Secondary Logo

Journal Logo


A multilevel intervention to reduce stigma among alcohol consuming men living with HIV receiving antiretroviral therapy: findings from a randomized control trial in India

Singh, Roopal J.a; Sarna, Avinaa; Schensul, Jean J.b; Mahapatra, Bidhubhushana; Ha, Toanc; Schensul, Stephen L.c

Author Information
doi: 10.1097/QAD.0000000000002604



In this paper by Singh et al. , three authors were unintentionally omitted. In addition to the authors listed, the following should also have been included:

1. Sushma S Gaikwad, M.D., T.N. Medical College & B.Y.L. Nair Ch. Hospital, Mumbai, India

2. Kavita Joshi, M.D., Seth G S Medical College, Mumbai, India

3. Rupal Padiyar, M.D., Lokmanya Tilak Municipal Medical College, Mumbai, India

AIDS. 35(3):527, March 1, 2021.


Stigma profoundly affects HIV; a result of self-blame of the persons living with HIV (PLHIV) and the reaction to the disease by family, neighbours, service personnel and the general public who fear contracting the disease and blame the person living with HIV for becoming infected [1–3]. Stigma has serious negative consequences for PLHIV health and wellbeing, [4,5] affecting disease management, immunosuppression and mental health [1,4,6–8]. Higher levels of stigma prevent PLHIV from seeking testing, diagnosis and care, thus risking transmission and comorbidities [4,6]. Stigma and fear of disclosure can prevent PLHIV from attending hospital-based locations to collect their ART medication. Fear of unwanted disclosure may inhibit medication adherence. Stigma is directly associated with psychological distress in PLHIV [9] and is a known contributor to depression [10–13]. As a result, stigma is a major barrier to reaching the global and Indian national goal of 90–90–90 in the test-and-treat continuum [14]. Addressing stigma then is critical to the physical and mental health of PLHIV, both before and during their ART treatment.

Researchers have identified several types of HIV-related stigma. These subdomains are ‘internalized stigma’ or self-blame for having contracted the disease ‘anticipated stigma’ in which PLHIV expect a negative reaction from family, peers and society and avoid situations where their expectations might be realized [6,8,15–17] and ‘enacted stigma’ involving discrimination experienced by PLHIV.

Despite the critical role of stigma, interventions that address the problem and related issues among PLHIV, particularly based on antiretroviral treatment centres are not common [18]. In low-and-middle-income countries, stigma-related interventions are either aimed at individual PLHIV [19], at serodiscordant couples [20] or at the general public [21,22]. There is growing recognition that HIV-related stigma are multifactoral and multilevel [23,24], requiring more complex and diverse interventions [25]. Multilevel models are based on social ecological theory derived from Bronfenbrenner's original work [26] and have been utilized in a variety of public health interventions [24,27]. Multilevel models can be tested simultaneously with matched populations or sequentially in the same population using crossover designs [28]. However, approaches that combine multiple levels of intervention are rarely implemented [1] and there are currently no studies that utilize a multilevel approach to test the effectiveness of interventions to reduce stigma reduction at the individual, group and community levels.

The data for this article are drawn from the project entitled ‘Alcohol and ART adherence: Assessment, Intervention, and Modeling in India’, an Indo-US collaboration that implemented and tested a multilevel approach to intervention to improve ART adherence for male PLHIV. This article examines the effectiveness of three differently sequenced interventions: individual counseling (IC), group intervention (GI) and collective advocacy (CA), against a control in reducing stigma among PLHIV receiving ART in hospital-based centres in Maharashtra, India. The aim of the article is to identify the optimal sequence of the three interventions on stigma overall and on its subdomains.


Study site selection

The study was a five-arm cluster randomized controlled trial, conducted in government ART Centers (ARTCs) in Mumbai and Navi Mumbai in Maharashtra, India between 2014 and 2018. The study team, in close consultation with the National AIDS Control Organization (NACO) and the State AIDS Control Society identified 13 ARTCs offering core services in the greater Mumbai area. Of these ATRCs, the project selected five, based on caseload and geographic spread. With the nodal medical officers in attendance, a blinded randomization process assigned a control site and each sequence to one of the three intervention sites. A final ARTC was assigned as the site for a pilot intervention based on selected components of each of the three interventions. For this article, only the three intervention ARTCs and the control site have been included in the analysis.

The interventions

The three interventions in this project were individual counseling, group intervention and collective advocacy (see Annexure 1, Previous research and intervention by the principal investigators in low-income communities in Mumbai provided an experiential base for the development and implementation of these interventions [29]. Individual counselling and group intervention were based on formative research with the study population, the existing government and NGO guidelines, and infrastructure for individual counselling at ART centres and group counselling at Community Support Centers (CSCs). Both interventions included an alcohol risk-reduction module and a significant component to address stigma that included addressing sources of internalized, external and relational stigma including disclosure. The third intervention was based on the successful HIV advocacy efforts of the Network of Maharashtra Positive People (NMP+), a collaborating project partner, as well as prior research in India by the principal investigatorss on community level advocacy and structural change and literature on the role of activism in reducing HIV-related stigma globally [30–36].

Study design and hypotheses

The study employed an intervention crossover design, in which one intervention was followed by two other different interventions with sequence varying by site. The ATRC 1 sequence was GI–IC–CA; the ARTC 2 sequence was IC–CA–GI; and the ARTC 3 sequence was CA–GI–IC. Each additional intervention was hypothesized to generate an additive effect, with different impacts on stigma hypothesized for each combination as follows:

  • (1) For ARTC 1 (GI–IC–CA), group intervention would reduce both internalized and anticipated stigma in a supportive context, whereas individual counselling would focus on internalized stigma, both of which would provide the basis for successful reduction of enacted stigma through collective action.
  • (2) For ARTC 2 (IC–CA–GI), individual counselling would reduce internal stigma making it easier for participants to engage in group actions to change structural barriers for PLHIV, thus reducing both enacted and anticipated stigma, which would be reinforced by group intervention.
  • (3) For ARTC3, it was hypothesized that collective action group intervention would address all three forms of stigma, thus establishing a strong base for reinforcement of reduction in internalized and anticipated stigma across the group and individual interventions.


Participants were interviewed at preintervention baseline (T1), at the end of cycle 1 (T2), cycle 2 (T3) and cycle 3 (T4); interviews were conducted within a window of 80–120 days after completing the last session of each intervention in each cycle. Data were collected using a structured instrument, administered Orally in Hindi and Marathi by trained research investigators and recorded digitally on tablets using an Android-based app.


Sociodemographic data collected at baseline included age, number of years of formal education, marital status, employment and number of months since on ART. The outcome, HIV-related stigma, was assessed using an adapted Berger's HIV stigma scale [37] with 16 items (see Annexure 2,; Cronbach's alpha = 0.81). The scale was adapted from the widely validated 40-item Berger scale (Cronbach alpha = 0.96) [37,38], and the adapted scale was validated among male and female PLHIV in Kenya [38]. Each item was scored on a four-item Likert scale (strongly disagree = 1, disagree = 2, agree = 3 and strongly agree = 4). The scale's four domains include: personalized or ‘enacted stigma’ (five items, scores 5–20); disclosure concerns or ‘anticipated stigma’ (four items, scores 4–16); negative self-image or ‘internalized stigma’ (three items, scores 3–12) and concerns with public attitudes towards people with HIV or ‘anticipated stigma’ (four items, scores 4–16). The total score ranges from 16 to 64 for all subscales. The scale assesses HIV-related stigma without a recall period, thus limiting bias.

Participant eligibility

Inclusion criteria involved men who reported consumption of at least one drink of alcohol in the previous 30 days, over 18 years of age and on ART for at least 6 months. Identification and recruitment of eligible PLHIV involved an anonymous screener. Research investigators approached PLHIV in the ART Center waiting rooms with the screener instrument that included inclusion/exclusion criteria and related questions. Once determining eligibility, the research investigators described the consent form orally, PLHIV reviewed the written form and those that agreed to participate signed the consent form. The sample size was set at 188 participants per ARTC (N = 752), based on baseline ART medication adherence, taking the intraclass correlation coefficient of 0.01 or 0.02 for four clusters and attrition rates of 10% in each cycle of intervention, or 30% overall.

Treatment as usual

All participants, both at intervention and control sites, received routine care approved by NACO. The protocol called for PLHIV at each visit to see the ART counsellor who reviews ART adherence to ART, positive prevention and healthy living. Newly diagnosed PLHIV were to receive longer counselling sessions for treatment preparedness and ART initiation. NACO guidelines require that PLHIV then be checked by the medical officer with assistance from the nurse, review CD4+ count and ART medicines and collect a 30-day supply to last until the next visit. CD4+ testing was conducted approximately every 6 months per the NACO treatment protocol.

Statistical analyses

All analyses were done using STATA version 16.0 (Stata Corporation, Texas, USA). Descriptive statistics assessed the sociodemographic characteristics of the study population. One-way analysis of variance (ANOVA) compared means for continuous data and associations between nominal variables used in the chi-square test. Linear Mixed Model (LMM) regression with random intercepts and random slopes tested for the effectiveness of the multilevel interventions comparing each intervention sequence with the control ARTC sample, while controlling for the other sites, using the following model: 

where i = T1;j = T2, T3 and T4;sites = control, site 1, site 2 and site 3;Xi = control variables: age, education, marital status and time since on ART;εij = error term.

The HIV-related stigma score, treated as continuous, was analyzed as a longitudinal panel to adjust for nonindependence of observations. In the LMM regression model, variance was controlled for age (continuous), education (categorical); marital status (categorical); and months on ART (continuous). All analyses involved ‘Intention-to-Treat’. Each intervention and intervention sequence were compared with the control at each cycle. Paired T-tests identified significant differences from one timepoint to the next for each intervention sequence.

Within group scores in each of the four subdomains of the stigma scale compared T1T2, T2T3 and T3T4 using paired t-tests. In addition, each intervention and sequence was assessed using LMM regression comparing with the control group for each of the four domains of stigma, namely – personalized stigma, disclosure concerns, negative self-image and concerns with public attitudes towards people with HIV, for T1T2, T1T3 and T1T4.

Ethical approvals

The study was approved by the Ethical Review Boards of the Health Ministry's Screening Committee, Indian Council of Medical Research, the Institutional Review Boards of the University of Connecticut, School of Medicine, the Population Council in New York, the International Center for Research in Women in Washington DC and from the Ethics Committees of three teaching hospitals in which the ART Centers were located. All participants provided written informed consent.


Baseline results

During 2015–2016, 9954 male PLHIV on ART were screened across the five ARTCs. Of these, 1792 (20%) were found to be eligible. Five hundred and seventy-three (32%) declined to participate in the study and a total of 762 participants (188 per site) in the four centres (three experimental and one control) were recruited (Fig. 1); Overall, only 22.8% (156/752) were lost to follow-up and 586 participants (77.2%) in the control and three intervention sites completed all four rounds of the study evaluation.

Fig. 1
Fig. 1:
Participant flow diagram and implementation plan.

Study participants were on average 43 years of age, mostly educated up to secondary level and most were currently married (Table 1). The average length of time on ART was 54 months (SD: 32.6 months). The sites differed from each other at baseline in sociodemographics and characteristics (Table 1), which was controlled for in multivariate analysis.

Table 1
Table 1:
Background characteristics of male persons living with HIV on antiretroviral therapy at baseline, Mumbai, 2015–2016.

Berger 16-item stigma score analyses

As a result of the first intervention cycle (T1T2), regression analysis shows a significant decline in stigma across all three intervention sites (Table 2), whereas stigma scores increased at the control site (+0.7). The largest reduction was observed among participants who received the collective advocacy intervention, followed by group intervention and individual counselling (collective advocacy: −9.0; group intervention: −4.4; individual counselling: −3.6; C: +0.7; ANOVA F = 49.1; p < 0.001). LMM regression analyses showed significant stigma reductions in all three interventions compared with the control group change in scores: group intervention: βcoeff = −5.20, p < 0.001; individual counselling: βcoeff = –4.42, p < 0.001 and collective advocacy: βcoeff = −9.74, p < 0.001.

Table 2
Table 2:
Change in stigma scores over four assessment points across study sites.

After the second cycle (T2T3) of interventions (ARTC 1: GI+CA; ARTC 2: IC+GI; ARTC 3: CA+IC), none of the intervention sites showed any significant change in mean stigma scores from the previous assessment (Table 2), whereas mean stigma scores increased significantly at the control site (+1.3; p = 0.002).

At the end of the third and final cycle (at T4) (ARTC 1: GI+CA+IC; ARTC 2: IC+GI+CA; ARTC 3: CA+IC+GI) mean stigma scores again declined significantly as compared to scores at T3. For the full sequence of interventions (from baseline T1 to T4), stigma scores reduced across two of the three sites. The sequence CA+IC+GI was most effective in reducing stigma (βcoeff = −10.20, p < 0.001) followed by the GI+CA+IC sequence (βcoeff = −8.19, p < 0.001). The IC+GI+CA sequence was not significant in reducing stigma (βcoeff = −0.52, p = 0.537) compared with the control.

Subdomains of stigma

Table 3 outlines the evaluation results of each of the interventions and sequences in tackling the four subdomains of stigma. Disclosure concerns were significantly lowered by the collective advocacy intervention in the first cycle (βcoeff = −0.74, p = 0.011). At the end of the second cycle, both at ARTC 2 and ARTC 3, disclosure concerns were significantly reduced compared with the control group. At endline, the CA+IC+GI sequence was the only that had a significantly lower score in this subdomain compared with the control group (βcoeff = −2.23, p < 0.001). Negative self-image as a concern was successfully tackled by each intervention (group intervention, individual counselling and collective advocacy) alone, and each sequence of interventions. All three interventions group intervention, individual counselling and collective advocacy addressed the subdomain of public attitudes in the first cycle and this effect was maintained through the second cycle of interventions. At the end of the last cycle of interventions, both ARTC 1 and ARTC 3 had a significant reduction in public attitudes’ subdomain score than the control group, but not at ARTC 2 (ARTC 1: βcoeff = −1.03, p < 0.001; ARTC 2: βcoeff = −0.06, p = 0.798; ARTC 3: βcoeff = −2.32, p < 0.001). The concerns around personalized stigma were successfully addressed by each of the three interventions in the first cycle, and this effect sustained through the second cycle. However, at the end of the last cycle, the sequences at ARTC 1 and ARTC 3 maintained a lower score than the baseline in this domain, compared with control (ARTC 1: βcoeff = −5.74, p < 0.001; ARTC 2: βcoeff = −0.68, p = 0.08; ARTC 3: βcoeff = −3.1, p < 0.001).

Table 3
Table 3:
Change in mean scores of four subdomains of stigma over four assessment points across study sites.


To our knowledge, this article is the first to examine stigma reduction using a multilevel intervention approach among alcohol-consuming men treated in ART centres in India. The study population consisted of men who had been on ART for an average of four or more years and were for the most part adhering to their ART medication. However, despite the effectiveness of their treatment, their experience with stigma persisted, as shown in the baseline findings. In cycle 1, all three interventions were effective in reducing overall stigma as well addressed subdomains of personalized stigma, negative self-image and public attitude concerns. Disclosure concerns were successfully addressed by interventions delivered in a group setting (group intervention and collective advocacy). Thus, program and policy makers can feel confident that any of the interventions is likely to have a positive impact on stigma and that selection can focus on the best fit for their ARTCs.

At the end of the second cycle, there was no additive change in the scores in IC+GI, CA+IC or GI+CA; for overall stigma scores or the subdomain scores, except for GI+CA, where the negative self-image concerns increased slightly. Although this may be disappointing from an implementation point of view, it was encouraging to find that overall, improvement shown in the first cycle was for the most part sustained across all three intervention arms, while scores in the control group indicated increased stigma.

At the end of the third and last cycle, there were differences in the impact of the three intervention sequences on stigma. ARTC 3 (CA+IC+GI sequence) showed the maximum decline in stigma scores from baseline when compared with the control group. This sequence also produced significant effects on all four domains of stigma. The first site (sequence GI+CA+IC) also gained substantially in terms of stigma reduction overall and in three of the four subdomains. The third sequence (IC+GI+CA) showed a significant decline in stigma and its subdomains after the first cycle, only, sustained the effects in the second cycle but showed a significant reversal in the third cycle, to finish with no significant difference compared with control.

In terms of the hypotheses concerning effectiveness, hypothesis 1 and 3 (sequences GI+CA+IC and CA+IC+GI) were strongly supported. Hypothesis 2 (IC+ GI+CA) showed the same effect on stigma in cycles 1 and 2 as the other two sequences. However, the impact of the first two interventions did not carry over into collective advocacy, where stigma increased. This unexpected result cannot be explained by the conduct of the intervention team, which had successfully administered two prior rounds of collective advocacy intervention. It could be explained by the shift from the protected environment of individual counselling and group intervention to a less protected and more confrontative approach required in collective advocacy, which might have brought about an increase in worry about stigma. The results suggest that collective advocacy is more effective when administered earlier in the sequence.

The results also show that the second cycle serves to sustain the first in all three sequences but does not improve overall or subdomain stigma scores. This may be a function of intervention fatigue although skipping the second intervention may result in a deterioration of the positive impact of overall and subdomain stigma scores. An additive effect in a cycle may not be necessary as long as the effect of one cycle is sustained by the next and the overall slope of change is significantly positive. It is notable that there are further improvements in two of the sites post third intervention administration (cycle III). Interventions are assessed by results, and the time and socioeconomic costs of implementation for both implementers and participants. Interventionists and policy makers must determine the costs and benefits of administrating one versus all three interventions, and whether the booster effect of cycles II and III can be obtained in other ways.

The strengths of the study include, firstly, that this is a unique multilevel multifactoral intervention that is tested among alcohol consuming male PLHIV. Further, this study is so designed that it is enabled to test the cumulative effectiveness of the interventions using the crossover design. There are some limitations. Firstly, the study included only men who had consumed alcohol in the last month. Thus results may not be generalizable to a nondrinking population. Secondly, respondents had been on ART on average for 4–5 years. It is possible that we would have seen different outcomes with participants who had recently initiated ART. Thirdly, there were some differences among the ARTCs in sociodemographics and other characteristics at baseline; however, we controlled for these differences by including them as covariates in the analyses. Lastly, we only tested three sequences (of the possible six permutations) of the interventions and it is possible other sequences could have provided different results.

The results presented in this article provide empirical evidence that leads to several conclusions. First, the most productive of the sequences began with groups (group intervention or collective advocacy) that engage participants collectively in addressing all forms of stigma. These group interventions are more reliable and effective in addressing stigma as a major source of stigma involves interaction and negative relationships that can be countered by supportive groups. Although individual counselling is a familiar mode in a clinical context, our experience in delivery of this intervention was that it mostly far exceeded the time allocated to counseling in the typical functioning of the ARTC, whereas the group interventions could be accomplished more efficiently. From the point of view of stigma reduction, we would argue that group interventions, coupling group intervention and collective advocacy with a lower cost booster in between might have a greater impact on stigma than the sequence that begins with individual counselling or has individual counselling in the second cycle.

Stigma is a major issue among PLHIV, contributing to a wide range of negative outcomes. Its importance calls for specially focused interventions that can address all aspects of stigma. This article has demonstrated a significant impact on stigma and its multiple manifestations using both single and a sequence of interventions. It provides the evidence that stigma can be addressed for the well being of PLHIV in ARTCs in Mumbai, India and beyond.


We are grateful to our partners at BYL Nair Hospital, KEM Hospital, LTMG (Sion) Hospital, Vashi Municipal General Hospital and Ulhasnagar Central Hospital. We are thankful to our investigators: Deepak Sonawane, Triloki Nath, Sachin Sakpal, Purna Kumal, Santosh Surve, Dnyaneshwar Khandekar, Santosh Salunke, Gopal Gujar, Jagdeep Bansod, Vijay Pawar; and our facilitators: Ramchandra Salunke, Ashok Dhokle, Vaijnath Mane and Ashok Gheware. This study could not have been implemented without the support of Maharashtra States AIDS Control Society. Lastly, we would like to thank study participants who willingly gave us their time and attention.

Source of funding: NIH/NIAAA grant: U01-AA021990R.

J.S. and A.S. conceptualized the article. R.J.S. conducted the data analysis and wrote the article. B.M. and A.S. guided the data analysis and reviewed the manuscript. A.S., T.H., S.S. and J.S. reviewed and edited the manuscript.

Presentation: These data have not been presented at any national or international meeting or conferences.

Availability of data and material: As per the Data Sharing agreement, project data will be available on February 1, 2023 from Professor Stephen L. Schensul PD/principal investigators ([email protected]).

Clinical Trial Registration Number: Retrospectively registered. NCT03746457. Registry: Clinical Trial.Gov.

Conflicts of interest

There are no conflicts of interest.


1. Stangl AL, Lloyd JK, Brady LM, Holland CE, Baral S. A systematic review of interventions to reduce HIV-related stigma and discrimination from 2002 to 2013: how far have we come?. J Int AIDS Soc 2013; 16: (3 suppl 2): 18734.
2. Brown L, Macintyre K, Trujillo L. Interventions to reduce HIV/AIDS stigma: what have we learned?. AIDS Educ Prev 2003; 15:49–69.
3. Jacoby A. Felt versus enacted stigma: a concept revisited: evidence from a study of people with epilepsy in remission. Social Sci Med 1994; 38:269–274.
4. Peitzmeier SM, Grosso A, Bowes A, Ceesay N, Baral SD. Associations of stigma with negative health outcomes for people living with HIV in the Gambia: implications for key populations. J Acquir Immune Defic Syndr 2015; 68:S146–S153.
5. Earnshaw VA, Smith LR, Chaudoir SR, Amico KR, Copenhaver MM. HIV stigma mechanisms and well being among PLWH: a test of the HIV stigma framework. AIDS Behav 2013; 17:1785–1795.
6. Kay ES, Rice WS, Crockett KB, Atkins GC, Batey DS, Turan B. Experienced HIV-related stigma in healthcare and community settings: mediated associations with psychosocial and health outcomes. J Acquir Immune Defic Syndr 2018; 77:257–263.
7. Pulerwitz J, Michaelis A, Lippman S, Chinaglia M, Diaz J. HIV-related stigma, service utilization, and status disclosure among truck drivers crossing the Southern borders in Brazil. AIDS Care 2008; 20:764–770.
8. Earnshaw VA, Chaudoir SR. From conceptualizing to measuring HIV stigma: a review of HIV stigma mechanism measures. AIDS Behav 2009; 13:1160.
9. Herek GM, Saha S, Burack J. Stigma and psychological distress in people with HIV/AIDS. Basic Appl Soc Psychol 2013; 35:41–54.
10. Simbayi LC, Kalichman S, Strebel A, Cloete A, Henda N, Mqeketo A. Internalized stigma, discrimination, and depression among men and women living with HIV/AIDS in Cape Town, South Africa. Soc Sci Med 2007; 64:1823–1831.
11. Epstein RS. The role of shame in understanding and treating depression. Psychiatry 2001; 64:240–241.
12. Li L, Lee S-J, Thammawijaya P, Jiraphongsa C, Rotheram-Borus MJ. Stigma, social support, and depression among people living with HIV in Thailand. AIDS Care 2009; 21:1007–1013.
13. Breet E, Kagee A, Seedat S. HIV-related stigma and symptoms of posttraumatic stress disorder and depression in HIV-infected individuals: does social support play a mediating or moderating role?. AIDS Care 2014; 26:947–951.
14. Ekstrand ML, Bharat S, Srinivasan K. HIV stigma is a barrier to achieving 90-90-90 in India. Lancet HIV 2018; 5:e543–e545.
15. Scambler G, Hopkins A. Being epileptic: coming to terms with stigma. Sociol Health Illness 1986; 8:26–43.
16. Link BG. Understanding labeling effects in the area of mental disorders: an assessment of the effects of expectations of rejection. Am Sociol Rev 1987; 52:96–112.
17. Markowitz FE. The effects of stigma on the psychological well-being and life satisfaction of persons with mental illness. J Health Soc Behav 1998; 39:335–347.
18. Sengupta S, Banks B, Jonas D, Miles MS, Smith GC. HIV interventions to reduce HIV/AIDS stigma: a systematic review. AIDS Behav 2011; 15:1075–1087.
19. Neema S, Atuyambe LM, Otolok-Tanga B, Twijukye C, Kambugu A, Thayer L, McAdam K. Using a clinic based creativity initiative to reduce HIV related stigma at the Infectious Diseases Institute, Mulago National Referral Hospital, Uganda. Afr Health Sci 2012; 12:231–239.
20. Nuwaha F, Kasasa S, Wana G, Muganzi E, Tumwesigye E. Effect of home-based HIV counselling and testing on stigma and risky sexual behaviours: serial cross-sectional studies in Uganda. J Int AIDS Soc 2012; 15:17423.
21. Young SD, Konda K, Caceres C, Galea J, Sung-Jae L, Salazar X, et al. Effect of a community popular opinion leader HIV/STI intervention on stigma in urban, coastal Peru. AIDS Behav 2011; 15:930–937.
22. Saad A, Lekhraj R, Sabitu K, AbdulRahman H, Awaisu A, AbuSamah B, et al. An HIV-STI risk reduction program among undergraduate students at a northern Nigerian university: a randomized controlled field trial. J Public Health 2012; 20:549–559.
23. Lewis MA, Fitzgerald TM, Zulkiewicz B, Peinado S, Williams PA. Identifying synergies in multilevel interventions: the convergence strategy. Health Educ Behav 2016; 44:236–244.
24. McCormack L, Thomas V, Lewis MA, Rudd R. Improving low health literacy and patient engagement: a social ecological approach. Patient Educ Counsel 2017; 100:8–13.
25. Chakraborty B, Collins LM, Strecher VJ, Murphy SA. Developing multicomponent interventions using fractional factorial designs. Stat Med 2009; 28:2687–2708.
26. Bronfenbrenner U. The ecology of human development: experiments by nature and design. Cambridge, Massachusetts: Harvard University Press; 1979.
27. Doran K, Resnick B, Kim N, Lynn D, McCormick T. Applying the social ecological model and theory of self-efficacy in the worksite Heart Health Improvement Project-PLUS. Res Theory Nurs Pract 2017; 31:8–27.
28. Schensul JJ, Trickett E. Introduction to multi-level community based culturally situated interventions. Am J Community Psychol 2009; 43:232–240.
29. Schensul SL, Saggurti N, Singh R, Verma RK, Nastasi BK, Mazumder PG. Multilevel perspectives on community intervention: an example from an Indo-US HIV prevention project in Mumbai, India. Am J Community Psychol 2009; 43:277–291.
30. Parker W, Schubert J, Owusu-Ansah S, Asante-Afari K, Edem Assisi S, Adiku ED, Pennas T, Evans Williams S. Addressing community-level stigma toward key populations: communication insights from action research in Ghana. Afr J AIDS Res 2020. 1–8.
31. Sultan A, Mažeikienė N. Living with HIV in post-Soviet states: Rejecting individual stigma through social activism. Int Soc Work 2019. 0020872819858746.
32. Sprague L, Afifi R, Ayala G, El-nasoor ML. Participatory praxis as an imperative for health-related stigma research. BMC Med 2019; 17:32.
33. Dhungana N, Blankenship KM, Biradavolu MR, Tankasala N, George A. Kerrigan D, Barrington C. No-one-size-fits-all: addressing the social and structural dimensions of sex worker vulnerability to HIV through community mobilization in Avahan. Structural dynamics of HIV: risk, resilience and response. Cham: Springer International Publishing; 2018. 67–96.
34. Earnshaw VA, Rosenthal L, Lang SM. Stigma, activism, and well being among people living with HIV. AIDS care 2016; 28:717–721.
35. Weeks MR, Convey M, Dickson-Gomez J, Li J, Radda K, Martinez M, et al. Changing drug users’ risk environments: peer health advocates as multilevel community change agents. Am J Community Psychol 2009; 43:330–344.
36. Spieldenner A, Sprague L, Hampton A, Smith-Davis = M, Peavy D, Bagchi AD. Kellett PM, et al. Consumer to community based researcher: lessons from the PLHIV Stigma Index. Lexington Books, Narrating patienthood: engaging diverse voices on health, communiciation and the patient experience. NY:2019.
37. Berger BE, Ferrans CE, Lashley FR. Measuring stigma in people with HIV: psychometric assessment of the HIV stigma scale. Res Nurs Health 2001; 24:518–529.
38. Kaai S, Bullock S, Sarna A, Chersich M, Luchters S, Geibel S, et al. Perceived stigma among patients receiving antiretroviral treatment: a prospective randomised trial comparing an m-DOT strategy with standardof-care in Kenya. SAHARA-J 2010; 7:62–70.

HIV; multilevel intervention; stigma

Supplemental Digital Content

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc.