Deaths among humanitarian workers

Summary of published paper

A group of researchers recently collected information from the records of aid agencies and other organisations on the death of humanitarian aid workers between 1985-98¹. Thirty two organisations provided data on 392 fatalities. Deaths were categorised by demography, occupational factors and circumstances of death. The team had hoped to calculate risk rates and ratios for nationals and expatriate workers but information about staffing levels that was needed to construct denominators, was only available from a few organisations.

Intentional violence was recorded as the cause of 68% of deaths, and unintentional violence in 7% of cases. Motor vehicles accounted for 17% of cases and other causes for 8%. Death from intentional violence was most common between 1992-5. Fifty eight percent of deaths were among local staff and 42% amongst expatriates. Mean age at death was 38.2 for locals and 39.9 for expatriates. Men accounted for 90% of deaths.

Information on types of occupation was only available for 227 workers who died:

Drivers 13%
Guards 12%
Office staff 21%
Field staff 22%
Medical staff 9%
Peace keepers 19%
Consultative staff 4%

Intentional violence was the cause of 76% of deaths among guards, 57% among medical staff and 56% among field staff. Occupations with the highest proportion of deaths caused by motor vehicle accidents were drivers, peacekeepers and office staff.

Deaths peaked in 1994 at the time of events in Rwanda. Since 1994 reported deaths among UN staff have decreased whereas deaths amongst NGO workers has continued to increase. The Great Lakes and Horn of Africa accounted for nearly half of all deaths.

While the number of deaths has increased, so have the number of relief agencies and the number of humanitarian workers. Without denominators for field staff the researchers could not calculate risk or rates making it difficult to ascribe the increased number of deaths to increased risks.

The large number of deaths due to intentional violence contrasts with death among Peace Corps and development workers and emphasises the violent circumstances in which humanitarian workers now operate. Descriptions of death included victims being robbed and killed while at an office, residence or roadblock or killed during a carjacking. Unintentional deaths were in some cases related to carelessness like running into a spinning airplane propeller. Chronic and acute disease figured prominently among death from other causes. Cerebral malaria, a preventable cause of death, was noted frequently. One third of deaths among NGO workers were due to 'other causes' compared with 5% in workers from the UN. This represented a wide range of both acute and chronic medical conditions. One reason could be that health screening is less stringent for NGOs hiring temporary staff than it is for UN organisations. Health screening for local staff is likely to be cursory in emergency situations.

The mean age at death of nearly 40 contradicts perceptions that deaths are mainly among young people who are ill prepared. The ratio of deaths among nationals to those among expatriates was 4 to 3. Most relief organisations report usual staff ratios in field operations of 7 to 1 or 8 to 1. Nearly a third of all deaths occurred in the first three months of duty with one of every six deaths occurring in the first month. This was unrelated to extent of previous field experience. Even allowing for short term contracts, common during emergencies, new arrivals may not be prepared for the dangers present, including driving risks.

The authors of the study suggest that actions to lessen deaths can be taken. These include accurate understanding of risks, better briefings and guidelines, providing helmets and protective jackets where appropriate, improving driving skills, managing stress better, and handling cash and protecting assets in other ways. Another action would be to limit aid in high risk situations, a decision many humanitarians find difficult to make.

The authors concluded that in order to fully understand the risks of death and the potential for prevention a comprehensive prospective approach to data collection and monitoring is needed and that much could be learned from such a database.

Post Script

A news item in the BMJ² reported the cessation of OLS flights in the wake of bombing raids against relief facilities. Roger Winter of the US committee of refugees claimed that "these bombings are clearly deliberate and that the Sudanese government is targeting southern Sudanese civilians and relief workers".

The air raids followed a series of intemperate articles in the Sudanese press alleging that aid agencies have been helping to arm and feed the rebel SPLA

¹Sheik et al (2000): Death among humanitarian workers. BMJ volume 321, 15th July 2000

²One Million threatened by suspension of Sudan's relief lifeline. BMJ vol. 321 19-26th August page 470