Clinical malaria case definition and malaria attributable fraction in the highlands of western Kenya

Background: In African highland areas where endemicity of malaria varies greatly according to altitude and topography, parasitaemia accompanied by fever may not be sufficient to define an episode of clinical malaria in endemic areas. To evaluate the effectiveness of malaria interventions, age-specific case definitions of clinical malaria needs to be determined. Cases of clinical malaria through active case surveillance were quantified in a highland area in Kenya and defined clinical malaria for different age groups. Methods: A cohort of over 1,800 participants from all age groups was selected randomly from over 350 houses in 10 villages stratified by topography and followed for two-and-a-half years. Participants were visited every two weeks and screened for clinical malaria, defined as an individual with malaria-related symptoms (fever [axillary temperature >= 37.5 degrees C], chills, severe malaise, headache or vomiting) at the time of examination or 1-2 days prior to the examination in the presence of a Plasmodium falciparum positive blood smear. Individuals in the same cohort were screened for asymptomatic malaria infection during the low and high malaria transmission seasons. Parasite densities and temperature were used to define clinical malaria by age in the population. The proportion of fevers attributable to malaria was calculated using logistic regression models. Results: Incidence of clinical malaria was highest in valley bottom population (5.0% cases per 1,000 population per year) compared to mid-hill (2.2% cases per 1,000 population per year) and up-hill (1.1% cases per 1,000 population per year) populations. The optimum cut-off parasite densities through the determination of the sensitivity and specificity showed that in children less than five years of age, 500 parasites per mu l of blood could be used to define the malaria attributable fever cases for this age group. In children between the ages of 5-14, a parasite density of 1,000 parasites per mu l of blood could be used to define the malaria attributable fever cases. For individuals older than 14 years, the cut-off parasite density was 3,000 parasites per mu l of blood. Conclusion: Clinical malaria case definitions are affected by age and endemicity, which needs to be taken into consideration during evaluation of interventions.