Chapter 5 - Cholera
Diagnosis and Treatment Outside the Hospital
Cholera spread from Asia during the nineteenth century in successive waves of epidemics. Several pandemics extended from the Middle East following commercial trading routes, troop movements, and pilgrimages. The reduction in transport times and the increase in population movements contributed to the spread of the disease (Bourdelais and Dodin, 1987, p. 33). Mecca, which is a focal point for pilgrims from all continents, became a hub for Vibrio cholerae O1. The disease seemed under control at the beginning of the twentieth century, thanks mainly to improvements in hygiene and water management in large industrial cities. Nothing foretold the emergence of a seventh pandemic in 1961, which was just as lethal as the six preceding identified periods of transmission and is still active today. From the departure point in the Sulawesi Islands, cholera spread through the Far East, arriving in the 1970s in Western Europe through Spain, and Africa through Guinea. Following the Niger River, the disease affected neighboring African countries and spread across the continent.
The pandemic involved a new strain of the cholera bacteria, named El Tor, discovered in 1905 in a lazaretto (quarantine station) of the same name in Mecca. A biotype of the O1 strain, it is found today on all continents, replacing the previously dominant O1 classic biotype. Following an epidemic outbreak in Peru in the early 1990s, the pandemic spread through South America, which had previously been spared.
Cholera is currently present in almost fifty countries. Africa is the most affected continent and accounts for 95% of all reported cases, followed by Asia, and, to a lesser extent, the Americas, Europe, and Oceania, the latter two reporting almost uniquely imported cases. In 2005, 131,943 cases were reported worldwide, including 2,272 deaths, a specific mortality rate of 1.72%. Real numbers are thought to be much higher; it is estimated that only 5% to 10% of all cases are actually reported, and that there are other shortcomings with surveillance systems (World Health Organization [WHO], 2006). A strain belonging to a new serogroup, Vibrio cholerae O139, appeared in the Gulf of Bengal in 1992 and has remained localized in Asia for the time being (WHO, 1996; WHO, 2004b).
Robert Koch identified the Vibrio cholerae O1 bacillus in 1883, beating Pasteur, whose research methods were based on the study of immunity mechanisms rather than on identifying the causal agent. During epidemics, direct inter-human transmission is the most frequent mode of infection, whether it be from a sick patient, a healthy carrier, an infected cadaver, or from feces-contaminated drinking water or food. In non-epidemic periods, aquatic plants, crustaceans, and mollusks are the reservoirs for the bacteria. These are environments where it can survive indefinitely under the right conditions. Clinical features of the disease mainly include profuse watery stools, sometimes associated with vomiting, and subsequent rapid dehydration. The natural course of the disease leads to death in 50% of symptomatic cases. Survivors may recover in four to six days, with an acquired immunity after one week, which might last several months. The main treatment is rehydration, either orally or by intravenous infusion, depending on the clinical state. Antibiotic therapy is sometimes used. The treatment is well known and has an undeniable effect on mortality, but mortality remains high in countries where health systems are not mobilized when faced with epidemics.
Research progresses slowly, and a useful vaccine has yet to be found for this disease that essentially affects poor populations.
The Cholera Camp, a Curative Innovation Based on New Organizational and Logistical Set-ups
Creation, Objectives, and Organization of the Cholera Camp
Cholera spread throughout Africa in the 1980s. MSF intervened regularly in refugee and displaced persons camps, known as “closed settings,” which were favorable to cholera propagation. Such was the case in Korem, Ethiopia; the Ethiopian refugee camp in Wad Kaoli, Sudan in 1985; and in 1988 in Malawi in the Mankhokwe Mozambican refugee camp. Then, from 1988 to 1993, again in Malawi, a series of epidemics affected other Mozambican refugee camps, where over seven thousand cases were treated. From July to August 1994, during the mass grouping of Rwandan refugees in eastern Zaire (now Democratic Republic of Congo), a major epidemic occurred, with fifty-eight thousand cases and 4,200 deaths. Here the specific mortality rate reached 22% in the first three days (Brown et al., 2002; Goma Epidemiology Group, 1995).
In March 1985, MSF had been assisting a displaced population of around forty thousand people in Korem, North Wolo Province, Ethiopia, for over a year. Living conditions were deplorable: plastic sheets for shelter and no blankets despite night temperatures approaching zero. One of the doctors remembers a group of corrugated barracks acting as a hospital where around seven hundred patients were treated. The average weight of hospitalized adults was thirty-three kilograms. It was in these conditions that almost two thousand people were hit by a cholera epidemic lasting over two months—an attack rate of nearly 5%, common in this kind of setting where crowding, hygiene, and the weakened state of the people favor disease transmission in general. The MSF teams, inexperienced in this type of situation, tried to organize a tent isolation camp to treat sick patients. They set up the necessary sanitation systems to prevent the spread of disease (i.e., water-source protection), and systematically carried out identification of sick patients in family shelters. The WHO recommended mass prophylaxis treatment with a long-acting, single-dose sulphamide, Fansadil, which was distributed by the teams. Neither the WHO, the Centers for Disease Control and Prevention (CDC), nor MSF was certain as to the effectiveness of this treatment.
This experience presented the teams with the organizational challenges of caring for a high number of patients over a short period of time. The medical protocols were simple but unwieldy; fifty to one hundred patients could be receiving intravenous treatment at any one time, assuming sufficient medical supplies.
These initial operational strategies in response to closed-setting epidemics were then applied in Malawi. Between 1986 and 1988, around four hundred thousand Mozambican refugees settled in villages and camps along the border. There were a total of nine hundred thousand refugees in Malawi by 1990. MSF had been providing medical care in the south of the country since 1986, and, in March 1988, the first cases of the epidemic occurred in the Mankhokwe camp, with a population of nearly thirty thousand people. The epidemic continued until May, and 951 cases were treated by MSF in collaboration with Mozambican medical staff and the Malawian authorities. Response strategy included identifying the cases and carrying out patient care in a cholera treatment center (CTC), or cholera camp. Preventive measures were also employed, including market and public site closures, contact prophylaxis, and water chlorination (MSF, 1988). The first field involvement of Epicentre, in 1988, also showed the relevance of improving coordination of epidemiology and emergency medical action.
The creation of the CTC model remains MSF’s most significant contribution to cholera epidemic response. Patient care and epidemiology were the two priorities, and the CTC provides rapid medical care to a large number of patients while isolating them from hospital structures where disease may spread. This strategy requires significant logistical resources so that medical care, treatment and recovery, as well as shelter, food, and water supply are all provided in simple, functional, and autonomous conditions.
In practice, isolation, supply, and hygiene were the three key elements of a functional CTC, and opening a structure was indicated once more than five new cases per day were identified in a closed setting such as a camp, a prison, or a social or sanitary institution (MSF, 1995). In terms of isolation, the cholera camp is closed and different from other care structures. It includes four separate entities, so as to regulate patient and caregiver movement, thus limiting the risk of transmission. One unit acts as a triage and observation center for suspected cases, another for hospitalization or isolation, including disinfection measures. A convalescence unit treats patients receiving oral rehydration therapy. Finally, the “neutral” zone houses the kitchen, supplies, and changing rooms. In terms of supplies, the center must be able to maintain sufficient medical and non-medical stocks to provide care (for cholera, but also for common associated diseases such as malaria, etc.), as well as lodging and food for all patients, to which must be added sixty liters of water per day per person to guarantee the center’s autonomy. The teams use tools developed by MSF, Epicentre, and MSF-Logistique to plan cholera center management and supplies, notably the cholera guide and cholera kits. Furthermore, in bigger camps, oral rehydration units are created in the periphery so as to begin treatment immediately, before transfer to the CTC, if indicated.
MSF thus developed emergency cholera response expertise from 1985 to 1989 in smaller camps. In the 1990s this knowledge was developed to use on a completely different scale, for epidemics in camps containing hundreds of thousands of people. These methods are still in use and pertinent today.
Medical Practice in the CTC
Two major controversies concerning medical protocols within CTCs have stimulated debate among specialists. The first involves intravenous fluid rehydration. Since the middle of the twentieth century, it has been common practice to rapidly administer large volumes of intravenous fluids to treat severe cases. This practice is not without complications, however (hypoglycemia, pulmonary edema in older patients and children, and hypokalemia), and was only administered in hospitals or health centers, whereas CTCs are deliberately created outside existing hospital structures. At the same time, a process of active case-finding in closed settings replaced passively waiting for patients to turn up. To achieve this, clinicians accustomed to hospital settings and uncomfortable with these new strategies were sometimes given specific training. Around this time, however, the WHO and the United Nations Children’s Fund (UNICEF) questioned the use of large-volume intravenous rehydration strategies and recommended oral rehydration, which in part explained the differences between the WHO and MSF cholera kits.
The second controversy concerned doxycycline. According the WHO, the cholera bacillus is commonly resistant to this drug, and prophylaxis only recommended for contacts of sick patients where at least one secondary case has occurred in a family of five. Despite this, doxycycline is still part of the MSF and WHO cholera kit. In an article in 2002, MSF detailed the limits of antibiotic therapy: “Antibiotic use is only envisaged for the most severe cases. It is useful for reducing diarrhea duration and volume, and germ carriage time” (Brown et al., 2002). The same article restated the absolute priority of rehydration: “Deaths in cholera camps are essentially due to delays in quality rehydration treatment: if this treatment is applied rigorously, specific mortality drops to less than 0.5% (in refugee camps). Standardizing antibiotic therapy risks inducing a false sense of security in medical staff and distracting from essential rehydration. … Preventive antibiotics may be recommended in cholera epidemics affecting populations in closed settings (e.g., prisons).” So MSF did not recommend curative antibiotic treatment during cholera epidemics—with exceptions for severe cases and situations of very high population density.
Cholera Guides and Kits: Tools for Emergency Responses
The idea of a kit for immediate responses during cholera epidemics was proposed a few weeks after the disastrous situation and shortages in Korem in 1985. Cholera cases cause concern, which may sometimes lead to importation problems. That’s why the cholera kit was called the “001 kit,” because cholera was the first disease on the WHO obligatory declared diseases list. The kit is designed for around six hundred patients and essentially includes intravenous material (Ringer’s Lactate), oral rehydration salts in sachets, an antibiotic (doxycycline), and chlorination material. With this kit, MSF contributed to the standardization of Ringer’s Lactate, and the large quantities provided avoid stock shortages of this essential item.
The kit contents caused and still cause controversy involving MSF and the WHO. MSF teams working with epidemics aim first to treat the severest cases. The WHO approaches cholera from a more theoretical angle, based notably on epidemic response strategies developed by the Bangladesh International Center for Diarrhoeal Disease Research (ICDDR), which predict an attack rate of 0.2% for the entire population during epidemic outbreaks in an open setting, of which around 20% are severe. MSF’s experience in Mozambican refugee camps in Malawi showed attack rates of 1% to 2% or higher. Differences in opinions resulted in an MSF kit which contains 80% intravenous treatments and 20% oral, and a WHO kit with contents of inverse proportions (Bitar, 1991b). To adapt to open-setting epidemics affecting populations spread over large areas, the 001 kit may be split into mini-kits, each treating twenty severe cases, for smaller peripheral treatment centers.
The first reference document was written by field teams in 1980 to provide clear and practical objectives for cholera epidemic response (MSF, 1980). The first guide recounting MSF experiences in closed settings was published in 1995, and stresses the cholera camp model, rehydration protocols, and water chlorination. Revision of the guide began in 1997, and was completed in 2004. The revised guide included strategies for epidemic control in open settings.
The WHO published Guidelines for Cholera Control in 1993. While there are no references to MSF publications or associated authors (WHO, 1993), it nevertheless includes the cholera camp model, which is used as a reference by many health actors.
Curative Care in Open Settings
MSF-Belgium and MSF-Holland were already present in Peru when the seventh cholera pandemic struck the continent in the early 1990s. In 1991, at the height of the epidemic, five operational sections developed a common response mission (Belgium, France, Holland, Switzerland, and Spain). Within three months of the arrival of the cholera bacillus on the continent, the WHO created the Global Task Force on Cholera Control, with the main objectives of promoting disease control and research and development initiatives. At the same time, MSF opened an international emergency response base in Central America, initially in El Salvador, then in Costa Rica, which eventually closed in 2000. The base offered technical and training support in cholera epidemic responses for both MSF teams and regional medical staff.
Lima, with a population of seven million, was struck by the disease, but quickly benefited from national and international humanitarian organizations that used their previous experience in urban settings. MSF then decided to intervene in peripheral rural areas where help was harder to obtain: first on the coast, then in the mountains, and finally in the jungle. These interventions involved eight provinces and continued for over a year.
Epidemic control in these contexts required new operational strategies. The purely medical aspects had been defined and would not change (rehydration and antibiotic protocols). Moreover, no effective vaccine then existed for prevention, nor were there any rapid diagnostic tests available. The changes made were in the organization of preventive and curative care.
Affected populations were spread out and difficult to access in mountainous and jungle zones, so large CTCs could not cater to needs. The objective was to get as close to affected populations as possible, using existing structures. Care was decentralized to mini-CTCs known as cholera treatment units and oral rehydration solution (ORS) points, supported by mobile teams. These structures were associated with official health centers, and on-site personnel were trained in cholera care and received appropriate supplies. Each center had a capacity to treat four to five severe cases at any one time. Supply and supervision were provided by regular visits that also served to update epidemiological data. MSF used the UNICEF child diarrhea rehydration program, an existing network from which operations for the population as a whole could be developed.
Peru was MSF’s first large-scale, open rural setting cholera mission. In the 1990s, following several similar interventions and in-house reflection on how to adapt response strategies, Epicentre published a report setting out the first recommendations for these kinds of epidemics (Dorlencourt, 1997). It specified that an initial, thorough, pluridisciplinary evaluation (exploratory mission) was necessary to develop responses. The report also stressed that the epidemic must officially be declared by health authorities conforming to WHO definitions, that initial suspect cases must be confirmed in reference laboratories, but also that the medical and sanitary context must be evaluated before deciding upon an appropriate intervention strategy. Clinical patient care must be carried out and documented using patient treatment cards from cholera kits.
It was also necessary to advertise that health care is free of cost, with the help of local authorities. The recommendations also insisted upon the necessity of establishing a simple functional surveillance system in treatment centers, ORS points, and health structures in the intervention zone. The cholera camp model remains MSF’s major contribution in terms of organizational, logistical, and curative innovation. Its clear effect on mortality inside but also outside camps has been recognized, and the model reproduced.
Other Measures Associated with Curative Care
Coupling of Epidemiology and Emergency Medical Action
In 1988, MSF associated an epidemiological survey with curative care during the cholera epidemic affecting Mozambican refugees. Epicentre conducted an investigation aimed at describing epidemic characteristics regarding timing, locations, and people affected, to better define epidemic breadth, severity, and evolution. This also allowed the definition of more appropriate curative response strategies, the modification of actions where needed, the identification of potential risk factors, and the development of recommendations aiming to reduce epidemic spread (Moren, Stefanaggi, Antona, et al., 1991).
In general, this kind of survey is based on a simple field-data collection system, involving the location, date, number of cases, their age distribution, and number of deaths. These parameters allow the calculation of the main epidemic surveillance indicators, which are attack rates and case fatality rates. These data are collected by health agents locally and by sections heads in displaced persons camps, but also in health structures and in cemetery surveys.
Several Epicentre reports summarize MSF’s care experience in a dozen epidemics, notably in Malawi and in Peru (Bitar, 1991a, 1991b; Wuillaume, 1992; Dorlencourt, 1997).
In short, cholera epidemic control measures in refugee camps were and still are technically simple and easy to master, and well supported by guides and kits produced by MSF. These measures also aimed to demonstrate the possibility of reducing mortality rates to less than 0.5% in closed settings thanks to CTCs. MSF has achieved this result not just once but repeatedly.
Furthermore, interventions in refugee camps have led to the analysis of closed-setting epidemics. The dramatic increase in the number of cases reaches a peak in a short time period, mostly within two to four weeks. The epidemic rarely lasts more than a few weeks, probably because of population density, whereas the average duration of open-setting epidemics is four to five months. Closed-setting attack rates are usually higher than in open settings, and may reach or exceed 5% (Brown et al., 2002).
Epidemiological monitoring in open settings developed from indicators used in refugee camps. The monitoring established in Peru aimed to anticipate epidemic evolution using a classical database, contamination sources, and new-case mapping, but also led to uncertainty due to the difficulties in defining populations moving freely over geographical areas with unclear boundaries. We must therefore remain very cautious if we are confidently to predict a cholera epidemic and its development, and many examples exist as reminders of how epidemic cycles are poorly explained and hard to control.
What distinguishes MSF from other medical actors today is its capacity to mount standardized medical practices and surveillance systems (cases and deaths) in a very short space of time, at scale and with or without the support of epidemiologists.
Diagnosis and Prevention
MSF participated in the development of rapid diagnostic tests, and MSF and Epicentre both took part in their field validation. The Pasteur Institute developed a rapid test for O1 and O139 Vibrio cholerae strains (the One-step Immunochromatographic Dipstick) starting in 2000, and performed an initial validation study in 2002 in collaboration with the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B) in Dhaka. Results were encouraging in endemic situations (Nato et al., 2003). In 2003, the CDC launched a study with the ICDDR and the Pasteur Institute to examine the precision and feasibility of use of three rapid cholera detection tests (SMART, Medicos, and the Pasteur Institute tests). Results showed that the Pasteur Institute test was the most appropriate for detection of the O1 strain in areas where health personnel skills were limited (Kalluri et al., 2006). Another study performed in 2004 by various partners, including MSF, confirmed the performance of the Pasteur Institute test in epidemic field conditions. A new format for this test (one stick with bands for O1 and O139 strains) was validated in the field by Epicentre (Alberti, 2007; Page, Alberti, 2008).
In addition to rapid tests, the cholera strain must be confirmed by a specialized laboratory, often in another country; correct sample-taking and transport conditions must be observed. MSF uses the filter paper transport medium developed in collaboration with the Pasteur Institute, but its efficacy has not yet been proven. Epicentre performed a study comparing filter paper disks, then limited to certain laboratories, to the reference transport medium, Cary-Blair (Page, Alberti, 2008). The question is of increasing importance in current contexts, where there is evidence of antibiotic resistance and new strains, so it is essential that samples arrive safely at reference laboratories.
The effect of cholera vaccination on prevention is still the subject of controversy. The parenteral vaccine developed at the beginning of the twentieth century has not been shown to be effective in prevention, as induced immunity is partial and short-lived (Mosley, Aziz, et al., 1972). In 1973, following a resolution at the Twenty-sixth World Health Assembly, the WHO no longer recommended vaccination for cholera epidemic control (WHO, 1992).
MSF and Epicentre participated in several studies of oral cholera vaccines in the 1990s and early 2000s (Legros, Paquet, Perea, et al., 1999; Epicentre, 1998a; Paquet, 1999; Dorlencourt, Legros, Paquet, 2000; Epicentre, 1998b; Naficy, Rao, Paquet, et al., 1998). In 1999, the WHO concluded in agreement with MSF that vaccinating with the oral bivalent killed whole-cell bacteria vaccine associated with recombinant toxin B subunit (B-Subunit- Whole Cell, or BS-WC, and rBS-WC) was useful in some specific emergency situations, such as refugee camps or high-density slums (WHO, 1999).
In 2001, a senior WHO staff member published an article confirming the importance of vaccination in the prevention of epidemics, in addition to other traditional preventive measures. The author also underlined the importance of further studies to refine vaccination criteria and identify very high-risk populations that might benefit from preventive mass vaccination campaigns (Chaignat, 2001).
The efficacy of the killed whole-cell vaccine with B subunit has been proven (WHO, 2004a), but with just 70% protection over one year; given the efforts required for mass vaccination, its usefulness is limited.
As for preventive vaccination of populations, MSF has participated in working groups to help identify situations in which vaccination would be a valid option in the interest of public health.
Water, Hygiene, and Sanitation
During the epidemic in the Chifunga camp in Malawi in 1990, MSF and Epicentre recommended extending therapeutic care to include prevention (Bitar, Brodel, Gastellu-Etchegorry, et al., 1992).
Evaluations showed the benefits of health education, which means providing clean water and soap to prevent cholera and other hygiene-related illnesses, such as simple and bloody diarrheas and skin and eye diseases, in the camp. Water risk prevention includes bucket chlorination and the banning of suspect water sources. Good water and sanitation programs also avoid contamination of the water table by excreta. Furthermore, all public meeting places were carefully checked or temporarily closed. Finally, corpses were sprinkled with chlorine (orifices in particular) before being enclosed in sealed sacks and buried.
The value of preventive measures remains controversial. For some, the effect is often taken for granted, when in fact it has been only partially proven. Subjective belief in the benefits of prevention overrides the fact-based approach.
Although the effect of clean water supplies on cholera-related mortality is recognized, and targeted actions such as water chlorination and water trucking have been shown to be effective in certain settings, questions remain about the effects of these actions on epidemic occurrence and evolution (Fewtrell et al., 2005). Eradication of the disease in industrialized societies is certainly due to improvements in water supply and hygiene, but also demonstrates the fundamental role of economic, sanitary, and social developments, which cannot be replaced by isolated preventive measures. As for health education in epidemic contexts, effects are doubtful, and benefits unproven.
MSF observed that the hospital-based model of treatment of cholera epidemics was poorly adapted to field situations due to the inability to accept large numbers of patients in acceptable conditions of hygiene. The organization developed a new curative care model, the cholera camp, which has had a proven effect on mortality.
The example of cholera epidemic response in confined areas (camps, prisons) illustrates the creative potential of humanitarian medicine when confronted by the inapplicability of classical treatments in precarious field circumstances. Unfortunately, the majority of all cholera-related deaths occur outside the closed universes described above, and the challenge remains to develop vaccines and other original approaches applicable to the majority of patients.