School of Public Health

Patricia B. Pavlinac

Potential etiologies of pediatric acute diarrhea & assessment of current treatment guidelines in rural western Kenya


Diarrhea is the second leading cause of death in children under 5 years of age and diarrhea case fatality rates are highest in sub-Saharan Africa where the HIV burden is concentrated and where laboratory infrastructure is lacking. Children who are HIV-infected or whose mother is infected with HIV but who are not infected (HIV-exposed uninfected [HEU]) may be more susceptible to specific high-risk enteric pathogens such as those associated with death and other lethal complications. Current syndromic algorithms for presumptive diarrhea diagnosis and treatment may not capture these high risk pathogens and may require updating to reflect the most recent evidence in diarrhea causes and consequences. 


Nested within a surveillance study of enteric and bloodstream infections in Western Kenya, we conducted sub-studies to determine the association among HIV infection, HIV-exposure, and pathogen prevalences (chapter 1) and to determine the diagnostic performance of the World Health Organization’s Integrated Management of Childhood Illness (IMCI) syndromic guidelines at classifying suspected Shigella spp. against the gold standard of stool culture (chapter 2).


Risk of infection with specific pathogens differed depending on HIV infection and exposure. Over 1000 children with acute diarrhea were enrolled and at least one potential pathogen (bacterial or parasitic) was identified in 45.8% of the children. HIV-infected children were more likely to be infected with enteropathogenic Escherichia coli  (EPEC) (10.7% vs. 3.6%, prevalence ratio [PR]: 2.95, P=0.008). HEU children were more likely to have Cryptosporidium species (spp.) identified than HIV-unexposed children (9.9% vs. 3.3%, PR: 2.98, P=0.003). Associations were independent of measured confounders. WHO IMCI guidelines did not perform well in correctly classifying children who may benefit from antibiotic therapy. Among the 51 children with Shigella-positive stools in whom we ascertained history of bloody stool, only 7 had a reported history of bloody stool meeting the classification for antibiotic therapy by IMCI (sensitivity: 13.7% [95% CI: 5.7% 26.6%] and PPV: 9.9% [95%CI: 4.0%-19.0%]). Among the 1,006 children without microbiologic isolation of Shigella, 942 did not have a history of bloody stools (specificity: 93.6% [95%CI: 91.8-95.0%] and NPV: 95.5% [95%CI: 94.1% - 96.7%]). 


EPEC and Cryptosporidium were more frequently detected in HIV-infected and HIV-exposed children, respectively. These data may explain recently reported increased mortality attributed to these two pathogens. Current international diarrhea guidelines do not consider suspected EPEC or Cryptosporidium infections as indication for treatment or intense follow-up. Guidelines such as the IMCI do however indicate treatment for suspected Shigella infections, but the currently used suspected Shigella criteria identified few children infected with Shigella. In the absence of laboratory diagnosis, these children may have missed the opportunity to receive potentially life saving, and transmission-reducing, antibiotic treatment. Therefore, reductions in childhood deaths due to diarrhea in sub-Saharan Africa may require increased attention to HIV infection in children and their caregivers as well as possible updates to current syndromic treatment and management guidelines.