Research

Changing etiologies of childhood acute febrile illness: implications for management in an era of declining malaria

Alastair Matheson | 2016

Advisor: Joel D. Kaufman

Advisor: Judd L. Walson

Research Area(s): Epidemiologic Methods, Global Health, Infectious Diseases, Social Determinants of Health

FULL TEXT


Background: Acute febrile illness (AFI) is a common syndrome among children presenting to a healthcare facility in resource-limited settings and a substantial contributor to childhood mortality. In sub-Saharan Africa, malaria has historically been a leading cause of childhood AFI but malaria incidence and mortality have declined dramatically over the past decade. At the same time, other emerging infections may increasingly be replacing malaria as causes of AFI. For example, dengue virus has spread rapidly throughout the world and is now established in Africa. It is unclear how shifts in pathogen distribution and transmission impact AFI incidence and how these changes impact the management practices of healthcare workers (HCWs). This dissertation sought to describe recent time trends of malaria and non-malarial AFI (Aim 1), to evaluate the contribution of dengue to AFI in western Kenya (Aim 2), and determine whether seasonal variations in malaria incidence influence antibiotic prescribing for children with AFI (Aim 3). Methods: All three aims were addressed using data from a large cohort in western Kenya that has been followed for over eight years. In particular, data from Lwak Mission Hospital formed the basis for these analyses. For Aim 1, quasi-Poisson regression was used to calculate the annual decline in visits from children with malaria, non-malarial AFI, and overall AFI. For Aim 2, serum specimens were obtained from patients of all ages with AFI who had either no known cause of their fever or a diagnosis for which dengue fever is often mistaken. Specimens were tested for the presence of dengue virus using reverse-transcription polymerase chain reaction (PCR). In addition, specimens were also tested for evidence of a recent infection, indicated by the presence of IgM anti-dengue virus antibodies in an enzyme-linked immunosorbent assay. For Aim 3, we used logistic regression, adjusting for a child’s malaria smear result, to assess whether seasonal variations in malaria were associated with HCWs’ adherence to Integrated Management of Childhood Illness (IMCI) guidelines on antibiotic prescriptions among children with AFI. Results: For Aims 1 and 3, approximately 11,400 childhood AFI visits to LMH between January 1, 2009 and December 31, 2014 informed the analyses. Visits from both non-malarial AFI and malaria declined substantially over this period (9.47% and 8.55% per year, respectively). Declines in malaria over this time were not statistically significant, likely due to a plateauing of malaria visits in the latter part of the study. However, among children with malaria there were opposing trends; visits where the child had malaria parasitemia alone declined significantly by 16.15% per year whereas the number of visits from children with malaria parasitemia and another clinical diagnosis of an AFI etiology remained largely the same. For Aim 2, 615 serum specimens were obtained from visits during two rainy periods (September–December 2011 and March–July 2013). There were no positive results for dengue infection using either PCR or IgM anti-dengue virus testing. For Aim 3, HCW management of AFI appeared to be associated with perceived risk of malaria. Overall, HCWs adhered to IMCI guidelines for antibiotics in 7,853 (69.0%) of included childhood AFI visits. However, periods of high malaria incidence were associated with fewer prescriptions of antibiotics in children who met criteria for antibiotic treatment and periods of low malaria incidence were associated with greater use of antibiotics in children who did not meet criteria for antibiotic treatment. Conclusion: The clinical and epidemiological pictures of childhood AFI in sub-Saharan Africa are becoming more complicated, even as malaria and non-malarial AFI both decline. While malaria continues to be a major driver of AFI seen at this outpatient setting in rural western Kenya, an increasing proportion of febrile children with malaria receive additional diagnoses of other causes of fever including upper respiratory tract infections and pneumonia. Dengue virus does not appear to be present in this part of western Kenya at this time but the risk of spread to the region from the coastal region remains. Seasonal variation in malaria was associated with non-adherence to IMCI guidelines for antibiotic prescriptions. Ongoing comprehensive surveillance of etiologies of AFI and investigations into the impact of changing patterns of disease burden on the management practices of HCWs’ may lead to opportunities for improved prevention and management of childhood AFI.