|
Methods: This research will identify
factors contributing to social, health and educational risks in a longitudinal
study of 3000 12-18 year olds in South Africa. Children from urban and
rural areas with high HIV prevalence will be recruited and followed for two
years. Study sites will be urban and rural areas in Kwa-Zulu Natal, Mpumalanga,
and the Western Cape will be recruited and
followed for two years. At recruitment, youth will be categorized with
reference to their primary parental caregiver’s AIDS-related morbidity and
mortality: as AIDS-orphaned (primary parental caregiver has died due to AIDS),
parental AIDS-affected (primary parental caregiver living with HIV/AIDS), or
parental AIDS-unaffected (primary caregiver alive and not living with
HIV/AIDS). We will also classify other-orphans (youth orphaned by causes other
than parental AIDS) or other parental illness (youth whose parents have other
chronic diseases).
At each time point we will assess children’s mental and physical health,
educational, sexual risk and social outcomes. We will collect information on
parental health and mortality, child caring, household composition and economic
conditions, access to social grants and health services, school attendance,
children’s social networks, family and peer dynamics. Interviewer-guided
questionnaires have been designed and pre-piloted with children.
Sampling: Carers and children will be sampled utilizing stratified systematic
random sampling of South African Census enumeration areas (EAs). Within EAs,
GIS mapping will identify random routes (in dense informal areas are unsuitable
for traditional street-based sequential sampling), using methodology developed
for SA Census collection in rural and informal areas (Stoker, 1985).
Determination of adult AIDS-illness and death will be made using WHO clinical
criteria and the verbal autopsy (VA) method. The VA method has been validated
in Southern African community samples for determining AIDS-related deaths in
high-prevalence areas where clinical evidence is unavailable or unreliable
(Hosegood et al., 2004, Lopman et al., 2006), and showed sensitivity of 83% and
specificity of 75% (Lopman et al., 2006). Our previous research has
successfully used this international standard to distinguish AIDS-orphaned and
other-orphaned children in South Africa. Fieldwork staff will be rigorously
trained and supervised on use of this method.
Measurement tools: Scales and items have been selected based on prior use with
this population, standardisation and strong psychometric properties. Many have
been tested in our prior studies (Cluver et al., 2007, Cluver et al., in
press-c, Cluver et al., 2008, Cluver et al., in press-b, Cluver et al., in
press-a). They have been further informed by our ongoing qualitative study in
the Western Cape. Where possible child self-report will be supported by data
from school and health records. Tools will be translated and backtranslated
into isiXhosa siSwati, isiZulu, isiNdebele and Sepedi. Children will complete
face-to-face 1 hour interviews in the language of their choice. The survey will
assess psychological health (i.e. depression, anxiety, PTSD and behavioural
problems) physical health (i.e. contraction of TB, nutrition, upper respiratory
tract infections), education (i.e. school enrolment, attendance, performance),
and social functioning (i.e. peer relationships, family functioning). Potential
mechanisms for child outcomes will include poverty, parental illness, extent
and timing of caregiving tasks undertaken by children, stigma, parental
monitoring, state support and succession planning.
Ethics: Ethical processes and full informed consent are essential. Interviewers
will be isi-Xhosa-speaking auxiliary social workers or community health
workers, trained in all aspects of the research protocol (recruitment, ethics,
survey administration, data safety management), empathetic and experienced in
working with HIV-affected children. The planning of this research was informed
through consultations with SA Department of Social Development officials, NGO
representatives, local social service providers, and an advisory group of
AIDS-affected children. For the ongoing qualitative stage, ethical approval has
been granted by Oxford University, University of Cape Town, Cape Town Child
Welfare and National Department of Social Development. For the proposed
quantitative stage, ethical approval will be sought from all these bodies, as
well as the Mpumalanga Health REC.
|
