Malaria in Africa

The extent of the challenge is illustrated by the map of Africa (Figure 1-1), showing a vast swath between the northern coastal countries and the most southerly ones (excluding most of the horn of Africa) where malaria transmission is intense and malaria is ever-present. This area is ringed by a region prone to seasonal malaria transmission or periodic epidemics—with often devastatingly high mortality rates. More than half a billion people— more than two-thirds of the African population—live with malaria year in, year out, most of them where transmission is intense. The vast majority of the million or so deaths from malaria occur in Africa, mostly among children who have yet to acquire sufficient immunity to protect them from heavy malaria infections. The prevalence of malaria varies with levels of endemicity, but it averages to at least one acute clinical episode per year for which some treatment is sought, or on the order of half a billion treated episodes.

FIGURE 1-1

Distribution of Endemic Malaria in Africa. SOURCE: Reprinted with permission from MARA/ARMA, Copyright 2001. Taken from Craig, MH, Snow RW, le Sueur D. 1999. A Climate-Based Distribution Model of Malaria Transmission in Sub-Saharan Africa. Parasitology (more...)

It is worth remembering that the malaria map would have looked very different half a century ago, and that control measures are largely responsible for shrinking the highly-endemic zone. The low- and no-transmission areas of southern Africa (including Namibia, Swaziland, South Africa, Botswana, and Zimbabwe) were previously highly endemic—and could become so again, if control measures fail.

The map lays out the boundaries of a very large problem, but does not tell the whole story. Who becomes sick and who dies is what is important. Once the discussion moves beyond acknowledging that malaria is a “big” problem, however, estimates of the relevant numbers—malaria cases, deaths from malaria—and the distribution among the population, geographically and by age, vary so widely that they can appear unusable for policy and planning. This lack of specificity is at least part of what makes tackling the problem so unsettling.

Malaria in Asia

There are about 27 million cases and 30,000 deaths from falciparum malaria in Southeast Asia each year, with no perceptible decline over the last decade (WHO Regional Office for South-East Asia,). Unlike Africa, where acquired semi-immunity protects most adults from severe clinical disease and death, intermittent infections in Asian endemic areas are usually symptomatic, not uncommonly with very high parasitemias (parasite numbers in the blood) and a risk of death. Apart from the sheer numbers— not approaching African levels, but substantial—Southeast Asia has been the epicenter of drug-resistant malaria precisely because of this pattern of infection and treatment of nonimmune individuals with heavy parasite loads. Some of the most difficult problems arise in border areas, where refugees have gathered after political conflicts. Isolated tribal groups in various countries, with little or no access to health care, also suffer disproportionately from malaria.

Malaria in the New World

North America and most of the Caribbean are free from malaria transmission, a result of eradication efforts that began in earnest in the 1950s. But 200 million people live where there is still some risk of transmission, in 21 countries of Central and South America, and the Caribbean, about 20 percent of them in areas considered “high risk” for the region. About one million cases and 200 deaths were recorded in 2001. Drug resistance is becoming widespread, and changes in drug policy have already been made in some places. And the same factors that can disrupt control in Africa and Asia—migration, civil unrest, poor economic conditions—occur in the Americas.

The Varied and Far-Reaching Effects of Malaria²

Malaria is not simply a matter of episodes of illness and deaths, as enormous as those burdens are. Effects vary from place to place, but nearly all people living in endemic areas will at some stage in their lives become infected with malaria through the bite of a mosquito, and after the parasite has multiplied in numbers, they will suffer an attack of fever and other symptoms when large numbers of parasites are released into the bloodstream. Untreated, these episodes may progress in severity and naturally resolve, or the person may become overwhelmed and die. Or antimalarial drugs may be used to quell the infection. Functional immunity develops in those who survive the early episodes, if they are continually exposed to malaria from birth, preventing most symptomatic recurrences at older ages.

But that is not the whole story. There are less obvious, but equally serious, consequences of the chronic infections and repeated reinfections that characterize life in high-transmission areas, including most of sub-Saharan Africa. Chronic, subclinical infections can cause anemia and pre-dispose to undernutrition. These processes may further increase the chances of severe malaria developing with a subsequent infection, and possibly of more severe outcomes of infections with other pathogens. Unlike other adults in endemic areas, pregnant women are themselves more susceptible to malaria's most severe effects—including death—and asymptomatic infection of the placenta significantly reduces the weights of their newborn children, reducing their chances of surviving infancy.

Some proportion of those who survive severe malaria—with or without effective treatment—are left with permanent, serious effects, including epilepsy and spasticity. More subtle consequences have also been described and include behavioral disturbances and cognitive impairment. Currently, it is not possible to do much more than mention this spectrum of effects. They cannot yet be quantified, which is not surprising when even the most concrete of end points—the number of deaths from malaria—is only roughly estimated (see Chapter 7).

Study Description

TEHIP studied the circumstances leading to deaths from malaria in the stable perennial malaria transmission belt that runs along the coast of Tanzania and up the Rufiji and Kilombero River basins, from January 1999 through the end of 2001. The Rufiji Demographic Surveillance Site (DSS) covers 85,000 people in 17,000 households in 31 villages, registering all births, deaths, in-migrations, out-migrations, pregnancies and other vital events. A “verbal autopsy” is conducted for each person who dies in the population. Following routine DSS procedures for verbal autopsy,³ a trained health officer interviews the next of kin (or other close “informant”) about the person who died and the events leading to the death. Questions are asked about the use of health facilities during the fatal illness, reasons for using or not using a particular health facility, and confirmatory evidence of cause of death, if available. The interviewer assesses the information and assigns a tentative cause of death. A physician reviews the entire record and assigns the final cause.

From the full details of each case, treatment can be divided broadly into “modern care”—which includes biomedical, Western, pharmaceutical, professional, official or formal health care—and “traditional care”—which includes traditional medicine, traditional healers, traditional providers, lay providers, traditional practices, or folk care. (Use of “modern care” does not necessarily mean that the care was appropriate, or used correctly.)

Results

Over the 3-year period, 3,023 people died and 2,953 (97.7 percent) verbal autopsies were conducted. Out of these, 722 were suspected to be directly or indirectly due to malaria. Just under half (46 percent) of the 722 who died were male, 44 percent were under 5 years of age, and 39 (5 percent) had had convulsions indicating possible cerebral malaria. Findings of particular interest include:

• Treatment most often started at home using antipyretics and antimalarials from local shops or left over from previous episodes.
• 626 (87 percent) sought care at least once before death, while 96 (13 percent) did not, or could not, seek care. These figures were similar for those under and over age 5.
• Of the group that did not have convulsions, 91 percent used modern care first, and by the second treatment, more than 99 percent had used modern care. Hence, traditional care delayed modern care for 9 percent of cases.
• Of those who did have convulsions, a similar proportion (90 percent) used modern care first, but only 63 percent continued with modern care for a second treatment when the first had failed.

This pattern is considerably better than was seen in the mid-1980s in Bagamoyo, when 55 percent of children who died had not utilized any modern care (Mtango et al., 1992).

TEHIP is continuing to refine the information from this study and will be analyzing the narrative portion of the verbal autopsy questionnaires to look at some specific aspects of the health care received, including:

• reasons for delay in seeking modern care (e.g., tried to treat at home without antimalarials, no transport, poor recognition of severity, lack of confidence in modern care, no power to decide, insufficient finances);
• reasons for delay in receiving modern care (e.g., arrived after working hours or on the weekend, long queues); and
• reasons modern care was ineffective (e.g., poor communication, no referral, drugs not available, abusive health worker, noncompliant providers, poor adherence of patients).

Treatment through the Formal Sector

Most malaria treatment takes place at home with drugs from the informal sector, but formal facilities bear a substantial burden. In sub-Saharan Africa and rural Asia, fever is the reason for 20 to 40 percent of all clinic and dispensary visits. These outpatient facilities generally do not offer the life-saving emergency treatments required when severe malaria develops, however. Only hospitals or health centers with inpatient facilities—where malaria's share of admissions ranges between 0.5 and 50 percent (Chima et al., 2003)—are able to provide this care.

The pattern of treatment-seeking and treatment itself is different in countries outside of Africa—mainly Asia and the Latin America/Caribbean region. Malaria is a significant problem in a number of countries, but it is generally more localized to specific areas, and everywhere is much less common than in Africa. Therefore, the burden on individuals as well as the health care system is much less. In Asia and Latin America, people with malaria are treated in general medical facilities, or in special malaria clinics. Laboratory diagnosis is much more widely used, and where these facilities are not available on-site, patients with suspected malaria often are treated presumptively for the immediate symptoms, and given further treatment if their malaria is confirmed.

Treatment through the Informal Sector

Use of the informal sector for malaria treatment is very common throughout the world, even where malaria is less common. Most of this involves the purchase of antimalarials from shops, where people may ask for a “product” by name, rather than a “service” including diagnosis and advice. In Thailand, for example, “ya chud”—a packet with a mixture of medicines including antimalarial drugs sold in private outlets—is used frequently (one study found that around 90 percent of respondents had used ya chud at least once in their lifetime) (McCombie, 1996).

The role of traditional healers and the use of traditional medicines for uncomplicated malaria varies, but generally is considered to be of less importance than other sources of care (McCombie, 1996), though people may underreport use of these services because of perceived disapproval. However, in various parts of Africa, traditional healers often may be consulted for severe malaria. In one site in rural Tanzania in the early 1990s, 90 percent of the children under 5 who died from acute febrile illness with seizures, died at home. Most (85 percent) had been seen by some kind of traditional healer, and only about half had been to a formal health facility before they died (de Savigny et al., 1999).

Weaknesses of Malaria Treatment

People seeking treatment for malaria through any of the channels described here often are frustrated by their experiences. In both the public and private sectors, patients may be faced with low quality of care, lack of drugs or poor quality drugs, and unpredictable costs, to name just a few of the problems. All of these problems affect patients, but they also are concerns for health care systems. Some of the key problems are enumerated below.

Issues in Diagnosis

• Where diagnosis is based on symptoms alone, overdiagnosis, with resultant overuse of antimalarial drugs and real causes of illness possibly left untreated (Stein and Gelfand, 1985; Olivar et al., 1991; Guiguemde et al., 1997).
• Where parasitologic diagnosis is attempted, poor equipment, lack of supplies, limited training and supervision of laboratory staff, and inapropriate use of diagnostic results in treatment choices (Palmer et al., 1999; Barat et al., 1999).
• Lack of accurate microscopy, and for rapid diagnostic tests, the difficulty of determining whether a parasitic infection that is detected is the cause of the current illness. Especially in highly endemic areas, diagnosis based on symptoms may be equally valid.

Lack of Effective Drugs and Inappropriate Use of Drugs

• Chloroquine remains the first-line drug, and sometimes the only antimalarial stocked in health centers and hospitals, even in some countries with very high rates of chloroquine resistance.
• Health facilities frequently are out of stock of essential drugs including antimalarials, and the malaria peak in some countries coincides with the end of the financial year, exacerbating the problem at a key time.
• Second- and third-line drugs may not be made available to lower level health facilities, despite the high frequency of treatment failure with the first-line remedy.
• “Polypharmacy”—prescribing of multiple, unnecessary drugs in addition to the antimalarial is common. As well as being wasteful, it may increase the risk of side effects.
• Where patients have to pay for drugs, they may buy only some of the products prescribed, or incomplete doses, including the antimalarial itself. They may buy a complete course, but take only the first part (particularly if they feel better), storing the remainder for a subsequent episode, leading to systematic underdosing and encouraging the emergence of resistance.
• Expensive (and possibly more dangerous) formulations may be used in place of cheaper ones: in a study in Ghana, 42 percent of chloroquine prescribed to outpatients was administered by injection rather than orally (Ofori-Adjei and Arhinful, 1996).
• Poor quality drugs are common in retail pharmacies in Africa and Asia, due both to lack of quality control in manufacture and to degradation during storage (Shakoor et al., 1997; Maponga and Ondari, 2003).
• Counterfeit drugs are an enormous problem in tropical countries. Fake antimalarials have been a particular problem recently in Southeast Asia. For example, until a recent publicity campaign, more than half of the antimalarial drugs available in the private sector in Cambodia were fake. Widespread counterfeiting of artesunate has led to erroneous reports of resistance (Newton et al., 2001).

When patients buy their own drugs through shops or drug sellers, they may use their own judgment about how much to take (or how much they can afford to buy), and they also may get information from the sellers which may or may not be accurate. As an example of what takes place:

• In a survey in Kenya, only 4 percent of children given store-bought chloroquine got an adequate dose, and only half of those children received this dose over the recommended 3-day period (Marsh et al., 1999).
• In the same survey, aspirin was widely used and 22 percent of children received potentially toxic doses (Marsh et al., 1999).

How Quality Issues Affect Patients

Patients factor in what they know from past experience and general community knowledge when they decide on malaria treatment. In one rural area in Tanzania, the most common reason people gave for not using government services was the poor drug supply. At the same time, people report that public clinic staff are often rude and insensitive, and there are long waiting times to be seen. The facilities themselves often are in poor condition, discouraging people from coming (Gilson et al., 1994).

Costs are sometimes unpredictable in public facilities. Providers may add on charges for drugs that ought to be covered by the consultation fee. If there are no drugs in the clinic, the patient may have to pay for the consultation, and then be told to purchase the drugs privately.

Private clinics and other outlets have their problems too. While staff attitudes may be better and waiting times shorter, private facilities may not have as wide a range of equipment or trained staff as in the public sector (Silva et al., 1997; Mutizwa-Mangiza, 1997). And patients are aware they are paying higher prices for private services, which may make them skeptical about the motivations of private providers, balancing the welfare of patients against their own profits (Silva et al., 1997; Smithson et al., 1997).

Public Sector Drug Flows

Private Sector Drug Flows

Drug Flows through the Illegal Sector

Senegal

The public network provides four antimalarial drugs (amodiaquine, chloroquine, quinine, and SP), which come in nine different presentations, forms, and dosages. In the private sector, there are 13 compounds sold in 89 presentations, forms, and dosages.

Zambia

In the face of widespread chloroquine resistance, in 2003, Coartem was named as the official first-line drug. Coartem has been introduced into pilot districts, but because of cost, SP (which is cheaper), is available in most other places. Although no longer provided through the public sector, there are still considerable sales of chloroquine in the private sector. In addition, more than a hundred different brands, presentations, and forms of antimalarials are available in the private market as branded products, generics, and branded generics, representing about 14 different compounds. Artemisinin products are freely available without prescription in this sector, as monotherapies and combinations.

Cambodia

More than 100 brands and forms of antimalarials are found in Cambodia's private sector outlets, where patients receive treatment based on their ability to pay. The public sector distributes mainly generic products, about 8 antimalarials in various forms and strengths. In total, 30 antimalarials are officially registered.

A combination of artesunate and mefloquine (A+M) was introduced as the first-line drug in 1999, with a recommendation that all first-line drugs be based on a diagnosis by either microscopy or a rapid dipstick test. The combination is available as “A+M” in the public sector and is socially marketed in the private sector as “Malarine.” Dipsticks also are socially marketed in the private sector.

Drug Pricing

In all three countries, in addition to transport, clearing charges and any taxes that may be added to manufacturing prices (in Senegal this is 10 percent and 1.5 percent), drug importers add a margin, which includes distribution costs. In Senegal, the markup by the public wholesaler was 13 percent, and by the private wholesaler, 9.9-24.6 percent, depending on the type of medication. District store markup in the public sector was typically 15 percent. In Zambia, wholesalers usually added a markup of 20-30 percent, and in Cambodia, this ranged from 1.82-49.57 percent. In Senegal, the Bamako pharmacies add margins of 30.5 percent in the public sector, while pharmacies and retailers in the private sector add margins of 9.9-56.2 percent. In Zambia, private pharmacies and retailers generally add a margin of 30-40 percent; in Cambodia these ranged from 1.85-2300 percent. These mechanisms, based on cumulative margins, lead to an almost doubling of the cost of entry prices (Table 1-2 lists consumer prices for selected antimalarials in the three countries, in some public and private sector outlets.)

TABLE 1-2

Consumer prices of selected antimalarials in three countries for a typical adult course of treatment (in US$).

Drug Financing

The RBM External Evaluation

In early 2002, several of the “core” RBM partners commissioned an evaluation of the organization and its progress by a seven-member team of experts, led by Richard Feachem (who shortly thereafter was named head of the Global Fund for AIDS, TB and Malaria). The evaluation, published in August 2002 (Feachem, 2002), documented both achievements and “serious constraints” that threaten progress in phase 2, defined as mid-2002 to 2007.

The accomplishments include raising the profile of malaria and increasing global funding—funding for malaria control doubled in RBM's first 3 years. The constraints centered on a lack of progress in getting activities under way quickly at the country level. RBM may not be able to achieve a “significant reduction in the global burden of malaria by 2007” given the pace of progress, which could undermine the partnership if not rectified.

Three major reforms and two tactical changes were recommended:

The Research Area

TEHIP studied the circumstances leading to deaths from malaria in the stable, perennial malaria-transmission belt that runs along the coast of Tanzania and up the Rufiji and Kilombero River basins. The transmission risk there is typical of the majority (75 percent) of Tanzania and of sub-Saharan Africa in general (MARA Collaboration, 2003). The population of Rufiji District is 203,000, divided among several ethnic groups. The district occupies 14,500 km² and is entirely rural, with 94 registered villages.

There are 57 formal health facilities: two hospitals (one government and one NGO), five health centers with inpatient facilities (all government), and 50 dispensaries (46 government). Over-the-counter drugs are available from many private shops and kiosks in the villages. People also consult traditional healers and traditional birth attendants.

Integrated management of childhood illness (IMCI) services, intermittent presumptive treatment (IPT) of malaria in pregnancy, and 1st, 2nd, and 3rd line antimalarial services are available at all formal health services. Within the District, insecticide treated nets (ITNs) are distributed through social marketing.

The Rufiji Demographic Surveillance Site (DSS) covers part of Rufiji District, including 1,800 km² north of the Rufiji River and west of the Rufiji Delta, with a population of 85,000 in 17,000 households in 31 villages. The DSS registers all births, deaths, in-migrations, out-migrations, pregnancies, and other vital events. Events are detected by 150 village “key informants,” and verified by DSS staff. Twenty-eight full-time enumerators update the population register every 4 months, based on household surveys.

Verbal Autopsies

The information available for each death is gathered during the process of “verbal autopsy” (VA). Once a key informant documents a death in the study area, a DSS VA supervisor, who is also a trained clinical officer or health officer, conducts the VA interview using standard questionnaires that are designed specifically for deaths of: a) children under 31 days of age; b) children under 5 years but 31 days and older; and c) population aged 5 years and above. The questionnaires and responses are in Kiswahili, the common language of the area. Open-ended and closed questions are asked on the history of events leading to death, as well as about previously diagnosed medical conditions, and signs and symptoms before death. Use of health facilities prior to death, reasons for using or not using a particular health facility, and confirmatory evidence of cause of death (if available) are asked about and recorded. A typical bereavement interview for a VA takes 45 to 60 minutes.

The VA supervisor establishes a tentative cause of death from signs, symptoms, sequence, and severity, as well as any confirmatory evidence. A physician reviews the material and makes the final determination of cause of death that is subsequently entered in the database.

Malaria-Related Deaths

This study covered the period January 1999 through the end of December 2001, during which 3,023 deaths occurred and 2,953 (97.7 percent) verbal autopsies were conducted in the DSS. Of these deaths, 722 which fell under the following codings were suspected to be directly or indirectly due to malaria:

• acute febrile illness 1-4 weeks
• acute febrile illness ≤7days
• acute febrile illness including malaria
• acute febrile illness with seizures
• acute febrile illness with anemia
• cerebral malaria
• fever plus malnutrition
• malaria
• malaria confirmed
• unspecified acute febrile illness

Among these 722, 45.8 percent were male, 54.2 percent female; 44.3 percent were under 5 years of age, 55.7 percent were 5 years or older; 39 (5.4 percent) had convulsions and possible cerebral malaria, while 94.6 percent reported no convulsions.

As a cautionary note, VA methods have distinct limitations. Specifically, not all the cases identified as “malaria” in this series truly represent malaria, especially those with unspecified acute febrile illness at an older age. And undoubtedly, some malaria deaths were coded as nonmalaria. For example, life-threatening anemia, likely due to malaria, is prevalent in young children over 6 months of age in the study area, yet VA coded deaths due to anemia with malaria are infrequent among recorded malaria deaths.

Types of Care Sought

The term modern care describes what conventionally includes biomedical, Western, pharmaceutical, professional, official, or formal health care. Traditional care describes what conventionally includes traditional medicine, traditional healers, traditional providers, lay providers, traditional practices, or folk care. From the VAs, 13 possible sources of treatment were collapsed into six subcategories of provider type:

1.

Government

2.

Home/Shops

3.

Non-Government

4.

Traditional Medicine at Home

5.

Traditional Medicine at Practitioner

6.

No Care

and into three broader subcategories:

1.

Modern Care

2.

Traditional Care

3.

No Care

Of the 722 individuals whose deaths were ascribed to malaria, 626 (87 percent) sought care at least once before death, while 96 (13 percent) did not, or could not, seek care. Treatment usually starts at home using antipyretics and antimalarials from local shops or left over from previous episodes. Malaria is perceived by adult caregivers as a mild disease, and if it becomes serious or life threatening, the presumptive diagnosis changes from malaria to something that is more often treated with traditional medicine or practices. Yet these beliefs are not rigid. Every case is subject to a process of continuing debate and reevaluation within the family. As a result, modern pharmaceuticals also are sought, albeit with delay, when convulsions fail to resolve, or recur after traditional medicine (Oberlander and Elverdan, 2000). Significant switching from one provider to another occurs over time; this phenomenon is most apparent when comparing treatment of malaria without convulsions to malaria with convulsions.

Key Observations

• Cases with convulsions were significantly (p<0.05) more likely to receive care.
• Of the group that did not have convulsions, 91 percent used modern care first, and by the second treatment, more than 99 percent used modern care. Hence traditional care delayed modern care for 9 percent of cases.
• Of the group that had convulsions, 90 percent used modern care first, but only 63 percent continued with modern care for a second treatment.
• Government health facilities and shopkeepers were the main source of antimalarial drugs. This pattern held up over broad age groups, sex, and socioeconomic status. (At the time of the study, all government providers had adequate drug supplies, which could be a factor in their popularity. The first-line drug was chloroquine, however, at a time when drug resistance was already widespread in this area.)

Other Relevant Factors

In this study, modern care was more popular than previous reports would suggest. But some part of that modern care involved self-treatment and a variety of private providers. Knowledge of appropriate treatment regimens is lacking on the part of the public as well as the private providers (Brugha et al., 1999; Marsh et al., 1999), and under-dosing in home-based care is common. These may account for at least some treatment failures.

The only other study of health-seeking patterns in a large series of VAs is from the mid-1980s from the Bagamoyo District of Tanzania, another district in the Coast Region (Mtango et al., 1992). In that series, malaria deaths were not analyzed separately, but government providers were the choice in only 45 percent of all deaths. A preference for traditional healers was cited by 41 percent of mothers as the reason for not using government services. At that time, it was also true that government providers were often without adequate drug supply.

Why Did They Die?

The study results to date cannot answer that question, but they begin to narrow the possibilities. Some form of care was sought in the majority of malaria deaths (88 percent). No care was sought for 12 percent of child deaths. Included among these were sudden deaths following apparently mild illness, which may have been impossible to prevent under the circumstances. This pattern is considerably better than was seen in the mid-1980s in Bagamoyo, when 55 percent of children who died had not utilized any modern care (Mtango et al., 1992), and better than for all deaths in the same area during the same period when 20 percent of all-cause deaths had no prior health-seeking events (Tanzania Essential Health Interventions Project and Ministry of Health of Tanzania, 2002).

With belief systems for malaria treatment-seeking now firmly on the side of modern care, there is obviously something still failing in the transaction to obtain this care, or in the quality of the care and referral once it is accessed. TEHIP is continuing to refine the information from this study, and will be analyzing the narrative portion of the verbal autopsy questionnaires to look at some specific aspects of the health care received, including:

• delay in seeking modern care (e.g., tried to treat at home without antimalarials, no transport, beliefs, poor recognition of severity, lack of confidence in modern care, no power to decide, insufficient finances);
• delay in receiving modern care (e.g., after working hours, weekends, long queues, satisfaction); and
• ineffective modern care (e.g., poor communication, no referral, drugs not available, abusive health worker, non-compliant providers, poor adherence of patients).

All of these factors could have contributed to failed treatment. The study, as yet, can provide no direct information on the possible contribution of drug resistant malaria to these deaths.

The current results suggest that policies and efforts aimed at improving early recognition of symptoms and danger signs at home, prompt treatment or treatment-seeking, compliance with the full course of treatment, and the quality of the antimalarial available are highly justified.