Evaluation of a “Stamp Out Sleeping Sickness” campaign in Uganda to control human African Trypanosomiasis (2004 - 2009)

Evaluation of a “Stamp Out Sleeping Sickness” campaign in Uganda to control human African Trypanosomiasis (2004 - 2009)

Herbert Mukiibi^1,2,&, Charles Waiswa³, Peter Waiswa², Enock Matovu², John David Kabasa², Susan Olet^1,2, Margaret Loy Khaitsa⁴

¹Department of Veterinary and Microbiological Sciences, North Dakota State University, Fargo, ND, USA, ²College of Veterinary Medicine, Animal Resources and Biosecurity, Makerere University, P.O BOX 7062, Kampala, Uganda (East Africa), ³Coordinating Office for Control of Trypanosomiasis in Uganda (COCTU), The Uganda, Trypanosomiasis Control Council (UTCC), Ministry of Agriculture, Animal Industry and Fisheries, Kampala, Uganda, ⁴Department of Pathobiology and Population Medicine, College of Veterinary Medicine, Mississippi State University, Mississippi State, MS, USA, ⁵Department of Pathobiology and Population Medicine, College of Veterinary Medicine, Mississippi State University, Mississippi State, MS, USA

^&Corresponding author
Herbert Mukiibi, College of Veterinary Medicine, Animal Resources and Biosecurity, Makerere University, P.O Box 7062, Kampala, Uganda

Introduction    

Human African Trypanosomiasis (HAT), or sleeping sickness, is a vector-borne disease caused by protozoa belonging to the Genus Trypanosoma. HAT is caused by infection with one of two parasites: Trypanosoma brucei rhodesiense (T.b. rhodesiense) or Trypanosoma brucei gambiense (T.b. gambiense). These organisms are extra-cellular protozoan parasites that are transmitted by insect vectors in the genus Glossina (tsetse flies). HAT is restricted to sub-Saharan Africa, in the range of the tsetse vector (Figure 1). These trypanosomes are transmitted to humans by the bite of an infected tsetse fly [1-3]. The tsetse fly acquires its infection by feeding on animals or humans harboring the parasites. The parasites can be distinguished through molecular methods [3], but not parasitologically; the geographic range of the parasites has been a key component of the differential diagnosis of HAT, as T.b. gambiense occurs in West and Central Africa, and T.b. rhodesiense occurs only in East Africa, albeit concerns that an overlap of their ranges in northern Uganda may have occurred. T.b. rhodesiense is a zoonosis with a number of wildlife and domestic animal species known to act as reservoirs [3]. Where wildlife is not abundant, domestic species, particularly cattle, are the main reservoir [3], with livestock demography driving outbreaks [3]. T.b. gambiense is generally not considered zoonotic - it can be isolated from animal hosts [3], but large-scale control campaigns targeting only the human reservoir (active screening and treatment of human cases) are able to locally eliminate transmission [3], and thus from an epidemiological perspective, the presence of animal hosts is unlikely to mean they serve as a reservoir of infection for humans [3]. The transmission of HAT occurs primarily in rural areas, in areas at the furthest extremities of the formal health system, creating particular problems for patients to access health care and for control campaigns to have an effective outreach [3]. Cattle are the most common reservoirs of HAT in Uganda [4], and that 75% of the trypanosomes in cattle in the infected areas affect humans [5, 6]. In sub-Saharan Africa, HAT takes two forms, acute or chronic, depending on the parasite involved. Uganda is unique because it is the only country where both forms exist [7]. The two HAT forms (chronic and acute) are caused by two genetically different but morphologically indistinguishable trypanosome strains [7,8]. T.b. gambiense causes the chronic form (gambiensis) of HAT while T.b. rhodesiense causes the acute form [4]. The two strains can be differentiated in the laboratory by polymerase chain reaction [3, 9]. Both trypanosome strains are transmitted by tsetse flies. Humans are the carriers of the causative trypanosomes for gambiensis while cattle are the major reservoir for rhodesiensis infections [4]. Control of gambiensis requires active surveillance because humans can harbor the trypanosomes for more than twenty years without showing clinical signs [9].

In Uganda, gambiensis occurred in the north-west, near South Sudan while rhodesiensis was confined to south-eastern Uganda (Figure 1). In 1980s, there was a lapse in HAT control due civil unrest in northern Uganda followed by movement of humans and cattle, and subsequent restocking when the region became peaceful [10]. Acute HAT spread north as a result, posing a real risk for the two forms of the disease to overlap, and thereby complicating the diagnosis and treatment for HAT [11]. Stamp Out Sleeping Sickness (SOS) campaign, a Public-Private Partnership was launched in Uganda on 11th October 2006 to keep the two forms of HAT from merging. A detailed description of SOS activities has been published elsewhere [11, 12]. Briefly, the project treated over 200,000 cattle in the target districts to clear cattle of trypanosomes so that when the tsetse flies fed on them, they would not acquire the infection. A Restricted Application Protocol (RAP) treatment was then used to spray deltamethrin to the lower part of cattle bodies transforming them into “live baits” for tsetse flies, in order to suppress the overall population and prevent possible re-infection. The local communities were sensitized about HAT to try and dispel the myth that HAT was caused by witchcraft, and to increase their awareness of the disease in order for them to report cases and to continue seeking treatment [11, 13]. These measures were expected to reduce the vector population for HAT, reduce the population of trypanosomes in cattle thereby reducing infectivity of tsetse flies and improve attitudes of people so they could seek treatment. All these activities would ultimately reduce HAT incidence in people. Over the years in Uganda, many international and national tsetse control programs have been carried out, but despite all these interventions, HAT outbreaks continued to occur [9]. At global platforms, HAT has been identified as one of the climate-sensitive diseases likely to spread globally beyond Africa, where it is endemic [14]. The United Nations Forum for Climate Change Convention (UNFCCC, 1998) affirmed that global warming is real, and climate change affects distribution of vectors for infectious diseases [15, 16]. The 2010 World Bank report estimated the cost of climate-sensitive diseases and their impact on health at 9% of the gross domestic product in some countries [17]. All these factors underscore the need to control HAT. This study evaluated the impact of SOS (phase I) on incidence of HAT: 1) in Kaberamaido district, from 2004 to 2009, 2) reported by sub-county in relation percent of cattle treated, 3) in relation to distance traveled from residence to HAT reporting center, and 4) in relation to weather.

Methods     

Study area: SOS phase I campaign covered five districts in northern Uganda (Kaberamaido, Dokolo, Apac, Amolatar and Lira) in what´s referred to as Teso and Lango sub regions of Uganda. Teso agro-ecological system (the study area) received bimodal rainfall with annual rainfall of 1,000 to 1,500 mm [18]. The dry season (December to March) was longer. The region had sandy-loamy soils of medium to low fertility with short grasslands of moist Combetrum/Butyrospermum and grass savannas ideal for grazing. Mixed agriculture (crops and livestock) was practiced and cultivation by oxen was the main agricultural technology. Livestock were kept extensively in areas that were tsetse-fly free. Only one health center, Lwala hospital, in Kaberamaido district served as the sentinel center for HAT surveillance in humans in Teso sub-region. Also, of 529 total HAT cases reported at Lwala hospital during the study period (2004-2010), 360 (68 %) were from Kaberamaido. Additionally, Kaberamaido and Dokolo had recently been infected with sleeping sickness [7] therefore were a good study area for impact of SOS compared to HAT endemic districts that had confounding factors. Therefore, the study area was confined to Kaberamaido district in northeastern Uganda (coordinates 1°, 47' N and 33° 09' E). The Uganda Bureau of Statistics estimated a total Kaberamaido human [19], and cattle [20] population of 195,400, and 70,000, respectively. The major economic activities were cattle keeping (for beef production) and farming where cattle also provide necessary drought power. In 2002 to 2004 Kaberamaido was under political insurgencies and people were displaced [10]. The United Nations High Commission for Refugees [21], estimated that internally displaced persons (IDP) camps in Lango region were established between 2002 and 2004, and war in northern Uganda displaced more than 1.8 million Ugandans [21]. By 2005 there were 242 camps hosting 1,842,500 IDPs who had been displaced because of the Lord´s Resistance Army and Ugandan Government forces war. About 92 percent of 466,000 IDPs in Lango had returned to their homes in 2005. During political insurgency, Lango was covered with bush that provided a good habitat for tsetse flies. People returning IDPs camps were immunologically naive to HAT, and prior to SOS, HAT control programs were absent. Also, these districts were restocked with new cattle due to return of previously migrated populations. Kaberamaido district experienced a bimodal rain fall (March to May, peaking during April/May, and July-October, peaking in September). A brief dry spell occurred in June, and the longer dry spell occurred from November to February/early March. Average rainfall was 1225 mm per year. The highest temperatures up to 31°C occurred during the long dry season while the lower temperatures occurred during the rains [22].

Data source: individual patient records (patient´s name, age, sex, village, parish subcounty and district of residence, date of admission and discharge) of confirmed HAT cases in Kaberamaido district from 2004 to 2010 were collected from Lwala hospital. Names of patients were replaced with serial numbers to protect patient confidentiality. Data on SOS intervention were obtained from Professor Charles Waiswa at the College of Veterinary Medicine, Animal Resources and Biosecurity (CoVAB), Makerere University, Uganda. The distances in kilometers from residence to Lwala hospital were determined using an established “ruler and thread” method on Google Maps [23]. Weather data from Soroti district were used because Kaberamaido was a new district formed in 2000 out of Soroti district. Data on total number of cattle in Kabaramaido and weather data (from the Meteorology Department, Ministry of Water and Environment) were obtained from the Uganda Bureau of Statistics abstract [20]. The distribution of cattle by parish in Kaberamaido district for the year 2006, were obtained from the SOS team (led by Professor Charles Waiswa). These proportions were used to calculate the proportion of cattle by parish based on the Cattle census conducted in 2007 by Uganda Bureau of Statistics [20]. The proportion of cattle treated by parish (treatment coverage) was computed as number of cattle treated in each parish divided by the total number of cattle in the parish) and the incidence of HAT cases by parish was computed by dividing the number of HAT cases reported in each parish by the total number of people in that parish. For purposes of data analysis, parishes that did not report any case of HAT were given a HAT incidence of 0.5 cases/100,000 people.

Data analysis: data were entered in Excel and exported into Access and SAS version 9.2 for further analysis. Both descriptive and inferential analyses were performed. Geographical information System (GIS) Arc-Info was used to display Kaberamaido district HAT cumulative incidence, HAT cases by subcounty, cattle population, and percent of cattle treated by subcounty. For data analysis purposes, 2004-2005, 2006, and 2007-2010 were considered as pre-intervention, intervention, and post-intervention periods. Using SAS, Relative Risk (RR) of HAT incidence (cases per 100,000 people) by parish and by year relative to the intervention year and associated confidence intervals were estimated. The Attributable Fractions (AF) of each variable to HAT incidence was also computed. All analyses were performed at a 5% level of significance. AF and RR were used to quantify the HAT risk associated with each variable. The AF represents the proportion of HAT cases saved when comparing the reference year (2006) to the population exposed during the year in question. A high AF indicates a high protective effect in the intervention year when compared to the year of interest. SOS effect on HAT incidence was evaluated by comparing HAT incidence in 2006 with that of pre-intervention (2004-2005) and post-intervention (2007-2010) using ‘proc procedure´ of SAS. A Generalized Estimating Equation (GEE) assuming a Binomial distribution was used to evaluate effect of percent of cattle treated by subcounty and the effect of distance from parish of residence to HAT reporting center on HAT incidence. Effect of weather on HAT incidence, was assessed using a linear association between average temperature and rainfall. This research was cleared by the North Dakota State University Institution Review Board (IRB Determination AG12016).

Results     

Descriptive statistics: over all, there were 362 cases 345 of which (95.3%) reported an age and 4.7% (17) did not. Among the 345 HAT cases with age data, the age range was 0.3 to 85 years, median 25 years and mean 29 years (standard deviation =18.42). Of 362 total HAT cases, 352 (97.2%) reported gender and 10 (2.8%) did not. Of 352 HAT cases with gender data, 174 (49.4%) were female and 178 (50.6%) male with age range of 1-73 years, median 27.5, mean 29.2, standard deviation 17.6 for females and a range of 0.3 to 85 years, median 24.5, mean 28.7, standard deviation 19.3 for males.

Effect of SOS on the incidence of HAT in Kaberamaido District, Uganda: Table 1 summarizes pairwise comparisons of incidence of HAT cases per 100,000 people during pre-intervention (2004-2005) and post-intervention (2007-2009) with that of SOS intervention (2006). HAT incidence during pre-intervention (2004 and 2005) and post-intervention year 2009 were statistically significantly different (p < 0.05) from HAT incidence during intervention (2006). The risk of reporting HAT cases in 2004 and 2005 was 2.1 and 3.0 times greater, respectively, than in 2006 (p < 0.05). HAT incidence in 2007 and 2008 were similar to that of 2006 (p > 0.05) while HAT incidence in 2009 was twice as high as that of 2006 (p=0.0006). Data for 2010 were incomplete so were omitted from the analysis. Attributable fraction (% incidence of HAT saved due to SOS) during pre- and post-intervention followed a similar trend to Relative Risk (Table 1). The Attributable Fraction represents the number of people (potential cases) saved from contracting HAT by the intervention. A high percentage of Attributable Fraction shows a high protective effect in the intervention year when compared to the year of interest. Figure 2 summarizes the seasonal distribution (average monthly incidence of HAT cases) during pre-intervention (2004-2005), intervention (2006) and post-intervention (2007-2009). The year 2010 had only January to May data so was omitted from the analysis. Overall, HAT incidence during pre-intervention was higher than that of 2006 (intervention year). Also, HAT incidence was lower during post-intervention (2007-2009) compared to both intervention (2006) & pre-intervention (2004-2005) (Figure 2). In general, the incidence of HAT was lower during the months of April to July and higher during the months of November to January.

Effect of cattle treatment coverage (%) on the incidence of HAT cases: the number of cattle treated in 2006 in Kaberamaido was 27,990 out of total population of 71,741 (39.0%). Figure 3 summarizes cattle distribution and percent of cattle treated by subcounty. Percentage of cattle treated by subcounty varied from 24% (2,419/10,022) for Bululu (lowest) to 55.6% (5,998/10,779) for Ochero (highest), with a median coverage of 42% in Anyara and Alwa (Figure 3, Figure 4 and Figure 5). During pre-intervention (2004-2005) the average HAT incidence by subcounty varied from 1, 329 to 721 HAT case per 100,000 people in Kobulubulu, Alwa and Otuboi subcounties, respectively (Figure 3). During post intervention (2007-2009) the HAT incidence varied from 1 to 328 and 413 cases per 100,000 people in Ochero, Otuboi and Alwa subcounties, respectively (Figure 3). During intervention (2006), the average HAT incidence ranged from 1 HAT case per 100,000 people in 3 subcounties (Bululu, Kobulubulu and Ochero) to 126 and 329 HAT cases/100,000 people in Otuboi and Alwa subcounties, respectively. The average HAT incidence in Anyara, Kalaki and Kaberamaido subcounties was 21, 51 and 93 cases per 100,000 people, respectively. The impact of percent of cattle treated by subcounty on HAT cases reported was assessed using Generalized Estimating Equation (GEE). Adjusting for the proportion of cattle treated, time over the study period was associated with a decrease in HAT cases as implied by the negative parameter estimate (-0.0424) although the difference in number of cases over the years was not statistically significant (p = 0.3025). Overall, time (in years) and proportion of cattle treated were not significantly associated with HAT cases (p = 0.1758 and 0.5242, respectively). Adjusting for time, the proportion of cattle treated was not statistically significantly associated (p=0.4639) with HAT cases. A positive parameter estimate (1.2749) implied that HAT cases were more where more cattle were treated. The cumulative incidence of HAT cases in Kaberamaido district over the study period (2004-2009) is summarized by subcounty and cattle population (Figure 4) and by percent of cattle treated (Figure 5).

Effect of distance from residence to HAT reporting Lwala medical center on the incidence of HAT cases: the distance travelled from residence of cases to medical center (Lwala hospital) where HAT cases were reported ranged from 3.2 Km in Lwala parish in Otuboi subcounty to 73.5 Km in Kagaa parish, Ochero subcounty. The mean distance travelled was 24.7 Km. The majority of cases (95%) lived within a range of 19.2 km and 30.2 Km from Lwala hospital. The impact of distance (Kilometers) from residence to HAT reporting medical center on the incidence of HAT cases was assessed using Generalized Estimating Equation (GEE) and the empirical standard error estimates were evaluated. Distance was statistically significantly associated (p =0.0002) with a decrease in HAT cases as implied by the negative parameter estimate (-0.0256, 95% CI -0.0389, -0.0123). The greater the distance from Lwala hospital, the fewer the number of cases reported. Overall distance was found to be statistically significantly associated (p=0.0049) with the number of HAT cases reported.

Effect of weather in study area on incidence of HAT: the average monthly temperature during the study period (2004-2009) was fairly constant and ranged from 23.7 in August to 26.7 in February. The average rainfall during the same period ranged from a minimum of 25 mm in December to a high of 154 mm and 162 mm in the months of August and April, respectively. The overall rainfall distribution was bimodal with two peaks in the months of April and August. The linear correlation between average incidence of HAT cases by month during pre-intervention period (2004-2005) and average rainfall (2004-2009) was inverse with a moderate fit (r = -0.55). The HAT incidence cases were lower during high rainfall months and vice versa. The linear correlation between average temperature and HAT incidence cases was not assessed because there was no variability in average temperature over the study period.

Discussion     

During the study period, HAT incidence was higher in the most active age group (average age groups affected = 25-29 years). The % of males affected was slightly higher than that of females possibly due to different gender roles; males were more likely to come into contact with tsetse flies when grazing cattle in bushy areas. SOS significantly reduced HAT incidence during post-intervention years (2007 and 2008). This was expected as SOS procedures were expected to lower infection rates of cattle with trypanosomes and also reduce the density of tsetse flies, thereby reducing HAT incidence. However, HAT incidence during post intervention (2007 to 2008) remained similar to that of the intervention year (2006) for a period of 2 years before it increased significantly in 2009. This indicates that SOS procedures need to be complemented with community based sustainability programs for a longer period than the free public good intervention period. If not, free interventions should be administering at least every 2 years to keep HAT incidence as low as the intervention year [24]. The 2 years period could be matching with the time needed for tsetse flies population to grow given the life cycle [1, 2]. This period also matches the cyclical pattern observed in the population of trypanosomes in the blood due to antigenic variation and evasion of the host immune system [25]. HAT incidence in 2004 and 2005 (pre-intervention) was higher than 2006 (intervention year) by two and three times, respectively. However, for the immediate years post-intervention (2007 and 2008), there was not much protection (AF = 21% and AF = 9%, respectively). This implies that implementing SOS activities in 2007 or 2008 was not cost effective. Minimal activities such as spraying of cattle during this time could be protective. Except for Ochero subcounty, the subcounties that reported the highest percentage of cattle treated with trypanocides (Otuboi, Alwa and Anyara) also reported most HAT cases. The fewer HAT cases in subcounties with less cattle treated could be a reflection of fewer hosts in those subcounties for tsetse flies to feed on and subsequently transmit HAT. Since Kaberamaido had just been habited by people returning from IDP camps, possibly some subcounties had not been inhabited yet for various reasons, including presence of tsetse flies. Traditionally, people from Kaberamaido or Teso region are agro-pastoralists who keep livestock and also practice farming. Also, it is possible that the high level of awareness created by and the importance attached to SOS activities could explain why places with high cattle treatment also had many HAT cases. Interestingly, Kalaki subcounty had the lowest cattle population but was among the subcounties with the highest cattle treatment indicating that other factors apart from cattle population explained the % cattle treatment. For instance, access due to good roads could have explained a higher % treatment of cattle in northern Kaberamaido compared to southern subcounties closer to Lake Kyoga. Ochero subcounty had a high cattle population and high cattle treatment coverage but the lowest HAT cases reported; possibly due to the long distance from the reporting center, and its location close to the lake, where the vegetation was marshy and not suitable for tsetse flies to inhabit. This observation warrants further investigation. Distance from parish of residence significantly influenced HAT incidence reported at Lwala medical center; the greater the distance the fewer the number of cases reported. A similar observation was made by other scientists who reported that geographical levels of accessibility of a surveillance center affected the number of HAT cases reported [26]; this was explained by the remoteness of the rural areas that are always found at the far end of the medical centers. The road network in the study area supports this observation. This finding calls for improvement in infrastructure for surveillance of HAT in the study area, including facilitation of transportation for HAT cases to medical reporting centers. A seasonal pattern in HAT incidence that correlated with rains was observed. possibly due to: 1) the life cycle of the tsetse fly which is affected by rainfall and temperature [1,2]; 2) human-animal interface at watering points during the dry season; or 3) bush fires that leave only shrubs around gardens to serve as tseste fly habitats and swampy areas to serve as watering points. For instance, Abalang swamp located in Alwa subcounty supplies water constantly to humans and animals but also contributes to high HAT cases.

Conclusion     

SOS significantly reduced HAT incidence from pre-intervention (2004-2005), and maintained lower HAT incidence during the intervention (2006), and two years (2007-2008) post-intervention. However, HAT incidence significantly increased in 2009 underscoring the need to implement SOS activities at least every 2 years or provide sustainability interventions in place. The association between % of cattle treated and HAT cases reported by subcounty was not statistically significant while distance travelled from residence to medical reporting center was significantly associated with HAT incidence - the longer the distance the fewer the cases reported. Fewer HAT cases were reported during the rainy months. These data contribute valuable information to HAT control.

Competing interests     

The authors declare no competing interest.

Authors´ contributions     

The primary author (Mukiibi) was a beneficiary of CIMTRADZ, and this research project was part of his MS-International Infectious Disease Management degree requirement. Dr. Susan Olet helped with data analyses, and all authors contributed to writing drafts of the article, reviewed several drafts, and provided intellectual content. All authors approved the version to be published.

Acknowledgments     

The authors acknowledge SOS team (Anne Holm Rannaleet and Martin Mitchell) for provision and validation of data on SOS activities conducted in Kaberamaido district, Anthony Walekhwa Wamono for assistance with GIS mapping and Abel Ekiri for final edits to the manuscript. Funding for this project was provided by United States Agency for International Development special research grant (Grant # FAR18479).

Table and figures     

Table 1: relative risk (RR) of incidence of HAT cases (number of cases per 100,000 people) by year (2003-2004 or pre-intervention period and 2007-2010 or post-intervention period) relative to the intervention year (2006)

Figure 1: map of Africa showing distribution of Human African trypanosomiasis cases by country: modified from: Eric M. Fèvre, Beatrix V. Wissmann, Susan C. Welburn, Pascal Lutumba (2008)

Figure 2: average incidence of HAT cases (number of cases per 100,000 people) by month during pre-intervention period (2004-2005), intervention period (2006) and post-intervention period (2007-2009)

Figure 3: percent of cattle treated by Subcounty and average incidence of HAT cases (number of cases per 100,000 people) before intervention (2004-2005) and post-intervention period (2007-2009)

Figure 4: distribution of cattle population and HAT cumulative incidence by Subcounty in Kaberamaido District of Uganda (2004-2009)

Figure 5: distribution of percent of cattle treated and HAT cumulative incidence by Subcounty in Kaberamaido District of Uganda (2004-2009)

References     

1. Buxton P. Natural History of Tsetse Flies: An Account of the Biology of the Genus Glossina (Diptera). London UK. H.K. Lewis & Co. 1955. Google Scholar


2. Horsfall WR. Arthropods and human diseases. Medical entomology. New York. Ronald press Co. 1962. Google Scholar


3. Fèvre EM, Wissmann BV, Welburn SC, Lutumba P. The Burden of Human African Trypanosomiasis. Neglected Tropical Diseases. 2008; 2(12): e333. PubMed | Google Scholar


4. Waiswa C, Olaho-Mukani W, Katunguka-Rwakishaya E. Domestic animals as reservoirs for sleeping sickness in three endemic foci in south-eastern Uganda. Annals of Tropical Medicine and Parasitology. 2003; 97(2): 149-55. PubMed | Google Scholar


5. Welburn SC, Coleman PG, Maudlin I, Fèvre EM, Odiit M, Eisler MC. Crisis, what crisis? Control of Rhodesian sleeping sickness. Trends in parasitology. 2006; 22(3): 123-8. PubMed | Google Scholar


6. Welburn SC, Fèvre EM, Coleman PG, Odiit M, Maudlin I. Sleeping sickness: a tale of two diseases. Trends Parasitol. 2001; 17(1): 19-24. PubMed | Google Scholar


7. Fèvre EM, Coleman PG, Welburn SC, Maudlin I. Reanalyzing the 1900-1920 Sleeping sickness epidemic in Uganda. Emerg Infect Dis. 2004; 10(4): 567-73. PubMed | Google Scholar


8. Batchelor NA, Atkinson PM, Gething PW, Picozzi K, Fèvre EM. Spatial predictions of Rhodesian Human African Trypanosomiasis prevalence in Kaberamaido and Dokolo, Two Newly Affected Districts of Uganda. Neglected Tropical Diseases. 2009; 3(12): 563. PubMed | Google Scholar


9. Wastling SL, Picozzi K, Wamboga C, Wissmann B, Amongi-Acup C, Wardrop NA, Stobard JR, Kakembo A, Welburn SC. Latent Trypanosoma brucei gambiense foci in Uganda: a silent epidemic in children and adults? Parasitology. 2001; 138(12): 1480-7. PubMed | Google Scholar


10. Nathan N, Tatay M, Piola P, Lake S, Brown V. High mortality of displaced populations of northern Uganda. Lancet. 2004; 363(9418): 1402. PubMed | Google Scholar


11. Waiswa C, Kabasa JD. Experiences with an in-training community service model in the control of zoonotic sleeping sickness in Uganda. Journal of veterinary medical education. 2010; 37(3): 276-281. PubMed | Google Scholar


12. Kabasa JD. Public-private partnership works to stamp out. sleeping sickness in Uganda. Trends Parasitol. 2007; 23(5): 191-2. PubMed | Google Scholar


13. Morton J. The innovation trajectory of sleeping sickness control in Uganda: research knowledge in its context. Research into Use Discussion Paper. Accessed on 2010 Oct; 8.


14. Moore S, Shrestha S, Tomlinson KW, Vuong H. Predicting the effect of climate change on African trypanosomiasis: integrating epidemiology with parasite and vector biology. Journal of the Royal Society Interface. 2012 May 7; 9(70): 817-30. PubMed | Google Scholar


15. United Nations Framework Convention on Climate Change. Climate change: impacts, vulnerabilities and adaptation in developing countries. 2007. Accessed 10 August 2012.


16. Evengard B, Sauerborn R. Climate change influences infectious diseases both in the Arctic and the tropics: joining the dots. Global Health Action. 2009 Nov 11; 2. PubMed | Google Scholar


17. World Bank. Development and climate change. World development report. 2010. Accessed 4 July 2012.


18. Waiswa C, Rannalette A. Entrepreneurship Initiatives in the control of sleeping sickness: Experiences of the stamp out sleeping sickness in Uganda. Journal of Small Business and Entrepreneurship. 2010; 23(4): 555-464. PubMed | Google Scholar


19. Ministry of Natural Resources. State of the environment for Uganda Ministry of Natural Resources. National Environment Information Center, Government of the Republic of Uganda. 1995. Google Scholar


20. Uganda Bureau of Statistics. The National Live Stock Census. A summary Report of the National Live Stock Census 2008. 2009. Accessed 15 August 2012.


21. United Nations High Commission for Refugees (UNHCR). Uganda's IDP camps start to close as peace takes hold, September 2007. 2007. Accessed 30 December 2012.


22. Opio OC. Diagnosis of the relationship between weather variables and occurrences of measles in Soroti and Serere districts. 2011. Accessed 24 July 2012.


23. Leveson DJ. Using map scales. 2002. Accessed 31 December 2012.


24. Odiit M, Coleman PG, McDermott JJ, Fevre EM, Welburn SC, Woolhouse ME. Spatial and temporal risk factors for the early detection of Trypanosoma brucei rhodesiense sleeping sickness patients in Tororo and Busia districts, Uganda. Transactions of the Royal Society of Tropical Medicine and Hygiene. 2004; 98(10): 569-576. PubMed | Google Scholar


25. Pays E, Vanhollebeke B. Human Innate Immunity against African Trypanosomes. Current opinion in immunology. 2009; 21(5): 493-498. PubMed | Google Scholar


26. Uganda population and housing census, years: 1980, 1991, 2002. Kampala: Government of Uganda. 2002. Google Scholar