h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5

H O S T E D BY Available online at www.sciencedirect.com

ScienceDirect

journal homepage: http://ees.elsevier.com/hsag/default.asp

HIV stigma experiences and stigmatisation before

and after an intervention

H. Christa Chidrawi a, Minrie Greeff a,*, Q. Michael Temane a,

Colleen M. Doak b
a
Africa Unit for Transdisciplinary Health Research (AUTHeR), North-West University, Private Bag X6001,
Potchefstroom, 2520, South Africa
b
Section of Infectious Disease, Department of Health Sciences, Vrije University, Amsterdam, Netherlands

article info abstract

Article history: This study focuses on one aspect of a more extensive SANPAD-funded HIV stigma
Received 5 February 2015 reduction research project. The study addresses not only the continuous burden of HIV
Accepted 16 November 2015 stigma, but more specifically on the low rate of participation in healthcare opportunities
and HIV stigma reduction interventions by people living with HIV (PLWH) This study tested
both change-over-time in HIV stigma experiences of PLWH and change-over-time in the
Keywords: HIV stigmatisation behaviour of people living close to them (PLC) in an urban and rural
Change-over-time setting in the North-West in South Africa. These aspects were measured before and after
Community-based intervention the comprehensive community-based HIV stigma reduction intervention. A quantitative
HIV stigma experiences single system research design, with a pre-test and four repetitive post-tests, and purposive
HIV stigmatisation voluntary and snowball sampling were used. Findings did not indicate significant differ-
ences between urban and rural settings, but demonstrated some significance in change-
over-time in the HIV stigma experiences of PLWH as well as the HIV stigmatisation
behaviour of PLC after the intervention. Recommendations include the continuation of this
intervention, following the same guidelines that were implemented during the study.
Copyright © 2015, The Authors. Production and hosting by Elsevier B.V. on behalf of
Johannesburg University. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

accurately describe stigma. They suggested that a term similar

1. Introduction to racism be found that would strongly portray public disap-
proval of discriminatory and stigmatising behaviour.
HIV infection remains a globally stigmatised condition (Greeff HIV stigma manifests as the assertion of a socially con-
et al., 2008) and HIV stigma remains a complex concept structed “undesired differentness” through acts of ostracism,
(Mbonye et al., 2013) associated with blame, shame, disgrace discrimination, social control, marginalisation and social
and social unacceptability (Mandal, 2013). Fifty years ago domination (Herek, Saha, & Burack, 2013). Earlier authors
Goffman (1963) described stigma as a deeply discrediting have described it as a disempowerment of PLWH through
personal phenomenon and forty years later, Deacon and labelling, stereotyping, separation, diminishing and discrimi-
Stephney (2007) argued for an even stronger term to more nation (Link, Yang, Phelan, & Collins, 2004). Moreover,

* Corresponding author. Tel.: þ27 18 299 2092; fax: þ27 18 299 2088.
E-mail addresses: Minrie.Greeff@nwu.ac.za (M. Greeff), c.m.doak@vu.nl (C.M. Doak).
Peer review under responsibility of Johannesburg University.
http://dx.doi.org/10.1016/j.hsag.2015.11.006
1025-9848/Copyright © 2015, The Authors. Production and hosting by Elsevier B.V. on behalf of Johannesburg University. This is an open
access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5 197

according to Kohi et al. (2006), stigma manifests in the viola- is important to understand that stigma may in some instances
tion of human rights which may take the form of refusing to also arise from PLC participating in stigmatising behaviour
care for PLWH within health facilities, verbal and physical towards PLWH.
abuse, food deprivation, denial of employment or income The experiences of PLWH with the above types of HIV
opportunities, denial of leadership positions and a so-called stigma directed at them can be devastating with life-inhibiting
justified breach of confidentiality regarding the HIV status of emotional, psychological, relational or material outcomes.
PLWH. Further aspects of psychological distress experienced Repeated experiences of abuse and discouragement to
by stigmatised PLWH include increased physical distance, participate in treatment programmes and attend care facil-
awkward social interaction, indifference, avoidance, blaming, ities could lead to diminished physical and mental health (Greeff
exaggerated kindness, aggression, exclusion, excessive hy- et al., 2008). Isolation or decreased social participation of PLWH
gienic measures and being told to disclose or not (Stutterheim stems from living while fearing stigma and thus minimising
et al., 2009). their exposure to others (Gilbert & Walker, 2010) which in turn
These complexities led to Holzemer et al. (2007) proposing leads to poor participation in healthcare as well as in personal
a four-dimensional process model to facilitate a better un- and employment relationships (Greeff et al., 2010). A further
derstanding of HIV stigma in Africa. The model served as a outcome of HIV stigma for PLWH relates to self-isolating
basic framework for this study as it acknowledges that HIV behaviour to prevent HIV-transmission to others and to mini-
stigma occurs within an interactive context where environ- mise secondary (associated) stigma directed at those living
ment, the healthcare system as well as people all play a role in close to them (Salter et al., 2010). PLWH often experience
the everyday stigma reality. In this model there are four personal and emotional frailty coupled with internalised self-
interactive processes of HIV stigma, namely triggers, behav- stigmatisation, shame and a compromised self-efficacy
iours, types and outcomes of HIV stigma. This suggests that an (Naidoo et al., 2007). This leads to social avoidance, real or
HIV test can trigger certain behaviour, lead to a specific type of perceived loss of friends, perceived discomfort of those they
stigma and result in a definable outcome of stigma for the are in contact with, symptoms of depression and feelings of
PLWH. anxiety, hopelessness and unattractiveness (Cahill & Valadez,
The types and outcomes of HIV stigma are broadly covered in 2013). All these aspects contribute to a decreased quality of life as
the literature and are summarised in Fig. 1 to enhance un- an outcome of HIV stigma for PLWH and poor disclosure
derstanding. Internal stigma seems to refer to self- practices, poor sexual choices or quality-of-life decisions
stigmatising (Holzemer et al., 2007) by PLWH or their insider driven by a debilitating, internalised fear of losing significant
view (Rensen, Bandyopadhyay, Gopal, & Van Brakel, 2011). relationships and losing their source of income/livelihood.
Received stigma (Holzemer et al., 2007) relates to concepts like Other fears include the fear of losing out on marriage, child-
perceived/anticipated, outsider-view/external or felt stigma bearing, family care, hope, self-worth and reputation
as mentioned by Mak et al. (2007), Weiss et al. (1992) and Herek (Kasapoglu, Saillard, Kaya, & Turan, 2011; Rensen et al., 2011).
et al., (2013) respectively, and as such experienced by PLWH as However, HIV stigma could be turned around and lead to
directed towards them. Lastly, associated stigma (Holzemer positive outcomes if PLWH become involved in actively
et al., 2007) or secondary stigma (Ogden & Nyblade, 2005) re- reducing stigma, accept their HIV status, demonstrate a
lates to stigma that stems from someone's association with pleasant disposition, choose positive interpretations of social
PLWH and often includes PLC (people living close to PLWH). It interactions, refuse the victim role and focus on health,

Fig. 1 e Types of HIV stigma.

198 h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5

resilience, strengths and capacity and thereby flourish in so- Similarly, Uys et al. (2009) used three basic elements in their
ciety (Shih, 2004). intervention, namely accurate information, personal contact
However, the reality is that the South African Medical with the stigmatised and skills to cope with stigma. Ongoing
Research Council (Visser, 2007) some seven years ago found research started promoting multi-pronged approaches such
no evidence of HIV stigma reduction in sub-Saharan Africa. as a combination of sharing information and building skills, or
Mbonye et al. (2013) recently conducted a longitudinal study in education and empowerment combined with personal con-
which it was found that there was a decrease in HIV- tact (Brown et al., 2003; Holm-Hansen, 2009). An example of
associated stigma with commencement of anti-retro viral this approach is the Cross, Heijnders, Dalal, Sermrittirong, and
treatment. But the stigma increased again after 18e30 months Mak (2011) matrix which was used as a framework for situa-
on treatment and seemed to then persist in the long term. tional stigma interventions and strategic stigma reduction
Even HIV status disclosure behaviour decreased as the visible guidelines. This model was based on the intrapersonal,
HIV signs and symptoms of the illness decreased. interpersonal, and organisational and community or govern-
Furthermore, it is interesting to observe differences of HIV mental levels of stigma and cross-referencing components
stigma and stigmatisation between urban PLWH and PLC and like labelling, stereotyping, separation, status loss and
their rural counterparts. The literature points to complexity discrimination.
and even disagreement among researchers in this regard. This
study did not find significant differences between the HIV
stigma experiences of PLWH or stigmatisation behaviour 2. Problem statement and research objective
within urban and rural communities. The complexities of
continuous migration of community members between urban HIV stigma, HIV stigmatisation and the need for the reduction
employment and rural homesteads were, however, noted. (and eradication) of both have become a growing international
Voeten, Egesah, and Habbema (2004) suggest that this is one of concern. Community-based change needs to reflect specific
the reasons why a full understanding of HIV and HIV stigma outcomes like a measurable decrease in the HIV stigma ex-
has been delayed in rural areas. Naidoo et al. (2007) found that periences of PLWH as well as in stigmatisation. Such change
urban PLWH faced more received stigma than their rural will ideally also reflect the empowerment of PLWH and the
counterparts and that urban infrastructure facilitates higher relational enhancement between PLWH and PLC. This study
reporting on HIV stigma. Heckman et al. (1998) found that focused on a specific intervention that responded to three
urban and rural PLWH in the USA did not differ in age, edu- research questions. Firstly, is there a difference in the stigma
cation, employment, income or HIV symptomatology but that reduction experiences between urban and rural PLWH and
rural PLWH rated the severity of barriers to competent and PLC following the comprehensive community-based HIV
compassionate care and care facilities higher than urban stigma reduction intervention? Secondly, will the compre-
PLWH. These barriers included a shortage of competent hensive community-based HIV stigma reduction intervention
health professionals, long distances to medical facilities, reduce the stigma experiences of PLWH? Thirdly, will the
inadequate public transport and painful experiences of stig- comprehensive community-based HIV stigma reduction
matisation. Ankrah (1993) added a lack of privacy, anonymity intervention reduce HIV stigmatisation by PLC?
and confidentiality as common aspects of HIV stigma in rural The research objective for this article was thus to observe
areas. According to Mswela (2009), unsympathetic and harsh change-over-time in the HIV stigma experiences of PLWH and
treatment of sick relatives appears to be more common in the stigmatisation by PLC in both urban and rural settings,
rural areas. following the comprehensive community-based HIV stigma
This study tested an intervention for the reduction of reduction intervention.
stigma and stigmatising. According to a literature review,
studies have been conducted and the outcomes have been
reported of historic and recent community-based HIV stigma 3. Research design and methodology
reduction interventions. During the first 25 years of the AIDS
pandemic, the interventions have reported limited success in A quantitative single system design (De Vos, Strydom, Fouche,
alleviating the effects of HIV stigma on community, national, & Delport, 2005) with a pre-test and four repetitive post-test
and global levels. However by the end of the nineties, Corrigan measures (01  02 03 04 05) was implemented. Both an
(2000) implemented the attribution model which focused on urban and a rural setting were included.
replacing incorrect attributions/beliefs with correct ones. The
operative elements of this model were protest against inac- 3.1. Sample
curate information and myths, insurance of responsible HIV
education, and facilitation of contact between stigmatised The sample for this study comprised two groups representing
and stigmatising people (Corrigan, 2000). their respective communities. These groups were represen-
Generally, interventions were based on intrapersonal, tative of an urban and a rural community each, with each
interpersonal, community, institutional and governmental having PLWH and PLC groups. This study formed part of a
strategic levels (Mahajan et al., 2008). The critical elements of SANPAD-funded study aimed at promoting a deeper under-
such interventions were identified. Brown, Macintyre, and standing of HIV stigma and how people cope with the stigma
Trujillo (2003) mentioned a set of four elements. These were related to HIV. The study thus also focused on the strength-
information to the public, personal contact with PLWH, coping ening of relationships between PLWH and PLC as well as the
skills for dealing with stigma and applied counselling. activation of leadership by both the PLWH and PLC towards
 h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5 199

the reduction of HIV stigma. The therapeutic nature (Thorne, namely the a) sharing of information on HIV stigma and
2008) of the intervention limited the number of participants coping with HIV stigma, b) the equalising of relationships
because meaningful interaction in small groups was required between PLWH and PLC through increased interaction and
and non-probability sampling methods were thus used. contact among them by grouping them together, and c) the
Mediators in already trusting relationships with PLWH empowerment of members of both groups towards leadership
from the identified urban and rural settings were sourced with in HIV stigma reduction through practical knowledge and
the help of existing NGOs and healthcare clinics. These me- experience of project planning regarding HIV stigma reduc-
diators helped the researcher find 18 PLWH by means of tion and implementation in their communities.
purposive sampling. This was followed by snowball sampling The comprehensive community-based HIV stigma reduc-
to identify PLC for the study. tion intervention primarily involved three processes (see
The inclusion criteria for PLWH were as follows: Partici- Fig. 2) and ran over a five-month period in both the urban and
pants had to be over 18 years old, conversant in either Afri- rural settings. Firstly, there was a two-day presentation and
kaans, English or Setswana, HIV-positive for a minimum of six activity-based workshop for PLWH only. It focused on their
months, and willing to give informed consent for participation personal understanding of HIV stigma, identification of their
and recording of discussions. PLWH participants were also personal strengths and teaching responsible disclosure man-
required to openly share in HIV-status disclosure workshops agement to prepare them for the rest of the workshops in the
with other PLWH. Furthermore, they were expected to be intervention. This workshop for PLWH was followed by a se-
willing and able to nominate PLC for participation in various ries of six three-day workshops for each group of PLC: first the
workshops of designated groups involving both PLC and group of spouses/partners, then the group of children over 15,
PLWH. Eventually, 10 PLWH from the Potchefstroom urban then family members, friends, spiritual leaders and, lastly,
district and eight PLWH from the rural Ganyesa district of the neighbours or community members. The PLC workshops
North-West volunteered (see Table 1). All the PLWH were occurred two weeks apart and were led by two facilitators (one
black South Africans. Twelve finished school with Grade 10 or HIV-infected and one non-HIV infected person) for each
higher, seven held a post-school certificate and one a diploma. group. These workshops were attended by all PLWH. The first
Ten persons had no post-school education. day of these workshops focused on an understanding of and
The PLWH participated in snowball sampling for the next coping with HIV stigma and the relationship between PLWH
part of the study and could nominate six PLC, one from each of and PLC. The second day focused on learning and practising
six designated categories e a spouse or partner, a child over 15 the planning of an HIV stigma reduction project with a group
years of age, a family member, a close friend, a spiritual leader similar to their specific designated group, e.g. partners with a
and a community member. Six designated groups were thus community group of partners. Each of the 12 groups was given
formed. The inclusion criteria for nominated PLC corre- a month to implement their projects in their community while
sponded with that of PLWH, except that PLC did not have to be receiving support from the facilitators. In the third one-day
HIV-positive. Not all PLWH were able to nominate a suitable workshop, the original designated group invited community
person for each designated group but a total of sixty (n ¼ 60) members as guests and then presented feedback on their
PLC were identified (see Table 1). There were 23 urban and 37 projects. Small prizes were awarded by the research team.
rural Setswana-speaking participants from the same province.
The PLC sample included 93.3% black, 4.8% coloured and 0.3% 3.3. Data collection
Indian participants. Of the 60 PLC, 83.3% had passed Grade 10
or higher, 55% had no post-school education, 41.7% had ob- The data collection process made use of two structured, valid
tained a post-school certificate and 3.4% a diploma or degree. and reliable measuring instruments, namely the Perceived
AIDS Stigma Instrument PLWA (HASI-P) (Holzemer et al., 2007)
3.2. The intervention and the AIDS-Related Stigma Measure for Community HIV
Stigma (Maughan-Brown, 2004). A pre-test and four post-tests
The intervention was adapted from the validated intervention were conducted on a three-monthly basis over a one-year
manual of Uys et al. (2009) and was based on three tenets, period for PLWH and PLC in urban and rural settings to test

Table 1 e Sample distribution.

Urban Rural Total
Female Male Sub-total Female Male Sub-total
PLWH
9 1 10 5 3 8 18
PLC
Partners 0 2 2 1 0 1 3
Children 3 1 4 5 2 7 11
Family 2 0 2 4 1 5 7
Friend 2 0 2 6 0 6 8
Spiritual Leader 2 4 6 7 3 10 16
Community Member 6 1 7 8 0 8 15
Total 15 8 23 31 6 37 60
200 h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5

Fig. 2 e The comprehensive community-based stigma reduction and wellness enhancement intervention.

the change-over-time in the HIV stigma experience of PLWH Maughan-Brown (2004) reported that initial reliability was
and the stigmatisation by PLC. established by factor analysis with an alpha coefficient of 0.76
The HIV/AIDS Stigma Instrument PLWH (HASI-P) is a 33- for the factor relating to behaviour intention stigma, 0.59 for
item instrument developed by Holzemer et al. (2007) and the factor relating to symbolic stigma, and 0.55 for the factor
measures six dimensions of HIV and AIDS-related stigma relating to instrumental stigma. These indexes tested reliable
(verbal abuse, negative self-perception, healthcare neglect, in this study with Cronbach alpha scores of 0.54, 0.69 and 0.53
social isolation, fear of contagion and workplace stigma) respectively.
experienced by PLWH. It was validated with a sample of 1477 Symbolic stigma (SS) refers to a moralistic, value-based
respondents from five African countries. Holzemer et al. (2007) position or a prejudice-based position for what HIV symbol-
reported a Cronbach reliability coefficient of 0.94 for the total ises for the PLC (Maughan-Brown, 2004). Instrumental stigma
scale. The Cronbach alpha value for subscales of the HIV/AIDS (IS) relates to the personally useful stigmatising thoughts or
Stigma Instrument PLWA (HASI-P) for PLWH was 0.62 for actions that the PLC use for self-protection. For instance, a
healthcare neglect (HCN) and alpha values for the other four personal fear of contagion may lead to a person to refuse to
subscales ranged between 0.77 and 0.85. As the results on all share cups or cutlery, to avoid touch or refrain from intimacy
subscales for the 18 PLWH were not statistically significant, (Maughan-Brown, 2004). General stigma (GS) suggests
effect sizes were compared to analyse potential Cohen d- improvement (change over time) in the general stigmatisation
values and determine practical significance. The community- by PLC following the intervention. The third research question
based HIV stigma intervention was implemented after time e whether the comprehensive community-based HIV stigma
one. Time-one scores (see Table 2) thus reflect pre- reduction intervention would reduce HIV stigmatisation by
intervention scores on five different HIV stigma dimensions PLC e was thus confirmed.
(subscales) for PLWH. In describing effect sizes, time-one Only three of the five indexes of this scale for stigmatisa-
scores are compared with scores on timelines two, three, tion by PLC were used in the analysis because the Behavior
four and five for each subscale. The 33 item four-point Likert Index (BI) refers to policy issues with regard to HIV stigma in
scale used the following scores: 1 ¼ no HIV stigma, 2 ¼ some the community and were removed from the personal experi-
HIV stigma, 3 ¼ definite HIV stigma, and 4 ¼ high prevalence of ences of participants in the particular intervention.
HIV stigma. The comprehensive community-based HIV stigma reduc-
The AIDS-related Stigma Measure for Community HIV tion intervention was preceded by training to prepare the
Stigma is a 39-item instrument that measures AIDS-related research assistants for their task. They were taught how to
stigma for community and was developed by Maughan- conduct the interviews, use the instruments and ensure ac-
Brown (2004). With factor analysis, four indices, namely curate reporting of the process. As the names of the partici-
policy/resource-based stigma (PI), behaviour intention stigma pants became available through the mediators, a research
(BI), symbolic stigma (SS) and instrumental stigma (IS), are assistant made appointments with them and facilitated the
measured on subscales and a combined score is then administering of the relevant instruments. Participants were
computed to yield a fifth index indicating general stigma (GS). transported to and from the North-West University campus
 h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5 201

Table 2 e HIV stigma dimensions experienced by PLWH.

Dimensions Mean scores Effect sizes
Time 1 Time 2 Time 3 Time 4 Time 5 MSE p 1 with 2 1 with 3 1 with 4 1 with 5
VA 12.1 9.8 8.2 8.2 8.3 1.25 0.1 2.06 3.49 3.49 3.4
NSP 7.39 6.8 6.19 6.69 6.32 4.76 0.26 0.27 0.55 0.32 0.49
HCN 7.6 7.1 7.04 7.1 6.97 0.03 0.03 2.89 3.23 2.89 3.64
SI 6.7 5.75 5.51 5.51 5.42 1.78 0.38 0.71 0.89 0.89 0.96
FC 7.7 6.63 6.31 6.42 6.08 0.34 0.03 1.84 2.38 2.2 2.78
Tot 43.61 37.94 35.01 35.47 34.69 12.61 0.02 1.6 2.42 2.29 2.51

NB: VA ¼ Verbal abuse; NSP ¼ Negative self-perception; HCN ¼ Health care neglect; SI ¼ Social Isolation; FC ¼ Fear of contagion; Tot ¼ Total;
MSE ¼ means square error.

for data collection and workshops. They were also offered a

light lunch during the workshops. 5. Results and discussion

Hierarchical linear modelling was used to estimate variability

3.4. Data analysis between urban-rural groups while taking into account the
dependency on data collected from specific persons over time
The data analysis for this quantitative data was computed (McCoach, 2010). The modelling indicated no statistical sig-
with the Statistical Package for the Social Sciences (SPSS) nificance in the interaction effects of the urban and rural
(Version 21; IBM Corp., 2012). Descriptive statistics, namely groups in any of the analyses, and no statistical significance
mean, standard deviation, mean square error, p values and was seen in the main effect of urban versus rural groups. The
effect sizes, were calculated. Hierarchical linear modelling urban/rural results were therefore pooled and the answer to
was used to estimate variability between urban-rural groups the first research question e whether there would be a dif-
while taking into account the dependency on data collected ference between urban and rural PLWH and PLC following the
from specific persons over time (McCoach, 2010). comprehensive community-based HIV stigma reduction
intervention - was thus no as no significant difference was
found. The HIV stigma experiences of PLWH and the stigma-
tisation by PLC in the pooled data demonstrated change over
4. Ethical considerations time.

Ethical approval was obtained from the School of Nursing

Science as well as from the North-West University ethics 5.1. Stigma experiences of PLWH
committee (NWU-OOO 11-09-A1) (30/03/2009e29/03/2014).
Permission was also obtained from the North-West Provincial The comparison of scores reported over time did not give
Department of Health as well as the district health authorities. statistically significant results. But there were indications of
Proceedings were guided by basic ethical principles, like practical significance when effect changes, Cohens ‘d’ scores
respect for human subjects and benevolence and justice as as such were reflected. Although not statistically significant
described by Botma, Greeff, Mulaudzi, and Wright (2010). Thus (p ¼ 0.10), the numeric value of the mean score on verbal abuse
the participants were provided with information regarding (VA) decreased from m ¼ 12.1 at time one to m ¼ 9.8 at time
the criteria for their inclusion on a voluntary basis; they were two and m ¼ 8.3 at time five. This indicated a decreasing
informed that their privacy and anonymity would be guar- tendency by PLWH to experience verbal abuse. Moreover, the
anteed by means of computer coding, and that their identities effect sizes between times one and two, one and three, one
would not be linked to the collected data, analysis or study and four as well as one and five were all larger than 0.5. These
report. Partial confidentiality was ensured through a group d-values or effect sizes also indicated practical significance
contract with the group members. Participants were also with regard to the change-over-time in the HIV stigma expe-
informed that they were free to withdraw at any time. They riences of PLWH. The negative self-perception (NSP) scores of
then signed a consent form. The participants were also pro- the PLWH were not statistically significant (p ¼ 0.26), but the
vided with knowledge and clarification about HIV stigma; they effect sizes of the decrease between times one and three and
identified their personal strengths and PLWH were taught times one and five indicated practical significance with values
responsible disclosure management. Their best interests were of d ¼ 0.55 and d ¼ 0.49 respectively. The healthcare neglect
pursued at all times by enhancing relationships between (HCN) subscale indicated a statistical significant improvement
PLWH and PLC, educating them about HIV stigma reduction in healthcare of PLWH with p ¼ 0.03 while these were
and how to cope coping with stigma, as well as providing them confirmed by the four effect sizes ranging between d ¼ 2.89
with basic skills for managing similar projects in the com- and d ¼ 3.64 and indicated practical significance as well. The
munity. Fair treatment of participants was important and subscale of social isolation (SI) did not show statistical sig-
therefore possible risks to them were identified and managed. nificance but demonstrated practical significance with effect
Counselling was made available to all of them should they sizes between 0.71 and 0.96 (d > 0.5) on the four timeline
need it. comparisons to time one. This was an indication of
202 h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5

improvement in how different social isolation linked to HIV personal fear of contagion could justify someone's refusal to
stigma was experienced by PLWH after the intervention. share cups or cutlery, or avoidance to touch or refrain from
Lastly, fear of contagion (FC) indicated clear statistical signif- intimacy (Maughan-Brown, 2004). The instrumental index (IS)
icance with p ¼ 0.03 in change-over-time in the experiences of at time one pre-intervention was m ¼ 10.7, and increased to
HIV stigma by PLWH. All five dimensions of HIV stigma on the m ¼ 11.6 at time five. The p vale of 0.01 indicated a statistical
HASI-P scale indicated improvement (change-over-time) in significant difference and the effect size of times one and five
the HIV stigma experiences of PLWH after the intervention. (d ¼ 0.50) gave a practical significant result. This could possibly
The summation of the scores of each of the timelines indicate that the tenets of the intervention (HIV knowledge
introduced the opportunity of a total stigma score for each sharing, relationship equalisation and personal empower-
(see Table 2). The score of p ¼ 0.02 thus indicated statistical ment) helped to replace older stigmatising patterns, thoughts
significance of change-over-time after the intervention on the and actions.
total scores. The effect sizes for indication of practical sig- The general stigma (GS) index for PLC also offered a sta-
nificance on total HIV stigma experiences of PLWH indicated tistical significant result, with an increase at time one
even stronger differences between times one and three of (m ¼ 38.5) to m ¼ 42.1 at time five and p < 0.01. Again, there
verbal abuse (VA), negative self-perception (NSP) and health- were also an indication of practical significance as well, as all
care neglect (HCN). It also revealed large effect sizes between four effect sizes were larger than d ¼ 0.50. These results sug-
times one and five of the social isolation (SI) and fear of gested a change-over-time in the general stigmatisation
contagion (FC) scales. In terms of the second research ques- behaviour by PLC following the intervention. It was thus found
tion it was found that stigma experiences of PLWH were that the comprehensive community-based HIV stigma
reduced following the community-based stigma reduction reduction intervention did indeed reduce HIV stigmatisation
intervention. by PLC.

5.2. Stigmatisation by PLC

6. Conclusions
The AIDS-related stigma measure for community HIV scale
was used to measure symbolic stigma (SS), instrumental The urban and rural groups did not demonstrate significant
stigma (IS) and general stigma (GS) of HIV stigmatisation in differences over time. While not proven, this could have been
PLC. Results as seen in Table 3 showed statistically significant due to the homogeneity and the inherent cultural similarities
increases for all three mentioned subscales. in the mainly Setswana-speaking population in both the
As symbolic stigma (SS) referred to an almost immovable urban and their rural counterparts. There was HIV stigma
moralistic, value-based position or a type of prejudice for though. The PLWH responses on each first measure of the
what HIV symbolises in the mind of the PLC (Maughan-Brown, subscales confirmed the presence of HIV stigma which
2004), ideally PLC should have become capable of moving from showed consistently on all five dimensions measured. The
their prejudice as a result of the intervention. The mean score study offered an opportunity to measure and interpret data in
for symbolic stigma (SS) at time one (m ¼ 9.2) reflected the pre- terms of traditional statistical significant results as well as a
intervention measure for PLC and increased to m ¼ 10.6 at method of demonstrating practical significant results where
time five. This gave a statistically significant result (p < 0.01) effect changes justified this. This methodology contributed to
and indicated change-over-time in the stigmatisation by PLC. the inclusion of change-over-time in experiences and stig-
In addition, effect sizes between the compared timelines on matisation observed in interaction but not always measured
symbolic stigma (SS) demonstrated practical significance as in statistics. Results indicated the decline of HIV stigma ex-
the d-scores between times one and three and one and four periences of PLWH on all the dimensions: verbal abuse,
were exceptionally large, measuring 0.70 and 0.65 respec- negative self-perception, healthcare neglect, social isolation
tively. There was thus a change-over-time in symbolic HIV by others and fear of contagion. The change-over-time in the
stigma experiences and the potential shift in PLC prejudice overall stigma experiences of PLWH indicated by the total HIV
was confirmed by practical significant results. stigma score showed a statistical significant result which thus
The scale for instrumental stigma (IS), similar to the above indicates that the community-based stigma reduction inter-
scale, needed to demonstrate meaningful shifts away from vention was successful. The third timeline measure, about
certain personally useful stigmatising thoughts or actions three months after the intervention, revealed large effect size
(instruments) used by PLC for self-protection. For instance, a changes which could be indicative of personal benefits

Table 3 e AIDS related stigma measure for community HIV (Stigmatisation by PLC).
Dimensions Time 1 Time 2 Time 3 Time 4 Time 5 Estimate p Effect sizes of each time with 1
residual
Mean Mean Mean Mean Mean 1 with 2 1 with 3 1 with 4 1 with 5
SS 9.16 10.4 10.96 10.82 10.59 6.56 <0.01 0.19 0.27 0.25 0.22
IS 10.66 10.81 11.53 11.34 11.61 3.59 0.01 0.04 0.24 0.19 0.26
GS 38.46 41.44 42.42 42.85 42.06 33.19 <0.01 0.09 0.12 0.13 0.11

NB: PLC ¼ People living close to PLWH; SS ¼ Symbolic stigma; IS ¼ Instrumental stigma; GS ¼ General stigma.
 h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5 203

derived from the community-based HIV stigma reduction interventions. Mbonye et al. (2013), however, have warned
intervention. The linear correlation among the five HIV stigma that it is possible that HIV stigma and stigmatisation can re-
dimensions indicates the tangency and even complexity of turn once PLWH reach the next level of seeking parity and
HIV stigma where variability between urban-rural groups is equality in labour and reproductive issues.
estimated while working with data collected from the specific
people over time. As a result, a multi-dimensional approach
was taken for the intervention and research.
7. Limitations of the study
Three of the five indices for stigmatisation by PLC indicated
statistical significant change. Similar to the findings of the
The sample size of PLWH was small. It was deliberately kept
experiences of PLWH, large effect size scores featured at the
small in order to accommodate the therapeutic nature of the
third measure about three months after the intervention as
intervention which required small group interaction and the
PLC started to internalise the effects of the intervention. The
building of personal relationships among PLWH and PLC. It
symbolic enhancement is indicative of a change in the
also required pairs of well skilled facilitators of whom one had
moralistic behaviour of PLC and fewer prejudices. PLC also
to be HIV-positive and the other not, so as to model a positive
realised that they probably did not need the symbolic behav-
relationship between PLWH and PLC. The snowball sampling
iour to protect themselves from becoming infected. This could
method in recruiting PLC was also limited as it was dependent
be due to the increased contact between PLWH and PLC. The
on the PLWH. This resulted in some uncertainty regarding the
type of prejudice-based HIV stigma intent measured by sym-
eventual numbers of PLC since availability of potential par-
bolic, instrumental and general stigma scales included the
ticipants for the six specified categories (spouses, children,
opinion that HIV was punishment for sleeping around, refusal
family members, friends, spiritual leaders, neighbours/com-
to admit HIV-positive children to public schools and fear of
munity members) could not be established beforehand.
touching someone with HIV. All these stigmatising thoughts
and actions of the PLC showed reduced incidences. It can thus
be concluded that stigmatisation by PLC was reduced through
the intervention. 8. Recommendations
The change-over-time in the HIV stigma experiences of
PLWH occurred concurrently with the change-over-time in The comprehensive community-based HIV stigma reduction
the HIV stigmatisation behaviour by PLC. These changes in intervention could serve as a useful tool in communities. In
stigma experiences and stigmatisation were sustained over a future interventions, the snowball sampling of PLC could be
one-year period after the intervention, showing that the carried out without the restriction of designated categories.
impact on both PLWH and PLC relatively long lasting. The An aspect of culture sensitivity could be added to cater for
preparation of PLWH to understand HIV stigma, manage their diversity in communities. The basic tenets, methodology,
disclosure responsibly and identify their strengths laid a participation, ethical considerations and programmatic
foundation for the rest of the workshops with PLC. Using a expertise should, however, be retained. Groups should be kept
team of infected and non-infected facilitators as well as small enough to ensure therapeutic benefit and PLWH should
bringing both PLWH and PLC together in the same workshop never be exposed to intervention content without being well
underlined the importance of equal relationships. In this way prepared. A set of guidelines should be compiled for future
equality, acceptance and working together could be modelled implementation of the intervention. It would be helpful to
practically. The fact that both the stigma experiences and have the intervention tested in a variety of cultures and lo-
stigmatisation changed, show that the approach of the inter- cations with a view to building community-based networks
vention and its content were effective at bringing about and structures to eradicate HIV stigma and enhance wellness
changes. The interaction and contact between participants in the community at large. Such actions could perhaps also
normalised the social interaction, offered opportunities to address the issues mentioned by Mbonye et al. (2013) whereby
share experiences and activated support for each other. The HIV stigma and stigmatisation could return after some time of
projects that were undertaken by the PLWH and the PLC initial decline. A booster type intervention at time three may
together as leaders in stigma reduction in their own com- also contribute to long-term sustainability of change-over-
munity could have led to the reduction of fear of contagion. time in the reduction of HIV stigma experiences and stigma-
Both PLWH and PLC regained some control after HIV became a tisation after a successful intervention.
reality in their lives. If all PLC of PLWH could be included in
such programmes as this one, more understanding of the
stigmatisation process could be effected thus leading to a Funding
change in attitude in the entire community.
The intensity and time that the intervention required over The authors received research funds and student bursaries
a four-month period required much energy from the PLWH at from SANPAD.
times, but it did lead to positive results. There is no doubt that
participants benefitted from the therapeutic nature of the
intervention that provided skilled facilitators to a small
number of participants within a structured environment. The Acknowledgements
community-based HIV stigma reduction intervention helped
to bridge the gap in the quest for successful HIV reduction  SANPAD for the student bursary and financial support.
204 h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5

 Dr. Suria Ellis for statistical consult, direction and Holm-Hansen, C. (2009). Stigma reduction: promoting greater
support. understanding of mental health. Wilder Research Snapshot,
 Poncho Molaudzi for research assistance, dedication and 1e4.
Holzemer, W. L., Uys, L., Makoae, L., Stewart, A., Phetlhu, R.,
enthusiasm.
Dlamini, P. S., et al. (2007). A conceptual model of HIV/AIDS
 Fieldworkers for their graceful contribution and human stigma from five African countries. Journal of Advanced Nursing,
connection. 58(6), 541e551. http://dx.doi.org/10.1111/j.1365-
 The PLWH and PLC for their selfless sharing and honest 2648.2007.04244.x.
participation. Kasapoglu, A., Saillard, E. K., Kaya, N., & Turan, F. (2011). AIDS
related stigma in social relations: a qualitative study in
Turkey. The Qualitative Report, 16(6), 1496e1516.
Kohi, T. W., Makoae, L., Chirwa, M., Holzemer, W. L.,
references Phetlhu, D. R., Uys, L., et al. (2006). HIV and AIDS stigma
violates human rights in five African countries. Nursing Ethics,
13(4), 405e414. http://dx.doi.org/10.1191/0969733006ne865oa.
Ankrah, E. M. (1993). The impact of HIV/AIDS on the family and Link, B. G., Yang, L. H., Phelan, J. C., & Collins, P. Y. (2004).
other significant relationships: the African clan revisited. AIDS Measuring mental illness stigma. Schizophrenia Bulletin, 30(3),
Care. Psychological and Socio-Medical Aspects of HIV//Aids, 5(1), 511e541.
5e22. Mahajan, A. P., Sayles, J. N., Patel, V. A., Remien, R. H., Ortiz, D.,
Botma, Y., Greeff, M., Mulaudzi, F. M., & Wright, S. C. D. (2010). Szekeres, G., et al. (2008). Stigma in the HIV/AIDS epidemic: a
Research in health science. Cape Town, South Africa: Pinelands. review of the literature and recommendations for the way
Brown, L., Macintyre, K., & Trujillo, L. (2003). Interventions to forward. AIDS, 22(2), 67e79. http://dx.doi.org/10.1097/
reduce HIV/AIDS stigma: what have we learned? AIDS 01.aids.0000327438.13291.62.
Education and Prevention, 15(1), 49e69. http://dx.doi.org/10.1521/ Mak, W. W. S., Cheung, R. Y. M., Law, R. W., Woo, J., Li, P. C. K., &
aeap.15.1.49.23844. Chung, R. W. Y. (2007). Examining the attribution model of
Cahill, S., & Valadez, R. (2013). Growing older with HIV/AIDS: new self-stigma on social support and psychological well-being
public health challenges. American Journal of Public Health, 3(3), among people with HIVþ/AIDS. Social Science & Medicine, 64(8),
7e15. http://dx.doi.org/10.2105/AJPH.2012.301161. 1549e1559. http://dx.doi.org/10.1016/j.socscimed.2006.12.003.
Corrigan, P. W. (2000). Mental health stigma as social attribution: Mandal, A. (2013). AIDS-stigma. News medical. Retrieved from
implications for research methods and attitude change. http://www.news-medical.net/health/AIDS-Stigma.aspx.
Clinical Psychology: Science and Practice, 7(1), 48e67. http:// Maughan-Brown, B. (2004). Measuring HIV/AIDS stigma: Working
dx.doi.org/10.1093/clipsy.7.1.48. Paper No. 74. Cape Town: Centre for Social Science Research
Cross, H., Heijnders, M., Dalal, A., Sermrittirong, S., & Mak, S. [CSSR], University of Cape Town. Retrieved from http://cssr.
(2011). Interventions for stigma reduction e Part 1: theoretical uct.ac.za/sites/cssr.uct.ac.za/files/pubs/wp74.pdf.
considerations. Disability. CBR & Inclusive Development, 22(3), Mbonye, M., Nakamanya, S., King, R., Seeley, J., Birungi, J., &
62e70. http://dx.doi.org/10.5463/dcid.v22i3.70. Jaffar, S. (2013). Stigma trajectories among people living with
De Vos, A. S., Strydom, H., Fouche, C. B., & Delport, C. S. L. (2005). HIV (PLHIV) embarking on a life time journey with
Research at grass roots: For the social sciences and human service antiretroviral drugs in Jinja, Uganda. BMC Public Health, 13(1).
professions (3rd ed.). Pretoria, South Africa: Van Schaik. http://dx.doi.org/10.1186/1471-2458-13-804.
Deacon, H., & Stephney, I. (2007). HIV/AIDS, stigma and children. McCoach, D. B. (2010). Hierarchical linear modelling. In
Cape Town: HSRC Press. G. R. Hancock, & R. O. Mueller (Eds.), The reviewer's guide to
Gilbert, L., & Walker, L. (2010). My biggest fear was that people quantitative methods in the social sciences (pp. 123e140). New
would reject me once they knew my status: stigma as York, NY: Routledge.
experienced by patients in an HIV/AIDS clinic in Mswela, M. (2009). Cultural practices and HIV in South Africa: a
Johannesburg, SA. Health and Social Care in the Community, 18(2), legal perspective. Potchefstroom Electronic Law Journal, 12(4),
139e146. http://dx.doi.org/10.1111/j.1365-2524.2009.00881.x. 172e360. http://dx.doi.org/10.4314/pelj.v12i4.50053.
Goffman, E. (1963). Stigma: Notes on the management of spoilt Naidoo, J., Uys, L., Greeff, M., Phetihu, R., Holzemer, W.,
identity. Englewood Cliffs, NJ: Prentice Hall. Makoae, L., et al. (2007). Urban and rural differences in HIV/
Greeff, M., Uys, L. R., Holzemer, W. L., Makoae, L. N., AIDS stigma in five African countries. African Journal of AIDS
Dlamini, P. S., Kohi, T. W., et al. (2008). Experiences of HIV/ Research, 6(1), 17e23. http://dx.doi.org/10.2989/
AIDS stigma of persons living with HIV/AIDS and nurses 16085900709490395.
involved in their care from five African countries. Africa Journal Ogden, J., & Nyblade, L. (2005). Common at its core: HIV-related
of Nursing and Midwifery, 10(1), 78e108. stigma across context. Washington DC, WA: International
Greeff, M., Uys, L. R., Wantland, D., Makoae, L., Chirwa, M., Center for Research on Women (ICRW). Retrieved from http://
Dlamini, P., et al. (2010). Perceived HIV stigma and life www.icrw.org/files/publications/Common-at-its-Core-HIV-
satisfaction among persons living with HIV infection in five Related-Stigma-Across-Contexts.pdf.
African countries: a longitudinal study. International Journal of Rensen, C., Bandyopadhyay, S., Gopal, P. K., & Van Brakel, W. H.
Nursing Studies, 47(4), 475e486. http://dx.doi.org/10.1016/ (2011). Measuring leprosy-related stigma e a pilot study to
j.ijnurstu.2009.09.008. validate a toolkit of instruments. Disability and Rehabilitation,
Heckman, T. G., Somlai, A. M., Peters, J., Walker, J., Otto-Salaj, L., 33(9), 711e719. http://dx.doi.org/10.3109/
Galdabini, C. A., et al. (1998). Barriers to care among persons 09638288.2010.506942.
living with HIV/AIDS in urban and rural areas. AIDS Care: Salter, M. L., Go, V. F., Le Minh, N., Gregowski, A., Ha, T. V.,
Psychological and Socio-medical Aspects of HIV/AIDS, 10(3), Rudolph, A., et al. (2010). Influence of perceived secondary
365e375. http://dx.doi.org/10.1111/j.1748-0361.1998.tb00615.x. stigma and family on the response to HIV infection among
Herek, G. M., Saha, S., & Burack, J. (2013). Stigma and injection drug users in Vietnam. AIDS Education & Prevention,
psychological distress in people with HIV/AIDS. Basic and 22(6), 558e570. http://dx.doi.org/10.1521/aeap.2010.22.6.558.
Applied Social Psychology, 35(1), 41e54. http://dx.doi.org/ Shih, M. (2004). Positive stigma: examining resilience and
10.1080/01973533.2012.746606. empowerment in overcoming stigma. The ANNALS of the
 h e a l t h s a g e s o n d h e i d 2 1 ( 2 0 1 6 ) 1 9 6 e2 0 5 205

American Academy of Political and Social Science, 591(1), 175e185. afroaidsinfo.org/MRCWeb/portlets/AutonomyRetrieval/

http://dx.doi.org/10.1177/0002716203260099. autosuggest.jsp?username¼e8e9e6e5f9e0f8&threshold¼
Stutterheim, S. E., Pryor, J. B., Bos, A. E. R., Hoogendijk, R., 20&numresult¼10&defaultlogo¼i_html.gif&display¼774&url¼
Muris, P., & Schaalma, H. P. (2009). HIV-related stigma and 22195&links¼HIV,AID,STIGMA,VISSER&command¼
psychological distress: the harmful effects of specific stigma getoriginal.
manifestations in various social settings. AIDS, 23(17), Voeten, H. A. C. M., Egesah, O. B., & Habbema, J. D. F. (2004).
2353e2357. http://dx.doi.org/10.1097/QAD.0b013e3283320dce. Sexual behaviour is more risky in rural than in urban areas
Thorne, S. (2008). Interpretive description. Left Coast, CA: Walnut among young women in Nyanza Province, Kenya. Sexually
Creek. Transmitted Diseases, 31(8), 481e487.
Uys, L. R., Chirwa, M. L., Kohi, T. W., Greeff, M., Naidoo, J. R., Weiss, M. G., Doongaji, D. R., Siddhartha, S., Wypij, D., Parhare, S.,
Makoae, L. N., et al. (2009). Evaluation of a health setting-based Bhatawdekar, M., et al. (1992). The explanatory model
intervention. AIDS Patient Care and STDs, 23(12), 1059e1066. interview catalogue (EMIC): contribution to cross-cultural
http://dx.doi.org/10.1089/apc.2009.0085. research methods from a study of leprosy and mental health.
Visser, P. (2007). HIV/AIDS and stigma. South African Medical The British Journal of Psychiatry, 160, 819e830. http://dx.doi.org/
Research Council [MRC]. Retrieved from http://www. 10.1192/bjp.160.6.819.