The mix of methodologies was very useful in both urban and rural settings, and in the public and private facilities. It allowed for both flexibility and cross-validation of data. The FGDs were particularly informative. The study yielded a large amount of additional qualitative data. Wherever possible, this information has been used. The in-depth interviews, which were carried out in both rural and urban settings, yielded not only information on factors facilitating or constraining adherence but also more general information on trends in the use of ARVs in Uganda. Although originally the research team believed that, for ethical reasons, treatment observations may not be possible, the researchers were allowed by staff and patients to observe while consultations were going on.