Sentinel Panel of Districts (SDP)

IHI have made good progress in developing the facility based and population based arms of the Sentinel Panel of Districts (SPD). A new information platform, SPD, will generate facility-based and population-based health statistics that are representative of Tanzania Mainland as a whole.

The SPD is designed as a new, nationally representative data collection system with two arms to it. It will operate in 27 districts selected to be nationally representative. The facility based arm will collect data on health outputs delivered in all health facilities in the selected districts using the standard Health Management Information System - HMIS (20% of health facilities in Mainland Tanzania) for example it captures data on births in institutions, antenatal sessions provided, HIV tests conducted, etc. There is also scope to collect additional data on an ad hoc basis.

The population based arm will collect data on a sample of villages in the chosen districts, it captures births, deaths, and cause of death through use of SAVVY (Sample Vital Registration with Verbal Autopsy) methodology.

SAVVY will provide annual, age/sex and cause-specific mortality estimates, including mortality attributable to HIV/AIDS in the 18-59 year age group. The SAVVY will be based on a sentinel population distributed across Tanzania so as to provide a nationally-representative estimate. The total sample size should be large enough to detect a relative change of 20% in HIV/AIDS mortality over five years.

A two-stage probability sampling approach was employed and 23 districts were selected with probability proportional to size (PPS); and within these, we randomly selected 10 villages/enumeration areas, comprising around 7,000 households (approximately 35,000 people). This will yield a total national SAVVY population of approximately 805,000, which is expected to detect 600 – 1,000 HIV/AIDS deaths per annum. Reporting of deaths will be done by supervised community informants, followed up by verbal autopsy and coding of cause of death using the international standard verbal autopsy approach and ICD10 classification. Denominators will be determined by conducting censuses during year 1 and year 4, following the SAVVY methodology. Denominators for year 2, 3 and 5 will be estimated using established populations projections methods. The system will permit annual estimation of the national number, rate and proportion of HIV/AIDS deaths in the 18-59 year age group, stratified into urban/rural and male/female sub-groups. In addition to population-based estimates of AIDS mortality, the system will provide a platform that may be employed for diverse HIV/AIDS research and other health initiatives.

Project implementation will be led by Ifakara Health Institute (IHI), supported by the National Bureau of Statistics, the National Institute of Medical Research (NIMR-Mwanza) and in close consultation with the Ministry of Health and Social Welfare. IHI will take overall responsibility for project planning, management and financial oversight. National Bureau of Statistics will provide technical assistance on sampling methodology and data analysis, quality-assurance of SAVVY reports and supportive supervision from NBS. NIMR-Mwanza will provide HIV-AIDS technical expertise and will supervise sample districts lying within geographical reach. Project implementation will be planned in phases, starting in year 1 (2009/10) and reaching full scale in year 5 (2013/14).