Cholera toxin–producing Vibrio cholerae cause the characteristic life-threatening gastroenteritis, cholera. At least 7 pandemics of the disease, originally designated “Asiatic cholera,” have occurred in recent history. The ongoing pandemic has seen the emergence of O139 V. cholerae, as a non-O1 pandemic strain and, importantly, the emergence and spread of drug-resistant O1 strains. The current focus of the cholera pandemic is Africa, which has seen two thirds of all cholera outbreaks in the last decade (5). The primary treatment for cholera is rehydration. Most patients will overcome the infection if they are rehydrated promptly and properly, even if they do not receive antimicrobial drugs. Antimicrobial drugs, however, shorten the course of infection and prevent person-to-person transmission, which may be crucial for slowing outbreaks because organisms from infected persons may be more virulent than those acquired in the wild (6). Antimicrobial agents may also be life-saving for malnourished and other immunocompromised patients who have cholera.

Tetracycline was the empiric drug of choice for cholera in Africa and elsewhere for many years. At the end of the 1970s, however, incompatibility group C tetracycline-resistant plasmids were isolated from V. cholerae isolates in Tanzania, Kenya, and other parts of Africa (7,8). In each case, resistance emerged during an ongoing epidemic where tetracycline was being used intensively for prophylaxis as well as treatment. Tetracycline has sequentially been replaced by trimethoprim-sulfamethoxazole and, more recently, quinolones, because of the emergence and spread of resistant strains. Molecular evaluation of more recent resistant V. cholerae isolates typically found regionally conserved plasmids, some of which carried class 1 integrons bearing multiple resistance cassettes (9,10). A chromosomally integrated transferable resistance element, SXT, has also spread worldwide and has been recently reported from Africa (9).

Emergence of resistance in V. cholerae has been linked to increased mortality rates in recent African outbreaks. Similar experiences have also been reported with Shigella dysenteriae type 1, another enteric pathogen that causes life-threatening disease and has epidemic potential. The impact of resistance in both pathogens is illustrated by an overwhelming outbreak in July 1994 at the Goma camp, which resulted in the deaths of ≈12,000 Rwandan refugees (11). More recently, Dalsgaard et al. (12) observed a marked increase in case-fatality rate during the 1997–1998 phase of a Guinea-Bissau cholera outbreak, compared to an overlapping 1996–1997 outbreak. A major feature of the latter wave of cholera was the presence of strains simultaneously resistant to ampicillin, erythromycin, tetracycline, furazolidone, aminoglycosides, trimethoprim, and sulfamethoxazole. These multidrug-resistant strains were not present during the first wave and probably arose following the acquisition of a 150-kb–pair resistance plasmid bearing a class 1 integron and genes encoding resistance to all antimicrobial agents commonly used in empiric management of cholera.

The emergence and spread of multidrug-resistant Salmonella enterica subsp. Typhi worldwide has had important consequences for mortality rates from typhoid fever (13). There are very few reports from Africa; nonetheless, available data suggest that although the problem may not be as intense as in other parts of the world, resistance has emerged, and alternatives to current treatment protocols are often not available or unaffordable. Multidrug-resistant nontyphoidal Salmonella spp. (NTS) have emerged as a global public health threat. In industrialized countries, they are most commonly associated with foodborne gastroenteritis. In parts of sub-Saharan Africa, however, NTS are important causes of life-threatening bacteremia. Studies from Kenya have found that community-acquired NTS are among the top 3 causes of death among children <5 years of age (14,15). Moreover, pulsed-field gel electrophoresis data suggest that most life-threatening disease is caused by isolates that are clonal in origin (14). In a recent study, children from poor slums of Kenya were significantly more likely to be infected with multidrug-resistant NTS than were children from middle-income families (14). The patterns of resistance among these strains suggest that third-generation cephalosporins should be the drug of choice for empiric management of these infections, but in most cases, these drugs are too expensive.

Antimicrobial drug resistance is a large and growing problem among organisms that cause diarrheal disease. Although most diarrheal diseases are self-resolving and should not be treated with antimicrobial agents, invasive or protracted infections require chemotherapy and are typically managed empirically. Recent data from Gabon, Kenya, Nigeria, Senegal, and Tanzania suggest that resistance among causative organisms of these infections, such as enterotoxigenic, enteropathogenic, and enteroaggregative Escherichia coli, is high and appears to be rising (16–18). Although oral rehydration therapy has drastically reduced deaths from the disease, prolonged infectious bouts of diarrhea have long-term consequences for physical and cognitive development. Very few reports have examined the epidemiology of diarrheal pathogens and even fewer have looked at drug resistance. Notable drug-resistant enteropathogenic E. coli outbreaks and sporadic cases have been reported from several African countries, including Kenya and Tanzania (16,19). The more recently defined enteroaggregative E. coli are typically multidrug-resistant and are one of the most common causes of childhood diarrhea, particularly persistent infections (20). Antimicrobial drug–resistant diarrheagenic E. coli pathotypes, including enteroaggregative E. coli, are also emerging as important diarrheal pathogens in AIDS patients (21).

Surveillance in healthy populations has demonstrated that commensals constitute a rich reservoir of genetic material from which pathogens can readily acquire resistance on mobile elements. A long-term study in Nigeria showed that resistance of commensal E. coli to almost all agents studied increased rapidly over time (22). Additionally, urban residents in Nigeria, Ghana, and Zimbabwe were more likely to carry multidrug-resistant E. coli than were rural or provincial residents (23,24). This finding has important consequences in light of the rapid rate of urbanization in these countries and other parts of the continent. Travel networks have become more efficient and are more extensively used. Therefore, just as Africa has had to deal with imported resistant organisms, resistant strains that emerge or are amplified in Africa will be exported (25,26).

Two overlapping problems are worsening the situation regarding diarrheal disease within Africa: the failure to control the spread of diarrheal pathogens, due to unclean water, poor sanitation, and malnutrition; and the failure to contain resistant organisms and resistant genes so that, when infections occur, they produce more adverse consequences. It is perhaps obvious, if unaddressed, that poor and displaced persons in Africa are least likely to be able to access potable water, safe sanitation, and other factors to prevent fecal–oral infection and that public health facilities need to be strengthened to protect the poor (27,28). However, the poor also disproportionately bear consequences from drug resistance, and interventions to curb the current trend are sorely needed. Economists describe a situation in which the decision to use a commodity that produces a deleterious byproduct that imposes costs on other persons, who are not part of the decision, as a negative externality. Because resistance is an externality, the poor are victims of the resistance-predisposing activities of the affluent, including the presence of visitors from other countries. Furthermore, resistance-promoting activities and the consequences of resistance more often than not occur at a different time and place. It is therefore unrealistic to expect that the poor will mount resistance-curbing interventions as a priority without prompting and support.

There are difficulties associated with monitoring resistance in many parts of Africa, but sufficient published data exist to suggest that resistance rates are high and rising (22,29). In 2001, the World Health Organization issued a strategy for resistance containment (30). Most developing countries, particularly those in sub-Saharan Africa, have yet to implement any of the recommended interventions in spite of a mounting resistance crisis. Special risk factors for resistance, and roadblocks for evaluating and implementing interventions, are linked to patient poverty and health system poverty. This is evident in sub-Saharan Africa, where we study resistance, but also in other developing countries.

Figure

Figure. Itinerant medicine vendor in Oja-tuntun marketplace, Ile-Ife, Nigeria.

Industrialized countries that are actively addressing community-acquired resistance have typically prioritized those interventions predicted to lower total antimicrobial agent consumption and therefore selective pressure (31,32). An important concern for poor countries is not the total amount of antimicrobial agents consumed, whose need exceeds the resources available, but the way these drugs are used. Diagnostic imprecision spurs overprescribing, particularly of broad-spectrum agents, and low antimicrobial diversity promotes emergence and spread of potentially epidemic resistant clones. In Africa, medicines are available from unorthodox sources and persons earning low daily wages often procure them 1 dose at a time. Sources include itinerant vendors, often encountered on buses in many African countries, who cannot always be located should a patient choose to purchase more doses later (Figure). Poor patients have little personal incentive to purchase more medicine than is needed to produce short-term relief. Additionally, regimen fragmentation comes without sanction and allows for large price mark-ups and consequent exploitation. Therefore, distributors have every incentive to encourage antimicrobial drug misuse.

Resistance is encountered with virtually every infectious disease. Few proven mechanisms exist for resistance control, and almost none have been validated in the developing-country setting. The dearth of intervention study data is particularly acute in the context where infectious disease prevalence is high and access to antimicrobial agents is low, which best describes the situation faced by low-income persons in Africa. Most intervention studies in developing countries have focused on relatively inexpensive and easily piloted educational interventions. Educational interventions push against the strong influence of unregulated distribution, sometimes accompanied by unscrupulous counteradvertising, and their value has not been evaluated in the long run. Importantly, although educational interventions typically yield positive results, these results are modest (31). Other methods may in fact be more cost effective, or might boost the value of education. Strategies that have been evaluated and found to deal with the problem of resistance need to be further diversified in poor countries. Safe water and sanitation, addressing the imbalance between antimicrobial drug supply and demand, and building realistic infrastructure for rational antimicrobial use are priority areas for resistance control that could address the short- and long-term disease effects on the poor.