Demographic characteristics

Table 2 displays the demographic characteristics of the respondents for the health behaviour survey. Follow-up rates from the original study population were 55% amongst Alur (68 individuals) and 59% amongst Bagungu (91 individuals), and from the second sample, 47% amongst Alur (41 individuals) and 53% amongst Bagungu (46 individuals). Thirty-seven individuals from the second sample lived in a region of Booma known locally as Wantembo; this area is remote from the main village and approximately 2 km away from the lakeshore. Since previous systematic water contact observations suggest that individuals living in Wantembo have very little contact with the water (Pinot de Moira et al., 2007), these individuals were excluded from the current study; this resulted in a final sample size of 209, consisting of 78 (37%) Alur and 131 (63%) Bagungu. Comparison of the demographic characteristics of the respondents for the health behaviour survey with those of the whole Booma community suggest that the study sample was representative of the Booma community, given the age and duration of residency restrictions that were applied in the original sampling frame (results not shown).

Table 2.

Demographic characteristics of individuals participating in the health behaviour survey

^a If born elsewhere.

^b 95% CI=95% confidence interval.

^c Received or still receiving.

^d 100% of housewives were female.

^e One (3%) Alur female and two (4%) Bagungu women reported ‘fisherman’ as their occupation.

The health behaviour survey questioned participants on whether they had received some formal education, and if so, to what level. Although this is a fairly crude means of measuring level of education, and fails to convey any information as to the quality of literacy or education, there were clear, significant differences between the two ethnic groups. Significantly more Bagungu than Alur had received formal education (χ²=6.3, p=0.01), and, although not significant, amongst those receiving an education, more Bagungu than Alur had proceeded to secondary or higher education. Free primary education for all school-aged children was made available in Uganda in 1997 (UNICEF/Government of Uganda, 1997), but this has only affected younger members of the Booma study. Stratifying data according to whether individuals were children or adults in 1998 (the year in which the Booma study began) reveals quite an interesting picture. The percentage of children with formal education was not significantly different in the two ethnic groups: 27 (90%) Alur vs 37 (97%) Bagungu children (p=0.20). Amongst the adults, however, 27 (56%) of Alur vs 73 (78%) of Bagungu (p=0.01) had received formal education. More interestingly, this disparity in education status between adults of the two ethnic groups was most significant amongst women. Only four (19%) of Alur women questioned had received a formal education, compared with 21 (57%) of Bagungu women (p=0.01). Although more Bagungu men than Alur men had received an education, this difference was non-significant (24 [86%] educated Alur vs 52 [93%] educated Bagungu, p=0.29).

It is not clear whether these disparities in receiving a formal education are specific to Booma, or whether the Alur population of Uganda have a generally lower prevalence of formal education. It has been reported that the Alur as a whole tend to be viewed as ‘primitive’ (Leopold, Reference Leopold2006), although this may be because they are generally disliked by other ethnic groups (Leopold, Reference Leopold2005). Nevertheless, other sources have commented on the low standards of literacy in the Alur (Southall, Reference Southall1953). It should also be noted that the Alur participating in the current study are not recent migrants: as detailed above, the study was restricted to individuals that were either born in the village or who had been resident for 10 or more years.

Also displayed in Table 2 are the occupations of the two ethnic groups. The groups shared similar occupations; the major occupation of males was fishing (35 [80%] Alur and 52 [69%] of Bagungu men were fishermen), whereas among females, the major occupations were housewife (fourteen [41%] Alur females and 21 [38%] Bagungu females), fish trader (eight [24%] Alur females and nine [16%] Bagungu females] or not working (five [15%] Alur females and ten [18%] Bagungu females).

Fig. 1.

Age-adjusted proportion of all reported contacts involving each of the various activities, amongst females (a) and males (b), by ethnic group. Activities were classified as domestic (fetching water, or washing clothes or utensils), bathing (washing body), recreational (playing, swimming or socializing; only playing was reported), economic (buying/selling fish, cleaning fish, fishing or boat-related) or other (fording, hyacinth picking, cutting reeds).

All Bagungu men participating in the FGDs were fishermen. One Alur man was employed to pull in boats from the lake; all other Alur men were also fishermen, but one man supplemented his income by cutting poles for construction. Bagungu women were a mixture of housewives, fish traders and grass cutters. Alur women were employed in fishing and trading fish.

Reported water contact behaviour

During FGDs, all groups emphasized how essential the lake was to their survival. Many groups described the lake as their ‘garden’: one Alur man stated, ‘our whole life entirely depends upon it’, and similarly one Mugungu woman explained, ‘without the lake we cannot survive’.

Female activities

The activities discussed during the focus groups by Alur and Bagungu women strongly mirrored those observed during the systematic water contact observations (Pinot de Moira et al., 2007). Alur women reported fetching water from the lake, catching small fish for bait, buying and cleaning fish, washing clothes and utensils, bathing and collecting snails to sell to chicken feed manufacturers. In contrast, there was no mention of washing clothes and utensils at the lake amongst Bagungu women, and only one woman said she might bathe at the lake if she were feeling tired. The major activities mentioned by Bagungu women were buying fish and fetching water; cutting grass and cleaning/processing fish were also mentioned.

As is to be expected, there was evidence that individual water contact behaviour had altered since the original observations were made in 1998. Focus groups revealed changing behaviour with age, marriage, biomedical knowledge of schistosomiasis and changes to the lake environment. In addition, the establishment of the Murchison Falls National Park, and the subsequent introduction of laws prohibiting the hunting of wild game, had led to an increase in lake-dwelling wild animals such as crocodiles and hippopotami, and an increased ‘fear’ of the lake. In light of this, it was deemed inappropriate to compare observed and reported contacts directly and instead overall patterns were compared. Figure 1a displays age-adjusted proportion of contacts for each type of activity reported, relative to total reported frequency of contact, amongst females, by ethnic group. In contrast to observed water contact behaviour (Pinot de Moira et al., 2007), boat-related activities were not reported, perhaps because they were not perceived as being ‘contact with the lake’. ‘Fording’ was also not reported, despite being frequently observed during water contact observations, especially amongst Bagungu females. This may be because it is not a distinct activity or because it is no longer possible: an increase in the lake level was alluded to by the Bagungu participants during FGDs and was also visibly evident.

Differences between Alur and Bagungu females' contacts clearly mirrored those apparent from observations (Pinot de Moira et al., 2007) and focus groups (Figure 1a and Table 3). Reported frequency of ‘washing body’, ‘cleaning fish’ and ‘washing clothes and utensils’ was greater in Alur women compared with Bagungu women (z=3.8, p=0.0001; z=2.4, p=0.02; z=3.4, p=0.001, respectively), and the frequency of ‘fetching water’ and ‘buying and selling fish’ was not significantly different between females of these two ethnic groups (z=0.51, p=0.61 and z=−0.2, p=0.88, respectively).

Table 3.

Geometric mean (95% confidence interval) reported frequency of water contact activities, by sex and ethnic group

^a Geometric mean (GM) reported monthly number of contacts; 1 was added to each measurement prior to calculating GMs.

For males, excludes two males either not having contact with the lake or unable to recall frequency of contact.

When probed about possible differences in behaviour between Alur and Bagungu women in FGDs, many agreed there were differences. Most Alur women related these differences to the fishing of ‘Ragogi’ (small silver fish), which was practised by many Alur women, and even Alur children. Unfortunately, the fishing of small fish by Alur women was not revealed by questionnaire data: although 36% of Alur women over the age of 14 were observed fishing during water contact observations, only one female (representing 3% of Alur women questioned) reported fishing. This discrepancy in the Alur women's observed and reported fishing may indicate that they view their ‘fishing’ activity as a mere support to the ‘real’ fishing of the men. The involvement of Alur women in fishing is in contrast to Bagungu women, for whom only 5% of women were observed fishing, and one female (representing 2% questioned) reported fishing.

One Alur woman reported: ‘It differs because Alur ladies fish a lot in the lake and spend most of their time in the lake’. Alur women also said they traded more in fish at the lake than Bagungu females, and also used the lake more for transport. Bagungu women were rather more disparaging of Alur women. They also suggested fishing was only conducted by Alur women and not Bagungu women, but also that Alur women wash their clothes in the lake and bathe in the lake, ‘even if they’re educated'. One Bagungu woman claimed that, ‘Alur women even bathe in swamps’. It was suggested Alur women did not care about the risk of bilharzia, that they did not fear the lake, but that Bagungu women did. One woman said: ‘Alur women don't care about bilharzia – they go back and also allow their children to go’. The Bagungu women also claimed that the Alur women would bring their children to the lake when bathing or fishing: ‘They don't care about the young ones – when they go to wash their clothes they also bring their children and place them in the water. It's the same when they're fishing Ragogi.’ Further derogatory remarks included: ‘If the child has a distended abdomen and you tell them the child has bilharzia, they will say “no, the child has been charmed”’ and, ‘They even drink the lake water directly without boiling and also give this to their children’.

It was also suggested by Alur and Bagungu men that Alur women participated more in fishing activities than Bagungu women. A possible explanation for their increased involvement was provided by one Alur woman: ‘We Alur ladies buy clothes for ourselves and it forces us to spend most of our time in the lake, while Bagungu ladies get clothes from their husbands.’ When probed further on their economic dependency on their husband, there did appear a greater dependency amongst Bagungu women, which was also apparent from male focus groups. Although there was a general consensus amongst Bagungu that women did assist their husbands with income generation, there were exceptions; in one female Bagungu focus group it was suggested that Bagungu women sometimes refused to assist, as it was the husband's responsibility to provide for the family. In contrast, many Alur women said they were the main economic providers for the family; a tendency for Alur males to spend money on alcohol rather than the family was also hinted. Other Alur women said that they worked in partnership with their husband to provide for the family. Therefore it seems whereas Bagungu women are either completely or partly reliant on their husbands for economic status, Alur women are either partly reliant on their husbands or economically independent.

Male contacts

Activities mentioned by Alur men during focus groups were primarily fishing and bathing, but cutting reeds and pulling in boats were also mentioned. The predominant activity reported by Bagungu men was fishing; bathing and processing fish were also reported.

Figure 1b displays age-adjusted proportion of contacts for each type of activity reported in the health behaviour survey, relative to total reported frequency of contact, amongst males, by ethnic group. Table 3 tabulates GM reported frequency of water contact activities amongst males, by ethnic group.

Similarly to females, ‘launching boat’ was the only boat-related activity reported by males, and this was under-reported compared to observed boat-related activities (Pinot de Moira et al., 2007); it is likely, however, that many boat-related activities were reported as ‘fishing’. Recreational activities were under-reported, whereas ‘fishing’ was reported more frequently than it was observed amongst Alur males. Again, despite these disparities between observed and reported contacts, similar differences in water contact activities between Alur and Bagungu males emerged from reported and observational data. Reported frequency of ‘bathing’ was greater (z=3.5, p=0.001) and reported frequency of ‘buying and selling fish’ lower (z=−2.3, p=0.02) amongst Alur males than amongst Bagungu males, whilst the reported frequency of ‘washing clothes and utensils’ was not significantly different for Alur and Bagungu males (z=1.2, p=0.22). However, in contrast to observational data, although the frequency of ‘fetching water’ was greater in Bagungu males than Alur males, this difference was not significant (z=−0.3, p=0.78); reported frequency of ‘cleaning fish’ was also significantly greater amongst Alur males than amongst Bagungu males, despite observational data suggesting no difference (z=2.2, p=0.03). More importantly, there was no significant difference in reported frequency of fishing by Alur and Bagungu males (z=0.4, p=0.66).

The health behaviour survey asked fishermen to estimate the duration of time they spent out on the water during a typical fishing expedition; in Fig. 2, box plots compare estimated duration of fishing reported by Alur and Bagungu males. Contrary to observational data presented previously (Pinot de Moira et al., 2007), reported time spent fishing was greater for Alur males compared with Bagungu males (GM duration 609.6 min per day [95% CI: 504.7, 736.3] vs 314.3 min per day [95% CI: 228.5, 432.2] for Alur and Bagungu males respectively; z=2.8, p=0.01). This indicates that many Alur males' fishing-related contacts may have gone unobserved during water contact observations, as has been previously suggested (Pinot de Moira et al., 2007). This became more apparent when men were probed about their relative fishing practices during FGDs. It appeared that Alur men would spend longer fishing, sometimes fishing for many days at a time; however, it was also apparent that many of the Alur males' fishing activities were conducted in deeper waters. More Bagungu men than Alur men reported fishing in the lagoon area, whereas, although some Alur men did report using seine nets, which are used in shallower waters to trap small fish, relatively more reported fishing in deeper waters.

Fig. 2.

Box plots displaying reported duration of fishing activities by ethnic group. Plots display minutes spent fishing per day, which is a function of reported frequency of fishing and duration spent fishing per expedition.

Fishing at night and in deeper waters are both likely to pose less of a risk of infection than fishing during the day or nearer to shore (due to the circadian pattern of cercarial shedding by Biomphalaria snails and evidence that the deeper fishing waters surrounding Booma are not inhabited by Biomphalaria snails; Francis Kazibwe, personal communication). This may explain why levels of re-infection were similar among Alur and Bagungu males (Table 1) and highlights the fact that, for a truly holistic approach to control and research, environmental as well as behavioural elements need to be considered. This is particularly the case in Booma; from malacological studies it is known that there are two Biomphalaria species that transmit S. mansoni at Booma, Lake Albert: B. sudanica and B. stanleyi (Kabatereine et al., Reference Kabatereine, Kariuki, Kazibwe, Kemijumbi, Sturrock, Ornbjerg and Madsen1999; Kazibwe et al., Reference Kazibwe, Makanga, Rubaire-Akiiki, Ouma, Kariuki and Kabatereine2006). These two snails have been shown to be allopatric in their distributions (Kabatereine et al., Reference Kabatereine, Kariuki, Kazibwe, Kemijumbi, Sturrock, Ornbjerg and Madsen1999; Kazibwe et al., Reference Kazibwe, Makanga, Rubaire-Akiiki, Ouma, Kariuki and Kabatereine2006). Biomphalaria stanleyi is also known to be a more permissive host (Kazibwe et al., Reference Kazibwe, Makanga, Rubaire-Akiiki, Ouma, Kariuki and Kabatereine2010). This is important because the Alur and Bagungu live at geographically distinct ends of Booma and observe geographically distinct spatial patterns of water contact. This has been found to influence their exposure to S. mansoni cercariae (Pinot de Moira et al., 2007) and their risk of infection (Pinot de Moira et al., Reference Pinot De, Fulford, Kabatereine, Ouma, Booth and Dunne2010).

Bagungu men admitted that the Alur men earned more money and were ‘more business minded than the Bagungu’. However, like the Bagungu women, Bagungu men were derogatory of Alur men; they suggested Alur men were more likely to be heavily infected with bilharzia because they spent so long in the water. Referring to the research programme, they stated that, ‘they refuse bleeding because they’re not aware of bilharzia'. One Bagungu man described Alur men as ‘primitive’, and went on to say: ‘when their bellies are big they think they’ve gained weight. It's difficult to convince them they have bilharzia. They believe they're charmed.' The tendency of the Bagungu to speak negatively of the Alur may in part relate to the recent ethnic tensions experienced in Booma. Although many of the Alur living in Booma were born in the village, they are not indigenous to the area and originate either from the West Nile region in Uganda, or from the Democratic Republic of Congo (Francis Kazibwe, personal communication). A period of sustained migration by the Alur from the Democratic Republic of Congo led to increasing discomfort among local authorities and populations. As the Alur grew in number, they also gained more power, and some Alur became local chairman. This led to fears amongst the indigenous people, including the Bagungu, that the Alur could eventually dominate in number and challenge their authority. In 2005, local politicians, including the Local Council 3 (LC3) chairman, encouraged the Bagungu to evict the Alur, accusing them of ‘land grabbing’ and becoming over active in local politics. As a result, many Alur were forced to leave Booma; some have now returned after receiving assurances from the central government of their safety.

Fetched water

Because fetched water could be an additional source of infection, health behaviour surveys questioned individuals on their uses of fetched water and also alternative sources of water. Table 4 tabulates uses of fetched lake water by ethnic group. A significantly larger proportion of Alur than Bagungu used fetched lake water for washing utensils and as drinking water (p=0.03 and p<0.0001 respectively), whereas significantly more Bagungu used fetched lake water for washing clothes (p=0.004). Since the Booma study began in 1998, Gravity Flow System taps (free from S. mansoni cercariae) have been installed towards the south-west end of the village, where most of the Bagungu are resident. Consequently, taps were a major alternative source of water for the Bagungu, but not the Alur (122 [93%] of Bagungu respondents used taps as an alternative source of water vs five [6%)] Alur; χ ²=154.2, p<0.0001). A major alternative source of water for the Alur was the River Waki (67 [86%] of Alur obtained water from the River Waki vs eleven [8%] Bagungu; χ ²⁼125.5, p<0.0001). Table 5 displays uses for this alternatively sourced water. A significantly larger proportion of Bagungu than Alur used alternatively sourced water for washing utensils, washing clothes, bathing, cleaning fish and cooking (p≤0.02); in both ethnic groups, 100% of individuals questioned used alternatively sourced water as drinking water.

Table 4.

Uses of fetched lake water amongst the Alur and Bagungu

n=number of respondents who reported using fetched lake water for a given task; percentage of respondents is given in parentheses.

Chi-squared tests were used to compare the Alur and Bagungu people's reported uses of fetched water.

Table 5.

Uses of water obtained from alternative water sources amongst the Alur and Bagungu

n=number of respondents who reported using alternatively sourced water for a given task; percentage of respondents is given in parentheses.

Chi-squared tests were used to compare the Alur and Bagungu people's reported uses of alternatively sourced water.

Biomedical knowledge of schistosomiasis (bilharzia)

The majority of individuals interviewed in the health behaviour survey had heard of bilharzia (n=206, 99% overall; 97% of Alur and 99% of Bagungu). Most individuals had learnt of bilharzia through the research project: 71% of individuals (72% of Alur and 71% of Bagungu) listed the research project as one of their sources of information, and 45% of individuals (53% Alur and 41% Bagungu) listed this as their only source of information. Significantly more Bagungu than Alur had learnt about bilharzia through family and friends (11% Alur vs 24% Bagungu, χ ²=5.7, p=0.02); this remained significant after adjusting for age and sex in a logistic regression analysis (OR=3.2 [95% CI: 1.3, 7.9], p=0.01). Other sources of information included school (19%), the health centre (20%) and the radio (0.5%).

Of the 206 individuals claiming to have heard of bilharzia, 88% (n=181) listed either distended abdomen (80%), diarrhoea (34%), passing of blood in the stools (28%) or persistent stomach pain (12%) as symptoms of bilharzia; other symptoms listed that could also be attributable to schistosomiasis included weight loss (n=74, 36%), bloated stomach (n=2, 1%), itchy skin (n=6, 3%) and fatigue (n=5, 2%). Significantly more Bagungu listed these symptoms than Alur (62 [82%] Alur vs 124 [95%] Bagungu, χ ²=10.4, p=0.001) and this remained significant after adjusting for age and sex (adjusted OR=4.9, p=0.003). Of the 186 listing these attributable symptoms, 87 (47%) individuals also listed incorrect symptoms or symptoms less likely to be attributable to intestinal schistosomiasis such as: sight loss (n=2, 1%), hair loss (n=11, 6%), headache (n=12, 6%), vomiting (n=28, 15%), yellow skin/jaundice (n=40, 22%) and blood in the urine (n=2, 1%). Only 99 (48%) individuals interviewed claiming to have heard of bilharzia described only commonly attributable symptoms. A significantly higher proportion of Bagungu (12 [16%] of Alur vs 45 [35%] of Bagungu, χ ²=8.5, p=0.004; adjusted OR=2.9, p=0.01) and more males than females (15 [17%] of females vs 42 [35%] of males, χ ²=7.7, p=0.01; adjusted OR=2.6, p=0.01), were able to identify snails as the intermediate host.

The results of the health behaviour survey suggest, therefore, that biomedical knowledge of intestinal schistosomiasis in Booma is high, and that this is likely to be as a direct result of the research project. However, these data were collected using a Knowledge, Attitude and Practice (KAP) style approach, which tends to have a narrow focus with regards to knowledge. For this reason, KAP surveys often neglect or overlook other types of knowledge. Other illnesses may be recognized within communities with overlapping or similar signs and symptoms. For example, in Panyimur, Parker et al. (Reference Parker, Allen and Hastings2007) identified a local illness termed awola with symptoms almost identical to the biomedical syndrome ascites. However, treatment of awola with biomedical drugs was strongly discouraged and perceived as dangerous, which may have consequences for drug uptake in schistosomiasis control programmes. Focus group discussions were intended to supplement data obtained from the health behaviour surveys and compensate for their narrow focus. These revealed that prior to the research project, bilharzia was widely known as kodollo (relating to a distended abdomen) and believed to be caused by charms (burogo in Lugungu, awola in Alur), spirits (mizimu in Lugungu, tipo in Alur) or to be inherited in some way (since it affected whole families). Treatment was sought from witch doctors (mufumu in Lugungu, ajoga in Alur) and mainly consisted of herbal remedies, which at most, brought short-term relief of symptoms, but would not prevent people from dying. Unfortunately, FGDs did not explore whether additional local illnesses with similar symptoms to bilharzia existed concurrently with bilharzia. In Panyimur, however, Parker et al. (Reference Parker, Allen and Hastings2007) thought it likely that bilharzia was likely to be an alternative explanation for the symptoms associated with awola rather than one that is held simultaneously. From discussions it was apparent that current beliefs in Booma concerning bilharzia are very much biomedical in their outlook. Although there was evidence for some confusion as to how schistosomiasis is transmitted – for example, individuals from both ethnic groups suggested boiling lake water for drinking could limit transmission and some Alur also suggested eating fresh fish was involved in transmission – discussions generally revealed a high level of awareness of intestinal schistosomiasis. For instance, other methods for limiting transmission mentioned by Alur included boiling bathing water, using running river water rather than lake water, the construction of taps or boreholes, the burying of faeces and the construction of pit latrines. They also requested more drugs, alternative water sources and laws to enforce pit latrines. The Bagungu suggested the provision of gumboots, ‘hip boots’ and gloves, and the construction of pit latrines and boreholes to limit transmission. They also requested more drugs, and in addition the provision of gumboots and gloves. Therefore, whereas the emphasis for limiting transmission was on improving water supplies and sanitation amongst the Alur, there was a greater emphasis on limiting water contact amongst the Bagungu.

Focus group discussions explored whether the increased awareness of schistosomiasis had had any influence on behaviour. All groups agreed that it was impossible to avoid the lake; as one Mugungu woman described it: ‘we fear the lake, but we can't leave it completely; we are forced to return, it's our garden.’ Both ethnic groups, however, described some changes. Some Alur males said they now boiled water for bathing. Alur women said that they tried to fetch water in the morning, and avoided the lake in the afternoon; this suggests that they are correctly aware that the risk of infection increases at certain points in the day. Both Bagungu men and women suggested that many Bagungu men had changed their method of fishing as a result of their increased awareness of bilharzia, and that they now no longer fished on foot but instead used boats. Some Bagungu men also said they had reduced the amount they fished, although for others the greater fear was wild animals, and as a result fished during the day rather than at night. Reductions in the amount of time spent bathing were also mentioned by Bagungu men, and some Bagungu women said they had stopped bathing in the lake after they learnt about bilharzia. Like Alur women, Bagungu women described changes to their fetching patterns: they said they tried to reduce the number of times they fetched water, fetched water ‘quickly’, and avoided the lake in the afternoon, when ‘bilharzia attacks’. Both ethnic groups found the water ‘itchy’, especially in the afternoon; contact with the lake, where possible, was therefore avoided around this time. Some Bagungu identified the source of the itching as ‘bilharzia trying to penetrate the body’; this would be considered within the scientific world as precise knowledge of schistosomiasis.

When asked about whether they tried to educate others about schistosomiasis, all groups said that they tried to educate the children. However, many Alur said the children ‘didn't listen’; there was a strong consensus that Alur children today were less controlled than previous generations, that they did not worry about bilharzia, and that they spent more time bathing and playing; it was also mentioned that Alur children earned money fishing. Bagungu children, in contrast, were said to be more controlled; it was suggested that Bagungu children today had less contact with the lake than previous generations because of increased awareness of bilharzia, the increased number of lake-dwelling wild animals and increased time spent at school.