Page 114 - SAHCS HIVMed Journal Vol 20 No 1 2019

Page 3 of 7 Original Research

pre-intervention training on the specific content of the interviews, to evaluate the changes linked to the integrated key
integrated key population training. Local government and population sensitisation training. Similar methods have been
civil society organisations providing services in the provinces used to assess changes in healthcare workers’ knowledge,
where the training was provided, nominated the healthcare attitudes and practices related to training in African contexts,
workers who participated in the training. about HIV treatment and also mental health. 36,37

This article presents the findings from a mixed-method study Quantitative component
evaluating the outcomes of the key populations’ sensitisation Healthcare workers who received the intervention were
training in selected locations. The evaluation objectives were invited to complete anonymous paper-based pre- and post-
to (1) assess changes in the perceptions, attitudes and training assessment questionnaires. The questionnaires
knowledge of healthcare workers regarding HIV-related assessed previous training and experience working with key
issues affecting MSM, sex workers and PWUD, and (2) to populations, knowledge about key populations’ health
assess changes in healthcare workers’ attitudes, for example, needs, as well as attitudes, opinions and beliefs pertaining to
stigma and discrimination towards MSM, sex workers and key populations and their behaviours. All of the completed
PWUD who access health and related services.
paper-based questionnaires were entered into an Excel
5
Methods spreadsheet and imported into Stata 14. Quantitative
analysis included comparison of proportions and frequencies,
Background to the evaluation research as well as t tests for independent proportions to determine
Of the five provinces that received training, two provincial if there were significant differences between pre- and
38
capitals were selected for evaluation: Bloemfontein post-training responses.
(Mangaung Metropolitan District, Free State Province), in
which 84 healthcare workers received the sensitisation Qualitative component
training, and Mafikeng (Ngaka Modiri Molema District Prior to the training intervention, baseline in-depth
Municipality, North-West Province), in which no training interviews (IDIs) were conducted with eight healthcare
intervention was implemented. These locations were selected workers purposively sampled from four government health
because they are characteristic of the public health system facilities in Bloemfontein and four in Mafikeng. Three months
outside of major South African metropolitan areas (poor after the training intervention was implemented in
infrastructure, under-resourced, operating with inadequate Bloemfontein, follow-up IDIs were conducted with three
health personnel, and with significant financial and healthcare workers in Bloemfontein and one in Mafikeng (the
geographic barriers to accessing healthcare), 33,34 and because variance in sample size at follow-up was because of
implementation and scale up of key population HIV respondents declining follow-up interviews). Both baseline
programmes in these areas were lacking at the time the and follow-up IDIs explored healthcare workers’ narratives
research was completed. The number of locations was limited of their experience of working with key populations viz.
by the resources available for the evaluation. attitudes and views of the behaviour and social vulnerabilities
of this group and of their own capacity to provide appropriate
Compared to other South Africa provinces, both North-West and nonjudgmental services to them. In addition, views with
and the Free State have low population densities, each regard to the training of healthcare workers in this field were
containing 7% (N = 3 856 174) and 5% (N = 2 866 678) of the elicited from the interviewees.
National population, respectively. The South African
3
National AIDS Council estimated that in 2018 there were Interviews were conducted by a female socio-behavioural
1228 female sex workers in Mangaung and 1753 in the interviewer and a male research assistant, both competent in
Ngaka Modiri Molema District. MSM estimates for these English and local languages, and trained in the specific
districts are 3655 and 3779, respectively. HIV prevalence for implementation of the research tools. IDIs were conducted at
these districts is estimated to be similar: 53% for female the health facility of each participant, often in consultation
sex workers and 18% for MSM. PWUD estimates have not rooms during tea or lunch breaks. IDIs were conducted in a
4
been developed because of lack of data. At the time of the combination of English and local languages, largely
research, health services for MSM and sex workers were Sesotho and Setswana, and were audio-recorded with
limited (but have subsequently increased), no PWUD consent from the interviewees. Debriefing notes were
19
outreach programmes or harm reduction services existed captured after each IDI by the interviewer and research
35
and the availability of water-based condom-compatible assistant. Audio-recordings were transcribed and translated
lubricants was poor.
into English, and transcriptions were quality-checked by the
multi-lingual lead interviewer. NVivo 10 software was used
6
This study used a combination of qualitative and quantitative
methods to evaluate the intervention. We used quantitative pre- to code and manage qualitative data. Two coders read and
and post-training questionnaires, combined with qualitative coded the transcripts independently, and any discrepancies
were resolved through inter-coder consensus. Predetermined
3.StatsSA.
5.StataCorp, College Station, TX.
4.South African National AIDS Council Key Populations Cascade Stakeholder Group,
February 2018. 6.NVivo qualitative data analysis Software; QSR International Pty Ltd., Version 10, 2014.
http://www.sajhivmed.org.za 107 Open Access