Beyond emergency care: challenges to health planning in complex emergencies.

Abstract

The 1999 Nobel Peace prize awarded to Médecins sans Frontières (MSF) highlighted the value increasingly attributed to medical emergency care in acute and chronic crisis areas. MSF, however, is perhaps the most visible face of a much more generalized set of developments over the past 15 years. An increasing number of non-governmental organizations (NGOs) have become involved in humanitarian aid. Not only are a wide range of newly established NGOs active in such areas; but there has also been a trend towards development NGOs becoming increasingly involved in emergencies (25). Globalization, in particular the ‘CNN-ization’ of emergency situations, has brought humanitarian needs into the living rooms of millions of people with a concomitant rise in public and donor support for humanitarian activities. The tendency for donors to subcontract the implementation of humanitarian programmes and the large amounts of funds allocated to humanitarian crises are factors that have influenced the expansion in the number of humanitarian agencies operating in emergencies. Funds allocated to humanitarian assistance by the Organization of Economic Co-operation and Development (OECD) countries showed a steep rise and reached an all-time high of 3.5 billion USD in 1994, a 10-fold increase compared with 1980 (18). The NGOs, the Red Cross movement, several specialized United Nations agencies and bilateral donors form the international humanitarian system called upon during acute and chronic humanitarian crises. Although this system is often publicly praised, as evidenced by the Nobel Peace Prize award to MSF, critical comments circulate widely in the field. Emergency work is often regarded as too ‘short-term’, cowboy-like, creating dependence of potential beneficiaries, not sufficiently ‘developmental’, and lacking in accountability (9). Recognizing that many chronic conflicts are maintained as a result of the economic interests of the warring factions (10) has led to a more recent concern that emergency aid may prolong conflicts, although precise agreement on the depth of this critique and the circumstances in which it is most true remain contested. The lack of co-ordination between agencies and limited evidence of any organizational learning within the system are two other frequently articulated criticisms (1). Such critical remarks are usually aired in relation to the entire humanitarian system; the health sector within it has not been subjected to extensive debate and is often assumed to be ‘doing good’. What do we actually know about the impact of external medical assistance in crisis situations? What indicators are used to measure this impact? What evidence exists to support usual modes of operations? We have argued elsewhere that despite evidence of effectiveness in some settings, much of the response to emergencies is poorly evaluated (3). We briefly explore three different contexts in which the international community is involved in humanitarian health programming: refugee camps, control of epidemics and basic health care provision in conflict affected areas. We devote particular attention to the latter, which poses particular challenges to health planning. We examine a number of factors that make planning difficult in this context and consider developments that may contribute to overcoming these difficulties. The International Committee of the Red Cross (ICRC) was founded in 1863, MSF in 1971, but the main boom in establishing NGOs in medical emergency work dates from the 1980s and early 1990s. They are virtually all grounded on the two ‘inspirational’, essential principles of humanitarian action, humanity and impartiality (11). Humanity is the desire ‘to prevent and alleviate human suffering wherever it may be found … to protect life and health and to ensure respect for the human being’. Impartiality means that an organization makes no discrimination as to nationality, race, religious beliefs, class or political opinions. It endeavours to relieve the suffering of individuals, being guided solely by their needs, and to give priority to the most urgent cases of distress. Impartiality means that an organization makes no discrimination as to nationality, race, relegion, beliefs, class or political opinions. It endeavours to relieve the suffering of individuals, being guided solely by their needs, and to give priority to the most urgent cases of distress (21). Typically one would find a phrase to underline these principles in a charter of an agency (13). Other principles, like neutrality and independence, or about the way an organization works, are usually added, but may vary between agencies. It has proved most straightforward to realize humanitarian action in medical work in refugee camps, because of the structured and relatively easily regulated operating environment they represent. Indeed, many of the guidelines and other publications in recent years in the field of medical humanitarian work relate specifically to work inside refugee camps (15). One influential tool in planning for refugee health work has been the determination of ‘excess mortality’ (23). Excess mortality refers to mortality in excess of the baseline (‘normal’) mortality in a given population and is expressed as the number of deaths per 10 000 population per day. A crude mortality rate exceeding 1 death per 10 000 per day is considered an emergency and warrants an intervention until mortality has dropped to ‘normal’ rates. The ‘excess mortality’ concept introduced a hard criterion with which to both start an intervention and to monitor its impact. It is now generally accepted that the application of a specific set of public health interventions, including the provision of food, clean water, shelter and health care can bring mortality down rapidly, ideally within a couple of weeks (24). The health component of this package has typically included the establishment of a surveillance system, capacity to control outbreaks of infectious diseases, development of basic health care facilities, training for health care workers, and – very specifically – immunization against measles. Once mortality is reduced, a system needs to be kept in place to maintain this status quo, until such time that the camp ceases to exist. In practice, it has been accepted that as long as a camp exists and is supported by the international community, basic health services will need to be provided and no specific exit strategy need be thought of. Along with the work in refugee camps and famine relief setting, when populations move into denser concentrations, came the development of a set of technical guidelines [4]. There was a need for existing medical knowledge to be documented in such a way that it could be used under the often harsh conditions of emergency field work, while dealing with massive numbers of patients. For instance, health workers needed to know not only how to calculate fluid replacement for a patient with cholera, but at the same time how to deal with a cholera epidemic in a refugee camp, how to monitor the fatality rate and how to set up a cholera treatment centre with minimal means and to train staff in its efficient use (14). The growing ability of medical humanitarian agencies to control their own logistics meant that long supply chains could be maintained to bring in drugs, vaccines and other medical supplies. Agencies also learned to deal with insecurity. Through strict adherence to security protocols and policies, risk for staff could be minimized. These developments allowed agencies to gain more and more access to areas affected by chronic warfare, where frequently populations were encountered in appalling medical conditions. It was again demonstrated that once access could be achieved, excess mortality could be considerably lowered with relatively simple means, even under very basic conditions. A classic example is the kala azar epidemic that raged from the late 1980s in the middle of southern Sudan, the site of a chronic ongoing conflict for decades. This epidemic, largely unaddressed and exacerbated by the war, caused 100 000 deaths. Ensuring availability of appropriate drugs, such as sodium stibogluconate injections, made it possible to save tens of thousands of lives (20). Although precise data are usually lacking, in this particular case it was possible to measure outcome and a cost-effectiveness analysis demonstrated the relatively low cost of saving a life, despite the requirement for expensive logistics (6). There are numerous examples of situations of successful interventions by health agencies to control infectious disease outbreaks in war-affected areas, such as cholera, malaria and meningitis. All investment in an outbreak control programme, including extensive training of local staff, is typically primarily geared at curbing the epidemic, not at serving longer-term developmental goals. Programmes typically stop when the outbreak is under control. Conceptually much more problematic are situations in which humanitarian medical agencies enter areas affected by ‘complex political emergencies’, to mitigate the collapse of basic health care services. NGOs often substitute for such services, developing facilities in the remains of pre-existing health centres or rural hospitals. Such work typically includes physical rehabilitation of facilities, training of health workers, and provision of drugs and medical supplies. In such ‘open’ populations, in complex emergencies, health information systems are usually absent, logistics and security will often hamper other forms of data collection, while the size of the population is frequently a guess. Excess mortality is, therefore, less likely to be used as justification to intervene, whereas action is engendered by seeking to ensure ‘access to healthcare as a fundamental human right’ (13). But this notion is difficult to operationalize, does not have an end-point, or clear way to monitor its success. Models used in these chronic conflict environments usually follow a Primary Health Care approach, as propagated in many other, more stable areas. However, unlike the situation in refugee camps or outbreak control, it is much less clear what these health services are expected to achieve. That mortality may be very high in these situations has recently been shown by the International Rescue Committee (IRC), which reported an excess mortality of 2.5 million people, over 32 months from 1998 to 2001 in the eastern part of the Democratic Republic of the Congo. A minority (350 000) died as a direct consequence of the fighting, while the overwhelming majority of excess deaths were related to disease and malnutrition (19). This kind of work in complex emergencies, less in the limelight than work in acute emergencies, forms a major part of the work of most specialized health agencies. The kinds of environments in which such work is carried out can vary widely, and changes necessarily take place over time, providing further challenges to intervention models. In one country one may encounter serious disruption, active fighting and/or hardly any leadership from the (remnants of) local health authorities. In other countries areas of relative stability may be found. Also, over time, the instability in a particular area may fluctuate. In some areas there is very little hope for improvement in the near future, while other areas go through ups and downs, not infrequently leading to expectations that things will be better ‘tomorrow’ before the conflict worsens again. Countries where the threshold towards peace and stability has been crossed can be considered a special case, and early transition of the emergency health services into a – usually extensively reformed – regular health system needs to take place. Recent examples include Mozambique, Cambodia, Kosovo and East Timor. In all these different contexts, though, health planning efforts are confronted with similar problems, essentially related to lack of an exit strategy. Once a programme is established and immediate needs are covered, the focus of the assistance shifts towards rehabilitation and reconstruction of parts of a health system. It is usually not explicitly stated, but the assumption is that when conflict is over, the government that will then be in place will take the responsibility to ‘take over’ and will have sufficient means to do so. The term ‘developmental relief’ has been coined to describe humanitarian relief activities which have a longer-term development perspective in mind. However, numerous problems emerge as some degree of stability has been achieved; the new government may have a weak capacity for central planning and policy making, or may not wish ‘to take over’ the services established, or may not have the funds to do so. Furthermore, if instability continues for longer than expected, it becomes increasingly difficult to set planning targets. The end of such projects is more often determined by organization and donor ‘fatigue’ than by a deliberate exit strategy. Not only is the context extremely difficult in these open, chronic situations, but a range of additional compounding factors can be identified. Health service provision has some peculiarities not shared with other emergency subsectors. Health service delivery is complex, located in numerous health facilities with different levels of sophistication, which all need skilled staff and continuous highly specialized supplies, like drugs and other medical goods. Quality control through supervision is also indispensable and someone has to pay for these services. The beneficiaries are unable to do so – even under more ideal situations in stable countries, cost-recovery schemes rarely cover more than a few per cent of actual cost. Planning by an individual agency has limits, as it can only plan for itself. Only a government or other co-ordinating body could place such planning in a wider framework and assess relevance and complementarity of activities of various agencies. It is rare to find such overall frameworks in place in complex emergency situations. Co-ordination is notoriously problematic in humanitarian assistance in general and within the health sector, and good examples of co-ordinated efforts are difficult to trace. Some co-ordination around technical issues may take place, for instance joint discussion regarding how to address tuberculosis, but this is less apparent in relation to setting a general policy for health sector development. No single agency is equipped to do so, and to do such a task collectively, with a wide range of different sorts of agencies with different donors and different mandates, is difficult. At least a generally trusted, ‘neutral’ lead agency would be needed. Although WHO could potentially play this role, it often does not feel sufficiently equipped or mandated to do so (26). Another compounding factor is that most agencies do not ‘just’ deliver health services. They are aware that health work alone is of limited value, while the underlying conflict and its causes continue to be the greatest threat to health. Different agencies have different answers to this, resulting in different – largely self-imposed – mandates and modes of action. One agency may see the conflict as the result of underdevelopment, so will want the health programme to contribute to development. Another agency will focus on the responsibilities of local and international politics and will aim to reach politicians through advocacy. Yet another agency will want to use health work as a tool for (local) peace building. Health as a bridge to peace (27) is the description of one such initiative. Another example is the framework, widely promoted and well known as the ‘Do No Harm’ principles. The framework seeks to ensure that agencies analyse how their activities may influence conflict resolution (2). Finally, and highly relevant to health planning in complex emergencies, is a wider debate in the humanitarian field. This relates to ‘whether responding to needs or upholding rights should be the basic approach governing the work of humanitarians’ (5). This highly complex and rather abstract debate between the two different approaches continues to be lively but unresolved. A stand on either end of the spectrum will have concrete consequences in the design of health interventions, the exit strategy and the evaluation of the interventions. Most implementing agencies are to a greater or lesser extent dependent on donor funds. In this way, the availability of donor money, related donor policies and sector allocations will influence health planning. Donor policies are not only influenced by humanitarian needs, but clearly also by a host of other political and economic considerations. Given aid is increasingly subject to aid conditionalities (12). Predictability of availability of donor funding is low, while underlying donor policies are often not transparent. In particular at times of possible transitions to a different phase in the conflict, donor policies may rapidly switch. Donors typically have different mechanisms and time-scales for funding relief or development activities. The time-scale for relief funding may be extremely short, from 3 months to 1 year. Donors often wish to see some concrete manifestation of their support, and hence there is a propensity to fund infrastructural and building projects. What gets lost in all this are the lower-priority investments in such matters as human resources planning, policy making strengthening, and the development of health information systems. Concerns for quality within agencies themselves, as well as increasing demands for accountability with increasing donor expenditures in the humanitarian sector (1), caused a clear rise in the number of evaluations commissioned. Gradually, methodologies have been developed, based on development aid criteria, but more appropriate for emergency work. For instance, evaluation criteria of coverage, connectedness, coherence and appropriateness were added to relevance, effectiveness, efficiency, impact and sustainability, as used in development assistance (8; 17). However, further criteria also need to be considered in evaluating health care interventions in such complex settings: these are likely to include notions of protection, co-ordination and advocacy. Evaluation of the performance of a single agency or even a group of agencies has limits. Relevance and impact may only be fully understood when taking stock of the performance of other components of the humanitarian system. That would need system-wide evaluation, like the 1995 Joint Evaluation of Emergency Assistance to Rwanda [16]. Another drawback of the use of evaluations is the limited organizational learning and in particular system-wide learning that seems to take place based on the lessons learnt in an evaluation. One of the initiatives spinning off from the Rwanda evaluation was the Sphere project, a joint effort of 228 relief organizations which sought to consolidate ideas on good practice in a handbook, the first edition of which was published in January 2000 (22). The Sphere handbook contains a humanitarian charter and minimum standards, accompanied by key indicators, for five sectors of disaster response: water supply and sanitation, nutrition, food aid, shelter and site management, and health services. This impressive work, although critiqued by some as being potentially open to misuse (7), provides guidelines for work in emergency situations. However, as in the case of other guidelines from the past, the minimum standards for health are largely applicable to camp-like situations and are not directly relevant for health planning in open situations in chronic emergencies. It is envisaged that Sphere will continue to engage in activities to update the guidelines, enhance their evidence-base, and address historically neglected concerns such as gender. The field of humanitarian health work has not been the subject of much research to date. Work that has been carried out is mostly in the realm of disease and epidemic control (15). Instability, insecurity and quickly changing circumstances are a deterrent to research with its need for longer term planning, preparation and implementation. The context of every complex emergency is very different, posing difficulties for case study approaches, i.e. it is not always easy to extrapolate evidence or findings from one case to another. Furthermore, funding research has not been a priority of agencies most immediately involved in the implementation of humanitarian aid. However, calls to provide humanitarian health work with a broader evidence base have been made and gradually some initiatives are taking shape (3). We believe that especially research contributing to strengthening the health policy process is needed. In any such work, the different contexts will have to be taken into account, as will the different stages of the conflict, the mandates and capacities of external aid, and the views of the beneficiaries. In no way should this lead to blueprint approaches or rigid standards. But research may strengthen the process of health policy formulation itself and provide the evidence with which appropriate policies can be informed. Health planning by individual agencies as well as fund raising to finance concrete activities will be greatly facilitated by the formulation of an evidence-informed health policy, specific for the contexts in question and leaving sufficient room for individual agencies and donors to express their respective mandates.