Ecology and epidemiology of integrated malaria vector management in Dar es Salaam, Tanzania

Malaria remains one of the major contributors to the global burden of disease with
approximately 70% of the clinical malaria attacks occurring in sub-Saharan Africa. Sub-
Saharan Africa has the highest risk as ideal climatic conditions for transmission coincide with
occurrence of some of the most efficient malaria vectors, namely Anopheles gambiae s.s.,
Anopheles arabiensis and Anopheles funestus.. Even though it is estimated that by the year
2030 more than 50% of the African population will live in towns and cities, relatively little is
known about urban malaria epidemiology, larval ecology and adult mosquito behaviour.
Although integrated malaria control programs including environmental management and
larviciding have proven successful before the Global Eradication Campaign started in 1955,
they were neglected after the invention of DDT. Lately interest into these control measures
has revived but it remains to be determined whether they are feasible and cost-effective in
urban Africa.
The overall goal of the research presented in this thesis was to enhance current understanding
of urban malaria epidemiology and ecology and to take an in-depth look at the effectiveness
of larviciding with Bacillus thuringiensis (Bti) in the context of the Urban Malaria Control
Program (UMCP) in Dar es Salaam, Tanzania. Our findings are based on data derived from
the first 3 years of the UMCP, where data collection started in March 2004. The project area
includes 5 wards in each of the 3 municipalities which consist of 67 mitaa covering an area of
55 km2 in which 611,871 people lived during the population census of 2002. Achieving the
UMCPs objectives fundamentally relies on three component activities: 1) Mapping and
surveillance of potential Anopheles breeding sites, 2) Monitoring of adult mosquito densities,
and 3) Household surveys with questionnaires and blood smears testing for malaria parasite
infection. In the third year of the UMCP, beginning in March 2006, the routine application of
the microbial larvicides Bti in open habitats and Bs in closed habitats was initiated in 3 of the
15 wards in the study area, adding to existing interventions such as bednets, house screening,
ceiling boards, repellents, spray and coils. At the same time a detailed survey of mosquito
biting behaviour, human behaviour and domestic protection measures was conducted in 12
Ten Cell Units (TCU), the smallest subunit of local government in Tanzania, which presented
the highest An. gambiae s.l. densities during the early period of the UMCP surveillance
system. Human landing catch (HLC) was conducted in 216 houses on an hourly basis indoors
and outdoors from 6 pm till 7 am and residents were interviewed about their sleeping
behaviour, where they spend their evenings and what kind of preventive measures against
malaria they use. Personal protection of an insecticide treated net (ITN) was evaluated using
an extension of a recently developed mathematical model.
Overall An. gambiae s.l. exhibited a classical hourly biting pattern. In contrast one of the
complex’s component sibling species, namely An. arabiensis, had an early biting peak before
10 pm. Both sibling species, namely An. gambiae s.s. and An. arabiensis, as well as An.
funestus and An. coustani were highly exophagic. This behaviour led to a reduced personal
protection against exposure to An. gambiae s.s. by ITNs which conferred 59% reduction of
exposure in Dar es Salaam compared to 70% in rural Tanzania. An. arabiensis is a vector of
only modest importance in Dar es Salaam which is fortunate because ITNs only conferred
38% protection against exposure to this species of mosquito. ITNs conferred slightly less
protection against exposure to malaria vectors in good quality houses. This is mainly because
people living in good houses tend to spend more time indoors before they go to bed.
An. gambiae s.l. is the most important vector in Dar es Salaam , responsible for an EIR
(entomological inoculation rate) of 1.00 infectious bites per person per year whereas An.
funestus has an EIR of 0.13. Surprisingly, An. coustani also acts as a notable vector in Dar es
Salaam with an EIR of 0.20 infectious bites per person per year. Malaria transmission is
seasonal with two peaks of malaria prevalence during and after the two rainy seasons. Malaria
prevalence was only related to EIR in children under 5 years of age, with a classical ageprevalence
distribution similar to most of rural Africa. Malaria prevalence steadily declined
from 2004 onwards as the use of window screenings, ceiling boards and more effective drugs
like amodiaquine and artemisin-based drugs increased. ITNs (prevalence reduction estimate
20%, 95% CI 0%-36%; P=0.060; year 1) and ceiling boards (prevalence reduction estimate
22%, 95% CI 3%-38%; P=0.026; year 2) conferred modest personal protection and reduced
malaria prevalence by approximately one fifth. By comparison, a much greater reduction
(prevalence reduction estimate 50%, 95% CI 20%-64%; P=0.002) of malaria prevalence was
achieved by larviciding with Bti. This was mainly achieved through major reductions of An.
gambiae during July and August when most of the sporozoite infected mosquitoes were
caught, combined with all-year-round suppression of the secondary vectors, namely An.
funestus and An. coustani. This major achievement was only possible through the novel
surveillance and staff management procedures developed by the UMCP to enable effective
community based implementation in a decentralized manner. Standards of the surveillance
improved greatly after the onset of the program with realized reaction times to vector
surveillance at observations being one day, week and month at ward, municipality and city
level, respectively.
These results of changing biting behaviour of the main malaria vectors in urban settings and
the therefore lower but still useful personal protection offered by ITNs call for additional
complementary vector control methods such as environmental management or larviciding.
The UMCP demonstrated that major reductions in malaria prevalence can be achieved
through routine application of microbial larvicides with its new practical management and
surveillance system. As these represent the early results of the program, we expect substantial
improvement with time and investment. Here we demonstrated for the first time since before
the Global Eradication Campaign era, a success story of a malaria control program integrating
larviciding, which could be easily adapted by other African cities as a cost-effective option for
malaria prevention.