Document 2348
 DOCN  M94A2348
 TI    Verbal autopsy as a tool for assessing adult HIV-1 attributable
       mortality in a rural population.
 DT    9412
 AU    Kamali A; Wagner HU; Kengeya Kayondo JF; Nakiyingi J; Mulder DW; Medical
       Research Council AIDS Programme, Entebbe, Uganda.
 SO    Int Conf AIDS. 1994 Aug 7-12;10(1):324 (abstract no. PC0228). Unique
       Identifier : AIDSLINE ICA10/94370227
 AB    OBJECTIVE: To validate the use of verbal autopsies (VA) for the
       assessment of HIV-1 attributable adult mortality in a rural population.
       METHODS: A rural population cohort was enrolled in 1989/90 through a
       census and serological survey, and kept under surveillance through
       annual surveys. In addition, a village based death registration was
       established. Sera were tested for HIV-1 antibodies using two independent
       EIA systems. A VA questionnaire based on the Bangui AIDS case definition
       was administered by paramedical staff within 4-12 weeks of death to
       relative(s) of the deceased. Assessment of the VA forms was done by 2
       clinicians who, unaware of HIV, status, independently indicated likely
       causes of death. The clinicians' assessments were categorised according
       to whether HIV was mentioned as the likely cause of death. RESULTS: The
       HIV-1 seroprevalence rate in adults (13+ yrs) was 8%. VA assessement are
       available for 154 deaths: 76 seropositives (SP) and 78 seronegatives
       (SN) deaths. Of these, 82 (61 SP; 21 SN) were aged 13-44 years. In 83%
       of SP and 13% of SN the death was attributed to HIV-1. Agreement between
       the assessors was high. The table shows the specificity and positive
       predictive value (PPV) for the age groups 13-44 and 45+. Also shown are
       attributable mortality fractions (AMF) comparing estimates based on the
       use of VA with those from the general cohort surveillance. TABULAR DATA,
       SEE ABSTRACT VOLUME. CONCLUSION: Comparison of the surveillance and VA
       based HIV-1 AMFs suggests that VA is a useful tool to estimate, at low
       cost, the HIV-1 AMF among adults in populations with a prevalence rate
       of 8% or more.
 DE    Adolescence  Adult  Autopsy/*METHODS  *Cause of Death  Cohort Studies
       Human  HIV Infections/*MORTALITY  HIV Seroprevalence/TRENDS  *HIV-1
       Middle Age  Population Surveillance  Rural Population/*STATISTICS &
       NUMER DATA  Truth Disclosure  Uganda/EPIDEMIOLOGY  MEETING ABSTRACT

       SOURCE: National Library of Medicine.  NOTICE: This material may be
       protected by Copyright Law (Title 17, U.S.Code).