Clinical predictors of malaria in Gambian children with fever or a history of fever.

Diagnosis of malaria in children is difficult without laboratory support because the symptoms and signs of malaria overlap with those of other febrile illnesses such as pneumonia. Nevertheless, in many parts of Africa diagnosis of malaria must be made without laboratory investigation. Therefore, a scoring system has been developed to assist peripheral health care workers in making this diagnosis. Four hundred and seven Gambian children aged 6 months to 9 years who presented to a rural clinic with fever or a recent history of fever were investigated. A diagnosis of malaria was made in 159 children who had a fever of 38 degrees C or more and malaria parasitaemia of 5000 parasites/microL or more. Symptoms and signs in children with malaria were compared with those in children with other febrile illnesses to identify features which predicted malaria. Symptoms and signs were incorporated into various logistic regression models to test which were best independent predictors of malaria and these regression models were used to construct simple scoring systems which predicted malaria. A nine terms model predicted clinical malaria with a sensitivity of 89% and a specificity of 61%, values comparable to those obtained by an experienced paediatrician without laboratory support. The ability of peripheral health care workers to diagnose malaria using this approach is now being investigated in a prospective study.

Levels of tumour necrosis factor and soluble TNF receptors during malaria fever episodes in the community.

The pyrogenic cytokine, tumour necrosis factor (TNF), is a mediator of malaria fever. Since high plasma levels of TNF are sometimes found in afebrile individuals with Plasmodium falciparum parasitaemia, it has been suggested that soluble forms of TNF receptors (sTNF-R55 and sTNF-R75) in the plasma may act to inhibit the pyrogenic effect of TNF. We have investigated plasma levels of TNF, sTNF-R55 and sTNF-R75 in relation to episodes of malaria fever detected in a cross-sectional study of 313 rural Gambian children during the malaria transmission season. Levels of TNF were significantly higher in the 20 children who had parasitaemia associated with fever than in 120 children who were afebrile despite malaria infection and 173 who had no detectable parasitaemia. In contrast, soluble TNF receptor levels did not differ between these clinical groups and, in a logistic regression model which included level of parasitaemia, we found TNF but not soluble TNF receptor levels to be associated with the presence of fever. These data support the role of TNF in malaria fever but suggest that soluble TNF receptors are not a major factor in modulating the fever.

An efficacy trial of the malaria vaccine SPf66 in Gambian infants--second year of follow-up.

In 1994, 630 Gambian infants were immunized with three doses of the synthetic polypeptide malaria vaccine SPf66 or with a control vaccine. No significant protection against first or total attacks of malaria was observed among the children who received SPf66. However, the period of follow-up was short. Thus, 532 children were followed for a second malaria transmission season during which 291 episodes of malaria were detected. Protective efficacies of SPf66 against first attacks of malaria and against all attacks of malaria were 8% [95% CI-20%, 30%] and 2% [95% CI-26% 24%] respectively. SPf66 did not provide any significant degree of protection to Gambian infants during a second year of follow-up.

The Gambian National Impregnated Bednet Programme: costs, consequences and net cost-effectiveness.

Clinical trials have indicated that treating mosquito nets with insecticide could be a potentially cost-effective method of preventing malaria. As malaria is one of the most common causes of death in children under five in developing countries, there has been substantial interest in whether such findings can be replicated for a country's control programme in practice. The cost-effectiveness of the Gambian National Insecticide-impregnated Bednet Programme (NIBP), from the viewpoint of providers (government and non-governmental agencies) and the community, has been calculated. Information was collected from existing records, interviews with NIBP personnel, observation and household surveys. Information is provided on the resource use consequences of the NIBP in terms of reduced expenditure on anti-malaria preventive measures, treatment in government health services, household financed treatment and "charity" (burial, funeral and mourning activities), as well as cash income lost as a result of child death. The annual implementation cost of the NIBP was D757,875 (US$91,864), of which 86% was recurrent cost. The estimated number of death averted was 40.56. The net implementation cost-effectiveness ratio per death averted and discounted life years gained were D3884 (US$471) and D260 (US$31.5), respectively. Adding the cost of all mosquito nets would increase the cost-effectiveness ratios by over five times, which is an important consideration for countries with a lower coverage of mosquito nets per capita. It is concluded that insecticide-impregnated mosquito nets are one of the more efficient ways of reducing deaths in children under 10 years in rural Gambia.

The Gambian National Impregnated Bed Net Programme: evaluation of effectiveness by means of case-control studies.

Two case-control studies, one on mortality and the other on malaria morbidity, were carried out in order to evaluate the impact of the Gambian National Insecticide Bed Net Programme during the second year of intervention and to explore the feasibility of such a study for the evaluation of programme effectiveness. For the mortality study, children 1-9 years old who died during the 1993 rainy season were matched by age and sex with 2 healthy controls from the same village. For the morbidity study, children 1-9 years old attending Fatoto or Jahalia Health Centres in The Gambia and who had fever and parasitaemia > or = 5000/microL were matched by age with a child attending the health centres without fever or parasitaemia. An additional healthy control was recruited from the case's village. No impact of insecticide-treated bed nets on mortality was detected and this was in keeping with the results obtained by prospective surveillance. A protective effect of insecticide-treated nets on malaria morbidity was detected when cases were compared with controls recruited at the health centres. However, this disappeared when cases were compared with controls recruited from the cases' villages. The mortality case-control study suggested that reducing the time between onset of disease and treatment may have an important impact on childhood mortality. In order to calculate programme cost-effectiveness, important for informed resource allocations to be made by health managers, it is essential to obtain evidence of effectiveness. This can be done by means of case-control studies, which are easier to carry out and require fewer resources than prospective surveillance. Nevertheless, it is necessary to be conscious of their pitfalls, particularly of the bias involved in the choice of cases and controls. The measurement of insecticide on the nets of the cases or controls is essential for such studies.

PIP: The impact of the Gambian National Impregnated Bed Net Program was evaluated during its second year of implementation through two case-control studies on malaria mortality and morbidity. The first study matched 167 children 1-9 years of age who died during the 1993 rainy season with two healthy controls (n = 334) of the same age and sex from the same village. The second study matched 143 children 1-9 years of age attending Fatoto or Jahalia Health Centers with fever or parasitemia with a child the same age attending the health centers without these symptoms and an additional healthy control from the same village. The mortality study failed to document any protective effect of insecticide-treated bed nets. Child survival was more closely linked to consultation with a village health worker during serious illness, household availability of chloroquine tablets, sponging, and maternal awareness of fever and convulsions as symptoms of malaria. A protective effect of the bed nets on malaria morbidity was revealed when cases were compared with health center controls, but this effect disappeared when cases were compared with controls from the same village. Going to bed early and father having a job other than working on the family farm were the two variables associated with significant protection against malaria in the multivariate analysis. Case-control studies are easier to conduct and require fewer resources than large prospective community-based studies. However, it is recommended that any future such studies of malaria control programs clarify the criteria used to select cases and controls and incorporate insecticide measurements of study nets.

The impact of a national impregnated bed net programme on the outcome of pregnancy in primigravidae in The Gambia.

In 1992, the Gambian national impregnated bed net programme (NIBP) introduced insecticide treatment of bed nets into half of the primary health care villages in The Gambia. One component of the evaluation of this programme was the determination of whether it had any impact on the outcome of pregnancy in primigravidae. From February 1992, 651 primigravidae were recruited into the study. Less than 50% of them used an insecticide-treated bednet. During the rainy season the prevalence of Plasmodium falciparum among primigravidae was lower, fewer babies were classified as premature, and the mean birth weight was higher in villages where treated bed nets were used than in control villages. Therefore, during the rainy season, despite the low use of insecticide-treated bed nets by Gambian primigravidae, the NIBP had some impact on the outcome of pregnancy, particularly on the percentage of premature babies, and this was probably due to the decreased risk of malaria infection achieved during this period.

Geographical perspectives on bednet use and malaria transmission in The Gambia, West Africa.

Insecticide-impregnated bednets are now widely accepted as an important tool in reducing malaria-related deaths in children in Africa. Defining the circumstances in which net treatment programmes are likely to be effective is essential to a rational development of this control strategy. In The Gambia a National Impregnated Bednet Programme was introduced into the primary health care system in 1992. Prior to its introduction baseline epidemiological and entomological studies were conducted throughout the country. These studies showed that in areas where mosquito biting nuisance was high, people protected themselves with bednets and that where mosquito densities (and therefore bednet usage) was low malaria prevalence rates were relatively high. Since the national programme is designed to assist only those people who already own a bednet (by providing the insecticide) an understanding of the factors which determine bednet ownership is needed to help evaluate the programme's effectiveness and provide guidelines for increasing bednet usage. Village scale bednet usage rates and malaria prevalence rates obtained from the baseline survey were correlated with certain geographical variables: dominant ethnic group, area, habitat, distance from the River Gambia and distance from the 'bluffline' (the interface between the sandstone soils and alluvial soils which border the river system). In a multiple regression analysis, bednet usage was independently associated with area (P < 0.001), ethnic group (P = 0.010), habitat (P = 0.006) and distance from the river (P = 0.013). A negative association of bednet usage with malaria prevalence persisted after allowing for the other variables. Malaria prevalence was not independently associated with area, ethnic group, habitat or distance from the river. Our analysis showed that the impregnated bednet programme is likely to be most effective in villages which are sited near to or on the alluvial soils in the middle and lower river zones. These villages, which were originally settled for easy access to the river (for transport) and its swampy margins (for rice production) are within the flight distance of mosquitoes that have their breeding sites on the poorly drained alluvial soils. Variation in malaria prevalence rates (after bednet usage has been taken into account) may be related to factors such as poverty and access to health care, and/or to localized differences in the ecology of The Gambia, which determine the duration and intensity of transmission. If the National Bednet Programme is to be effective throughout The Gambia it is vital to develop promotional activities which will encourage bednet usage in areas where nuisance biting by mosquitoes is low.

C-reactive protein and haptoglobin in the evaluation of a community-based malaria control programme.

When cross-sectional surveys are used to evaluate malaria intervention programmes in the community, the prevalence of morbidity is difficult to assess because of the fluctuating nature of malarial fever. We have therefore investigated the impact of bed net usage on 2 surrogate markers of malarial morbidity: (i) elevated C-reactive protein (CRP) (> 8 mg/L) plus detectable parasitaemia, as an indicator of malaria-induced acute-phase response; and (ii) reduced haptoglobin levels (< 180 mg/L), which in this population indicates malaria-induced intravascular haemolysis. Among 1505 Gambian children 1-5 years old, examined on a single occasion at the end of the malarial transmission season, 5% had parasitaemia plus fever, while 24% had parasitaemia plus elevated CRP, and 35% had low haptoglobin. The proportion of children who had parasitaemia plus elevated CRP was significantly lower among those who had slept under insecticide-treated bed nets than among those who did not use a bed net (16% vs. 34%, P < 0.003), and the proportion with low haptoglobin differed similarly (24% vs. 49%, P < 0.003). Use of an untreated bed net had a weaker effect on both indices (22% had parasitaemia plus elevated CRP, 34% had low haptoglobin). CRP and haptoglobin are simple and inexpensive to measure in large numbers of people, and these results suggest that they could be useful for the assessment of malaria intervention programmes.

A comparison of the efficacy of insecticide-treated and untreated bed nets in preventing malaria in Gambian children.

An evaluation of the Gambian national insecticide bed net programme, which has introduced insecticide treatment of bed nets into all primary health care (PHC) villages in The Gambia, provided an opportunity to compare the individual risk of malaria in children who slept under untreated or insecticide-treated bed nets. 2300 children 1-4 years old were selected for a survey at the end of the 1992 rainy season, 1500 from PHC villages and 800 from non-PHC villages. All malariometric indices were lower, and the mean packed cell volume was higher, in children who slept regularly under treated or untreated bed nets than in those who did not use a net. This study suggested that untreated bed nets provide some individual protection against malaria, although not as efficiently as that provided by insecticide-treated bed nets which were particularly effective at preventing infections accompanied by high parasitaemia.

Efficacy trial of malaria vaccine SPf66 in Gambian infants.

SPf66 malaria vaccine is a synthetic protein with aminoacid sequences derived from pre-erythrocytic and asexual blood-stage proteins of Plasmodium falciparum. SPf66 was found to have a 31% protective efficacy in an area of intensive malaria transmission in Tanzanian children, 1-5 years old. We report a randomised, double-blind, placebo-controlled trial of SPf66 against clinical P falciparum malaria in Gambian infants. 630 children, aged 6-11 months at time of the first dose, received three doses of SPf66 or injected polio vaccine (IPV). Morbidity was monitored during the following rainy season by means of active and passive case detection. Cross-sectional surveys were carried out at the beginning and at the end of the rainy season. An episode of clinical malaria was defined as fever (> or = 37.5 degrees C) and a parasite density of 6000/microL or more. Analysis of efficacy was done on 547 children (316 SPf66/231 IPV). No differences in mortality or in health centre admissions were found between the two groups of children. 347 clinical episodes of malaria were detected during the three and a half months of surveillance. SPf66 vaccine was associated with a protective efficacy against the first or only clinical episode of 8% (95% CI -18 to 29, p = 0.50) and against the overall incidence of clinical episodes of malaria of 3% (95% CI -24 to 24, p = 0.81). No significant differences in parasite rates or in any other index of malaria were found between the two groups of children. The findings of this study differ from previous reports on SPf66 efficacy from South America and from Tanzania. In The Gambia, protection against clinical attacks of malaria during the rainy season after immunisation in children 6-11 months old at time of the first dose was not achieved.

PIP: During December 1993 to November 1994, in the Upper River Division of The Gambia, 630 infants aged 6-11 months at time of first dose received three doses of malaria vaccine SPf66 or injected polio vaccine (IPV) and were followed up at home during the rainy season using active and passive case detection methods to determine the protective efficacy of SPf66 against clinical episodes of malaria due to Plasmodium falciparum. The researchers were able to use data on only 547 children (316 SPf66/231 IPV) to determine efficacy. The definition of clinical malaria was fever (37.5 degrees Celsius or higher) and a parasite density of at least 6000/mcl. The two groups were essentially the same in terms of mortality, health center admissions, parasite rates, or any other index of malaria. Health workers identified 347 clinical episodes of malaria during the three months of surveillance. SPf66 vaccine had a protective efficacy against first or only clinical episode of malaria of 8% (p = 0.5). Its protective efficacy was 3% against all clinical episodes of malaria. The results of this trial were different than those from earlier reports on SPf66 efficacy from South America and Tanzania. Possible reasons accounting for the different findings were a mistake in coding syringes for the third dose, substantially less intensity of malaria infection in The Gambia than South America, younger children in The Gambia than in Tanzania, genetic differences in the populations, and difference in length of follow-up. In conclusion, protection against malaria during the rainy season after immunization with SPf66 vaccine in infants aged 6-11 months did not occur.

Entomological evaluation of the Gambia's National Impregnated Bednet Programme.

Entomological studies were conducted in paired study villages in three of the five study areas used for the epidemiological assessment of the Gambia's National Impregnated Bednet Programme. Baseline data collected in 1991 were compared with post-intervention data from 1992, when one of each village pair (from areas II, III and V) was included in the treatment programme in which villagers' nets were dipped in permethrin. In a longitudinal study, indoor-resting densities of Anopheles gambiae s.l. in the treated villages were significantly reduced, when compared with their paired untreated village, in areas II (t = 3.32, 13 degrees of freedom, P = 0.006) and III (t = 3.71, 13 degrees of freedom, P = 0.003). However, this was not associated with higher outdoor-biting rates in the evenings in the treated villages than in the controls. The reduction in vector population was most evident in area II, where the treated village was relatively isolated and 74% of the population slept under a treated net. Anopheles gambiae s.l. were also collected during two fortnightly periods in area V, where the sporozoite rates in 1991 had been relatively high (7.65% and 6.07%). There was no significant decrease in the sporozoite rate in the treated village in area V in 1992, despite the fact that this village was over 5 km from its nearest neighbouring village. However, the proportion of villagers sleeping under a treated net in this village was < 50%. No clear evidence for an epidemiologically significant 'mass killing effect', resulting from the village-wide use of treated nets, emerges in this or any other study undertaken in The Gambia. However, such an effect is indicated by results from Salikene, where the village mosquito population was relatively isolated and where the majority of the population slept under a treated net. This situation is unlikely to be repeated in most Gambian villages in the primary-health-care scheme, since treated and untreated villages are generally interspersed and the mosquito populations move freely between them. We must therefore conclude that, on a national scale, the nets provide an exceptionally high level of personal rather than community protection. The mechanisms underlying the personal protection are, however, still unclear.

Mortality and morbidity from malaria in Gambian children after introduction of an impregnated bednet programme.

After the success of a controlled trial of insecticide-treated bednets in lowering child mortality, The Gambia initiated a National Insecticide Impregnated Bednet Programme (NIBP) in 1992 with the objective of introducing this form of malaria control into all large villages in The Gambia. Five areas (population 115,895) were chosen as sentinel sites for evaluation of the NIBP. During the first year of intervention a 25% reduction was achieved in all-cause mortality in children 1-9 years old living in treated villages (rate ratio 0.75 [95% CI 0.57-0.98], p = 0.04). If one area where the programme was ineffective was excluded, the reduction was 38% (0.62 [0.46-0.83), p = 0.001). A decrease in rates of parasitaemia and high-density parasitaemia, an increase in mean packed-cell volume (rate ratio 0.75 [95% CI 0.59-0.98], p = 0.04) and an improvement in the nutritional status of children living in treated villages were also detected. In a country such as The Gambia, where nets were widely used and which has a good primary health care system, it is possible to achieve insecticide-treatment of bednets at a national level with a significant reduction in child mortality; but at a cost which the country cannot afford.

Malaria prevalence is inversely related to vector density in The Gambia, West Africa.

Baseline epidemiological and entomological studies were conducted in 5 different areas of The Gambia before the introduction of a national malaria control programme, the objective of which was to treat all the bed nets belonging to people living in primary health care villages with insecticide. All malariometric indices used (parasite density, parasite rates, splenomegaly, and packed cell volume) indicated that malaria transmission was more intense in the east of the country than elsewhere. High transmission in the east was associated with a high sporozoite rate but not with the greatest vector abundance; the lowest malaria prevalence rates were found in villages which were close to very productive breeding sites of Anopheles gambiae s.l. Bed net usage was strongly correlated with vector density and the highest malaria rates were found in villages where bed net usage was relatively low. These results suggest that in The Gambia malaria prevalence rates are reduced where nuisance biting by mosquitoes is sufficient to encourage the population to protect themselves with bed nets.

Can malaria chemoprophylaxis be restricted to first pregnancies?

The harmful effects of malaria are most pronounced during first pregnancies and chemoprophylaxis is most effective when given at this time. However, restriction of chemoprophylaxis to first pregnancies might lead to enhanced susceptibility to malaria during second pregnancies. We have investigated this possibility by studying the outcome of second pregnancies in 165 Gambian women who had received either malaria chemoprophylaxis with Maloprim or placebo during their first pregnancy. Many of these primigravidae did not present until the third trimester of pregnancy so that some are likely to have experienced a malaria infection before they started medication. The prevalence of malaria infection of the blood and of the placenta during second pregnancies was similar in women who had received chemoprophylaxis during their first pregnancy and in those who had not, and the mean birth weights of babies born to women in each group were almost identical. Thus, in areas where the epidemiology of malaria is similar to that of The Gambia and where most women present relatively late in pregnancy, it may be possible to restrict malaria chemoprophylaxis to first pregnancies with consequent savings in cost and a reduction in drug pressure on Plasmodium falciparum.

Nationwide survey of bednet use in rural Gambia.

A nationwide survey was carried out to investigate the use of bednets in rural areas of the Gambia, particularly among children under 5 years old and among pregnant women. A stratified multistage design was used; 360 compounds in 60 villages were visited. Overall, 58% of beds had a net, with very little difference between villages with primary health care (PHC) facilities and those without (non-PHC). Bednet use was higher in the Central Region (76%) than in the Western and Eastern Regions (both 51%). It was highest among the Jola ethnic group (77%) and lowest among Sarahulis and other minority groups. Use of bednets was higher among target groups (such as infants, children under 5 years old, and pregnant women) than among the general population, and reached a level of over 90% in these groups in the Central Region. Use was associated with possession of a metal bed and a radio. Previous trials of impregnated bednets in the Gambia have been carried out in the Central Region, where bednet use is highest. The efforts of the National Impregnated Bednet Programme should therefore concentrate on the other regions and among the minority ethnic groups in order to increase bednet ownership.