D.A.P. Bundy

Re-assessing the global prevalence and distribution of lymphatic filariasis.

This paper estimates the global burden of lymphatic filariasis based on a review of the published literature on infection and disease surveys. A method for aggregating and projecting prevalence data from individual studies to national, regional and global levels, which also facilitates the estimation of gender and age-specific burdens, is presented. The method weights in favour of the larger, and hence presumbably more reliable, studies and relies on estimated empirical relationships between gender, age, infection and disease in order to correct studies with incomplete data. The results presented here suggest that although the overall prevalence of filariasis cases is 2.0% globally (approximately totalling 119 million cases), the disease continues to be of considerable local importance, particularly in India and Sub-Saharan Africa. Estimates by age and gender clearly show that, unlike other helminth infections, filariasis is mainly a disease of the adult and older age-classes and appears to be more prevalent in males. This work suggests that the derivation of more accurate estimates of the burden of filariasis will require a better understanding of both the epidemiology and the spatial aspects of infection and disease. It also suggests that filariasis is preventable based on a geographically targeted strategy for control.

Global mapping of lymphatic filariasis

Disease maps are becoming increasingly important in infectious disease epidemiology and control. For lymphatic filariasis, the development of such maps has been hampered in the past by the lack of data on the geographical distribution of levels of infection or disease. Here, Edwin Michael and Don Bundy present an atlas for this parasitic disease derived from a recently compiled geographical database. Focusing on mapping and analysis of case prevalence data at the global and regional levels, the authors show how mapping the geographical distribution is integral not only to assessing spatial patterns in the infection and disease distribution but also to stratifying endemic areas by infection and/or disease rate.

Towards an Atlas of Human Helminth Infection in sub-Saharan Africa: The Use of Geographical Information Systems (GIS)

The value of a geographical perspective to infectious disease epidemiology and control has long been recognized. However, the labour required to produce maps, and keep them up to date, has inhibited the development of this area, and very little is currently known about the spatial distribution of parasitic infections other than malaria, trypanosomiasis and onchocerciasis. A recent initiative by an international group of collaborators is attempting to redress the absence of detailed spatial information on the major helminth infections of humans. In this article, Simon Brooker and colleagues describe progress made by this initiative in mapping helminth infections in sub-Saharan Africa, highlighting the value as well as the limitations of this empirical mapping approach.

Morbidity and mortality due to ascariasis: re-estimation and sensitivity analysis of global numbers at risk.

This paper presents estimates of the global numbers of people at risk from morbidity related to infection with Ascaris lumbricoides and the numbers of deaths from this infection. Morbidity is classified into 4 types: deficits in growth and fitness which are contemporaneous with infection, or permanent, overt acute illness of mild to moderate severity, and complications involving hospitalization. The estimation of morbidity is based on theoretical models of parasite distributions developed in previous papers. A sensitivity analysis is carried out in which parameters of the model are varied using a Latin hypercube sampling technique. The results estimate ≈1300 million infections globally with 59 million at risk of some morbidity. The estimate for acute illness is 12 million cases per year with ≈10 000 deaths. Most morbidity is in children. Sensitivity analysis suggests that infection estimates will not vary greatly with changes in parameter values but that morbidity estimates may be highly variable.

Patterns of concurrent hookworm infection and schistosomiasis in schoolchildren in Tanzania

A cross-sectional study of 6897 schoolchildren in 59 out of the 155 primary schools in Magu District on the shores of Lake Victoria, Tanzania, was undertaken in 1997 to determine the prevalence of single- and multiple-species helminth infection. Schistosoma haematobium, hookworm (primarily Necator americanus) and S. mansoni were the most common helminth species infecting schoolchildren in the district. The prevalences of Ascaris lumbricoides and Trichuris trichiura were negligible (< 1%). Anaemia and stunting were highly prevalent and widespread. Hookworm and S. mansoni occurred more frequently in multiple infections with other helminths than as single-species infections, but triple-species infection was rare. Analysis of the frequency distribution of infection amongst schools showed that prevalences of S. haematobium and hookworm tended to be normally distributed, with medians 75% and 45%, respectively, while the distribution of S. mansoni was markedly skewed such that only 17% schools had a prevalence greater than 20%. An inverse association between S. mansoni and S. haematobium was observed. Geographical information system (GIS) analysis indicated that S. mansoni infection was highly prevalent only along the shore of Lake Victoria, whilst S. haematobium was homogeneously prevalent everywhere except the lakeshore. This pattern appears to reflect the distribution of schistosome species-specific snail intermediate hosts. The results imply that joint treatment for hookworm infection and schistosomiasis would be beneficial throughout the district.

Hookworm infection in pregnancy.

Hookworm infection is a recognized precipitating factor for iron deficiency anaemia (reviewed by SCHAD & WARREN 1990). In developing societies where hookworm infection is prevalent diets are marginal in quality and fertility rates are high it is probable that hookworm infection during pregnancy is a significant contributor to the burden of disease in women. Direct estimation of the disease burden is not possible using currently available data. This analysis pursues the simpler goal of providing an order of magnitude estimate of the number of women who are both pregnant and infected with hookworm burdens of a magnitude that is likely to be associated with disease. A basic sensitivity analysis is undertaken for data for sub-Saharan Africa before attempting a global estimate. (excerpt)

Morbidity and mortality due to Ascaris-induced intestinal obstruction

We examined epidemiological aspects of Ascaris-induced intestinal obstruction (AI-IO) through analysis of published reports on the subject. In 9 studies of > or = 100 patients admitted to hospital due to ascariasis, intestinal obstruction was the single most common complication and accounted for 38-87.5% of all complications (weighted mean 72%). The proportion of intestinal obstruction caused by ascariasis was identified in 14 studies from 7 countries with varying degrees of endemic ascariasis. Using relevant data on the duration of the study, the number of beds in the reporting hospital, and the number of hospital beds/1000 population in the area, the number of cases of AI-IO/year/1000 population was estimated from 11 studies. Both the proportion of AI-IO (range 0-0.71) and the number of cases of AI-IO/year/1000 population (range 0-0.25) were significantly related, in a non-linear manner, to the local prevalence of ascariasis (range 0.01-0.92). In 12 studies of > or = 30 patients with AI-IO, the case fatality rates ranged from 0 to 8.6% (weighted mean 5.7%). The mean age of patients with AI-IO was < or = 5 years in 6 of 7 studies in which age was specified.

Episodic adenolymphangitis and lymphoedema in patients with bancroftian filariasis

In order to explore the relationship between acute and chronic disease, age-specific data on the frequency and duration of episodic adenolymphangitis (ADL) in patients with 3 defined grades of lymphoedema in bancroftian filariasis were examined. The age distribution of grades I and II exhibited a convex age profile, but that of grade III showed a monotonic increase. The mean duration of oedema increased with its grade (grade I, 0.3 years; grade III, 9.9 years). The mean number of ADL episodes in the previous year for all cases was 4.2 and it increased with grade (grade I, 2.4 and grade III, 6.2). The mean duration of each ADL episode for all cases was 4.1 d and it was independent of grade and age. The mean period lost to ADL episodes in the previous year was 17.5 d; it increased from 9.4 d with grade I to 28.5 d with grade III. The results imply that there is a dynamic progression through the grades of lymphoedema and that the frequency of ADL episodes is positively associated with this progression. However, the study design could not separate cause from effect.

Schools for health: Focus on health, education and the school-age child

Mortality in children under five years old has been dramatically reduced through successful programmes of immunization and control of diarrhoeal diseases. UNICEF estimates that some 90% of children in developing regions now survive to reach school age. These survivors face new and continuing threats to their health, which can affect their physical development and may also prevent them taking full advantage of their only opportunity for formal education. The physical and mental growth of the 1000 million school-age children today will influence how the world is shaped for coming generations. Yet the health problems of this age group have received little attention. Recognizing the importance of this age group, a workshop funded by the Edna McConnell Clark Foundation was held 10-13 November 1994 in Fort Mitchell, Kentucky, USA, to review what is known about the health of school-age children, what is or can be done to improve their health, and what steps must be taken to find ways to improve the health and educational achievement of this important segment of the worlds population. Don Bundy and Helen Guyatt here report on the workshop, which had three major conclusions: (1) the school-age children of developing countries face health problems that remain neglected and poorly understood; (2) an important research need is to develop simpler means of monitoring the health status of school-age children and evaluating the impact of public health interventions in this age group; and (3) two strategies are available to address this public health problem. The first is to develop further and test programmes that appear, from available evidence and pilot studies, to offer effective means of improving the health of this age group at reasonable cost, and to be sustainable; and the second is, over a longer term, to develop the capacity within countries to assess the health problems of school-age children and devise cost-effective strategies to address these problems. This report attempts, in brief form, to survey what is known about the health status of school-age children, to discuss the possible benefits to health and learning that accrue from health interventions, and to suggest some avenues currently available for both research and application.

EPIFIL: the development of an age-structured model for describing the transmission dynamics and control of lymphatic filariasis.

Mathematical models of transmission dynamics of infectious diseases provide a useful tool for investigating the impact of community based control measures. Previously, we used a dynamic (constant force-of-infection) model for lymphatic filariasis to describe observed patterns of infection and disease in endemic communities. In this paper, we expand the model to examine the effects of control options against filariasis by incorporating the impact of age structure of the human community and by addressing explicitly the dynamics of parasite transmission from and to the vector population. This model is tested using data for Wuchereria bancrofti transmitted by Culex quinquefasciatus in Pondicherry, South India. The results show that chemotherapy has a larger short-term impact than vector control but that the effects of vector control can last beyond the treatment period. In addition we compare rates of recrudescence for drugs with different macrofilaricidal effects.