Estimating the malaria attributable fraction of fever in cohort studies through a before and after comparison of impact: Nagongera, Tororo, Uganda, 2011-2019

AO_SCPLOWBSTRACTC_SCPLOWO_ST_ABSBackgroundC_ST_ABSMalaria is an important cause of fever across much of sub-Saharan Africa and other places where Plasmodium falciparum infection is highly prevalent. Here, we estimate the fraction of fever that is attributable to malaria using data from two studies in Nagongera, Tororo, Uganda that followed cohorts of children and adults longitudinally from 2011-2019. The study included three years before and five years after indoor residual spraying (IRS) sharply reduced mosquito populations, malaria exposure, and the prevalence of malaria infection. MethodsWe estimate the malaria attributable fraction of fever (MAFF) by directly quantifying and comparing fever before and after IRS started. We compared subjective (i.e., self-reported) and objective fever during scheduled and unscheduled visits (i.e., to seek care) in young children (under 5 years old), older children (aged 5-10 years), and adults (over 18 years old). ResultsWe estimated that there were 78-90 total days per person, per year (pppy) with subjective fever during the pre-IRS baseline in young children; 52-58 in older children; and 38-46 days in adults. After IRS, sub-clinical fever declined to 5-6 days pppy with fever in young children to around 3 in older children, and around 1 in adults: a 94% reduction in young children, 95% in older children, and 99% in adults. Reductions in total fever prevalence for care seeking (during unscheduled visits) declined by around 50% in young children, 65% in older children, and 80% in adults. In the before vs. after comparison, malaria accounted for 88% of objective fever during scheduled visits in young children, 75% in older children, and 91% in adults. Total fever declined by 80-85% in children and 90-93% in adults. During care seeking, malaria accounted for around 44% of objective fever in young children, but no meaningful differences were observed at other ages. These patterns were accompanied by changes in care seeking and total fever. Over the first few months of the study, care seeking rates increased in all groups, but then care seeking rates started a decline that continued until the study ended. By the end of the study, care seeking rates had declined by more than 75% overall compared with the start. ConclusionsThe fraction attributed to malaria differed by age and context. In this study population with good access to care, fever was rare at the end of the study in the absence of malaria. Based on the before vs. after comparison, malaria was directly or indirectly responsible for most subjective fever in the clinical setting, and it was also the dominant cause of objective fever. Surprisingly, a large fraction of subjective fever that occurred before IRS, during both scheduled and unscheduled visits, occurred in people who tested negative for malaria. The study draws attention to the importance of sub-clinical disease as a contributor to the burden of health in malaria endemic settings. FundingThe PRISM studies (U19AI089674) were funded by the National Institutes of Allergies and Infectious Diseases (NIAID) as part of the International Centers of Excellence for Malaria Research (ICEMR).