Arbovirus circulation, epidemiology and spatiotemporal distribution in Uganda

Abstract

Background: Arboviruses are endemic in Uganda; however, little is known about their epidemiology, seasonality and spatiotemporal distribution. Our study sought to provide information on arbovirus outbreaks from acute clinical presentations.

Methods: Immunoglobulin M (IgM) and confirmatory Plaque Reduction Neutralisation Test (PRNT) results for arbovirus diagnosis of samples collected from patients attending sentinel sites from 2016-19 were analysed retrospectively. Demographic data were analysed with SaTScan and SPSS software to determine the epidemiology and spatiotemporal distribution of arboviruses.

Results: Arbovirus activity peaked consistently during March-May rainy seasons. Overall, arbovirus seroprevalence was 9.5%. Of 137 IgM positives, 52.6% were confirmed by PRNT, of which 73.6% cases were observed in central Uganda with Yellow Fever Virus had the highest prevalence (27.8%). The 5-14 age group were four times more likely to be infected with an arbovirus p=0.003, 4.1 (95% CI 1.3-12.3). Significant arboviral activity was observed among outdoor workers(p=0.05) . Spatiotemporal analysis indicated arboviral activity in 23 of the 85 districts analysed..

Interpretation: Our study shows that arbovirus activity peaks during the March-May rainy season and highlights the need for YFV mass vaccination to reduce the clinical burden of arboviruses transmitted within the region.

Figure 1

Sample collection and arboviruses distribution by region: A. Extract of the map of Uganda showing sample collection by region, further divided into districts. The orange represents the north, green for the west, red for the central, blue for the east, and grey shows areas with no sample collection. The number of samples collected from each region is represented as a percentage of the total collection. B. Regions with arboviruses confirmed by PRNT. The map divisions represent districts and are coloured to represent geographical regions. The uncoloured districts represent areas where no arbovirus has been confirmed. The maps are not drawn to scale.

Figure 2

Seasonality of arboviruses in Uganda: A. Number of arbovirus positives by month in Uganda from March 2016 to November 2019. The alphaviruses (red) and flaviviruses (blue) incidences peaked between March and May every year. The March-May 2016 and 2019 peaks coincide with the YFV outbreak in Masaka and Koboko districts. B. PRNT confirmed YFV (red) and other arboviruses (blue) observed in Uganda from 2016 to 2019. Four distinct YFV outbreaks were observed, with the first occurring in the central region March-May 2016 in the central and western regions. The second was in September-October 2017 (Masaka district) and the third in April 2018 (Wakiso district), while the fourth spread across the central, north, and west regions from March to September 2019.

Figure 3

Spatio-temporal distribution of arboviruses in Uganda in clock-wise direction. I. Spatial analysis of arboviruses showing their distribution in Uganda. Arboviral incidences were observed in 23 districts (black and red dots). Red circles represent clusters in districts from which more than two arboviruses were detected, indicated as loci A- F. Districts A-F are: A. Buikwe, B. Masaka, C. Sembabule, D. Ntungamo, E. Rubanda, and F. Rukungiri. Abbreviation: DRC- Democratic Republic of Congo. II. Space-time scan of arbovirus activity in Uganda. The blue, black and red spots show areas where arbovirus activity was observed. Two clusters E (Masaka) and F (Sembabule) had high arboviral rates and occurred from March-May 2016 in Masaka district. Cluster F overlapped with E but occurred from September to February 2017. The blue dots represent districts with low arboviral rates. A- Buikwe, B- Mukono, C- Kampala and D – Wakiso. III. Spatial Analysis of YFV activity. The blue, black and red spots show areas where YFV activity was observed. Cluster C (in Masaka) represent significant YFV activity, whilst clusters A (Kampala) and B (Wakiso) showed low activity. All but the eastern region showed YFV activity. IV. Spatial distribution of alphaviruses May 2016 - March 2019. The black, grey and red spots show the 16 locations where alphaviruses were detected. The red spot shows the Hoima district cluster with a high rate of alphavirus activity, whilst the blue spots represent areas of low virus rates in Kampala and Wakiso.