Enhanced passive screening and diagnosis for gambiense human African trypanosomiasis in north-western Uganda - Moving towards elimination

Abstract

Introduction: The incidence of gambiense human African trypanosomiasis (gHAT) in Uganda has been declining, from 198 cases in 2008, to only 20 in 2012. Interruption of transmission of the disease by early diagnosis and treatment is core to the control and eventual elimination of gHAT. Until recently, the format of available screening tests had restricted screening and diagnosis to central health facilities (passive screening). We describe a novel strategy that is contributing to elimination of gHAT in Uganda through expansion of passive screening to the entire population at risk.

Methodology / principal findings: In this strategy, patients who are clinically suspected of having gHAT at primary health facilities are screened using a rapid diagnostic test (RDT), followed by parasitological confirmation at strategically located microscopy centres. For patients who are positive with the RDT and negative by microscopy, blood samples undergo further testing using loop-mediated isothermal amplification (LAMP), a molecular test that detects parasite DNA. LAMP positive patients are considered strong suspects, and are re-evaluated by microscopy. Location and upgrading of facilities to perform microscopy and LAMP was informed by results of georeferencing and characterization of all public healthcare facilities in the 7 gHAT endemic districts in Uganda. Three facilities were upgraded to perform RDTs, microscopy and LAMP, 9 to perform RDTs and microscopy, and 200 to screen patients with RDTs. This reduced the distance that a sick person must travel to be screened for gHAT to a median distance of 2.5km compared to 23km previously. In this strategy, 9 gHAT cases were diagnosed in 2014, and 4 in 2015.

Conclusions: This enhanced passive screening strategy for gHAT has enabled full coverage of the population at risk, and is being replicated in other gHAT endemic countries. The improvement in case detection is making elimination of the disease in Uganda an imminent possibility.

Fig 1. The gHAT endemic region in north-western Uganda, showing the 4 health facilities that performed confirmatory diagnosis of the disease prior to 2013, and the cases detected between 2008 and 2012.

The gHAT region is geographically separated from the rHAT region. The geographic layers were obtained from CC-BY License compatible sources: GADM (http://www.gadm.org), RCMRD geoportal (http://geoportal.rcmrd.org/) and GeoNames database (http://www.geonames.org/).

Fig 2. Barplot of gHAT cases reported in Uganda from 2008 to 2012, by district in which the case was diagnosed.

Cases from the districts of Koboko and Maracha were referred to Arua for diagnosis and treatment.

Fig 3. The T. b. gambiense human African trypanosomiasis (gHAT) diagnostic workflow implemented in Uganda.

Gland puncture (GP); Acridine Orange–Fluorescence Microscopy; capillary tube centrifugation (CTC or mHCT).

Fig 4. Map showing health facilities in the gHAT focus of north-western Uganda that were characterised and equipped to perform RDTs, microscopy and LAMP in the initial phase between May 2013 and February 2014.

HCIV = level four health centre; HCIII = level 3 health centre. See details on the sources of the geographic data in Fig 1.

Fig 5. Cumulative distribution plots (as lines) of the distance of the population to a screening test.

The calculation of cumulative distribution is such that for each distance location on the x-axis, the corresponding percentage on the y-axis are within that distance. This is illustrated by district (A) prior to expansion of screening and (B) after expansion of screening. Accordingly, considering all districts, 50% of the population were within 23km of a screening facility prior to this programme, compared to 2.5km following the programme.

Fig 6. Density plot of the Euclidean distance to gHAT diagnostics (microscopy) for the diagnostic coverage prior to this programme and under this programme.

The top lines are density plots and the lines below represent the median and central 75% of the data. Density plots were calculated using a Gaussian kernel with a bandwidth of 3.17km. Note that this was calculated separately for the facilities on the west and the east of River Nile, assuming the Nile to be a barrier that would not be crossed.

Fig 7. Usage of HAT RDTs by month of the programme.

A red ‘X’ represents a gHAT case that was confirmed. The decline in numbers after August 2014 corresponds with the reduction in the number of health facilities using RDTs in Amuru and Adjumani districts, and the increase in September 2015 corresponds to the expansion to private clinics and re-engagement of refugee facilities.

Fig 8. Usage of HAT RDTs by facility between August 2013 and December 2015.

Note that one gHAT case was from South Sudan and is outside the map extent. The case locations correspond to the home village of the cases. See details on the sources of the geographical data in Fig 1.

Fig 9

Barplot of total HAT RDT usage to December 2015 (A) and monthly rate of RDT usage (B), by type of facility. Parasit’ = parasitology.