Am. J. Trop. Med. Hyg., 91(5), 2014, pp.

1002–1010
doi:10.4269/ajtmh.14-0060
Copyright © 2014 by The American Society of Tropical Medicine and Hygiene

Soil-Transmitted Helminths in Pre-School-Aged and School-Aged Children in an Urban Slum:

A Cross-Sectional Study of Prevalence, Distribution, and Associated Exposures
Stephanie M. Davis,* Caitlin M. Worrell, Ryan E. Wiegand, Kennedy O. Odero, Parminder S. Suchdev, Laird J. Ruth,
Gerard Lopez, Leonard Cosmas, John Neatherlin, Sammy M. Njenga, Joel M. Montgomery, and LeAnne M. Fox
Parasitic Diseases Branch, U.S. Centers for Disease Control and Prevention, Atlanta, Georgia; Data Management Activity, Division of Parasitic
Diseases and Malaria, U.S. Centers for Disease Control and Prevention, Atlanta, Georgia; International Emerging Infections Platform, Kenya
Medical Research Institute-Centers for Disease Control and Prevention, Nairobi, Kenya; Nutrition Branch, U.S. Centers for Disease Control
and Prevention, Atlanta, Georgia; Global Disease Detection, Centers for Disease Control and Prevention, Atlanta, Georgia; Eastern
and Southern Africa Centre for International Parasite Control, Kenya Medical Research Institute, Nairobi, Kenya

Abstract. Soil-transmitted helminths (STHs) are controlled by regular mass drug administration. Current practice
targets school-age children (SAC) preferentially over pre-school age children (PSAC) and treats large areas as having
uniform prevalence. We assessed infection prevalence in SAC and PSAC and spatial infection heterogeneity, using a
cross-sectional study in two slum villages in Kibera, Nairobi. Nairobi has low reported STH prevalence. The SAC and
PSAC were randomly selected from the International Emerging Infections Program’s surveillance platform. Data
included residence location and three stools tested by Kato-Katz for STHs. Prevalences among 692 analyzable children
were any STH: PSAC 40.5%, SAC 40.7%; Ascaris: PSAC 24.1%, SAC 22.7%; Trichuris: PSAC 24.0%, SAC 28.8%;
hookworm < 0.1%. The STH infection prevalence ranged from 22% to 71% between sub-village sectors. The PSAC
have similar STH prevalences to SAC and should receive deworming. Small areas can contain heterogeneous preva-
lences; determinants of STH infection should be characterized and slums should be assessed separately in STH mapping.

INTRODUCTION climactic and sanitation conditions. In each zone, SAC are

sampled in at least five schools chosen at random, providing
Soil-transmitted helminths (STHs) are associated with sub- one stool sample each.7,8 Prevalence results determine a uni-
stantial acute and chronic morbidity, particularly among chil- form frequency of deworming for the entire zone. Under
dren, in whom the highest intensities of infection are found.1 current practice, zones are typically the size of several dis-
The World Health Organization (WHO) estimates STHs tricts, allowing a small number of schools to represent a vast
may infect nearly two billion people2 and are associated with area. In particular, rural, informal urban settlement (slum),
at least 12,000 deaths each year.3 Many more individuals and non-slum urban schools can all represent each other.
suffer chronic and subtly debilitating effects of STH infec- However, there is evidence for heterogeneity across these
tion, including anemia, growth and cognitive impairment, and settings: urban areas in general may have lower infection
decreased productivity.4–6 Routine mass drug administration prevalences than rural ones,16 and although prevalences of
(MDA), the major tool used to control STH burden, has STH infection in slums vary widely,17–19 there is evidence
undergone expansion and received increasing guidance from that they are higher than those in the adjoining cities.9,20,21
WHO over the past decade.7,8 However, key issues in MDA Plausible mechanisms are sanitation and hygiene conditions
policy remain unsettled. that favor STH transmission.22,23 This susceptibility of slum
One such issue deals with the most appropriate target areas, which house an estimated one-third of the world’s
groups for routine deworming. Several demographic groups population,24 has been periodically recognized and addressed
are considered eligible, including pre-school age children through deworming campaigns21,25 and intervention trials14,26–28;
(PSAC) and school-age children (SAC). However, SAC are however, there is no explicit guidance or consistent practice
the focus of WHO’s most recent guidelines8 and many on distinguishing slums from other areas in agro-ecological
national programs target SAC. This is due in part to the zoning. A better understanding of geospatial prevalence varia-
logistical advantages of school-based treatment, and in part tions and their determinants is needed to better predict uni-
to a perception of greater benefit connected with deworming form prevalence zones.
in SAC.8 The majority of research defining burden has also Our objectives, to inform approaches to deworming, were
been performed in SAC,9 however PSAC can also have con- to investigate 1) STH infection prevalence and associated
siderable prevalence of STH infection10,11 and experience factors in PSAC and SAC, and 2) determinants of STH infec-
morbidity. Furthermore, evidence does exist that PSAC bene- tion distribution in an urban setting.
fit from deworming,12–14 and others have advocated for their Nairobi slum villages were chosen for the study because
inclusion in MDA.15 although nationally, Kenya has areas with a high burden of
Another key unsettled issue is demarcation of areas over STH infection,29,30 in Nairobi low prevalences have been
which a uniform frequency of MDA is appropriate. Current found among primary school students in the city as a whole31,32
guidelines for STH mapping recommend stratifying countries and it had not yet been included in the national deworming
into distinct “agro-ecological” zones defined by common program. We hypothesized that infection prevalences in both
PSAC and SAC would be higher than those found previously
elsewhere in Nairobi and would qualify for treatment under
*Address correspondence to Stephanie M. Davis, U.S. Centers for
current WHO prevalence-based recommendations. We also
Disease Control and Prevention, 1600 Clifton Rd. NE, MS E-04, hypothesized that the distribution of infection prevalence
Atlanta, GA 30333. E-mail: [email protected] would not be uniform, and that its determinants would include

1002
 SOIL-TRANSMITTED HELMINTHS IN AN URBAN SLUM 1003

sibling infection status and proximity to latrines. Finally, we paths, drainage patterns, and landmarks. Final sampling
assessed the sensitivity of testing varying numbers of stools weights are a product of the selection weight, the weighting
using the Kato-Katz technique for detecting STH infection class adjustment, and the post-stratification ratio.
from stool samples, using three stools as the “gold standard.” Data collection. Some household-level demographic vari-
Overall, STH infection prevalence and morbidity findings ables for each household, including sector of residence, were
for this study are discussed elsewhere33; here, we provide and provided by the IEIP registry. The remaining data were col-
discuss STH infection-associated factors and detailed preva- lected using Personal Digital Assistants (PDAs) programmed
lence, spatial distribution, and testing sensitivity results. in Questionnaire Manager.37 The study was performed during
April–June 2012. Participants (6 months–14 years of age) and
METHODS their parents/guardians were approached to participate in the
study. A questionnaire was used to capture household-level
Ethics statement. This study was approved by Institutional demographic factors and individual relevant history for the
Review Boards at the Kenya Medical Research Institute selected child, including deworming history in the past year
(KEMRI) and the U.S. Centers for Disease Control and Pre- and source of deworming. For SAC (5–14 years of age) the
vention (CDC). Participants provided written informed paren- questionnaire also collected information on school atten-
tal consent, and written assent from all participants 13 years of dance and school name; for PSAC (6 month–under 5 years
age or older. Test results for malaria and hemoglobin levels of age) it collected information on whether time was spent
were returned to participants at the time of testing, and stool regularly outside the home and location. (see Table 1). In
results were provided after examinations of all submitted stools addition, an observational survey was completed for their
were completed. Participants with anemia, malaria, STH infec- households that focused on socioeconomic status (SES)
tion, or moderate or severe undernutrition or anemia were assessment using questions from the Kenya Demographic
referred for free care at the International Emerging Infections and Health Survey (DHS) SES questionnaire. Water, sanita-
Program (IEIP) community disease surveillance reference tion, and hygiene environmental elements were included and
clinic as per the WHO/Kenya national guidelines. will be described separately elsewhere.
Study area and population. The study was conducted in Selected children were encouraged to provide three stool
two slum villages, Gatwekira and Soweto West, within the samples, the clinical diagnostic gold standard and an
Kibera urban slum area in southern Nairobi, using the established research approach, to mitigate limitations in
platform of the CDC-Kenya IEIP. This is a community- sensitivity seen with a single sample as a result of day-to-day
based surveillance platform run by KEMRI and CDC oper- variations in egg excretion.38 In the case of selected SAC,
ating in the villages. The IEIP methods have been described three stools were collected from their PSAC and SAC sib-
elsewhere34; briefly, the program enrolls all consenting lings as well, to assess sibling infection as a factor associated
adults and children with head-of-household consent who with infection. Stools were accepted only if produced after
have been living in all households in the study area for at midnight, and maintained after collection in cool boxes
least 4 months, thus attempting to enroll a census of all with ice packs before delivery to the laboratory for analysis
residents; the household refusal rate is under 2%.35 The before 2 PM that day. Stools were collected over 3 consecu-
IEIP conducts at least biweekly visits to all enrolled house- tive days where possible, or up to 15 days apart.
holds, to record interval self-reported morbidity and health Stool samples were processed in the laboratory of the
care usage data. The IEIP maintains an up-to-date registry Eastern and Southern Africa Centre of International Para-
by continually approaching new arrivals in all community site Control (ESACIPAC) based at KEMRI, using the stan-
households for enrollment and updating records to reflect dard Kato-Katz technique.39 Two slides were prepared for
outmigration. These villages are characterized by high popu- each sample, and were examined by different experienced
lation density, a lack of official city water or sewage services, laboratory technicians. Eggs seen per slide were quantified
and a proliferation of informal institutions such as schools by species and converted into eggs per 1 gram (EPG) of
not directly administered by city educational services (“infor- stool. Mean EPG for each child was used to categorize infec-
mal schools,” as opposed to city-run “formal schools”). tions as light, moderate, or heavy infection as per WHO
Study design. Sampling design. To avoid a household being species-specific definitions; the overall infection intensity for
selected into both the PSAC and SAC samples, a simple each child was that of his/her most intense species-specific
random sample without replacement of 25% of the IEIP infection. Results were double-entered into the study data-
participant registry was selected as eligible PSAC house- base and the two reads for the same sample were averaged.
holds; the remaining 75% were eligible as SAC households. Beginning 1 month after the study launch, 10% of all sub-
From each group, households were chosen with probability sequent slides were reread for quality control purposes40 by
proportional to size by the number of PSACs or SACs a third reader not involved in reading either slide originally.
living in the home,36 with collapsing at large-sized house- One blood sample per selected child obtained by finger
holds because of the rarity of such households. One child stick was tested for malaria by rapid diagnostic test kits (Pf/
was chosen at random with equal probability from each Pan, Bioline, Standard Diagnostics, Yongin-si, Korea). To
selected household. These children are referred to as “selected assess morbidities associated with STH infection, anthropom-
children” to distinguish them from their siblings on whom data etry, and finger stick blood testing for hemoglobin, and micro-
were also collected. Random selection was performed using nutrient deficiencies including iron and vitamin A were also
PROC SURVEYSELECT in SAS version 9.3 (SAS Insti- performed for these children; methodology and results for
tute, Inc., Cary, NC). A weighting class adjustment and post- these studies have been reported previously.33
stratification were performed based on IEIP sector, a Sample size. Sample size was limited by laboratory capacity
predefined division into one of 10 sub-areas demarcated by for daily processing of fresh stools. It was calculated to
1004 DAVIS AND OTHERS

Table 1
Demographic characteristics of analyzed PSAC and SAC*
PSAC (N = 205) SAC (N = 487)

Characteristic N (%) N (%)

Male 98 (47.8) 228 (46.8)

Female 107 (52.2) 259 (53.2)
Index child age at midpoint† 3.1 (STD 1.2) 11.2 (STD 2.7)
Sector of residence
1 39 (19.0) 98 (20.1)
2 25 (12.2) 74 (15.2)
3 7 (3.4) 17 (3.5)
4 6 (2.9) 19 (3.9)
5 22 (10.7) 43 (8.8)
6 25 (12.2) 44 (9.0)
7 10 (4.9) 35 (7.2)
8 32 (15.6) 62 (12.7)
9 12 (5.9) 27 (5.5)
10 27 (13.2) 68 (14.0)
Mean household size 5.3 (STD 1.8) 6.4 (STD 2.4)
Number of children < 15 years in household 2.9 (STD 1.6) 3.4 (STD 1.5)
Mean people per bedroom 4.8 (STD 1.6) 5.2 (STD 2.1)
Weekly food expenditure per HH member (US$) 5.42 (STD 2.81) 4.95 (STD 2.75)
Female head of household highest level of school attended Nursery– 0 Nursery–2 (0.4)
Primary– 143 (69.8) Primary–340 (69.8)
Vocational–1 (0.5) Vocational–1 (0.2)
Secondary –51 (24.9) Secondary–127 (26.1)
College 6–(2.9) College–4 (0.8)
University–1 (0.5) University–0
Unknown–3 (1.5) Unknown–13 (2.7)
Ethnicity Luo–137 (66.8) Luo–345 (70.8)
Luhya–42 (20.5) Luhya–84 (17.3)
Kisii–14 (6.8) Kisii–26 (5.3)
Other–12 (5.9) Other–32 (6.6)
Spends substantial time outside the home Yes –105 (51.2) Yes–480 (98.6)
No –100 (48.8) No– 7 (1.4)
Of those spending substantial time outside the home, setting:
School 4 (3.8) 452 (94.0) any school
191 (39.2) formal
224 (46.0) informal
Pre-primary 86 (81.9) 28 (5.8)
Relative 2 (1.9) 1 (0.2)
Babysitter 4 (3.8) −
Daycare 9 (8.6) −
Travel outside Nairobi in past year 76 (37.1) 264 (54.2)
Dewormed in past year 119 (58.9) 258 (54.1)
Of those dewormed in past year, setting:
School 7 (5.9) 142 (55.0)
Clinic, hospital or health center 48 (40.3) 51 (19.8)
Chemist shop 55 (46.2) 50 (19.4)
Other/Unknown 9 (7.6) 15 (5.8)
*PSAC = pre-school age children; SAC = school-age children. SAC includes school-age siblings of selected SAC.
†May differ by up to 8 weeks from age at anthropometric exam, which was applicable for selected children only and was used in the main analysis.

provide 80% power to detect an odds ratio (OR) of 2.0 for other a composite index made of the sum of the number of
one of the factors suspected to be associated with STH luxury/utility items owned by the household, using the
infection: having an infected sibling, for SAC. Sample size household ownership questions from the Kenya DHS.
was then increased 20% for nonresponse, for a total of Analyses were performed in R version 2.15.2 (R Core
293 PSAC and 899 SAC, out of an eligible population of Team, 2012) using the survey package. Associations were
4,056 PSAC and 11,731 SAC. considered significant at the 5% level. All analyses incorpo-
Statistical analysis. Criteria for inclusion of selected chil- rate the survey weights. Comparisons between SAC and PSAC
dren in the analysis consisted of providing a blood sample, were performed with c2 or Fisher’s exact test. Univariable and
at least one stool, and information on whether the child had multivariable analyses used Poisson regression to provide
been dewormed in the past year. For the siblings of selected prevalence ratios (PRs). Results between weighted and
SAC children, inclusion criteria were providing at least one unweighted regression models were compared and the results
stool and having the household’s selected child also be eligi- were nearly identical with no change in the model inferences;
ble for analysis. weighted models are presented. All analyses were run sepa-
Two variables to capture household wealth and income were rately for PSAC and SAC. Models explored associations
calculated and included in the analysis, one representing the between STH infection and demographic and predictive
household’s per capita weekly expenditure on food, and the factor variables. Predictors included in multivariable
 SOIL-TRANSMITTED HELMINTHS IN AN URBAN SLUM 1005

models were chosen by an automated procedure41 which to provide at least one analyzable stool. There were no sig-
uses the Bayesian information criterion42; tested variables also nificant differences between analyzed and non-analyzed chil-
included epidemiologically plausible interactions between dren in sex, age, or sector of residence.
covariates. For SAC, an alternative model was also run includ- All tables show results for selected children only, not sib-
ing SAC siblings of the selected SAC to gain a larger sample lings of selected SAC.
size; in this model, individual-level variables ascertained only Demographic characteristics. Demographic and confounder
for selected children were necessarily eliminated. In con- characteristics are shown, unweighted, in Table 1. Mean
trast, data from PSAC siblings of SAC was used only in household sizes were 5.3 (PSAC) and 6.4 (SAC) people,
the main analysis, in the determination of whether other roughly 70% of female household heads for both PSAC
children in the household had STH infection. and SAC completed primary school only, the population
Calculation of sensitivity using only the first and only was predominately of Luo or Luhya ethnicity, and house-
the first two stool samples relative to all three stools was holds had average per-capita weekly food expenditures of
performed under the assumption of 100% specificity. The US$ 5.42 (PSAC) and 4.95 (SAC). This population was highly
subset of analyzed children who provided three analyzable mobile with rates of travel outside Nairobi in the past year of
stools was included. 37.1% and 54.2% among PSACs and SACs, respectively.
Malaria was found in 11.2% of PSACs and 4.5% of SACs.
RESULTS Ninety-three percent of SACs (94.0% of the 98.6% spending
time outside the home) attended school; nearly half of those
Analyzed and unanalyzed children. In total, 236 PSAC, were informal schools. Of note, over 85% of PSACs were in
604 SAC, and 903 SAC siblings of selected SAC were either pre-primary or school programs. Children who spent
enrolled, and 205 PSAC, 487 SAC, and 685 SAC siblings of substantial time outside the home were relatively concen-
selected SAC were eligible for analysis, for a net response trated in a few institutions: the seven most common specific
rate of 58% among selected children and 64% among institutions attended by PSAC accounted for 37% of PSAC,
all children (Figure 1 shows the participant flowchart for whereas the seven most common institutions attended by
selected children, excluding siblings). Losses to enrollment SAC accounted for 65% of SAC. Slight differences in some
were primarily caused by inability to find household mem- values from those reported elsewhere33 are due to unweighted
bers who could consent or inability to find selected children presentation here and, in the case of recent deworming, var-
at home; losses to analysis were primarily a result of failure iant classification of an ambiguous answer.

Figure 1. Participant flowchart, selected children. *Almost entirely a result of failure to provide at least one analyzable stool sample.
1006 DAVIS AND OTHERS

Table 2 tals, or health centers (40.3%). Among SAC, 61.1% of those

Soil-transmitted helminth infection prevalences in PSAC and SAC,* not described as attending school had been dewormed in the
weighted to represent study area population. past year, compared with 63.8% of those in formal schools,
PSAC SAC 45.5% of those in informal schools, and 43.7% of those in
Infection type (N = 205) n (%) (N = 487) n (%)
private schools. However, as seen in Table 3, children attend-
Any STH 82 (40.5) 193 (40.7)
Ascaris 50 (24.1) 109 (22.7)
ing any school type had significantly lower STH infection
Moderate intensity Ascaris infection 33 (15.7) 60 (12.4) prevalence than those not attending school.
Heavy intensity Ascaris infection 6 (2.9) 5 (1.1) Factors associated with STH infection. Weighted unad-
Trichuris 47 (24.0) 135 (28.8) justed and adjusted associations of potential predictive factors
Moderate intensity Trichuris infection 2 (1.1) 3 (0.6) for STH infection are shown in Table 3; no interactions were
Heavy intensity Trichuris infection 0 (0.0) 0 (0.0)
Hookworm 0 (0.0) 1 (0.2) retained by the variable selection procedure. Adjusted asso-
Ascaris/Trichuris coinfection 15 (7.5) 51 (10.8) ciations are provided only for factors included in the final
Only light intensity infection† 41 (20.9) 126 (26.8) multivariable model. Sector of residence was highly signifi-
Any moderate intensity infection‡ 35 (16.9) 61 (12.6) cant on univariable analysis; Figure 3 shows weighted STH
Any heavy intensity infection§ 6 (2.9) 5 (1.1)
Provided 1 analyzable stool¶ 28 (13.7) 82 (16.8)
infection prevalence by sector, illustrating a range of 21.8%
Provided 2 analyzable stools¶ 33 (16.1) 102 (20.9) to 71.1% within an area in Kibera of 0.4 square kilometers.
Provided 3 analyzable stools¶ 144 (70.2) 303 (62.2) Travel outside Nairobi in the past year was not associated
*PSAC = pre-school age children; SAC = school-age children. with STH infection.
†Cutoffs: Ascaris: 1–4,999 eggs per gram (EPG); Trichuris: 1–999 EPG.
‡Indicates children with moderate infection intensity in at least one species, but no heavy
Significant factors after adjustment were age and pres-
infection. Cutoffs: Ascaris: 5,000–49,999 EPG; Trichuris: 1,000–9,999 EPG. ence of another infected child in the home (assessed only
§Cutoffs: Ascaris: > 50,000 EPG; Trichuris: > 10,000 EPG.
¶Not weighted as not applicable to study area population. for SAC); a history of being dewormed in the past year was
protective only among SAC. Household crowding was a
nonsignificant predictive factor among PSAC (PR = 1.1,
STH infection prevalences. Weighted prevalence and inten- 95% confidence interval [CI] = 1.0–1.2).
sity of STH infection and number of stools provided are shown In the alternative model for SAC (not shown), which
in Table 2. Overall, 40.7% of selected children (40.5% PSAC, included all SAC siblings and dropped all individual-level
40.7% SAC) had at least one STH infection; Ascaris and variables that were ascertained only for the selected chil-
Trichuris were responsible for almost all infections (23.0% and dren (e.g., deworming and school attendance), presence of
26.3%, respectively), hookworm infection was rare. Heavy another infected child in the home and sector of residence
infection was also rare (1.7%; N = 11; see Table 2 for species- remained significant.
specific intensity information). Almost two-thirds (64.8%) Sensitivity of Kato-Katz stool analysis. Sensitivity of the
of children provided three stools. Prevalences by age of any first one and two stools for Ascaris and Trichuris as com-
STH infection and Ascaris and Trichuris infection are shown pared with three stools, and prevalence of STH infection by
in Figure 2. The prevalence of STH infection became sub- number of stools received, are shown in Table 4. A drop in
stantial early in PSAC life; over 25% of PSAC had STH prevalence from 40.7% to 30.8% and in comparative sen-
infection by 1 year of age. If only the first stool had been sitivity to 80.4% would have been seen if only the first stool
available for each child (i.e., if the current programmatic had been available for each child.
mapping approach had been used), an STH infection preva- Quality control. Of 4,051 analyzable stool samples, 272
lence of 30.8% rather than 40.7% would have been seen. (6.7%) were reviewed. Discrepancies that would have changed
History of deworming. By respondent report (Table 1), the overall qualitative diagnosis of the child were found in
among children whose parents provided a yes or no answer two cases; changes were not made to results for data analysis.
about recent deworming, 58.9% of PSAC and 54.1% of SAC
had been dewormed in the last year. For SAC, major sources
for deworming were schools (55.0%), chemist shops (19.4%), DISCUSSION
and clinics, hospitals or health centers (19.8%); for PSAC, This urban slum environment is characterized by high STH
major sources were chemist shops (46.2%) and clinics, hospi- prevalence in children from an early age with dramatic spatial
heterogeneity on small geographic scales, modest inten-
sity, and frequent uncoordinated deworming from multiple
sources. Determining appropriate deworming coverage and
outcome goals is challenging in such a setting, and these
findings have several implications for the further develop-
ment of deworming guidelines.
First, PSAC in the study area had equivalent prevalence
of STH infection to SAC, despite reported high frequencies
of deworming. As we have reported elsewhere,33 they also
had more evidence for associated morbidity. Studies of
comparative infection prevalences in other settings would
help to determine how broadly these findings apply. If
they reflect a general reality about comparative burdens of
Figure 2. Prevalence of any soil-transmitted helminth infection
by child age: International Emerging Infections Program Area, Kibera, disease, PSAC should be actively targeted as deworming
Nairobi, Kenya. PSAC = pre-school-aged children (6 months to < 5 efforts expand. Though currently PSAC and SAC world-
years of age); SAC = school-aged children (5–14 years of age). wide receive similar proportional deworming coverage among
 SOIL-TRANSMITTED HELMINTHS IN AN URBAN SLUM 1007

Table 3
Association of any soil-transmitted helminth infection with predictive factors in PSAC and SAC*
PSAC (prevalence = 40.0%) SAC (prevalence = 39.6%)

PR Exposed vs. PR Exposed vs. PR Exposed vs. PR Exposed vs.

unexposed unexposed unexposed unexposed
Predictive factor (unadjusted) 95% CI (adjusted) 95% CI (unadjusted) 95% CI (adjusted) 95% CI

Age (per year) 1.31 1.13–1.51 1.24 1.04–1.49 0.94 0.91–0.98 0.92 0.89–0.96
Female sex 1.06 0.75–1.48 1.06 0.77–1.46 0.92 0.74–1.14 0.93 0.76–1.15
Ethnicity (reference group = Kisii)
Luhya 1.59 0.52–4.88 − − 1.52 0.71–3.23 − −
Luo 2.23 0.79–6.33 − − 1.79 0.88–3.64 − −
Other 2.05 0.60–7.02 − − 1.39 0.59–3.29 − −
Subvillage sector (shown: 10 vs. 1) 2.49 1.20–5.14 − − 2.36 1.53–3.63 − −
Education of female head of 0.95 0.88–1.02 − − 0.98 0.93–1.02 − −
household (per level)
Income proxy 1: food spending 0.97 0.91–1.03 − − 1.03 0.99–1.06 − −
Income proxy 2: item ownership 0.92 0.81–1.04 − − 0.95 0.88–1.02 − −
Mean Household Size 1.00 0.92–1.09 − − 0.94 0.89–0.99 − −
(increase of 1 person)
Number of children under 15 years in 0.99 0.9–1.09 − − 0.96 0.88–1.03 − −
home (increase of 1)
Crowding (people per bedroom) 1.10 1.00–1.21 − − 0.98 0.93–1.04 − −
PSAC living in the home 1.12 0.79–1.58 − − 0.96 0.69–1.32 − −
Any other child in home infected − − − − 1.59 1.25–2.01 1.66 1.31–2.1
Travel outside Nairobi (past year) 0.91 0.64–1.28 − − 0.84 0.67–1.05 − −
History of being dewormed within 0.99 0.70–1.40 − − 0.71 0.57–0.88 0.70 0.56–0.87
the past year
Malaria 0.90 0.51–1.61 − − 0.85 0.63–1.16 − −
Pica 1.21 0.85–1.71 − − 1.18 0.78–1.79 − −
Wearing shoes outside the home
(reference group = never)
Always 0.91 0.45–1.86 − − 0.89 0.47–1.68 − −
Usually 1.33 0.69–2.57 − − 1.18 0.64–2.18 − −
Sometimes 0.78 0.39–1.56 − − 1.25 0.68–2.32 − −
Current breastfeeding 0.43 0.23–0.83 0.64 0.31–1.34 − − − −
Type of school attended
(reference group = none listed)
Public − − − − 0.44 0.32–0.61 − −
Private − − − − 0.62 0.41–0.95 − −
In-/non-formal or unknown − − − − 0.65 0.49–0.86 − −
All predictive factors with P < 0.05 shown in bold. Adjusted values shown only for variables included in final multivariate model.
*PSAC = pre-school-aged children (6 months to < 5 years of age); SAC = school-aged children (5–14 years of age); PR = prevalence ratio.

children at risk (30.6% and 30.7%, respectively, in 2011 by PSAC were enrolled in pre-primary or nursery programs sug-
WHO estimates),43 continued scale-up of deworming with gests it might be possible to target these children by expanding
a focus on SAC may leave PSAC unprotected. Although from a school-based approach to a broader “institution-based”
some44,45 have raised broader questions about the effective- approach that would include such programs. The PSAC were
ness of current specific MDA approaches, it is widely more dispersed among multiple institutions (with 37% attend-
accepted that infected children do benefit from treatment,1 ing the seven largest institutions versus 65% of SAC), but a
and as such, actively including PSAC in the evolution of substantial proportion could still be reached even if smaller
deworming approaches is critical to controlling the global institutions were missed. This high involvement of young chil-
burden of STH infection. dren in institutional care could be a broader attribute of urban
Our experience in the IEIP villages of Kibera suggests areas related to availability of remunerative employment
that including PSAC in school-based deworming programs for their guardians; similar assessments elsewhere would be
may also be more feasible than it is often considered, at helpful. Efforts to reach out more broadly to institutions
least in urban areas. The finding that the vast majority of should also consider the importance of informal schools,
which enrolled a higher percentage of SAC than did formal
schools. Even school-based MDAs cannot achieve reasonable
coverage in such environments without actively including
informal institutions.
Even where an “institution-based” approach is not fea-
sible or sufficient, the high-density nature of urban slums
may lend itself more easily to community-based distribu-
tion; there is precedent for this approach.25At minimum,
for programs unable to target PSAC through institution- or
community-based approaches, an alternative might be assess-
Figure 3. Prevalence of soil-transmitted helminth infection in ing their infection prevalences before and after routine treat-
children 6 months to 14 years of age by sector: International Emerg- ment of SAC, to test the degree to which PSAC benefit from
ing Infections Program area, Kibera, Nairobi, Kenya. SAC deworming.
1008 DAVIS AND OTHERS

Table 4
Sensitivities and prevalences using Kato-Katz for diagnosis of soil-transmitted helminth infection with one and two stools as compared with
three stools*
Ascaris Trichuris Any STH

Children positive n Children positive n Children positive n

Stools examined % sensitivity (%) % sensitivity (%) % sensitivity (%)

First only 88.2 91 (20.4) 68.4 72 (17.1) 80.4 134 (30.8)

First two only 96.3 99 (22.2) 88.8 99 (22.3) 94.0 161 (36.6)
All three (ref) 105 (23.7) (ref) 115 (26.4) (ref) 176 (40.1)
*Among all children submitting three stools (N = 447).

Second, there are large STH infection prevalence differ- unexplored candidates are water, sanitation, and hygiene
ences between the study villages and the rest of Nairobi.32 factors,22,23 some of which were assessed in this study and
The STH prevalence among SAC in Nairobi was < 7% in one will be reported separately. The lack of a protective impact
study31 reported to constitute a randomly selected sample of deworming for PSAC may be in part caused by different
of Nairobi schoolchildren, though detailed methods are sources of medications for SAC and PSAC; it is possible that
not described. If these results are reproduced elsewhere, this medications vary in efficacy by source (depending on drug,
suggests that slums should be considered separate agro- dose, manufacturer,51 and storage practices.) Differential
ecological zones from their parent cities. (Furthermore, because force of infection, differential recall, and misidentification of
travel outside Nairobi was not associated with STH infection, medications could also play roles.
it seems likely that transmission of infections diagnosed in Third, as expected, our sensitivity and specificity from
Kibera is in fact taking place inside Kibera, not during visits stool testing were intermediate between those found in a
to rural areas; thus, these data provide no reason to expect higher prevalence (thus presumably higher intensity) set-
that MDA in rural areas would directly benefit slum village ting52 and in a lower intensity setting.38 The gain in sen-
residents.) Combining dispersed slums into a single agro- sitivity from additional samples was primarily seen with the
ecological zone may not be beneficial either; infection prev- addition of the second sample, consistent with others’ results
alences in the IEIP villages are also higher than those with respect to Ascaris and Trichuris.52 The magnitude of
found in another slum in western Kenya, where 16.2% of underestimation of prevalence can be expected to increase
children 10–18 years of age17 were reported to be infected. as STH intensity decreases with a successful control pro-
Results from other urban slums worldwide also show gram; this should be considered in treatment decisions, as it
wide variation in STH prevalences: 52.8% in Karachi20 and could lead to misclassification of areas into strata receiving
2.5% in Sao Paolo18 (all based on assessment of one stool lower frequency MDA or no MDA.
per child.) Several limitations are noted. The response rate was rela-
In the case of the Kibera villages, the presence of high tively low; however, analyzed and non-analyzed children
infection prevalence despite frequent deworming may be were similar on assessable demographic characteristics, and
due not only to a transmission-conducive environment,22,23 but the primary reason for nonenrollment (travel outside of
also to the uncoordinated nature of the deworming, occurring Kibera) was not associated with the primary outcome. We
at multiple times throughout the year in different sub- therefore feel the internal validity of significant findings
populations. In such a setting, the widespread simultaneous is high. However, as the study design was not powered to
drop in transmission pressure that a coordinated MDA would find predictive factor associations with ORs < 2, some asso-
have provided might not take place. Regardless, although ciations may have been missed. In addition, generalizability
current WHO guidance includes considering sanitation quality may be limited by differences between the IEIP area and the
in delineating ecological zones, this is not frequently practiced, rest of Kibera and other urban slums. These include ongoing
potentially leading to similar bundling of disparate transmis- free medical care provided by the study site clinic to IEIP
sion zones worldwide. participants, the relative social stability for which the area
Furthermore, the significant variations in prevalence within was originally selected to house IEIP, and a vibrant non-
the villages, between adjacent small areas, suggest that governmental organization (NGO) presence that underpins
treating even this slum as a single unit could lead to ineffi- much of the noted deworming. All of these factors, however,
cient use of deworming resources. Although it would not would be expected to decrease STH prevalence in the study
be practical for national programs to assess prevalences villages relative to other Kibera villages. Quality control reads
individually in smaller areas, as a research goal, identifying were performed for a smaller than expected proportion of
factors responsible for small-scale variations in prevalence samples. Finally, because of the rapid disappearance of hook-
could lead to more accurate prevalence categorization of worm eggs prepared by Kato-Katz, these rereads also could
small areas. not correct any errors in hookworm diagnosis. Hookworm
However, the predictive factors for infection we noted egg hatching between stool production and reading also
here were generally consistent with others’ findings46–50 and cannot be ruled out, though the practice of reading within
thus do not suggest specific changes to current approaches 14 hours of production, with cold box conditions for 2 to
for predicting infection distribution. Average household 6 of those hours, was designed to minimize hatching and
crowding could potentially be used as a predictor, though prevent error caused by egg degradation.
its association here fell short of significance. The strong role The findings in this study provide evidence that PSAC
of an infected sibling in the home is less helpful in school- should be included in national deworming program approaches,
based prevalence mapping and treatment. The most obvious that STH infection prevalences in slums should be assessed
 SOIL-TRANSMITTED HELMINTHS IN AN URBAN SLUM 1009

separately from those in their parent cities, and that hetero- Second edition. Geneva: World Health Organization. Available
geneity of infection prevalence is seen within ecological zones at: http://whqlibdoc.who.int/publications/2011/9789241548267_
eng.pdf. Accessed May 9, 2014.
as delineated under current practice. 9. Brooker S, Kabatereine NB, Smith JL, Mupfasoni D, Mwanje
MT, Ndayishimiye O, Lwambo NJ, Mbotha D, Karanja P,
Received January 27, 2014. Accepted for publication July 17, 2014. Mwandawiro C, Muchiri E, Clements AC, Bundy DA, Snow
RW, 2009. An updated atlas of human helminth infections:
Published online August 25, 2014.
the example of East Africa. Int J Health Geogr 8: 42.
Acknowledgments: We gratefully acknowledge the contributions of 10. Awasthi S, Verma T, Kotecha PV, Venkatesh V, Joshi V, Roy
Fredy Munoz, Michael Deming, Diya Surie, the ESACIPAC labora- S, 2008. Prevalence and risk factors associated with worm
tory staff, the CDC-KEMRI community interviewers, and the resi- infestation in pre-school children (6–23 months) in selected
dents of Soweto and Gatwekira, without whom this work would not blocks of Uttar Pradesh and Jharkhand, India. Indian J Med
have been possible. Sci 62: 484–491.
11. Anah MU, Ikpeme OE, Etuk IS, Yong KE, Ibanga I, Asuquo
Financial support: Funding for this study was obtained from USAID BE, 2008. Worm infestation and anemia among pre-school
(No. OG11-12021) through an inter-agency agreement with the children of peasant farmers in Calabar, Nigeria. Niger J Clin
Centers for Disease Control and Prevention and through the Nutri- Pract 11: 220–224.
tion Branch at the Centers for Disease Control and Prevention. 12. Garg R, Lee LA, Beach MJ, Wamae CN, Ramakrishnan U,
Disclosure: This paper is published with the permission of the Direc- Deming MS, 2002. Evaluation of the integrated Management
tor of the Kenya Medical Research Institute. of Childhood Illness guidelines for treatment of intestinal
helminth infections among sick children aged 2–4 years in
Disclaimer: The findings and conclusions in this report are the findings western Kenya. Trans R Soc Trop Med Hyg 96: 543–548.
and conclusions of the authors and do not necessarily represent the 13. Stoltzfus RJ, Chway HM, Montresor A, Tielsch JM, Jape JK,
official position of the Centers for Disease Control and Prevention. Albonico M, Savioli L, 2004. Low dose daily iron supplemen-
Authors’ addresses: Stephanie M. Davis, Caitlin M. Worrell, and tation improves iron status and appetite but not anemia,
LeAnne M. Fox, Parasitic Diseases Branch, Centers for Disease Con- whereas quarterly anthelmintic treatment improves growth,
trol and Prevention, Atlanta, GA, E-mails: [email protected], cworrell@ appetite, and anemia in Zanzibari preschool children. J Nutr
cdc.gov, and [email protected]. Ryan E. Wiegand and Gerard Lopez, Data 134: 348–356.
Management Activity, Division of Parasitic Diseases and Malaria, 14. Awasthi S, Peto R, Pande VK, Fletcher RH, Read S, Bundy
Centers for Disease Control and Prevention, Atlanta, GA, E-mails: DA, 2008. Effects of deworming on malnourished preschool
[email protected] and [email protected]. Kennedy O. Odero, children in India: an open-labeled, cluster-randomized trial.
Leonard Cosmas, John Neatherlin, and Joel M. Montgomery, Global PLoS Negl Trop Dis 16: e223.
Disease Detection and Emergency Response, CDC, Nairobi, Kenya, 15. Albonico M, Allen H, Chitsulo L, Engels D, Gabrielli AF,
E-mails: [email protected], [email protected], jneatherlin@ke Savioli L, 2008. Controlling soil-transmitted helminthiasis
.cdc.gov, and [email protected]. Sammy M. Njenga, Eastern in pre-school-age children through preventive chemotherapy.
and Southern Africa Centre for International Parasite Control, PLoS Negl Trop Dis 2: e126.
Kenya Medical Research Institute, Nairobi, Kenya, E-mail: SNjenga@ 16. Walson JL, Stewart BT, Sangaré L, Mbogo LW, Otieno PA,
kemri.org. Parminder Suchdev and Ruth Laird, Nutrition Branch, Piper BK, Richardson BA, John-Stewart G, 2010. Prevalence
U.S. Centers for Disease Control and Prevention, Atlanta, GA, and correlates of helminth co-infection in Kenyan HIV-1
E-mails: [email protected] and [email protected]. infected adults. PLoS Negl Trop Dis 30: e644.
17. Mehraj V, Hatcher J, Akhtar S, Rafique G, Beg MA, 2008.
Prevalence and factors associated with intestinal parasitic
REFERENCES infection among children in an urban slum of Karachi. PLoS
ONE 3: e3680.
1. Hall A, Hewitt G, Tuffrey V, de Silva N, 2008. A review and 18. Korkes F, Kumagai FU, Belfort RN, Szejnfeld D, Abud TG,
meta-analysis of the impact of intestinal worms on child Kleinman A, Florez GM, Szejnfeld T, Chieffi PP, 2009. Rela-
growth and nutrition. Matern Child Nutr 4: 118–236. tionship between intestinal parasitic infection in children and
2. World Health Organization, 2012. Soil-Transmitted Helminthiases: soil contamination in an urban slum. J Trop Pediatr 55: 42–45.
Eliminating Soil-Transmitted Helminthiases as a Public Health 19. Odiere MR, Opisa S, Odhiambo G, Jura WG, Ayisi JM, Karanja
Problem in Children: Progress Report 2001–2010 and Strategic DM, Mwinzi PN, 2011. Geographical distribution of schisto-
Plan 2011–2020. Geneva: World Health Organization. Available somiasis and soil-transmitted helminths among school chil-
at: http://whqlibdoc.who.int/publications/2012/9789241503129_ dren in informal settlements in Kisumu City, Western Kenya.
eng.pdf. Accessed May 9, 2014. Parasitology 138: 1569–1577.
3. World Health Organization, 2002. The World Health Report 20. Mumtaz S, Siddiqui H, Ashfaq T, 2009. Frequency and risk
2002. Geneva: World Health Organization. Available at: factors for intestinal parasitic infection in children under
http://www.who.int/whr/2002/en/whr02_en.pdf?ua=1. Accessed five years at a tertiary care hospital in Karachi. J Pak Med
May 9, 2014. Assoc 59: 216–219.
4. World Health Organization, 2005. Deworming for Health and 21. Staff Reporter, New mass de-worming drive at MCD schools.
Development: Report of the Third Global Meeting of the The Hindu. February 18, 2012. Available at: http://www
Partners of Parasite Control. Geneva, 29–30 November 2004. .thehindu.com/todays-paper/tp-national/tp-newdelhi/new-mass-
Geneva: World Health Organization. Available at: http:// deworming-drive-at-mcd-schools/article2905730.ece. Accessed
whqlibdoc.who.int/hq/2005/who_cds_cpe_pvc_2005.14.pdf. May 9, 2014.
Accessed May 9, 2014. 22. Ziegelbauer K, Speich B, Mäusezahl D, Bos R, Keiser J,
5. Hotez PJ, Fenwick A, Savioli L, Molyneux DH, 2009. Rescuing Utzinger J, 2012. Effect of sanitation on soil-transmitted hel-
the bottom billion through control of neglected tropical dis- minth infection: systematic review and meta-analysis. PLoS
eases. Lancet 373: 1570–1575. Med 9: e1001162.
6. World Health Organization, 2014. Intestinal Worms. Available at: 23. Strunz EC, Addiss DG, Stocks ME, Ogden S, Utzinger J,
http://www.who.int/intestinal_worms/en/. Accessed May 9, 2014. Freeman MC, 2014. Water, sanitation, hygiene, and soil-
7. World Health Organization, 2006. Preventive Chemotherapy in transmitted helminth infection: a systematic review and meta-
Human Helminthiasis: Coordinated Use of Anthelminthic Drugs analysis. PLoS Med 11: e1001620.
in Control Interventions: A Manual for Health Professionals 24. United Nations Human Settlements Programme, 2011. Global
and Programme Managers. Geneva: World Health Organiza- Report on Human Settlements 2011: Cities and Climate Change.
tion. Available at: http://whqlibdoc.who.int/publications/2006/ London: Earthscan. Available at: http://www.unhabitat.org/
9241547103_eng.pdf. Accessed May 9, 2014. downloads/docs/GRHS2011_Full.pdf. Accessed May 9, 2014.
8. World Health Organization, 2011. Helminth Control in School- 25. Wasif S. Deworming Campaign: Slum Kids to be Given Tablets
Age Children: A Guide for Managers of Control Programs. from Next Week. The Express Tribune. June 8, 2012. Available
1010 DAVIS AND OTHERS

at: http://tribune.com.pk/story/390381/deworming-campaign- effect of multiple stool sampling and use of different diagnos-
slum-kids-to-be-given-tablets-from-next-week/. Accessed May tic techniques. PLoS Negl Trop Dis 2: e331.
10, 2014. 39. Katz N, Chaves A, Pellegrino J, 1972. A simple device for
26. Khan MU, Shahidullah M, Barua DK, Begum T, 1986. Efficacy quantitative stool thick smear technique in Schistosomiasis
of periodic deworming in an urban slum population for para- mansoni. Rev Inst Med Trop Sao Paulo 14: 397–400.
site control. Indian J Med Res 83: 82–88. 40. Mathewos B, Alemu A, Woldeyohannes D, Alemu A, Addis Z,
27. Sarkar NR, Anwar KS, Biswas KB, Mannan MA, 2002. Effect Tiruneh M, Aimero M, Kassu A, 2014. Current status of soil-
of deworming on nutritional status of ascaris infested slum transmitted helminths and Schistosoma mansoni infection
children of Dhaka, Bangladesh. Indian Pediatr 39: 1021–1026. among children in two primary schools in north Gondar, north-
28. Sur D, Saha DR, Manna B, Rajendran K, Bhattacharya SK, west Ethiopia: a cross-sectional study. BMC Res Notes 7: 88.
2005. Periodic deworming with albendazole and its impact 41. Calgagno V, de Mazancourt C, 2010. Glmulti: an R package
on growth status and diarrheal incidence among children for easy automated model selection with (generalized) linear
in an urban slum of India. Trans R Soc Trop Med Hyg 99: models. J Stat Softw 34: 12.
261–267. 42. Schwartz G, 1978. Estimating the dimension of a model. Ann
29. Riesel JN, Ochieng’ FO, Wright P, Vermund SH, Davidson M, Stat 6: 461–464.
2010. High prevalence of soil-transmitted helminths in western 43. World Health Organization, 2013. Soil-transmitted helminthia-
Kenya: failure to implement deworming guidelines in rural ses: number of children treated in 2011. Wkly Epidemiol Rec
Nyanza Province. J Trop Pediatr 56: 60–62. 14: 145–152.
30. London School of Hygiene and Tropical Medicine. The Pre- 44. Taylor-Robinson DC, Maayan N, Soares-Weiser K, Donegan S,
dicted Distribution of Soil-Transmitted Helminthes in Sub- Garner P, 2012. Deworming drugs for soil-transmitted intes-
Saharan Africa: Predicted Prevalence. Available at: http:// tinal worms in children: effects on nutritional indicators, hemo-
www.thiswormyworld.org/maps/maps-continent/africa. Accessed globin and school performance. Cochrane Database Syst Rev
May 9, 2014. 11: CD000371.
31. Wamanda CA, 1986. Survey of the prevalence of intestinal para- 45. Godlee F, 2013. Treat the worms, but do other things too. BMJ
sites among some fees paying and non-fees paying primary 346: e8721.
schools in Nairobi, Kenya. M.Med. Dissertation, University 46. Gendrel D, Richard-Lenoble D, Kombila M, Baziomo JM,
of Nairobi Medical Library. Gendrel C, Nardou M, 1988. Breast-feeding and intestinal
32. Mwanthi MA, Kinoti MK, Wamae AW, Ndonga M, Migiro PS, parasites [in French]. Arch Fr Pediatr 45: 399–404.
2008. Prevalence of intestinal worm infections among pri- 47. de Souza EA, da Silva-Nunes M, Malafronte Rdos S, Muniz
mary school children in Nairobi City, Kenya. East Afr J PT, Cardoso MA, Ferreira MU, 2007. Prevalence and spa-
Public Health 5: 86–89. tial distribution of intestinal parasitic infections in a rural
33. Suchdev PS, Davis SM, Bartoces M, Ruth LJ, Worrell CM, Amazonian settlement, Acre State, Brazil. Cad Saude Publica
Kanyi H, Odero K, Wiegand RE, Njenga SM, Montgomery 23: 427–434.
JM, Fox LM, 2013. Soil-transmitted helminth infection and 48. Pullan RL, Bethony JM, Geiger SM, Cundill B, Correa-Oliveira
nutritional status among urban slum children in Kenya. Am J R, Quinnell RJ, Brooker S, 2008. Human helminth co-infection:
Trop Med Hyg 2: 299–305. analysis of spatial patterns and risk factors in a Brazilian
34. Feikin DR, Audi A, Olack B, Bigogo GM, Polyak C, Burke H, community. PLoS Negl Trop Dis 2: e352.
Williamson J, Breiman RF, 2010. Evaluation of the optimal 49. Roy EH, Haque AH, Siddique AK, Sack RB, 2011. Patterns
recall period for disease symptoms in home-based morbidity and risk factors for helminthiasis in rural children aged under
surveillance in rural and urban Kenya. Int J Epidemiol 39: 2 in Bangladesh. SAJCH 5: 78–85.
450–458. 50. Halpenny CM, Paller C, Koski KG, Valdés VE, Scott ME,
35. Breiman RF, Cosmas L, Njuguna H, Audi A, Olack B, Ochieng 2013. Regional, household and individual factors that influ-
JB, Wamola N, Bigogo GM, Awiti G, Tabu CW, Burke H, ence soil transmitted helminth reinfection dynamics in pre-
Williamson J, Oundo JO, Mintz ED, Feikin DR, 2012. school children from rural indigenous Panama. PLoS Negl
Population-based incidence of typhoid fever in an urban Trop Dis 7: e2070.
informal settlement and a rural area in Kenya: implications 51. Newton PN, Green MD, Fernández FM, 2010. Impact of poor-
for typhoid vaccine use in Africa. PLoS ONE 7: e29119. quality medicines in the ‘developing’ world. Trends Pharmacol
36. Lohr S, Sampling: Design and Analysis. Boston, MA: Brooks/ Sci 31: 99–101.
Cole Publishing Company; 1999. 52. Tarafder MR, Carabin H, Joseph L, Balolong E Jr, Olveda R,
37. 2009. Questionnaire Manager [computer program], version 2.9. McGarvey ST, 2010. Estimating the sensitivity and specificity
Guatemala City, Guatemala: Universidad del Valle. of Kato-Katz stool examination technique for detection
38. Knopp S, Mgeni AF, Khamis IS, Steinmann P, Stothard JR, of hookworms, Ascaris lumbricoides and Trichuris trichiura
Rollinson D, Marti H, Utzinger J, 2008. Diagnosis of soil- infections in humans in the absence of a ‘gold standard’. Int J
transmitted helminths in the era of preventive chemotherapy: Parasitol 40: 399–404.