Health and mortality survey among internally displaced persons in Gulu, Kitgum and Pader districts, northern Uganda, Jul 2005


Executive summary
I. Background

In northern Uganda, nearly two decades of conflict have resulted in the internal displacement of up to two million persons. In Gulu, Kitgum and Pader Districts, the most affected by violence, nearly 90% of the population had relocated to camps as of 2005. The Government of Uganda's national policy for internally displaced persons (IDPs) calls, among others, for provision of medical care and water and sanitation to all IDPs by both central and local government, and entitles them to security of person and property. The Office of the Prime Minister is tasked with advocating on behalf of IDPs and sharing information nationally and internationally on their plight. In 2005, the Ministry of Health of Uganda and the United Nations Childrens' Fund (UNICEF) requested assistance from the World Health Organization (WHO) to assess the health status of IDPs in the three Districts. The study was led by the Ministry of Health and WHO in partnership with the offices of the District Director of Health Services of Gulu, Kitgum and Pader, UNICEF, the UN World Food Programme, the UN Population Fund, and the International Rescue Committee. Study objectives were:

  • primary: to estimate crude mortality rate (CMR) and under 5 mortality rate (U5MR) in the period between 1 January 2005 and the survey date (July 2005) among populations now living in IDP camps, recognised or unrecognised at the time of the survey, in Gulu, Kitgum, and Pader districts.

  • secondary: to measure other demographic indicators (age/sex structure of the population; monthly mortality; self-reported causes and circumstances of death; total excess mortality, violent deaths and abductions); estimate measles vaccination coverage; investigate caregivers' management of episodes of illness among children under 5 in the two weeks prior to the survey; estimate bednet availability and coverage among children under 5; and measure water availability indicators (source of drinking water, quantity per person per day, time taken to collect water).

II. Methods

We conducted four separate surveys designed to be representative of all IDPs in Gulu District camps, Gulu Municipality unrecognised camps, Kitgum District camps and Pader District camps. Sample size was determined so as to measure a CMR and U5MR in excess of non-crisis levels. Households within each population were selected using a 30 cluster x 32 household design, with clusters allocated proportionally to camp population size (i.e. total 120 clusters). Teams of two interviewers administered a structured, pre-piloted questionnaire in the Acholi language (Luo) to adult household respondents. We analysed results separately for each population (stratum), and jointly for the entire Acholi region after weighting for stratum population size.

III. Main findings

Demographics and mortality:

  • Both CMR and U5MR were well above respective emergency thresholds (1 per 10 000 per day and 2 per 10 000 per day) in all four surveyed populations, and were four times higher than non-crisis levels in Kitgum and Pader Districts (Table 1).

  • There was no obvious monthly trend in CMR.

  • Malaria/fever and AIDS were the top self-reported death causes (Table 1); among children under 5, top causes were malaria/fever and two lango, a local illness concept encompassing oral thrush, malnutrition and diarrhoea.

  • Less than half of all deaths (Table 1), and 54.2% among children under 5, occurred in a health facility.

  • A total excess mortality of 25 694 (of which 10 054 children under 5) can be projected for the entire Acholi region between January and July 2005, namely almost 1000 excess deaths per week (Table 1).

  • Violence was the third most frequent cause of death (9.4%), occurring mostly outside of camps (68.8%) and health facilities (93.5%). Persons killed were mostly adult males (70.1%), but 16.9% were children under 15. We estimate that 3971 persons were killed in the study population between January and July 2005 (Table 1).

  • Age/sex population pyramids in Gulu District, Kitgum District and Pader District display an apparent deficit in males 20 to 30 years old.

(pdf* format - 2649 KB)