HIV Activist Identity, Commitment, and Orientation Scale (HAICOS): Psychometric Evaluation to Assess Clinician’s Propensity Towards HIV Activism in Malaysia

Abstract

HIV activism has a long history of advancing HIV treatment and is critical in dismantling HIV-related stigma. This study evaluated the psychometric quality of the HIV Activist Identity, Commitment, and Orientation Scale (HAICOS) to assess clinicians’ propensity towards HIV activism in Malaysia. From November 2022 to March 2023, 74 general practitioners and primary care physicians in Malaysia participated in the study. The exploratory factor analysis (EFA) extracted an internally consistent three-factor solution with 13 items: (1) HIV activist identity and commitment, orientation towards (2) day-to-day, and (3) structural activism. The Cronbach’s alpha value was 0.91, and intra-class correlation coefficient for test-retest reliability was 0.86. Stigma-related (prejudice and discrimination intent) and clinical practice (comfort in performing clinical tasks with key populations and knowledge about HIV pre-exposure prophylaxis) measures supported the construct validity of the scale. The study provided concise, structurally valid, and reliable measures to evaluate HIV activism among clinicians.

Introduction

HIV activism has been a cornerstone of the HIV response in areas of policy reform, HIV treatment, community development, research, and program implementation. Activism is also critical in dismantling the HIV-related stigma that perpetuates the devaluation and disenfranchisement of marginalized communities. HIV activism refers to actions and initiatives that propel societal change in the HIV response [1]. These initiatives aim to promote social, health, economic, and political justice to communities most vulnerable to HIV/AIDS (e.g., improved treatment, access to healthcare, and legal rights) [1–4]. Initially labeled as illogical, immature, deviant, or delusory, activism has become an accepted and integral component of individual liberty in an open and democratic society [5].

Members of key populations are disproportionately affected by HIV, many of whom have stigmatized identities and are unaware of their HIV status [6]. A large proportion of key populations with HIV are also not on antiretroviral medication or reached by HIV prevention programs, with at best 62% of men who have sex with men (MSM) and 22% of female sex workers (FSW) on antiretroviral therapy [7, 8]. The structural criminalization of consensual same-sex relationships, transgender identity expression, sex work, as well as drug use and possession creates a hostile environment towards populations with and affected by HIV [9–14]. Disparities in the HIV response among key populations (e.g., low HIV prevention program coverage and rates of HIV testing) remain persistent [7], resulting in increased incidence of late HIV disease presentation into care and poor treatment adherence [15, 16]. Previous work has also found significant prejudice, stereotyping, fear, and associative stigma predicting more substantial endorsement of discrimination intent towards key populations and people with HIV (PWH) among Malaysian doctors and medical students [17–21].

Activism exists on a broad spectrum of risk or danger levels (i.e., from an anticipated low risk to high risk, including radicalism) [22]. Low-risk activism are generally safe engagements in social movements, with a low to negligible likelihood of arrest, danger, or other harms (e.g., collecting information, sharing opinions, contesting stigmatizing statements). In contrast, high-risk activism refers to social actions that possibly involve physical harm, the criminal justice system, or property destruction (e.g., protests, civil disobedience) [23]. As opposed to activism, radicalism or terrorism are extreme, violent, and life-threatening actions (e.g., bioterrorism, hostage-taking) [24].

Many studies on HIV activism focus on people with lived experiences of oppression. HIV activism has benefits and costs for people with lived experiences of oppression. Preliminary findings suggest that PWH who identified as activists did better than those who did not on several measures of well-being, exhibited more active coping with prejudice, had more robust social network integration, and reported more life satisfaction [2]. Evidence also suggests that individuals with lived experience who identified as activists endorsed lower internalized stigma and were more stigma-resistant [25]. Interpersonal connections mediate these well-being indicators through shared grievances, emotions, and centrality of identities [26–28]. HIV activism also reportedly reduced tuberculosis-related stigma [29]. Activism has shaped historical contexts that drive the dynamics of networks, alliances, and countermovements. Yet, PWH who identified as activists reported relatively higher depressive symptoms than their non-activist peers, epitomizing the complexity of activism and well-being [2]. While progress can be attributable to activism, weighing the benefits and costs of HIV activism is challenging.

The lived experiences of oppression exist in individuals of all professions. However, there is a dearth of studies on allyship, especially among healthcare professionals. While activism has stewarded the HIV response in addressing social injustice and health inequity in Malaysia [30–33], negotiation for change is contingent upon whether the individual is sufficiently amenable to be empowered. Glenn Laverack [1] described this tension by recognizing that only individuals who choose to exercise their right to be empowered can drive the seizure of power.

Exploring the role of HIV activism in dismantling stigma contributes to the shift in the body of stigma research from deficit-focused (e.g., the emphasis on the detriments of stigma as leverage to advocate for social justice and equity) to empowerment-based (e.g., centering resilience, resistance, and empowerment to increase the agency of individuals in dismantling interlocking systems of oppression) [34–36]. Stigma draws upon power and dominance [37] in that it exists in three forms [38, 39]: power-from-within, power-over, and power-with. Harnessing community collectivism to engineer social change and resistance can inform stigma reduction interventions through a strength-based and social justice lens.

Measuring HIV Activism in Healthcare

HIV activist identity and orientation towards social change are multifactorial. The existing scales that measure one’s propensity towards activism are devised either from an issue-specific perspective (e.g., nuclear arms race activism [40], environmental action [41]) or applicable generally [5, 22]). These scales elucidate how activist identity and commitment are associated with measures of well-being and stratified into risk levels.

Furthermore, these scales measured intent to engage. Measuring intent to engage is, at best, a proxy for measuring tangible actions, in line with the theories of planned behavior [42–44] and reasoned action [44, 45]. Even so, these underpinning theories assume the decision-making process to be unifactorial (i.e., the intention is understood as a single factor that mediates the relationship between cognitive reasoning and behavior) [46] and do not account for time [47] or variability in behavior [48]. Lastly, Klar and Kasser’s [5] Activist Identity and Commitment Scale established intention to engage, but a measure of tangibility, i.e., measurable engagement in activism, may improve the scale quality.

The strengthening of HIV activism and allyship among clinicians is pivotal. An effective HIV response includes clinicians as community allies. Clinicians are likely to be more influential and considered figures of authority within healthcare institutions and beyond the scope of their professional services [49, 50]. They also play a central role in patient-care services [51]. HIV activism and allyship among clinicians can deepen their commitment to servitude and altruism. As part of their core competencies, clinicians are ethically responsible for the health needs of the individuals they serve, especially advocating for structural and sociopolitical change [52].

Yet, many medical schools and residency curricula do not include training in social justice and activism [53, 54]. Consequently, health professionals may regard intervening in sociopolitical determinants of health as actions beyond their domain of influence and expertise [55]. The hierarchical nature of health systems [56] and the anticipated risk towards employment could also deter clinicians from practicing activism [57]. Regarding oneself as apolitical may not be adequate in dismantling interlocking structures of power and oppression. Implementing change in the healthcare architecture is an intrinsically political process driven by dominant power structures [58, 59].

Current Study

Corning and Myers’ [22] and Klar and Kasser’s [5] scales evaluated the psychometric characteristics of scales that measured activist orientation. In assessing HIV activism among clinicians, measures should consider the epidemiology of HIV, manifestations of HIV-related stigma, appropriateness of use with clinicians, and clinical implications. We address this gap by evaluating the psychometric quality of the HIV Activist Identity, Commitment, and Orientation Scale (HAICOS) among clinicians to gauge their propensity towards HIV activism and social action.

Method

Design

Conducted from November 2022 to March 2023, the study recruited a sample of 78 primary care physicians and general practitioners as part of a randomized controlled trial to test the piloting of a stigma reduction intervention through a tele-training platform. The clinicians were from all eleven states and two federal territories in Malaysia who served in the government and private sectors. We collected the data twice at an 18-week interval through Qualtrics surveys. Baseline data were used to appraise data distribution, validity, and reliability. We subsequently performed an 18-week test-retest stability assessment.

Participants and Procedures

Data preparation was done in R. We recruited 78 participants according to the sample size calculation using the power package for single-level generalized linear models based on the recommendation from Cohen’s A Power Primer [60] for the randomized controlled trial. When conducting the EFA with a sample size of 78, we carefully considered several factors to ensure the validity and reliability of our findings. We acknowledge that the sample size falls on the lower end for factor analysis. However, it is essential to note that EFA can yield reliable results for samples well below 50, even in small distortions [61]. Such conditions require the data to have favorable properties such as ample variability. Given that our data stems from a RCT with a discernible underlying factor structure, we anticipate these qualities to be present.

With convenience sampling, we recruited general practitioners and family medicine specialists in this study based on the following eligibility criteria: (1) practicing general practitioners or family medicine specialists in Malaysia, (2) aged 18 and above, and (3) having internet access and literacy in English. We excluded clinicians who could not understand English. We defined general practitioners as clinicians who completed their internship training (i.e., housemanship) at the time of study and practicing at various hospitals or clinics throughout Malaysia, including but not limited to private practice and community-led health centers. Family medicine specialists are primary care physicians who have completed their training in family medicine and are serving at various government or private primary care centers in Malaysia. Participants provided their consent for their involvement in the study. Study procedures received IRB approval from the University of Delaware (1588354), the National Medical Research Register (21–788-59486), and the University of Malaya Medical Centre (202047–846). The participants were compensated for their time with RM75 (approximately USD 15).

Recruitment focused on the four main venues: the Family Medicine Specialists’ Association (FMSA), Malaysian Medical Association (MMA), Malaysian Primary Care Network, and Doctors Only Bulletin Board System (DOBBS). FMSA represents primary care physicians from across Malaysia. MMA is a professional body representing Malaysia’s largest number of medical practitioners. Malaysian Primary Care Network and DOBBS are online communities for Malaysian doctors where clinicians receive healthcare-related updates or activities.

Measures

At baseline, participants responded to sociodemographic questions and the questions from the HAICOS constructs. Participants responded to questions about their sociodemographic and clinical characteristics, including their age, ethnicity, gender identity, faith affiliation, country of medical training, clinical role, and service affiliation(s). In addition to the HAICOS subscales, participants responded to validated quantitative measures on stigma-related constructs (prejudice and discrimination intent), knowledge about HIV pre-exposure prophylaxis (PrEP), and comfort performing clinical tasks with members of key populations. After 18 weeks, participants responded to a similar set of subscales and variables.

HIV Activist Identity, Commitment, and Orientation Scale (HAICOS)

We asked the participants about their identity, commitment, and orientation towards HIV activism. We adapted and tailored items from the 7-item Activist Identity and Commitment Scale (AICS) by Klar and Kasser [5] and the 38-item Activism Orientation Scale (AOS) by Corning and Myers [22] to form a 16-item HIV Activist Identity, Commitment, and Orientation Scale (HAICOS). Klar and Kasser’s AICS assessed an individual’s propensity towards activism via the individual’s reported activist identity (Cronbach’s α = 0.96) and commitment towards activism (Cronbach’s α = 0.95). The response continuum for each item ranged from 1 (strongly disagree) to 5 (strongly agree).

Corning and Myers’ AOS, a 38-item validated measure, included items that represented constructs on social action, ideological stances, and social movements. Items were rated on an ordinal scale that ranged from 0 (extremely unlikely) to 3 (extremely likely). In its development, the study validated the AOS via construct, discriminant, and group differences criterion-related validity alongside exploratory factor analysis, the measure of internal consistency, and inter-item correlations. The AOS scale (Cronbach’s α = 0.96) loaded into two factors, namely (1) conventional activism (Cronbach’s α = 0.96) and (2) high-risk activism (Cronbach’s α = 0.91).

We applied the HIV context and geopolitical climate in Malaysia to the unrevised version of the 16-item HIV Activist Identity, Commitment, and Orientation Scale (HAICOS) (refer to Appendix A). Apart from tailoring 13 items from the two activism scales (AICS and AOS) to the HIV context, we also added three new items (“I share views or posts online in support of the legal rights of the communities most vulnerable to HIV/AIDS”, “I participate in an event (e.g., talk, webinar, march) for and/or to the communities most vulnerable to HIV/AIDS” and “I reach out to a community leader to address and repeal stigmatizing policies specific to the communities most vulnerable to HIV/AIDS”).

Regarding HIV activist identity and commitment, we asked participants the extent to which they agreed with the six items representing one’s social collective identity and motivations. Participants responded based on a Likert-type rating ranging from 1 (strongly disagree) to 5 (strongly agree). The HIV activism orientation constructs examined the participant’s propensity to engage in collective, problem-solving behaviors that ranged from day-to-day engagement to structural activism. Participants indicated their self-reported frequency of engagement on a continuum ranging from 1 (never) to 5 (always).

HIV-related Stigma Mechanisms: Discrimination Intent and Prejudice

We adapted validated measures of HIV-related stigma [62–64], with minor changes made based on our preliminary work [18, 20, 21] (Cronbach’s α = 0.70–0.95). The measures of HIV-related stigma evaluated the extent to which the participants endorse prejudicial beliefs and intent to discriminate towards five groups of individuals: PWH, MSM, FSW, people who inject drugs (PWID), and transgender women (TGW). We repeated the questions with each group of key population. The HIV-related stigma mechanisms scores were derived by averaging scale items.

The items measuring prejudice (e.g., “How do you feel about patients who are [name of group]?’ and “How comfortable would you be working with patients who are [name of group]?”) were rated on a feelings thermometer that ranged from 0 (very negative) to 100 (very positive).

The remaining Likert-type items reflected (1) strongly disagree to (5) strongly agree. The measures on discrimination intent had items such as “I am willing to work with [name of group]”, “I am willing to provide the same care to [name of group] as I do to other patients” and “I am willing to do physical exams on [name of group]”.

Comfort in Performing Clinical Tasks (Taking Sexual and Substance Use Histories, HIV Testing, and PrEP Prescription) with Key Populations

We adapted items from validated measures (Cronbach’s α = 0.95) that gauged the extent to which the participants felt comfortable performing clinical tasks such as history taking, HIV testing, and PrEP prescription with members of key populations [65, 66]. We included additional items that reflected practices that would increase individuals’ susceptibility to HIV acquisition (e.g., sexualized drug use, engagement in sex work, condom use). Participants indicated the degree to which they felt comfortable performing clinical tasks (e.g., asking about sexual orientation, sexual practices, screening for HIV or sexually transmitted infections, prescribing PrEP, providing risk reduction counseling) on a scale ranging from 1 (completely uncomfortable) to 5 (completely comfortable). The comfort scores were evaluated by averaging the scale items.

Knowledge of PrEP Prescription

We employed a validated measure (Cronbach’s α = 0.87) that assessed the participants’ familiarity and attitudes with PrEP prescribing practices [67]. The items were “I am familiar with current research on PrEP safety”, “I know the required labs for PrEP initiation and management,” and “I can determine if PrEP is indicated for my patients”. We added an item that evaluated clinicians’ familiarity with on-demand PrEP. Participants responded on a 5-point Likert-type scale ranging from strongly disagree to agree strongly, and a composite score was created by averaging responses.

PrEP Perceptions for Key Populations

The validated scale (Cronbach’s α = 0.71–0.84) provided us with an understanding of the participants’ views about the safety and effectiveness of PrEP across key populations (MSM, FSW, TGW, PWID) and PWH as well as if members of key populations can afford PrEP [68]. We repeated the group of questions with each group of key population. Items included “PrEP is effective for preventing HIV infection among [name of group]”, “PrEP is safe for [name of group]”, “[name of group] will have more sex partners if they start taking PrEP”, and “[name of group] may not be able to afford PrEP”. Participants responded on a 5-point Likert-type scale ranging from strongly disagree to agree strongly. We formed a composite score from the responses to the item that measured the perception of PrEP affordability by key populations.

Analysis

All analyses were conducted in SPSS (version 29). Descriptive statistics described the sociodemographic and clinical characteristics of the sample. Next, we used exploratory factor analysis (EFA) and the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy [69] to assess the scale’s factor structure in relation to the theorized construct structure of two activism scale sources for structural validity guided by Yong and Pearce [70]. KMO values that are between 0.8 and 1.0 suggest sampling adequacy. Deletion of items from the scale was in accordance with the following criteria: (1) Factor loading less than 0.5, (2) Commonalities (h²) of less than 0.5 [71], and (3) Cross loadings less than 0.1 [72]. Reliability testing included a measure of internal consistency with Cronbach’s α and test-retest reliability with intra-class correlation coefficient (ICC).

In terms of construct validity, we tested convergent and discriminant validity by examining bivariate correlation coefficients between HIV activism and conceptually related together with unrelated factors at baseline. We anticipated that the HAICOS would be at least moderately correlated with the stigma-related constructs (prejudice and discrimination intent), knowledge about HIV pre-exposure prophylaxis (PrEP), and comfort performing clinical tasks with members of key populations (i.e., sexual and substance use histories, screening for HIV and sexually transmitted infections (STIs), performing risk reduction counseling and prescribing PrEP). We expected the HAICOS to be less correlated with the perception of PrEP affordability by key populations.

Results

Clinician Sociodemographic and Clinical Characteristics

78 clinicians were recruited for the randomized controlled trial, but seventy-four clinicians participated in the baseline and the follow-up survey (100% retention rate). As shown in Table 1, participants were, on average, 37 years old, primarily identified as Chinese (44.9%) and Muslim (35.9%), with the majority (59%) identifying as women. Most participants were general practitioners or medical officers (47.4%) and primary care physicians (44.9%). Most participants (70.5%) received medical training in Malaysia and were affiliated with the Ministry of Health.

Table 1.

Sociodemographic and clinical characteristics of the sample (n = 74)

Characteristics	n (%) or Mean (SD)
Age	6.16 (37.36)
Gender
Man	31 (41.89)
Woman	43 (58.11)
Ethnicity
Malay	26 (35.14)
Chinese	35 (47.30)
Indian	12 (16.22)
Other	1 (1.35)
Faith affiliation
Muslim	26 (35.14)
Buddhist	21 (28.38)
Christian	12 (16.22)
Hindu	9 (12.16)
Others	7 (9.45)
Country of medical training
Malaysia	54 (72.98)
Other	20 (27.03)
Clinical Role
Medical Officer/General Practitioner	35 (47.30)
Specialist registrar (a registered doctor in specialization training)	4 (5.41)
Primary Care Physician	34 (45.95)
Consultant	1 (1.35)
Affiliation
Ministry of Health	52 (70.27)
Ministry of Education	3 (4.05)
Private Sector	19 (25.68)

Structural Validity of Revised Scale with Exploratory Factor Analysis (EFA)

The correlation matrix revealed patterns of association among the items in HAICOS, indicating that the construction of factor structure would benefit from an oblique rotation [71]. Bartlett’s test of sphericity revealed statistical significance (χ² = 711.63, df = 78, p < 0.001), and the Kaiser-Meyer-Olkin measure of sampling adequacy was 0.855, thus indicating that the sample size is sufficient and items were suitable to perform an EFA. Principal axis factors with iterations and oblique rotation (Promax) determined items with pattern coefficient factor loadings that ranged from 0.55 to 0.94 (see Table 2). Three items with commonalities (h²) of less than 0.5 were removed. There were no cross-loadings less than 0.1 (refer to Appendix A for the unrevised version of HAICOS). The three-factor solution described 76.15% of the variance and reflected valuable dimensions of HIV activism (HIV activist identity and commitment, orientation towards day-to-day HIV activism, and orientation towards structural HIV activism). In adherence to Kaiser’s criterion, we retained factors with eigenvalues exceeding one (refer to Appendix C for the scree plot illustrating eigenvalues in descending order).

Table 2.

Pattern of Coefficients and Communalities for the revised HIV Activist Identity and Commitment Scale and HIV Activist Orientation Scale (HAICOS), Exploratory Factor Analysis: Principal Factor Analysis with Oblique Rotation (n = 74)

Item	Mean (SD)	Factor loading values			h2
		Factor 1: HIV Activist Identity and Commitment	Factor 2: Orientation towards Day-to-day HIV Activism	Factor 3: Orientation towards Structural HIV Activism
I identify myself as a HIV/AIDS activist	3.53	0.88	0.01	− 0.14	0.72
Being a HIV/AIDS activist is an important reflection of who I am	3.45	0.94	0.01	− 0.06	0.85
Being a HIV/AIDS activist is central to who I am	3.28	0.91	− 0.3	− 0.05	0.78
I go out of my way to engage in HIV/AIDS activism	3.22	0.75	0.1	0.21	0.72
I take the time I need to engage in HIV/AIDS activism	3.42	0.74	-0.02	0.04	0.56
I am truly committed to engaging in HIV/AIDS activism	3.43	0.74	-0.11	0.06	0.69
I go out of my way to collect information on issues specific to communities most vulnerable to HIV/AIDS	2.82	0.01	0.55	0.32	0.57
I present facts to contest another person’s statement (s) about the communities most vulnerable to HIV/AIDS	2.89	-0.11	0.85	0.07	0.68
I confront jokes, statements, and/or innuendoes with intended or unintended stereotypes, prejudice and discrimination	2.99	0.08	0.84	− 0.02	0.78
I try to change a friend’s, relative’s or acquaintance’s mind about issues specific to communities most vulnerable to HIV/AIDS	3.30	0.01	0.91	− 0.19	0.71
I share views or posts online in support of the legal rights of the communities most vulnerable to HIV/AIDS	2.58	0.13	0.68	0.03	0.60
I attend regular planning meetings for organizations that affirm and support communities most vulnerable to HIV/AIDS	2.92	– 0.12	− 0.12	0.77	0.60
I reach out to a community leader to address and repeal stigmatizing policies specific to the communities most vulnerable to HIV/AIDS	2.04	0.09	0.08	0.87	0.72
Eigenvalues		6.68	1.93	1.29
Percentage of variance		51.41	14.83	9.90
Cumulative percentage of variance		51.51	66.25	76.15

Note:

h² = commonalities. Bold font indicates factor loadings over 0.45

Reliability Testing of Revised Scale (Internal Consistency and Test-retest Reliability)

We calculated the indicators of reliability for the 13-item scale and each subscale using the scores from baseline. The HAICOS was highly internally consistent with Cronbach’s α of 0.91. Additionally, each subscale, namely the HIV activist identity and commitment, day-to-day activism, and structural activism subscales, demonstrated high internal consistency with Cronbach’s α of 0.93, 0.89, and 0.78, respectively. The inter-item correlations suggest moderate to high association (refer to Appendix B) (r = 0.47–0.86, p < 0.001).

Test-retest reliability yielded a highly correlated intra-class correlation coefficient (ICC) score of 0.86, p < 0.001. Furthermore, clinician scores on HIV activist identity and commitment, engagement in day-to-day activism, and structural activism were correlated at 0.73, 0.66, and 0.60 (all ps < 0.001). The results of the internal consistency and test-retest reliability tests suggest that the HAICOS is an internally reliable scale.

Construct Validity: Convergent and Divergent Validity

Evidence of convergent validity.

The HAICOS was positively correlated with stigma-related constructs such as prejudice and discrimination intent (i.e., measured in feelings thermometers and willingness to provide healthcare services to PWH and key populations) (r = 0.33–0.58, p < 0.001). This finding suggests that clinicians who identified as HIV activists and were engaging in HIV activism were more likely to endorse lesser prejudicial beliefs and discrimination intent. HIV activist identity and engagement had a statistically significant high positive correlation with knowledge about PrEP and comfort in performing clinical tasks with members of key populations (r = 0.47–0.66, p < 0.001). Clinicians who identified as HIV activists and were engaging in HIV activism were more likely to be knowledgeable in PrEP prescribing and were more comfortable performing clinical tasks with clients from the key populations (see Table 3).

Table 3.

Construct validity: means, SDs, and correlations between HAICOS subscales and other variables

	Mean (SD)	HAICOS: total	HAICOS: HIV activist identity and commitment	HAICOS: Orientation towards day-to-day activism	HAICOS: Orientation towards structural activism
Prejudice: Feelings thermometer	74.58 (17.08)	0.58**	0.41**	0.53**	0.16
Discrimination intent	4.52 (0.49)	0.50**	0.38**	0.33**	− 0.03
Knowledge about PrEP	3.85 (0.61)	0.66**	0.56**	0.53**	0.15
Comfort performing clinical tasks with key populations	4.22 (0.52)	0.55**	0.41**	0.47**	− 0.01
Perception of PrEP affordability by key populations	3.38 (0.76)	0.12	− 0.01	0.5	0.05

Note:

^**p ≤ 0.001

Evidence of discriminant validity.

The HAICOS and measures of perception of PrEP affordability by key populations demonstrated a nonsignificant low association (r = − 0.01– 0.12, p > 0.01) (see Table 3).

Discussion

The current study evaluated the psychometric qualities of the HAICOS, a scale adapted to measure HIV activist identity, commitment, and engagement of Malaysian general practitioners and primary care physicians in a broad spectrum of actions geared towards HIV activism. Based on two scales that measured activism identity, commitment, and orientation [5, 22], HAICOS featured aspects of the HIV response and actions aimed at challenging HIV-related stigma that was otherwise uncaptured in the two scales. The EFA supported the expected factor structure and identified a 13-item scale with three dimensions of HIV activism: HIV activist commitment and identity, HIV activism orientation toward day-to-day activism, and HIV activism orientation towards structural activism. We examined internal consistency, test-retest stability, and construct validity (measures of convergent and discriminant validity) for the HAICOS scores. HAICOS was correlated with stigma-related measures (prejudice and discrimination intent), PrEP knowledge, and comfort in performing clinical tasks with members of key populations. Our analyses provided evidence that the factors of the HAICOS were related, as expected, to the dimensions of the HAICOS. Compelling preliminary findings from the reliability and validity of the HAICOS offered sound psychometric quality with items and factor structure adequate to estimate clinicians’ propensity towards HIV activism.

HIV activist identity and commitment were highly correlated with engagement in day-to-day HIV activism. Similar findings were demonstrated in a campus activism study [5]. However, it was found to be the opposite in a study on Black community activism wherein racial identity centrality alongside social responsibility beliefs and values had nonsignificant associations with low-risk, high-risk, and formal activism orientation [23, 73, 74]. An essential distinction in these differences in findings is that the measurement of HIV activism was done with individuals who do not necessarily hold similar lived experiences of oppression, while other studies involved participants who self-reportedly do. The theoretical underpinning of the current study suggests that lived experience of oppression among healthcare professionals may not be a strong determinant of propensity towards HIV activism because empathy, collectivism, and servitude are central denominators for healthcare professionals.

Unsurprisingly, HIV activist identity and commitment were strongly associated with day-to-day HIV-related activism. The relationship between identity centrality and activism orientation among individuals with lived experience is complex because individuals have varying identities and motivations to engage in sociopolitical action [75, 76]. Alternatively, studies on activism should address the intersectional nature of marginalized identities (e.g., race-based activism plausibly has differing associations compared to LGBTQ+ (lesbian, gay, bisexual, transgender, and queer) activism or HIV activism.

On the contrary, we found that engagement in structural activism did not yield similar patterns to day-to-day activism. Engagement in structural activism, interestingly, is not strongly related to any of the measures of stigma and clinical practice, including PrEP knowledge and perceptions. These findings reveal that, compared to associations for day-to-day activism, the extent to which clinicians pursue structural activism has a weaker association with how they see themselves as HIV activists or their sense of commitment toward HIV activism. Similar findings were found in activism studies in different contexts, such as in the Black Community Activism Orientation Scale [23], race-based collective action against racial injustice [73, 74], and college campus-based activism studies wherein scales that measured high-risk actions demonstrated low variance [5]. We theorized four reasons why these findings emerged. First, clinicians may find structural activism challenging because of competing clinical and non-clinical commitments. As a result, engagement in day-to-day activism could be the extent to which these clinicians have reasonable resources to employ. Second, clinicians potentially perceive structural activism as too high-risk, radical, not socially acceptable, or having significant negative medicolegal implications [9, 77, 78]. The study participants possibly have hesitated to provide sensitive data, which was misconstrued as illegal. Third, this group of clinicians could have been discouraged by the shortcomings of past engagement in structural activism, which results in a great sense of injustice and hopelessness, consistent with preliminary findings of higher depressive symptoms among HIV activists [2, 79] and HIV service providers who became wearied of providing continuous education in response to stigma [80]. Past shortcomings likely deter clinicians from engaging in future structural activism. Fourth, one’s shared identity as an HIV activist and an individual with lived HIV experience (or lack thereof) potentially limit the extent of engagement in macro-level sociopolitical reforms. People with lived experiences are best to lead the HIV response. Clinicians who do not share similar experiences (and interests) may not share similar motivations and commonalities in oppression faced by minoritized and underserved groups affected by HIV.

Key findings from the evaluation were compelling in support of the applicability of HAICOS in research involving allies. First, future research or interventions should consider the role of HIV activism in mediating the relationship between HIV-related stigma and disparities in the HIV response. HIV activism, when stewarded effectively, can address significant drivers of stigma, such as fear, anxiety, lack of knowledge, and perceived stigma, as well as disrupt stigma mechanisms, such as stereotypes, prejudice, and discrimination. Acts of HIV activism (e.g., contesting stigmatizing statements, jokes, or innuendos about PWH and key populations, sharing posts in support of PWH and key populations) challenge stereotypes and encourage perspective-taking, particularly in interpersonal interactions [81, 82]. Seemingly insignificant acts of activism (e.g., collecting information, giving a lecture, teaching) could enhance knowledge, leading to individuals questioning their stereotypes [83–86]. Lower stereotype and prejudice among Malaysian physicians and future healthcare providers have been associated with lower discrimination intent [18, 19].

Second, HAICOS has promising utility as a screening tool or post-intervention measure in study designs that track behavioral development. Differing propensities towards activism can be accounted as potential covariates to be statistically corrected in the relationship between exposure(s) and outcome(s). Role assignment or group allocation could also consider individuals’ activist orientation at baseline in group-based interventions.

Third, on a macro-level, the HAICOS can be utilized by researchers who are interested in studying how HIV activism impacts the social enablers in the HIV response, such as the UNAIDS 95–95-95 goals or the Sustainable Development Goals (SDG) 3.3 on ending the AIDS epidemic. HAI-COS is also consistent with frameworks of social action theory [87] and resource mobilization theory [88]. HAI-COS elucidates the pathway between individual orientation towards engagement in social action and proxies to clinical outcomes. Proxies to clinical outcomes, including comfort in performing clinical tasks with members of key populations and PrEP knowledge and perceptions, could be critical predictors that can be factored into how we study interventions that narrow disparities in the HIV cascade of care.

Strengths, Limitations, and Future Directions

The HAICOS has several strengths that may benefit behavioral and socio-political psychology research. HAICOS forefronts an empowerment-based agenda as opposed to deficit-focused perspectives in negotiating for change and challenging HIV-related stigma. The scale also provides a nuanced measurement to study HIV activism on a micro and macro level. It considers a broad spectrum of engagement in activism. Measuring HIV activism among Malaysian clinicians offers insights into allyship among healthcare professionals and the activism climate in Malaysia. The measure provides an opportunity to identify Malaysian clinician-activists before training and mobilizing them. These insights are timely and needed because of the stigma-charged sociopolitical environment perpetuating disparities in the HIV response [6, 10, 12, 13, 20, 21]. The majority of past research on activism is cross-sectional, based in regions where the sense of collectivism and political will are comparatively revered (e.g., the United States) [2, 3, 5, 22, 25], and included only participants with lived experience [2, 23].

Additionally, the HAICOS is short. It consists of only 13 items that capture the centrality of HIV activist identity, commitment, and orientation toward HIV activism. Therefore, it is easy and quick.

It is also essential to consider the limitations of the HAICOS. First, the responses through the survey were self-reported and possibly not wholly representative of the sample’s engagement in activism, prejudices, stereotypes, and their endorsement of discrimination. Biases such as recall bias and social desirability are potential threats to the validity of the responses. Furthermore, we did not collect sufficient information on stigmatized identities or other HIV-related lived experiences from the participants. The clinicians can have the lived experience of oppression (e.g., from their sexual orientation, gender identity, or substance use history). The participants who voluntarily participated in a HIV stigma-reduction intervention program for clinicians may also be likely to be biased or inclined to be interested in the study or HIV activism.

We also evaluated items that measured the engagement of clinicians in HIV activism on a non-numerical Likert-type scale, i.e., never, seldom, sometimes, often, always. This scale could be difficult for participants who were unsure of the quantitative representation of each response [89]. Analyses may have overlooked intervariable intervals [90–92]. While numerical scales have their advantage (i.e., concretely and conclusively providing values), recall bias remains a potential threat to the validity and reliability of a self-reported scale. Using non-numerical Likert-type responses should provide flexibility and opportunities for the respondents to reflect on their engagement, allowing them to reflect on the perceived frequencies they identify closest.

Although sufficient, the study included a relatively small sample of clinicians and did not test for generalizability. Additionally, we were cognizant of the complexity of the analysis and the risk of overfitting the model due to the smaller sample size. We adopted a conservative approach to mitigate this risk by keeping the analysis relatively simple and limiting the number of factors extracted [93]. We also exercised caution with factor rotation techniques, opting for more straightforward methods to avoid introducing unnecessary complexity [93].

Furthermore, we recognize that interpretation of the results becomes more challenging with smaller sample sizes, as estimates of factor loadings and communalities may be less precise. In light of this, we approached the interpretation of our results with caution, emphasizing the need for careful consideration and validation of the findings with future research.

While EFA indicated adequate sampling and yielded sufficient power with a medium effect size, the study should be replicated with more extensive and diverse groups of healthcare professionals, including accounting for existing lived experiences of oppression or concealable stigmatized identities (e.g., clinicians’ sexual orientation, gender identity, HIV status, substance use history). Larger samples would enable future studies to control for more moderating factors, thus providing insights into associations between HIV activism and clinical practice, including how specific sociodemographic and clinical characteristics (e.g., gender, affiliations, and years of clinical practice) or other factors influence these associations. Future research should also test for generalizability and model or external validity.

Conclusion

The study provides concise, structurally valid, and reliable measures to evaluate HIV activist identities, commitment, and orientation. The psychometric evaluation of the HAI-COS depicts the complexity of HIV activist identity and orientation towards social change. The associations between measures of HIV activism and clinical practice demonstrate the need to positively shift the dial on the HIV response among clinicians as allies and stigma changemakers. The measure can be further validated as HIV activism continues to be studied alongside capacity-building interventions among clinicians to encourage engagement in HIV activism and allyship. Measuring HIV activism increases the involvement of clinicians beyond their clinical duties to disrupt the cycle of inequity, call out injustice, and demand systemic changes in leadership, practices, and culture.

Data Availability

Data may be available from the corresponding author upon request.