fully protected against ceftriaxone resistant pneumococccal infections. Bacterial resistance patterns in the community should be brought under consideration when implementing such a model.

The prevalence of pneumococcal bacteremia in children with sickle cell disease has decreased in the past few years. In our analyses, we found older age and an elevated C-reactive protein level to be statistically significant predictors of a positive blood culture.

No relevant conflicts of interest to declare.