Page 186 - SAHCS HIVMed Journal Vol 20 No 1 2019

Page 3 of 7 Original Research

13
were digitally recorded and transcribed verbatim. The the trustworthiness of our study. The researchers held several
researcher (LD) facilitated the FGDs and individual meetings to debrief during data collection and analysis of
interviews and was assisted by an isiXhosa-speaking interviews. Transcripts were shared amongst the researchers
interpreter. isiXhosa transcriptions were translated to to check for quality and to check coding and formulation of
English. The researcher (LD) was a medical doctor, who themes. Disagreements were discussed until consensus on
worked in the HIV programme in City Health and conducted themes was reached. An independent person was used to
the current research towards her master’s degree. transcribe the interviews and FGDs. Translation from isiXhosa
to English was done by a first language speaker in isiXhosa,
The focus groups were divided according to gender and age with master’s level qualification in public health.
as follows:
Ethical considerations
• Females, 10–14 years old – 3 participants
• Males, 10–14 years old – 4 participants Ethics clearance for the study was provided by the University
• Females, 15–19 years old – 6 participants of the Western Cape Biomedical Research Ethics committee
• Males, 15–19 years old – 2 participants. (Registration number: 15/7/254) and approved by City
Health (ID number: 10537). All information was treated
Focus groups were age and gender aggregated, to allow confidentially, and all participants’ anonymity maintained.
free sharing within the group with a similar demographic Participation in the research was voluntary, and upon
or peer group. The FGDs were held in a meeting room in obtaining informed consent from all participants, and parents
the facility. or guardians (if adolescent was younger than 18 years).

Two participants from each FGD, who were identified to be Results
willing to share more valuable insights, were approached for
follow-up individual interviews. In the case of FGD 4, both In this study, several barriers and one facilitator of
participants agreed to be interviewed. The interviews and adolescents’ ART adherence were identified. A presentation
FGDs were semi-structured, with open-ended questions and of these factors follows below.
prompts. An interview guide was compiled for this purpose
in order to ensure standardisation across focus groups. Barriers to adherence

The reported barriers to adherence were school, social, health
In addition, two key informant interviews were conducted services, treatment- and patient-related factors.
with nurses who worked in the HIV programme in the clinic.
School factors
Data analysis School factors such as school (work) commitment,

The interviews were analysed manually making use of communication with school teachers and negative teacher
content analysis. Analysing the data started with reading attitudes were found to play a deterring role in accessing the
13
and re-reading the transcripts several times. This was clinic, disclosure and adherence to ART. Participants often
performed concurrently with reading the field notes, personal expressed feeling conflicted between school commitments
reflections and reading entries from my research diary. and the need to attend clinic appointments. Even though
there was a school adjacent to the clinic, many participants
On reading a transcript for the third time, the researcher attended school elsewhere:
(LD) made pencil notes in the margin of all the main issues ‘It would be nice for us to come at our own time, so that we do
that relate to adherence to treatment which came out from not have to miss our school work. That way we can be able to
the text. The researcher was as inclusive as possible and balance our life. Your school work doesn’t suffer because of the
also considered the things which were not being said such clinic appointments, and vice versa.’ (Group 4, male, 18)
as suggestive statements and links between statements in
different parts of the interview. Then, the list of all codes In addition, the need to communicate attending regular clinic
was transferred onto a separate page. In the next step, visits to teachers posed a significant barrier to attending
the researcher re-wrote the list of codes, but this time regular clinic follow-ups as they feared unintended disclosure
highlighting codes which were duplicated or emphasised which may potentially lead to stigma and discrimination:
by the participants. Similar codes were then grouped ‘Okay my life orientation teacher is not a friendly person. She
together, and in the last step, themes were developed. A likes to shout, beat and [is] always angry. When she is angry,
consensus was reached between the researchers (LD and she says a lot of things out of anger; imagine now if you tell her
BVW) on the themes and codes. about your status, and when she is angry she burst out in
front of everyone. The best way is to keep this to myself.’
(Group 4, male, 16)
Trustworthiness
‘It does not sit well with me, because people will be suspicious,
We followed Lincoln and Guba’s criteria for credibility, they will have questions about my frequent visits to the clinic.
transferability, dependability and confirmability to enhance That does not make me feel right.’ (Group 4, male, 18)

http://www.sajhivmed.org.za 179 Open Access