Some 98,000 Iraqi civilians have died since the US-led invasion began in March 2003, epidemiologist Les Roberts and his colleagues estimated by conducting a population-based survey. The report, in a recent issue of the Lancet, seemed to take the world by surprise. Many critics felt the estimate had to be exaggerated. Their surprise was misplaced. Surveys previously conducted by Roberts and the International Rescue Committee in the eastern part of the Democratic Republic of Congo (DRC) had estimated that 2.5 million civilians lost their lives in the intractable conflict that raged between 1998 and 2001. In other recent conflicts, from Bosnia to Sierra Leone, from Afghanistan to Rwanda, from Liberia to Chechnya, and from Cambodia to Colombia, civilians have been the principal victims.

The nature of war has changed considerably over the past century. Instead of armies in distinctive uniforms fighting across clearly drawn battle lines, modern conflict is characterized by guerillas or insurgents striking at seemingly randomly selected targets. This fundamental change is seen perhaps nowhere more clearly than in the character of victims. It has been estimated that during the First World War, more than 85 percent of those who lost their lives were soldiers, mostly young men doing battle for their countries. But in our day, 90 percent of those who die are civilians—most frequently women and children caught in the crossfire between fighting forces whose ultimate goal is no longer territorial conquest, but rather societal disruption. This suggests that despite the best efforts of politicians and humanitarian groups, war and public health remain inherently at odds.

Mortality in Complex Emergencies

The term “complex emergency” describes a situation in which a large civilian population is affected by a combination of war, civil strife, food shortages, and population displacements. Although there are a few exceptions, complex emergencies are characterized by substantially elevated mortality rates, especially in the acute phase. An arbitrary threshold, above which an emergency is said to exist, has been established at one death per 10,000 people per day, or about three per 1000 per month. This rate is approximately two to four times the baseline rate of mortality in developing countries. (This threshold is probably less relevant in developed countries, where baseline mortality levels are considerably lower.)

Calculating this figure allows measurement of the magnitude of complex emergencies, enabling objective comparisons between them. Serial measurements help to establish temporal trends of severity. For example, the DRC survey mentioned above reported a crude mortality rate of 5.4 deaths per 1000 per month from August 1998 to April 2001. This rate fell dramatically to 3.5 deaths per 1,000 per month in 2002, but the fact that it remained above the threshold, however arbitrarily defined that threshold may have beeen, indicated that a dire situation still prevailed.

Deaths in complex emergencies are often the direct result of violence. When populations are attacked in genocide or ethnic cleansing, civilians are intentionally killed. During the war in the former Yugoslavia, for example, the estimated deaths from violence ranged between 50,000 and 200,000 in Bosnia alone. The number of deaths due directly to violence during the 1994 Rwandan genocide was estimated from 500,000 to 800,000. As is evident from the wide range of these estimates and wide confidence intervals around Roberts’s estimate of mortality in Iraq (8,000 to 194,000), the accurate measurement of mortality during conflict is fraught with difficulty. In the Darfur region of Sudan, where genocide is widely believed to be occurring, a survey by EpiCentre, the epidemiological wing of the volunteer agency Doctors Without Borders, found mortality rates among internally displaced people residing in camps to be moderately elevated. Mortality in the population prior to their arrival in the camps, however, was 7.3 deaths per 10,000 people per day, with almost all of the deaths occurring in adult men and attributed to violence.

In most complex emergencies, however, most civilian deaths result not from violence but from the indirect consequences of war. One of the most obvious ways war affects civilians is by displacing them from their homes. From a legal perspective, refugees are people who flee their homes “because of a well-founded fear of persecution on account of race, religion, nationality, membership in a particular social group, or political opinion” and who are outside the country of their nationality. They are entitled to the international community’s protection under the 1951 Convention Relating to the Status of Refugees. Those displaced for the same reasons, but within their country, are termed internally displaced; their protection remains the responsibility of their government, although frequently it is precisely that government that persecutes them. The Global Internally Displaced Persons project of the Norwegian Refugee Council lists 11 conflict-affected countries as having more than 500,000 internally displaced people. At times, usually due to external pressure, governments allow the international community to provide assistance to the internally displaced, but in many countries, including Sudan, Burundi, and Cote d’Ivoire, tensions commonly arise between international relief agencies, either UN or non-governmental, and national authorities.

Most information we have regarding mortality and its causes in complex emergencies comes from refugee or internally displaced persons camps, although data is increasingly becoming available from non-camp settings such as those in war-torn areas in the DRC and Afghanistan. Camps are typically organized and administered by an UN agency or an host country government. Services are usually provided by a wide array of humanitarian assistance organizations from around the world. These operate, to a varying degree, with funds provided by the United Nations, other multilateral donors such as the European Union, or single-country donors. The level of technical assistance provided to people affected by emergencies is highly variable in terms of both quantity and quality, and both the level of financial assistance and its duration have proven unpredictable.

Mortality rates in refugees and internally displaced populations can vary considerably. At times, the arbitrary threshold of one death per 10,000 per day is barely, if at all, exceeded. At other times, crude mortality reaches as high as seven to ten times the baseline rate. When the Tutsi militia succeeded in overthrowing the Hutu genocidaires in Rwanda in 1994, between 500,000 and 800,000 Hutu refugees fled to the area surrounding Goma in neighboring Zaire (now DRC). During the first three weeks following their arrival, crude mortality rates soared to more than 30 per 10,000 per day, about 50 times the baseline rate. In that short period, 45,000 people, or between seven and nine percent of the refugee population, perished.