To define the etiologic agents of community acquired bacteremic syndromes (defined as septicemia, bacteremia, pneumonia and/or meningitis) in a malaria endemic setting.

Specific Aim 1. To define and characterize the etiologic agents of community acquired bacteremic syndromes in young Nigerian children Hypothesis 1a: the role of vaccine-preventable infections such as those caused by the Pneumococcus, Hib in the etiology of CABS is currently underestimated due to widespread use of non-prescription antibiotics. Hypothesis 1b: Salmonella species are the leading cause of persistent febrile illness (fever > 2 weeks) in young Nigerian children.

Specific Aim 2. To determine the role of respiratory viruses in promoting the severity of bacteremic and radio-logic pneumonia Hypothesis 2: Influenza viruses promote the severity of bacteremic pneumonia

Specific Aim 3. To determine acute host inflammatory response profiles in CABS associated with respiratory distress Hypothesis 3a: Bacteremia induces a pattern of host inflammatory profile that is distinct from those induced by malaria or respiratory viral infections. Hypothesis 3b: Clinical presentation and outcome (death, survival or prolonged hospitalization) is associated with pro-inflammatory and anti-inflammatory cytokine and chemokine dissonance.

The global burden of childhood deaths is largely borne by developing countries. Of the estimated 10 million deaths that occur in children aged less than 5 years, over 90% of these occur in developing countries and more than half of these in sub Saharan Africa. Causes of these deaths have been largely classified based on syndrome presentation, such as pneumonia, malaria, diarrhea, meningitis and septicemia and only recently have attempts been made to define the etiologic pathogens that are responsible for these deaths, an important step to implementing primary prevention. In sub Saharan Africa, malaria has for a long time been perceived as the leading cause of childhood deaths but recent studies from Kenya suggest that more deaths are in fact caused by bacteremia. However, disease burden varies by region and disease burden perceptions also vary, particularly in malaria endemic regions where clinical presentation of acute malaria, pneumonia and sepsis are often indistinguishable. The situation is compounded further in regions where malaria is present all-year round where children who are bacteremic may also have malaria parasitemia and healthy children may have asymptomatic parasitemia. These complex and overlapping features present a challenge for efficient clinical management and implementation of appropriate preventive measures.

In most developing countries, due to lack of local data, clinical management recommendations are often obtained from the World Health Organization for common conditions based on the integrated management of childhood illness approach 3. In such settings, diagnosis is based on a limited number of clinical syndromes. However, because seriously ill children often meet criteria for several clinical syndromes, each of which could be caused by several viral or bacterial pathogens, treatment is empiric and management approach is non-specific. A key component of clinical management in malaria endemic areas with high incidence of invasive bacterial disease is the initiation of antibiotic treatment and anti-malarials. While this approach provides an easy scheme, because it is not pathogen-specific there is a significant risk of promoting long-term antibiotic resistance due to over prescription of antibiotics and paradoxically, there is also the risk of increased mortality and morbidity due to inappropriate medication. An etiology-based management approach would be more effective and also provide data to explore primary preventive measures such as immunization.

Causes of bacteremia vary from reports in several sub-Saharan countries with high under-five childhood mortality. These include studies from Malawi and the Central African Republic where Non-typhoidal Salmonella (NTS) predominate and some reports from Nigeria where staphylococcus aureus has been the leading cause of bacteremia in young children.

There are only a few studies from Nigeria, the most populated country in sub-Saharan Africa, and these studies have not identified S. pneumoniae nor Hib as an important pathogen in the etiology of community acquired bacteremic illness or invasive bacterial disease in young children. The disparity in prevalence of different bacterial pathogens across countries is likely due to the different study entry criteria, age groups, bacterial culture methods and the prevalence of antibiotic exposure in the population. A high rate of antibiotic use is likely to select for resistant bacterial strains in such populations. It is notable that of all the studies that identified S. pneumoniae as a leading cause of bacteremia were studies that utilized the automated Bactec culture system. Thus to determine the true burden of vaccine-preventable bacterial disease in such a setting would warrant the use of culture-based and molecular microbiology techniques that would overcome the problem of poor sensitivity secondary to partially treated infections.

Of the childhood illnesses commonly associated with bacteremia, pneumonia is the leading cause of death, with Streptococcus pneumonia making a significant contribution to these deaths in reports from the Gambia but limited data from the West African sub region. Thus this study will focus on providing further understanding of the etiologic agents of bacteremia and pneumonia. Determination of the cause of pneumonia remains a challenge, since bacterial blood culture, the current "gold standard" is useful only when when pneumonia is associated with bacteremia and is positive only in about 10-15% of cases. The vast majority of pneumonias are not associated with bacteremia and to ascertain cause in these cases warrants the use of lung aspirates, a very specific but invasive procedure used only in research settings.

When a clinical diagnosis of pneumonia is made using the WHO guideline, which is very sensitive but poorly specific, only a fraction of these have any radio-logic abnormality and a variable smaller fraction still, have a confluent alveolar density that culminates in a lobar consolidation. When the lung consolidation is readily assessable to the chest wall, the more sensitive and specific approach for the etiologic diagnosis is the use of a percutaneous lung aspirate which is considered invasive and only used in research settings. This approach improved etiologic diagnosis from 18% with the use of blood culture alone to 52%. Since this procedure is invasive and reserved for only selected cases, the causes of most non-bacteremic pneumonia remains poorly defined.

At the population level, another approach for determining the burden of disease-attributable to an etiologic agent in a population is the use of an efficacious vaccine. For example, in The Gambia, the pneumococcal conjugate vaccine efficacy trial was associated with 37% reduction in radio-logic pneumonia and 16% reduction in all-cause childhood mortality; an important testament to the under-estimation of the true disease burden attributable to the pneumococcus and the poor sensitivity of the current diagnostic methods for the detection of invasive pneumococcal disease in young children. While the development and application of these new generation vaccines are a major advancement with a huge potential for saving several thousands of lives, they also provide a sensitive tool for probing the burden of vaccine-attributable preventable disease in a population but are very expensive.

The determination of the causes of pneumonia in childhood presents a further challenge as other agents such as viruses have often been identified from respiratory secretions but their role in the patho-genesis of severe pneumonia in developing countries is less well defined.

Understanding the host immune responses to infection is critical to providing insight into treatment modalities, since these responses and not necessarily the pathogen, determine outcome. This observation is further supported by the observation that prompt diagnosis and initiation of antibiotic treatment does not always avert death as recently reported from a study from the Gambia, where delayed mortality was reported within 28 days of hospital discharge after appropriate in-patient care at a hospital. The reason for these delayed deaths is unknown but prolonged immunologic dissonance (imbalance between pro-inflammatory and anti-inflammatory cytokines) is a likely contributing factor. This deserves further investigation since if these delayed deaths are due to immunologic dissonance, these deaths may be averted with the use of steroids or other immunomodulators, in addition to antibiotics. Characterization of host inflammatory responses among children with community acquired pneumonia may facilitate understanding of host inflammatory response, provide a prognostic tool for clinicians or identify potential therapeutic targets.

Children presenting with an acute febrile illness, with respiratory distress in a malaria endemic region may have severe malaria, bacteremia, sepsis and/or pneumonia. A number of studies have explored acute inflammatory markers such as C-reactive protein and procalcitonin but have reported significant overlap in serum concentration of these proteins in both conditions to justify a diagnostic use for in such settings. While the clinical presentation with respiratory distress in pneumonia and malaria may be similar, understanding the molecular basis for this clinical sign could provide valuable information for diagnosis and treatment.

In summary, community acquired bacteremic syndromes in malaria endemic regions of Africa presents a diagnostic conundrum for physicians and pose an important clinical management dilemma that warrants further evaluation. This study proposes to take advantage of new molecular and genomic technology in addressing this challenge with the ultimate goal of defining the etiologic agents of community acquired bacteremic syndromes in Nigerian children and generating information on host inflammatory responses that would be useful in developing primary and secondary preventative measures.

Inclusion Criteria:

• Children 0-14 years of age presenting with an acute febrile illness with a temperature of more than or equal to 38 degrees C or hypothermia (temperature less than 36 degrees C) with any of the following symptoms; convulsion, respiratory distress, prostration or non-traumatic circulatory collapse to the pediatric outpatient or emergency pediatric units of participating Nigerian institutions.

Exclusion Criteria:

• No consent obtained/withdrawal of consent, known congenital heart disease, bronchial asthma, or recent hospitalizations within the past 2 weeks.

Pediatric subjects 0-14 years of age