21. Conquering Malaria

Burden of Disease

In 2001, the World Health Organization (WHO) ranked malaria as the eighth-highest contributor to the global disease burden as reflected in disability-adjusted life years (DALYs), and the second highest in Africa (WHO 2002a). The DALYs attributable to malaria were estimated largely from the effects of P. falciparum infection as a direct 0000cause of death and the much smaller contributions of short-duration, self-limiting, or treated mild febrile events, including malaria-specific mild anemia and neurological disability following CM (Murray and Lopez 1996, 1997). The estimate assumes that each illness event or death can be attributed only to a single cause that can be measured reliably. Table 21.1 shows deaths and DALYs from deaths attributable to malaria and to all causes by WHO region (WHO 2002a). It does not include the considerable toll caused by the burden of malaria-related moderate and severe anemia, low birthweight, and comorbid events (Snow and others 2003). Sub-Saharan African children under four represent 82 percent of all malaria-related deaths and DALYs. Malaria accounts for 2.0 percent of global deaths and 2.9 percent of global DALYs. In the African region of WHO, 9.0 percent of deaths and 10.1 percent of DALYs are attributable to malaria.

Recent analysis of falciparum malaria morbidity concludes that 515 (interquartile range 298 to 659) million cases occur yearly (table 21.2). This figure is 92 percent higher than the 278 million malaria cases estimated by WHO for 1998, which also includes those attributable to P. vivax, and 200 percent higher than previous estimates for areas outside of Africa (Snow and others 2005). While the malaria incidence globally is 236 episodes per 1,000 persons per year in all endemic areas, it ranges from about 400 to 2,000 (median 830) episodes per 1,000 persons per year in areas with intense, stable (hyperendemic and holoendemic) transmission; these areas represent 38 percent of all falciparumendemic areas.

Background

Recent estimates of malaria deaths have varied from 0.5 million to 3.0 million per year (Breman 2001; Breman, Alilio, and Mills 2004; Snow and others 2003). Of the 10.6 million yearly deaths in children younger than 5 years 8 percent are ascribed to malaria (Bryce and others 2005). In 1998, an empirical analysis of malaria mortality undertaken on behalf of WHO used malaria risk maps to capture measures of disability, morbidity, and mortality associated with P. falciparum prevalence rates among African populations and yielded an estimate of about 1 million (Korenromp and others 2003; Snow and Marsh 2002; Snow and others 2003; Snow, Trape, and Marsh 2001). Each malarious country must now measure its own burden and progress toward decreasing that burden (WHO 2005).

Geographic and climate-driven (mainly rainfall) models of suitability for malaria transmission characterize the diversity of malaria transmission across the African continent (Craig, Snow, and le Sueur 1999; Snow and others 1999). Four distinct areas can be identified:

• class 1, no transmission (northern and parts of southern Africa)

• class 2, marginal risk (mainly in some areas of southern Africa and in high-altitude [>1500 meters] settings)

• class 3, seasonal transmission with epidemic potential (along the Sahara fringe and in highlands)

• class 4, stable and unstable malarious areas (most areas south of the Sahara to southern Africa and below an altitude of around 1,500 meters).

Direct Consequences of P. falciparum Infection

Two major syndromes, CM and moderate (hemoglobin less than 11 grams per deciliter) and severe (hemoglobin less than 8 grams per deciliter) malarial anemia, contribute directly and significantly to malaria mortality (tables 21.3 and 21.4). The differences in numbers derive from the different methods of calculating the burden. Children presenting with an acute febrile disease, peripheral parasitemia, and low hemoglobin concentrations account for the majority of inpatient admissions in areas with stable transmission. The arbitrary definition of 5 grams of hemoglobin per deciliter is prognostic for a fatal outcome and proves useful clinically as a criterion for transfusion. Lactic acidosis commonly coexists with hypoglycemia and is (with coma, repeated convulsions, shock, and hyperparasitemia) an important predictor of death from severe malaria (White and Breman 2005; WHO 2000b).

The vast majority of deaths in developing countries occur outside the formal health service, and in Africa, most government civil registration systems are incomplete (Breman 2001; Greenberg and others 1989). Newer demographic and disease-tracking systems are being used globally and should help rectify the woefully inadequate vital statistics available for malaria and other diseases (INDEPTH 2002).

Health personnel usually attribute causes of death during demographic surveillance system surveys through a verbal autopsy interview with relatives of the deceased about the symptoms and signs associated with the terminal illness. Both the specificity and the sensitivity of verbal autopsy vary considerably depending on the background spectrum of other common diseases, such as acute respiratory infection, gastroenteritis, and meningitis, which share common clinical features with malaria (Korenromp and others 2003).

In malarious Africa, some 30 to 60 percent of outpatients with fever may have parasitemia. Monthly surveillance of households will detect a quarter of the medical events that are detected through weekly surveillance, and weekly contacts with cohorts identify approximately 75 percent of events detected through daily surveillance (Snow, Menon, and Greenwood 1989). Given the predominance of fevers, malaria case management in Africa and other endemic areas usually centers on presumptive diagnosis.

Estimates of the frequency of fever among children suggest one episode every 40 days. If we assume that the perceived frequency of fever in Africa is similar across all transmission areas (and possibly all ages), African countries would witness approximately 4.9 billion febrile events each year. Estimates indicate that in areas of stable malaria risk, a minimum of 2.7 billion exposures to antimalarial treatment will occur each year for parasitemic persons, or 4.93 per person per year (Snow and others 2003). While these diagnostic, patient management, and drug delivery assumptions are debatable, they indicate the magnitude of the challenges malaria presents.

The case-fatality rates of CM are high, even with optimal management. Murphy and Breman (2001) report a mean case-fatality rate of 19.2 percent and Snow and others (2003) cite a figure of 17.5 percent. Those who succumb at home without optimal treatment will have higher case-fatality rates.

Studies of neurological sequelae after severe malaria indicated that 3 to 28 percent of survivors suffered from such sequelae, including prolonged coma and seizures (MungAla-Odera, Snow, and Newton 2004). CM is associated with hemiparesis, quadriparesis, hearing and visual impairments, speech and language difficulties, behavioral problems, epilepsy, and other problems (table 21.3). The incidence of neurocognitive sequelae following severe malaria is only a fraction of the true residual burden, and the impact of milder illness is unknown.

Studies of children presenting to hospital with malaria and hemoglobin of less than or equal to 5 grams per deciliter indicate a median transfusion rate of 80.1 percent. Thus, 275,400 to 442,290 surviving SMA admissions, newborn through 14 years, will be exposed to blood transfusion each year in Sub-Saharan Africa. As a result, each year 5,300 to 8,500 children, age birth through 14 years, living in stable endemic areas of Africa are likely to acquire HIV infection because of exposure to blood transfusion to manage SMA (Colebunders and others 1991; Savarit and others 1992).

Indirect and Comorbid Risks

The DALY model of malaria does not sufficiently take it into account as an indirect cause of broader morbid risks. Some consider anemia to be caused indirectly unless linked to acute, high-density parasitemia. Similarly, low birthweight may also be indirectly attributable to malaria, and a child's later under-nutrition and growth retardation linked to malaria infection enhances the severity of other concomitant or comorbid infectious diseases through immune suppression. Thus, malaria infection contributes to broad causes of mortality beyond the direct fatal consequences of infection and is probably underestimated (Breman, Alilio, and Mills 2004; Snow and others 2003).

In Africa, pregnant women experience few malaria-specific fever episodes but have an increased risk of anemia and placental sequestration of the parasite. Maternal clinical manifestations are more apparent in areas with less intense transmission, particularly in Asia. Estimates indicate that in Sub-Saharan Africa, malaria-associated anemia is responsible for 3.7 percent of maternal mortality, or approximately 5,300 maternal deaths annually.

Prematurity and intrauterine growth retardation resulting in low birthweight associated with maternal malaria account for 3 to 8 percent of infant mortality in Africa (Steketee and others 1996, 2001). Assuming an infant mortality rate of 105 per 1,000 live births, Snow and others (2003) calculate that in 2000, 71,000 to 190,000 infant deaths were attributable to malaria in pregnancy (table 21.3). Other studies indicate that malaria-associated low birthweight accounted for 62,000 to 363,000 infant deaths (Murphy and Breman 2001).

Anemia among African children is caused by a combination of nutritional deficiencies and iron loss through helminth infection, red cell destruction, decreased red cell production as a result of infectious diseases, and genetically determined hemoglobinopathies. Chronic or repeated infections, often associated with parasite resistance to drugs, are more likely to involve bone marrow suppression (Menendez, Fleming, and Alonso 2000).

Murphy and Breman (2001) estimate that 190,000 to 974,000 deaths per year in Sub-Saharan Africa are attributable to SMA. Children residing in areas where the prevalence of P. falciparum was more than 25 percent had a 75 percent prevalence of anemia. By modeling the relationship between anemia and parasite prevalence, Snow and others (2004) found that mild anemia rose 6 percent for every 10 percent increase in the prevalence of infection. Reducing the incidence of new infections through ITNs or the prevalence of blood-stage infections through chemoprophylaxis or IPT for children halved the risk of anemia.

Caulfield, Richard, and Black (2004) report that iron, zinc, and protein-calorie deficits are responsible for a considerable amount of malaria-related mortality and morbidity and indicate that 57.3 percent of deaths of underweight children under five are attributable to nutritional deficiencies. One striking feature of the global distribution of anthropometric markers of undernutrition is its congruence with the distribution of endemic malaria. Improved growth among young children has more recently been demonstrated in The Gambia and Kenya in a comparison of those protected and unprotected by ITNs (Ter Kuile and others 2003).

Early during the HIV epidemic, Greenberg and others (1988) and Greenberg (1992) demonstrated that malaria-associated anemia treated with unscreened blood transfusions contributed to HIV transmission. At the same time, two longitudinal cohort studies in Kenya and Uganda and one hospital-based case-control study in Uganda demonstrated that HIV infection approximately doubles the risk of malaria parasitemia and clinical malaria in nonpregnant adults and that increased HIV immunosuppression is associated with higher-density parasitemias (French and others 2001; Whitworth and others 2000). In pregnant women, the presence of HIV increases the rate and intensity of parasitemia and frequency of anemia (Ter Kuile and others 2004). The increasing incidence of HIV-associated febrile illnesses may lead to increased use of antimalarials. Some believe that the recommended use of trimethoprim-sulfamethoxazole for prophylaxis of bacterial pneumonia and other infections in HIV/AIDS patients may contribute to SP resistance and that monitoring is required. Yet, evaluation of trimethoprim-sulfamethoxazole for malaria prophylaxis in Mali did not show any increases in parasite resistance mutations specific for these drugs (Thera and others 2005).

Malaria accounts for 13 to 15 percent of medical reasons for absenteeism from school, but little information is available on the performance of parasitized schoolchildren (Holding and Kitsao-Wekulo 2004). A randomized placebo control study of chloroquine prophylaxis in Sri Lankan schoolchildren demonstrated an improvement in mathematics and language scores by those who received chloroquine but found no difference in absenteeism (Fernando and others 2003) As noted earlier, malaria may result in low birthweight, and low birthweight can lead to a range of persistent impaired outcomes, predominantly behavioral difficulties, cerebral palsy, mental retardation, blindness, and deafness. The recently launched studies of intermittent preventive treatments during infancy (IPTi) should provide a more precise means of examining the benefits of IPTi and consequences on learning and performance of infection early in life (Holding and Kitsao-Wekulo 2004; Rosen and Breman 2004; Schellenberg and others 2005).