In the above discussion on "collaboration between WHO/HQ programmes and NGOs", four conditions were listed as essential if information and decisions made at the international level are to flow down to national level and strengthen collaboration between NGOs, MoHs and WHO in drug distribution.

The country studies showed, however, that agreements and knowledge accumulated at international level do not necessarily flow down to country level. They also showed that WHO programmes focus extensively on MoHs and only to a limited extent on NGOs, even though NGOs are important in drug distribution and supply in all three countries. The clearest example of this approach was provided by the GTB programme, which focuses heavily on the MoH in Uganda, even though the German Leprosy Relief Association (GLAR) is heavily involved in distributing TB drugs in this country. GLAR funds leprosy activities in 33 districts and TB activities in 19 out of 39 districts. GLAR is satisfied with its role in providing services to the MoH, but closer interaction between GLAR, the GTB programme at HQ, and the WHO country office could have been expected.

The GLAR example is typical of the lack of interaction between WHO country offices and NGOs. In Kenya, Malawi and Uganda the following two questions were raised concerning the role of WHO:

• What can WHO offer NGOs? (Not many NGOs could answer this question.)

• Could WHO use its strong links with MoHs in order to get NGOs and MoHs to work more closely together, thereby helping to focus the work of MoHs and NGOs in the same direction?

The problems addressed by NGOs are recognized by WHO at country level. However, since there is no specific demand from programmes at HQ or from the African Regional Office (AFRO) to work with NGOs at country level, little is being done to encourage or initiate such collaboration. Interestingly, simply undertaking research in this area was seen to act as a catalyst for increased interaction between the stakeholders in drug distribution and supply. This was illustrated by the NDA in Uganda and KNDPIP in Kenya, who both used the research carried out for this analysis as a starting-point for activities of their own.

In Kenya and Uganda, NGOs and MoHs showed a clear interest in and capacity to undertake collaboration. Such interest was less apparent in Malawi. In all three countries, NGOs see WHO as a mediator who could bring NGOs, national drug programmes and MoHs together, thereby acting as a clearing-house for and stimulus to interaction. Several NGOs emphasized that they prefer clear and explicit rules for interaction and would appreciate the MoH and the national drug programme taking a leading role in initiating collaboration with NGOs.

The fora for collaboration between the stakeholders in the three countries are very diverse. They range from an actual meeting involving the stakeholders in Uganda, to the possibility of CONGOMA acting as a future catalyst for NGO involvement in Malawi, to the existence of structures for collaboration in Kenya which are currently unused. This diversity reflects the different attitudes displayed towards NGOs in the three countries. In Uganda, the MoH is open to NGOs. In Malawi, the MoH is more circumspect, while in Kenya, the MoH is generally positive towards NGOs. Furthermore, personal perceptions appeared to be a very important determinant of the degree of collaboration. The issue of whether NGOs were in official or informal relations did not seem to have any significance at country level.