Uganda has an under-five mortality of about 156/1000 per year, caused mostly by IMCI-related diseases. IMCI implementation began in 1996 in three districts and by 2000, IMCI had been introduced to 55 of the 56 districts in Uganda as a priority of the Ministry of Health. Because Ugandan health services are decentralized, districts can choose to implement IMCI independently. Successful implementation largely reflects district management priorities, capacities, and a district’s resources for planning, training and supervision.

The main objective of the evaluation was to determine whether full IMCI implementation would lower childhood mortality. Additional objectives were to document changes in process and outcome indicators at the facility, community and household levels, and to assess the costs associated with IMCI.

The design of the Uganda IMCI impact study was based on a dose-response model involving 10 districts. These districts were  initially selected in collaboration with the Government to represent a range of probable implementation strengths related to IMCI. In the first phase of the study (2000), a baseline demographic household survey was carried out to assess demographic and health indicators in 14,000 households selected from the ten districts. District surveys were used to monitor IMCI implementation at the health facility and community level and consisted of three health facility surveys, conducted in 2000, 2001 and 2002, and two community monitoring surveys conducted in 2001 and 2002. A series of IMCI adequacy indicators were estimated to determine if the level of IMCI implementation could be expected to have an impact on mortality rates. Examples of such adequacy indicators include coverage of health facilities with IMCI-trained health workers, health worker compliance with key IMCI performance standards, availability of IMCI drugs, vaccines and equipment/supplies, care-takers' health-seeking behaviour, and treatment compliance.

Preliminary analysis demonstrated that IMCI case management training was associated with significant improvements in the quality of care received by sick children visiting first level health facilities. Trends in performance over time were variable. Although the community component of IMCI was under development, as of 2002 coverage was still limited.  A national meeting convened to assess IMCI implementation in the 10 study districts concluded that coverage rates remained too low to expect a measurable impact on child mortality within the time frame of the MCE.  The data collection efforts were therefore stopped in 2003, and investigators focused their attention on the analysis and publication of the data collected up to that point.

This study was a collaborative partnership between Johns Hopkins University (USA) and Makerere University (Uganda), with initial funding under a grant from the United States Agency for International Development. The two main Makerere collaborators were the Institute of Public Health (IPH) and the Department of Population Studies (DPS) at the Institute of Statistics and Applied Economics. Active collaboration was maintained with the Ministry of Health, the Uganda office of UNICEF, WHO and WHO AFRO, as well as USAID Washington, and the USAID mission in Uganda throughout the study period, culminating in a major dissemination meeting held in 2003.