Geohelminth infection and re-infection after chemotherapy among slum-dwelling children in Durban, South Africa

Transcription

1 Annals of Tropical Medicine & Parasitology, Vol. 103, No. 3, (2009) Geohelminth infection and re-infection after chemotherapy among slum-dwelling children in Durban, South Africa C. C. APPLETON *, T. I. MOSALA *,{, J. LEVIN { and A. OLSEN 1 * School of Biological and Conservation Sciences, Westville Campus, University of KwaZulu- Natal, Durban 4000, South Africa { Epidemiology and Surveillance, National Department of Health, Private Bag X828, Pretoria 0001, South Africa { Biostatistics Unit, South African Medical Research Council, Pretoria 0001, South Africa 1 DBL Centre for Health Research and Development, Faculty of Life Sciences, University of Copenhagen, Thorvaldsensvej 57, DK-1871 Frederiksberg, Denmark Received 8 October 2008, Revised 3 December 2008, Accepted 8 December 2008 The prevalences and intensities of Ascaris lumbricoides, Trichuris trichiura and hookworm (probably Necator americanus) infection were measured in the young children (aged 2 10 years) living in 10 urban slums in Durban, South Africa. Re-infection was assessed at 4 6 and 12 months post-treatment. The baseline prevalences of A. lumbricoides and T. trichiura were 81.7% 96.3% and 54.5% 86.2%, respectively, and the corresponding geometric mean intensities were 960 and 91 eggs/g faeces. Most (85%) of the children found infected with A. lumbricoides and 23% of those found infected with T. trichiura had moderate heavy infections. A few of the children investigated had intensities of Ascaris and Trichuris infection that were considerably higher than those previously recorded in South Africa. The baseline prevalences of hookworm infection (0% 20% in individual slums, with a mean of 4.7%) and intensities of such infection (geometric mean517 eggs/g) were relatively low. Albendazole proved very effective against A. lumbricoides and hookworm but less so against T. trichiura. Re-infection by A. lumbricoides and T. trichiura reached pre-treatment prevalences by 4 6 months post-treatment in some of the slums and by 12 months in all the other slums. By 12 months post-treatment, the intensities of A. lumbricoides infection had reached their pretreatment levels while those of T. trichiura infection were higher than at baseline. Approximately 50% of children had moderate heavy T. trichiura infections at 12 months post-treatment compared with approximately 23% at baseline. Hookworm infections did not re-appear after treatment. The results show clearly that urban slums should be included in any future helminth-control programmes in South Africa. Most of what is known of geohelminth epidemiology in South Africa relates to the sub-tropical lowlands of KwaZulu-Natal province (Schutte et al., 1981; Appleton and Gouws, 1996; Appleton et al., 1999) and the low-lying environs of Cape Town in Western Cape province (Millar et al., 1989; Gunders et al., 1993). In terms of geohelminth infection, the most severely affected part of the entire country is the narrow, Reprint requests to: C. C. Appleton. appletonc@ukzn.ac.za; fax: z sandy coastal plain of KwaZulu-Natal, which includes the city of Durban. On this plain, children are co-infected with the three most common species Ascaris lumbricoides, Trichura trichiura and hookworm [identified as Necator americanus by Appleton et al. (1999)] at prevalences of 70% 100% (Appleton et al., 1999). Strongyloides stercoralis is also present, though nowhere common (Appleton et al., 1999). Neglected in previous helminth surveys in South Africa have been the country s burgeoning urban slums, for which even # The Liverpool School of Tropical Medicine 2009 DOI: / X398212

2 250 APPLETON ET AL. FIG. 1. Map of the Ethekwini municipal area showing the locations of the 10 study slums and Durban city centre and industrial areas. Also shown is the 150-m altitudinal contour that Mabaso et al. (2003) predicted to delimit hookworm transmission on the KwaZulu-Natal coastal plain, with no transmission at higher altitudes. basic epidemiological data are lacking. In South Africa, the word slum is synonymous with informal settlement and implies high-density housing that does not conform to building regulations and has little or no sanitation. At the time of the present study, approximately 500 slums were scattered within the Ethekwini municipal area (Fig. 1), together housing an estimated 180,000 families or 26% of the city s population (Mosala, 2001). The first geohelminth survey of a Durban slum was conducted more than 50 years ago, when Elsdon-Dew and Freedman (1952) investigated migrant labourers living in the overcrowded Cato Crest slum. They found that the prevalences of A. lumbricoides and T. trichiura doubled after 2 years residence, to

3 GEOHELMINTHIASIS IN SOUTH AFRICAN SLUMS % and 61.9%, respectively. The second such survey was conducted in Besters, where, as part of a nutritional study, Coutsoudis et al. (1994) recorded A. lumbricoides and T. trichiura prevalences of 59% and 61%, respectively, in children aged 3 6 years. Neither Elsdon-Dew and Freedman (1952) nor Coutsoudis et al. (1994) recorded hookworm infections or measured the intensities of the Ascaris or Trichuris infections that they observed. The aim of the present study was to assess geohelminth transmission in selected slums in Durban, by measuring the prevalences and intensities of A. lumbricoides, T. trichiura and hookworm infections among children. Re-infection rates for each of these parasites were also determined, 4 6 months and 12 months after treatment with albendazole. SUBJECTS AND METHODS Study Population, Area and Climate The present study was carried out, between 1998 and 2000, among 996 children aged 2 10 years in 10 established slums in Durban, KwaZulu-Natal province, South Africa. The slum lifestyle influenced selection of the target population, inasmuch as children aged.10 years were excluded because many did not attend school and each of those who did attend often went to several different schools. Sampling was conducted over weekends (to avoid these problems with school attendance) but, even then, children aged.10 years showed poor compliance and could not easily be located. As access to children aged,2 years required time-consuming visits to individual houses, such babies and infants were also excluded. Durban lies on the coastal plain and is hilly, with altitudes ranging from m above sea level. The climate is sub-tropical and most of the annual rainfall (mean mm) falls in the summer (Schulze, 1997). The mean daily mid-winter (July) temperatures range from uC whereas the mean mid-summer (January) temperatures fluctuate between 19.5 and 29.0uC. Selection of Sites Slums were eligible for inclusion in the study if the community leaders gave their consent, if there was no or only minimal dispute between the land-owners and slum communities regarding occupation of the land and if the slums were deemed safe for researchers, accessible by car, and established and stable (i.e. to have existed for.5 years and expected to exist until at least December 2000, when the present study ended). The 10 selected slums were Bottlebrush (29.909S, E), Kennedy Lower (29.819S, E), Lusaka (29.919S, E), Pemary Ridge (29.799S, E), Quarry Road West (29.809S, E), Simplace (29.779S, E), Briardene (29.809S, E), Smithfield (29.799S, E), Park Station (29.799S, E) and Canaan (29.819S, E). All 10 were situated within the residential area of Ethekwini Municipality (Fig. 1). Study Design This was a community-based, prospective, cohort study involving all children aged 2 10 years. At baseline, all such children were tested for geohelminth infection and then treated with albendazole, irrespective of their infection status. Drug efficacy was assessed 30 days post-treatment and children found positive were re-treated. All children were followed up 4 6 months and 12 months after the initial treatment. All children were treated again at the end of the project, immediately after the 12-month follow-up. Stool Collection and Examination At each survey, two stool samples were collected from each study child, over a week, by assistants who lived in the same slums. Each sample was examined, in

4 252 APPLETON ET AL. duplicate, as modified 50-mg Kato Katz thick smears (Archer et al., 1997; WHO, 1998). Smears were read for hookworm eggs within 1 h of preparation and again, 24 h later, for A. lumbricoides and T. trichiura eggs. Overall, 996 children were examined at baseline and the post-treatment (30-day) survey, 781 at the 4- to 6-month follow-up, and 947 at the 12-month follow-up. One slum, Canaan, did not participate in the 4- to 6-month follow-up because residents were then being relocated by the Durban Metro Council. Intensity of Geohelminth Infection Intensity of infection was measured as the number of eggs per 50-mg Kato Katz smear and expressed as the geometric mean (GM) egg output (with the egg-negatives included), in eggs/g faeces. Infections were classified according to Renganathan et al. (1995). Thus, light, moderate and heavy A. lumbricoides infections gave egg outputs of , ,999 and >50,000 eggs/g, respectively. The corresponding values for T. trichiura were 1 999, and >10,000 eggs/g, whereas those for hookworm infections were 1 999, and >5000 eggs/g. Anthelminthic Treatment As the baseline prevalences of A. lumbricoides and T. trichiura infection were found to be high (.60%) in all slums, albendazole (ZentelH; SmithKline Beecham, London) treatment was administered as a single 400- mg dose to each child in the slum who was aged >2 years of age. Younger children were each treated with 20 ml of ZentelH suspension (containing 400 mg albendazole) over 3 days, in three equal daily doses, with each first dose given by the researchers and the children s parents/guardians asked to supervise the remaining doses at home. The cure rate of the albendazole was estimated for each geohelminth, as the proportion of the treated children found positive for a given parasite at baseline who were found egg-negative for that parasite when checked 30 days post-treatment (WHO, 2002). Statistical Analysis Data were analysed using the SPSS software package (SPSS Inc., Chicago. IL). Prevalences and intensities were compared, between slums, using x 2 tests. In the case of low egg counts (i.e. those of hookworm), the results of the Pearson x 2 tests were confirmed using a likelihood-ratio x 2 statistic. A 5% level of significance was used for all the tests, with no adjustment for multiple testing. Ethical Considerations The study protocol was approved by the ethics committee of the University of Natal s Nelson R. Mandela School of Medicine, and by the Danish Central Medical Ethics Committee. Written informed consent for the participation and treatment of each child was obtained from the mother, father or guardian of the child. Children found to be excreting.100,000 A. lumbricoides eggs/g and/or.20,000 T. trichiura eggs/g were given letters of referral to the nearest Health Department clinic, so that treatment could be given under medical supervision. RESULTS The prevalences and intensities of the geohelminth infections observed at baseline and the 4-to 6-month and 12-month followups are summarized, for the 10 study slums, in Figures 2 and 3, respectively. Prevalences of the Geohelminth Infections at Baseline The baseline prevalences of A. lumbricoides and T. trichiura in individual slums were consistently high, at 81.7% 96.3% (mean5 89.2%) and 54.5% 86.2% (mean571.6%), respectively, whereas those of hookworm

5 GEOHELMINTHIASIS IN SOUTH AFRICAN SLUMS 253 FIG. 2. Mean prevalences of infections with Ascaris lumbricoides (&), Trichuris trichiuri (%) and hookworm (&) in the 10 slums at baseline, 30 days post-treatment, and at the 4- to 6-month and 12-month follow-ups. were low, at 0% 20.1% (mean54.7%; see Table and Figure 2). The between-slum differences in the prevalences of A. lumbricoides [x ; degrees of freedom (df)5 9; P50.003], T. trichiura (x ; df59; P50.001) and hookworm (x ; df59; P,0.001) were all statistically significant. Intensities of the Geohelminth Infections at Baseline Figure 3 shows the proportions of the baseline A. lumbricoides [Fig. 3(a)] and T. trichiura [Fig. 3(b)] infections that were categorized as light, moderate or heavy. Almost half (49.7%) of the A. lumbricoides infections were moderate whereas a similar percentage (47.7%) of the T. trichiura infections were light. As with the prevalences, the intensities of infection with A. lumbricoides (x ; df536; P,0.0001; with a geometric mean of 8812 eggs/g) and T. trichiura (x ; df536; P,0.0001; with a geometric mean of 128 eggs/g) varied significantly between the study slums. The intensities of the hookworm infections detected were generally very low, with 96.5% of such infections falling in the light category. Only 11 (1.1%) of the children investigated had moderate hookworm infections and just a single child was found to have a heavy hookworm infection. Despite the low outputs of hookworm eggs, the hookworm infections detected showed significant between-slum variation in intensity (x ; df59; P,0.0001, with a geometric mean of 17 egg/g), with no hookworm detected in the subjects from two of the study slums. Efficacies of Treatment The cure rates of albendazole against A. lumbricoides, as measured 30 days posttreatment, varied amongst individual slums (x ; df59; P50.043) from 90.2% 100%, with a mean (S.D.) percentage of 95.6 (3.3). The corresponding cure rates for T. trichiura were lower, at 44.0% 82.1%, with a mean (S.D.) percentage of 67.6 (11.8), but still varied significantly between the study slums (x ; df59; P,0.0001). Albendazole appeared 100% effective against hookworm and this parasite did not

6 254 APPLETON ET AL. FIG. 3. The intensities of Ascaris lumbricoides (a) and Trichuris trichiura (b) infection seen in the 10 slums at baseline, 30 days post-treatment, 60 days post-treatment (showing the data only for the children re-treated on day 30), and at the 4- to 6-month and 12 month follow-ups. Each child was categorized as uninfected (%) or having a light (&), moderate (&) or heavy (&) infection. No A. lumbricoides was detected in the re-treated children when they were checked 30 days after the re-treatments. re-appear during subsequent surveys. Those children found still infected 30 days after treatment were re-treated. When checked after another 30 days, a few (5%) of the retreated children were still found infected, albeit lightly, with T. trichiura. The intensities of the A. lumbricoides infections observed 30 days post-treatment showed a mean egg reduction rate (ERR) of 98.9% but the ERR for this nematode varied significantly between the study slums (x ; df527; P50.001). The intensities of the post-treatment T. trichiura infections also varied significantly between the slums (x ; df518; P,0.001), with a mean ERR of 91.1%.

7 GEOHELMINTHIASIS IN SOUTH AFRICAN SLUMS 255 TABLE. The baseline prevalences of Ascaris lumbricoides, Trichuris trichiura and hookworm infection among 996 children aged 2 10 years from the 10 study slums, and the levels of sanitation in the slums as given by Mosala (2001) No. and (%) of children: Sanitation Found positive for: Sanitation coverage Slum Examined Ascaris Trichuris Hookworm No. of dwellings No. and type of toilets (% of houses with toilet) (category) Bottlebrush (82.5) 145 (72.5) 0 (0.0) VIP latrines 100 High Kennedy Lower (91.8) 66 (54.1) 1 (0.8) VIP latrines 59 High Lusaka (88.6) 53 (75.7) 3 (4.3) traditional pit latrines 55 High Pemary Ridge (93.8) 48 (73.8) 2 (3.1) 56 Three traditional pit latrines 6 Low Quarry Road West (96.3) 60 (73.2) 5 (6.1) traditional pit latrines 11 Low Simplace (91.9) 95 (77.2) 22 (17.9) traditional pit latrines 5 Low Briardene (87.5) 89 (79.5) 0 (0.0) chemical toilets 10 Low Smithfield (86.2) 25 (86.2) 6 (20.7) traditional pit latrines 77 High Park Station (81.7) 51 (71.8) 6 (8.5) traditional pit latrines 7 Low Canaan (94.3) 81 (66.4) 2 (1.6) traditional pit latrines 8 Low All (89.2) 713 (71.6) 47 (4.7) VIP, Ventilated improved pit.

8 256 APPLETON ET AL. Frequencies of Re-infection with A. lumbricoides and T. trichiura At the 4- to 6-month follow-up (Fig. 2), the prevalences of A. lumbricoides ranged from 14.3% 100% (mean564.0%) in the 10 slums while those of T. trichiura varied from 20.2% 83.7% (mean543.6%). Clear between-slum differences in the frequencies of re-infection by this time were seen for both A. lumbricoides (x ; df58; P,0.001) and T. trichiura (x ; df58; P,0.001). At the 12-month follow-up (Fig. 2), the A. lumbricoides prevalences ranged from 25.0% 100.0% (mean587.5%) while those of T. trichiura varied between 37.6% and 95.1% (mean570.7%). In this, last followup, as at the 4- to 6-month follow-up, the between-slum variation in the prevalences of both A. lumbricoides (x ; df59; P,0.001) and T. trichiura (x ; df59; P,0.0001) was significant. Intensities of the Re-infections with A. lumbricoides and T. trichiura At the 4- to 6-month follow-up, the geometric mean intensities of the A. lumbricoides and T. trichiura (re-)infections were 188 and 12 eggs/g, respectively. At this time, only 4.0% of the A. lumbricoides-infected children and none of the T. trichiura-infected harboured heavy infections with these parasites [Fig. 3(a)]. The between-slum variation in the intensities of infection was significant for both A. lumbricoides (x ; df59; P,0.0001) and T. trichiura (x ; df59; P,0.0001). At the final, 12-month follow-up, the overall geometric mean intensities of A. lumbricoides and T. trichiura (re-)infection were 4834 and 212 eggs/g, respectively. Figure 3(b) shows the distribution of the intensities of the A. lumbricoides infections at this time, when, as at baseline, nearly half [mean (S.D.) percentage545.8 (22.5)] of such infections fell in the moderate category. Over half of the children found infected with T. trichiura at this follow-up (compared with only approximately 20% at baseline) were moderately to heavily infected. The between-slum variation in the intensities of infection was again significant for both A. lumbricoides (x ; df59; P,0.0001) and T. trichiura (x ; df59; P,0.0001). Multiple Geohelminth Infections At baseline, only 20.7% (206) of the investigated children were free from geohelminth parasites. Almost four percent (3.9%) had triple infections (i.e. A. lumbricoides, T. trichiura and hookworm) and most had double infections, with A. lumbricoides plus T. trichiura (66.4%), A. lumbricoides plus hookworm (4.3%), or T. trichiura plus hookworm (4.3%). At the 4- to 6-month follow-up, 20.4% (159) of the children investigated were free of geohelminth infection while 29.6% (295) were co-infected (all with A. lumbricoides and T. trichiura). At the 12-month follow-up, 12.5% (124) of the children checked were free of geohelminths but the percentage co-infected (again, all with A. lumbricoides and T. trichiura) had risen to 60.4% (603). DISCUSSION The results of studies in other parts of the world have shown the level of geohelminth transmission in urban slums to be variable often high (Rijpstra, 1975; Kan et al., 1989; Naish et al., 2004) but sometimes surprisingly low (Ferreira et al., 1994; Korkes et al., 2008). In the present study, of children living in Durban slums, A. lumbricoides and T. trichiura infections were very common at baseline (with prevalences consistently above 80%) and generally moderate or heavy in intensity. In contrast, the post-treatment prevalences of (re-)infection with these parasites, particularly those recorded at 4 6 months, were variable. Slums are a feature of most, if not all cities in developing countries, and are increasing

9 GEOHELMINTHIASIS IN SOUTH AFRICAN SLUMS 257 both in number and in the proportion of the world s population that they house. The evidence indicates that, in some areas at least, slums support intense parasite transmission and that this problem tends to be ignored by the public-health authorities. Parasite transmission in urban slums needs urgent review. It is important to see the results of the present study in the context of the future establishment of a national helminth-control programme in South Africa, since they demonstrate clearly the need for urban slums to be included something that was not done in the provinciallevel programme in KwaZulu-Natal (Appleton and Kvalsvig, 2006). The baseline prevalences in the present study were higher than recorded in the only other geohelminth surveys in Durban slums (Elsdon-Dew and Freedman, 1952; Coutsoudis et al., 1994). Intensities were also high, with small numbers of children voiding.100,000 A. lumbricoides and/or.20,000 T. trichiura eggs/g. The highest prevalences and intensities of A. lumbricoides and T. trichiura recorded in the present study exceeded those recorded in rural communities on the KwaZulu-Natal coastal plain, which is the area of South Africa most severely affected by geohelminths (Schutte et al., 1981; Appleton et al., 1999; Saathoff et al., 2004). In the present study, a single dose of albendazole was found to be very effective against A. lumbricoides and hookworm, with the observed cure rates falling within the ranges reported from three rural communities in South Africa. Among the subjects investigated by Evans et al. (1997), for example, in the Valley of a Thousand Hills (which lies in KwaZulu-Natal, about 40 km to the west of Durban), albendazole cured 64.7% of the A. lumbricoides infections, 20.7% of the T. trichiura, and 100% of the hookworm. Also in KwaZulu-Natal, on the Makathini Flats in the north eastern corner of the province, Saathoff et al. (2004) recorded cure rates of 96.4% against A. lumbricoides, 12.7% against T. trichiura, and 78.8% against hookworm. Adams et al. (2004) reported a corresponding rate of 23% against T. trichiura in the Overberg area of Western Cape province. Although the frequencies of post-treatment re-infection with Ascaris and Trichuris were higher in the Durban slums (present study) than those reported by Saathoff et al. (2004), in a rural area of north eastern KwaZulu-Natal, the pre-treatment prevalences recorded in the latter study, for both A. lumbricoides (19.4%) and T. trichiura (57.3%), were relatively low. The levels of re-infection recorded in the present study were similar to those reported by Albonico et al. (1995), from Pemba Island, Tanzania, in that, in both investigations, the prevalences and intensities of Ascaris and Trichuris infection had reached pre-treatment levels by 6 months post-treatment. High prevalences of A. lumbricoides re-infection were also recorded by Olsen et al. (2000) in villages in western Kenya, where, 12 months post-treatment, they exceeded baseline values by 50%. The prevalences of T. trichiura infection had, however, only reached 75% of their baseline values by this time. Using a geographical information system (GIS), Saathoff et al. (2005) reported an association between A. lumbricoides infection and normalized-difference vegetation indices (NDVI ) in rural northern KwaZulu- Natal, and identified reduced exposure to direct sunlight as important for the survival of the parasite s eggs here. In the case of urban slums, such as those in the present study, the shade cast by the tightly packed houses (which may be built no more than 2 m apart) might mimic the shading effect of vegetation, on the soil and any helminth eggs it might contain. The low prevalences and intensities of hookworm infection seen in the children of the Durban slums (present study) deserve comment. Although hookworm infections were detected at baseline in eight of the 10 study slums, prevalences never exceeded

10 258 APPLETON ET AL. 20%, most of the infections that were seen were light, and no hookworm-positives were recorded in the follow-up surveys. The results of previous studies on the coastal plain of KwaZulu-Natal (Appleton et al., 1999; Mabaso et al., 2003) indicated that most hookworm transmission occurred in sandy soils with a low (,15%) clay content at altitudes of,150 m above sea level. Since eight of the 10 study slums lay below 150 m (Fig. 1) and four were built on well drained soils with a low clay content of 3% 5% (data not shown), it was expected that more hookworm infections would be found. Given the general rarity of hookworm infection and the fact that the two slums supporting the highest baseline prevalences of such infection Simplace (17.9%) and Smithfield (20.0%) lie on poorly drained shale with a clay content of about 38% (Mosala, 2001), it seems likely that most of the hookworm infections detected at baseline had been acquired outside of the study slums. The present study complements one in Khayalitsha, a large slum in Cape Town, by Fincham et al. (1998) and Fincham (2001). In Khayalitsha, 94% of children had geohelminth infections (A. lumbricoides or T. trichiura or both) and 8% had infections that fell in the heavy category. Fincham et al. (1998) warned that the high T. trichiura prevalences reported from urban slums such as Khayalitsha posed a real risk of trichuriasis dysentery syndrome (TDS; Stephenson et al., 2000). Similarly, children with heavy A. lumbricoides infections are at risk of complications such as bolus, volvulus and hepato biliary involvement conditions regularly seen in the low-lying coastal areas of South Africa, including Durban (Lloyd, 1982; Wynne and Ellman, 1983; Madiba and Hadley, 1996). Environmental contamination is an important risk factor accounting for the rapid re-infection often observed following anthelmintic treatment in slums. In their study of excreta-disposal facilities and intestinal parasitism among urban slum dwellers in Botswana, Zambia and Ghana, Feachem et al. (1983) concluded that the provision and use of piped water and sanitation facilities may not protect families from infection if the overall level of faecal contamination in the environment remained high. Also in southern Africa, the results of a study by Muller et al. (1989), in a poor community in Maputo, Mozambique, cast further doubt on the effectiveness of either traditional or ventilated improved pit (VIP) latrines in reducing geohelminth transmission. In the present study, although both these types of toilet were in use in the slums investigated, no attempt was made to measure environmental contamination by recovering geohelminth eggs from soil samples. Mosala (2001), however, found that the slums investigated in the present study varied greatly in their sanitation coverage (i.e. the proportion of dwellings with a latrine). In the Table this coverage is divided into two broad categories: low (5% 11%) and high (.55%). The results of a preliminary comparison of these coverage data with infection status recorded, in the present study, at baseline and during the follow-up surveys indicate that relatively high levels of transmission at least those indicated by high prevalences of (re-)infection 4 6 months post-treatment are associated with low sanitation coverage (,11%), and vice versa. Although it therefore appears that high sanitation coverages (.50%) slowed re-infection with both A. lumbricoides and T. trichiura, this advantage disappeared over time, since sanitation coverage had little if any effect on the prevalences recorded 12 months post-treatment. In Chile, Navarrete and Torres (1994) were similarly able to associate the lowest infection rates with areas with.50% toilet coverage (although the prevalences they recorded were lower than seen in the present study). Although, even if well maintained, traditional and VIP latrines are not sufficient, by themselves, to prevent geohelminth infection, high latrine coverage in combination with regular anthelminthic treatment can

11 GEOHELMINTHIASIS IN SOUTH AFRICAN SLUMS 259 keep transmission low. As noted by Muller et al. (1989), Faecal pollution of the household environment is due more to promiscuous defaecation than to poor construction or maintenance of latrines. For this reason, communities characterised by geohelminth infections of moderate or high intensity should be targeted not only for anthelmintic treatment but also for additional measures notably health education to reduce environmental contamination. An additional factor in the epidemiology of geohelminths in slums may be the passive dispersal of ova by scavenging and coprophagous animals. In India, Traub et al. (2002) recovered viable A. lumbricoides and T. trichiura eggs from 31% and 25%, respectively, of the samples of dog faeces they investigated. Ascaris lumbricoides eggs, in various stages of development, were recently found in the faeces of 27% of the Rattus norvegicus (brown rat) and 9% of the Mus musculus (house mouse) trapped in the Cato Crest slum in Durban (C. C. Appleton and C. E. Archer, unpubl. obs.). Dogs and invasive rodents are common in urban slums and their role in A. lumbricoides transmission in these environments needs to be investigated. In conclusion, heavy transmission of A. lumbricoides and T. trichiura is confirmed for children living in slums (informal settlements) in Durban, South Africa. A single, 400-mg dose of albendazole was very effective against A. lumbricoides and hookworm but less so against T. trichiura. Posttreatment re-infection with A. lumbricoides and T. trichiura was common and rapid but might have been slowed, at least over the first 6 months post-treatment, in those slums in which.50% of houses had access to toilets. No hookworm infections were seen at follow-up and most of those seen at baseline were probably contracted elsewhere. The demonstration of high prevalences and intensities of Ascaris and Trichuris infection confirms that urban slums, such as those investigated in the present study, should be included in any future helminth-control programme in South Africa. ACKNOWLEDGEMENTS. The authors thank the DBL Centre for Health Research and Development (formerly the Danish Bilharziasis Laboratory), University of Copenhagen, Denmark, for financial support. They are also indebted to C. Archer (University of KwaZulu-Natal) for her support in this logistically difficult study, to Dr A. Robinson (Durban City Health Department) and his staff, for facilitating the research and supervising the treatment of the children, and Dr R. Mackey (University of KwaZulu-Natal) for help with the figures. REFERENCES Albonico, M., Smith, P. G., Ercole, E., Hall, A., Chwaya, H. A., Alawi, K. S. & Savioli, L. (1995). Rate of reinfection with intestinal nematodes after treatment of children with mebendazole or albendazole. in a highly endemic area. Transactions of the Royal Society of Tropical Medicine and Hygiene, 89, Adams, V. J., Lombard, C. J., Dhansay, M. A., Markus, M. B. & Fincham, J. E. (2004). Efficacy of albendazole against the whipworm Trichuris trichiura a randomised, controlled trial. South African Medical Journal, 94, Appleton, C. C. & Gouws, E. (1996). The distribution of common intestinal nematodes along an altitudinal transect in KwaZulu-Natal, South Africa. Annals of Tropical Medicine and. Parasitology, 90, Appleton, C. C. & Kvalsvig, J. D. (2006). A schoolbased helminth control programme successfully implemented in KwaZulu-Natal. Southern African Journal of Epidemiology and Infection, 21, Appleton, C. C., Maurihungirire, M. & Gouws, E. (1999). The distribution of helminthic infections along the coastal plain of KwaZulu-Natal province, South Africa. Annals of Tropical Medicine and Parasitology, 93, Archer, C., Appleton, C. C. & Kvalsvig, J. D. (1997). Diagnostic Methods for Use in a Primary Health Carebased Parasite Control Programme. Durban, South Africa: Centre for Integrated Health Research, University of Natal. Coutsoudis, A., Jinabhai, C. C., Coovadia, H. M. & Mametja, L. D. (1994). Determining appropriate

12 260 APPLETON ET AL. nutritional interventions for South African children living in informal urban settlements. South African Medical Journal, 84, Elsdon-Dew, R. & Freedman, L. (1952). Intestinal parasites in the Natal Bantu. South African Journal of Clinical Science, 3, Evans, A. C., Fincham, J. E., van Stuivenberg, M. E., Markus, M. B., Appleton, C. C., Kvalsvig, J. D., Lombard, C. J. & Benadé, A. J. S. (1997). Implications of Mass Deworming with Albendazole (ZentelH, Smithkline Beecham) at a School in the Valley of a Thousand Hills, KwaZulu-Natal. Cape Town, South Africa: Medical Research Council. Feachem, R. G., Guy, M. W., Harrison, S., Iwugo, K. O., Marshall, T., Mbere, N., Muller, R. & Wright, A. M. (1983). Excreta disposal facilities and intestinal parasitism in urban Africa; preliminary studies in Botswana, Ghana and Zambia. Transactions of the Royal Society of Tropical Medicine and Hygiene, 77, Ferreira, C. S., Ferreira, M. U. & Nogueira, M. R. (1994). The prevalence of infection by intestinal parasites in an urban slum in São Paulo, Brazil. Journal of Tropical Medicine and Hygiene, 97, Fincham, J. E. (2001). The Khayalitsha Task Team: building health partnerships that work. Epidemiological Comments, 4, 3 5. Fincham, J. E., Markus, M. B., Appleton, C. C., Evans, A. C., Arendse, V. J., Dhansay, M. A. & Schoeman, S. (1998). Complications of worm infestation serious, costly and preventable. South African Medical Journal, 88, Gunders, A. E., Cotton, M., Nel, E., Hendricks, M., Ebrecht, K., Hahne, H., Redecker, R., Shaw, M. L., van der Walt, J. & Williams, A. (1993). Prevalence and intensity of intestinal worm infections in creche attenders in urban and peri-urban settings in greater Cape Town. South African Journal of Epidemiology and Infection, 8, Kan, S. P., Guyatt, H. L. & Bundy, D. A. P. (1989). Geohelminth infection of children from rural plantations and urban slums in Malaysia. Transactions of the Royal Society of Tropical Medicine and Hygiene, 83, Korkes, F., Kumugai, F. U., Belfort, R. N., Szejnfeld, D., Abud, T. G., Kleinman, A., Florez, G. M., Szejnfeld, T. & Chieffi, P. P. (2008). Relationship between intestinal parasitic infection in children and soil contamination in an urban slum. Journal of Tropical Pediatrics, in press. Lloyd, D. A. (1982). Hepatic ascariasis. South African Journal of Surgery, 20, Mabaso, M. L. H., Appleton, C. C., Hughes, J. C. & Gouws, E. (2003). The effect of soil type on hookworm (Necator americanus) distribution in KwaZulu-Natal, South Africa. Tropical Medicine and International Health, 8, Madiba, T. E. & Hadley, G. P. (1996). Surgical management of worm volvulus. South African Medical Journal, 34, Millar, A. J. W., Bass, D. H & van der Merwe, P. (1989). Parasitic infestation in Cape Town children. South African Medical Journal, 76, Mosala, T. I. (2001). Geohelminth transmission among slum-dwelling children in Durban, South Africa. Ph.D. thesis, University of Natal, Durban, South Africa. Muller, M., Sánchez, R. M. & Suswillo, R. R. (1989). Evaluation of a sanitation programme using eggs of Ascaris lumbricoides in household yard soils as indicators. Journal of Tropical Medicine and Hygiene, 92, Naish, S., McCarthy, J. & Williams, G. M. (2004) Prevalence, intensity and risk factors for soil-transmitted helminth infection in a South Indian fishing village. Acta Tropica, 91, Navarrete, N. & Torres, P. (1994). Prevalence of infection by intestinal helminths and protozoa in school children from a coastal locality in the province of Valdivia, Chile. Boletino Chileno de Parasitologia, 49, Olsen, A., Nawiri, J. & Friis, H. (2000). The impact of iron supplementation on reinfection with intestinal helminths and Schistosoma mansoni in western Kenya. Transactions of the Royal Society of Tropical Medicine and Hygiene, 94, Renganathan, E., Ercole. E, Albonico, M., de Gregorio, G., Alawi, K. S., Kisumku, U. M. & Savioli, L. (1995). Evolution of operational research and development of national control strategy against intestinal helminths in Pemba Island, Bulletin of the World Health Organization, 73, Rijpstra, A. C. (1975). Results of duplicated series of stool examinations for all intestinal parasites by different methods in school children in East Africa with remarks on serological aspects of amoebiasis and schistosomiasis. Annales de la Société Belge de Médecine Tropicale, 55, Saathoff, E., Olsen, A., Kvalsvig, J. D. & Appleton, C. C. (2004). Patterns of geohelminth infection, impact of albendazole treatment and re-infection after treatment in schoolchildren from rural KwaZulu-Natal/South Africa. BMC Infectious Diseases, 4, 27. Saathoff, E., Olsen, A., Kvalsvig, J. D., Appleton, C. C., Sharp, B. L. & Kleinschmidt, I. (2005). Ecological covariates of Ascaris lumbricoides infection in schoolchildren from a small area within rural KwaZulu- Natal/South Africa. Tropical Medicine and International Health, 10, Schulze, R. E. (1997). South African Atlas of Agrohydrology and Climatology. Report TT82/96. Pretoria: Water Research Commission. Schutte, C. H. J., Eriksson, I. M., Anderson, C. B. & Lamprecht, T. (1981). Intestinal parasitic infections

13 GEOHELMINTHIASIS IN SOUTH AFRICAN SLUMS 261 in Black scholars in northern KwaZulu. South African Medical Journal, 55, Stephenson, L. S., Holland, C. V. & Cooper, E. S. (2000). The public health significance of Trichuris trichiura. Parasitology, 121 (Suppl.), S73 S95. Traub, R. J., Robertson, I. D., Irwin, P., Mencke, N. & Thompson, R. C. (2002). The role of dogs in transmission of gastrointestinal parasites in a remote tea-growing community in northeastern India. American Journal of Tropical Medicine and Hygiene, 67, World Health Organization (1998). Guidelines for the Evaluation of Soil-transmitted Helminthiasis and Schistosomiasis at Community Level. Document WHO/CTD/98.1. Geneva: WHO. World Health Organization (2002). Prevention and Control of Schistosomiasis and Soil-transmitted Helminthiasis, Report of a WHO Expert Committee. Technical Report Series No Geneva: WHO. Wynne, J. M. & Ellman, B. A. H. (1983). Bolus obstruction by Ascaris lumbricoides. South African Medical Journal, 63,