Sampling Procedure
The sample for the 2011 UAIS covered the population residing in households. A representative probability sample of 11,750 households was selected for the survey. The sample was constructed to allow for separate estimates for HIV/AIDS indicators for each of 10 geographic regions. The regions were created for the survey and do not represent administrative units of the country. Other than Kampala, each region comprised between 8 and 15 contiguous administrative districts of Uganda that share similar languages and cultural characteristics. Because of its unique character as an entirely urban district and capital city of Uganda,
Kampala comprised a separate region. The 10 regions were comprised of the following districts1:
• Central 1: Bukomansimbi, Gomba, Lwengo, Lyantonde, Kalangala, Kalungu, Masaka, Mpigi, Rakai, Ssembabule, and Wakiso.
• Central 2: Buikwe, Buvuma, Kayunga, Kiboga, Kyankwanzi, Luwero, Mityana, Mubende, Mukono, Nakaseke, and Nakasongola.
• Kampala: Kampala district.
• East-Central: Bugiri, Buyende, Iganga, Jinja, Kaliro, Kamuli, Luuka, Mayuge, and Namutumba
• Mid Eastern: Budaka, Bududa, Bukwa, Bulambuli, Busia, Butaleja, Kapchorwa, Kibuku, Kween, Manafwa, Mbale, Pallisa, Sironko, and Tororo.
• North East: Abim, Amudat, Amuria, Bukedea, Kaabong, Kaberamaido, Katakwi, Kotido, Kumi, Moroto, Nakapiripirit, Napak, Nora, Serere, and Soroti.
• West Nile: Arua, Adjumani, Koboko, Moyo, Nebbi, Maracha, Yumbe, and Zombo.
• Mid Northern: Agago, Alebtong, Amolatar, Amuru, Apac, Dokolo, Gulu, Kitgum, Kole, Lamwo, Lira, Otuke, Oyam, and Pader.
• South Western: Buhweju, Bushenyi, Ibanda, Isingiro, Kabale, Kanungu, Kiruhura, Kisoro, Mbarara, Mitooma, Ntungamo, Rubirizi, Rukungiri, and Sheema.
• Mid Western: Buliisa, Bundibugyo, Hoima, Kabarole, Kamwenge, Kasese, Kibaale, Kiryandongo, Kyegegwa, Kyenjojo and Masindi.
The sample was allocated equally across all 10 regions, so as to allow a sufficient size to produce reliable estimates in each region. Since the sample was not allocated in proportion to the size of each region, the UAIS sample is not self-weighting at the national level. Consequently, weighting factors have been applied to the data to produce nationally representative estimates.
The survey utilised a two-stage sample design. The first stage involved selecting sample points or clusters from a list of enumeration areas (EAs) covered in the 2002 Population Census. A total of 470 clusters was selected (47 in each region), comprised of 79 urban and 391 rural points. The second stage of selection involved the systematic sampling of 25 households per cluster from a list of households in each cluster that was produced by the Uganda Bureau of Statistics prior to the UAIS data collection.
All women and men age 15-59 years who were either permanent residents of the households in the sample or visitors present in the household on the night before the survey were eligible for interviews. All women and men who were interviewed were asked to voluntarily give a blood sample for testing. In addition, blood samples were drawn from children under age 5 after obtaining consent from their parents or caretaker.
(Refer Appendix A of the final survey report for detail sample design and implementation)