Out of the silos: identifying cross-cutting features of health-related stigma to advance measurement and intervention

Many health conditions perceived to be contagious, dangerous or incurable, to result in clearly visible signs, or to be caused by breaking taboos or immoral behavior share a common attribute – an association with stigma and discrimination. These health conditions are diverse in nature and include infectious diseases like HIV, tuberculosis (TB), leprosy and lymphatic filariasis, non-infectious chronic conditions such as epilepsy and cancers, and mental health conditions such as schizophrenia, depression, and substance abuse. Jones et al. [1] proposed six features, namely, (1) esthetics, (2) concealability, (3) course, (4) disruptiveness, (5) origin, and (6) peril, that help in recognizing and understanding why particular conditions are more vulnerable to health-related stigma, what factors would worsen or reduce a given stigma, and why some stigmas may be easier to address than others.

People often have co-morbidities and live with one or more of these health conditions and experience simultaneously different types of health-related stigma. Stigma is problematic because it affects people psychologically and restricts their social participation, and it can also create barriers to accessing healthcare, including retention in care for people living with HIV (PLHIV), relationships, education, and housing, thereby further marginalizing already vulnerable populations [2‐4]. While the etiology of stigma may differ between conditions and, sometimes, cultural settings, the manifestations and psychosocial consequences of stigma and discrimination are remarkably similar [3, 5, 6]. Regardless of the condition, stigma is a dynamic process enacted through structures and individuals, mediated by relationships of power and control that are constantly being produced and reproduced [7]. Similarities across conditions are most likely due to the fact that the core of stigma is social in nature and therefore a common problem based on common human interpersonal responses to differentness and the mechanisms by which these responses might be expressed [8, 9]. Nevertheless, responses to persons with the same condition may also differ in different locations, based on local differences in social determinants of stigma (e.g., religious beliefs). They may vary between conditions, depending on perceived cause and danger (e.g., in HIV or leprosy, people might avoid sharing a meal to avoid infection).

The cross-cutting nature of stigma is evidenced by the measurement methods used and the interventions that have been shown to be effective to reduce stigma or mitigate its impact across conditions [3, 10‐13]. In many of the disciplines dealing with stigmatized conditions, the problem has been recognized and is addressed to some extent, but often only in a condition-specific manner. One challenge is that the funding, research, assessment tools, and interventions often address stigma related to only one particular condition. If measurement tools and interventions that assess and address common dimensions of stigma were possible, the scarce resources to address stigma could be used more efficiently and healthcare providers could use the same tools and approaches, across conditions. Several theoretical models describing common elements of stigma have been proposed, including those by Scambler [14, 15], Link and Phelan [16], Pescosolido et al. [17], and Weiss [5].

In an attempt to ‘capture’ stigma, as well as in assessing its severity and monitoring and evaluating the impact of interventions to reduce stigma, a large number of instruments have been developed, often within specific fields such as mental health [41] and HIV [28]. In addition, tools have been developed for many of the different domains of stigma such as perceived or anticipated stigma, internalized stigma, public stigma, stigma by association, and healthcare provider-based stigma [3, 9]. For an extensive review of the types of stigma assessments as well as their use in evaluating changes in mental health-related stigma interventions, please see the recent report from the U.S. National Academy of Sciences/Institute of Medicine [42]. Unfortunately, most instruments are both condition specific and limited to a particular domain of stigma (e.g., internalized or public stigma). Despite these silos of tools, a detailed analysis of stigma assessments showed that many similarities exist in the approaches used across conditions and in the issues addressed in the items used in questionnaires and scales [3]. It is informative to pay particular attention to the instruments that have been used across several conditions, including the Social Distance Scale (SDS) [41, 43], the Berger Stigma scale [24], the Internalized Stigma of Mental Illness (ISMI) scale [44], and the Explanatory Model Interview Catalogue (EMIC) [45]. Some of these have also been used across domains to assess internalized stigma, public stigma, and healthcare provider-based stigma. Having shown applicability across different conditions, we might consider the aspects of stigma contained in these instruments to be ‘common’ elements of stigma across illnesses.

Information-based interventions are very likely the most common approach to addressing public stigma against any condition. However, they differ in content across conditions because they often address condition-specific knowledge gaps, stereotypes, fears, and other drivers of stigma; not infrequently, these are the only strategies used. However, while knowledge or education is often an essential part of stigma reduction, it is insufficient on its own [55‐57].

Many authors have reviewed stigma reduction strategies and interventions from either a disease-specific or generic perspective [11, 57‐63]. Evidence of effectiveness from well-designed studies using larger samples, particularly of longer-term impact, is scarce [58, 62]. However, available evidence suggests that stigma should be tackled at multiple levels, by using multiple strategies and the interventions must be context specific and continued or repeated to achieve a lasting impact [6, 8, 11, 64‐66].

Table 1 shows examples of stigma instruments that have been used across several conditions to measure attitudes and perceived and enacted stigma among the public or community. The SDS has a long history and was originally designed to assess willingness to associate with persons of different ethnic backgrounds [46]. Link et al. [90] used a version adapted for mental health to assess attitudes towards persons with mental health conditions. Lee et al. [91] assessed ‘victim blaming’ of persons with HIV or AIDS among US college students using the SDS. Peters et al. [43] used social distance as a proxy for respondent attitudes towards persons affected by leprosy in Indonesia, and a study in Germany assessed stigma against persons with obesity using the SDS [92]. The EMIC-CSS has been used across conditions most often, including in a study assessing attitudes and perceived behavior against persons with onchocerciasis [93], mental health conditions [49], Buruli ulcer [94], tuberculosis [95], and leprosy [43, 96, 97]. Additionally, the cultures were very diverse, including four countries in Africa and four in Asia.

In the same way, instruments used to assess stigma experienced by persons affected across a range of conditions are shown in Table 2. The Berger Stigma Scale, originally designed to measure perceived and experienced stigma among PLHIV [24], was successfully adapted for use in leprosy [98] and meticillin-resistant Staphylococcus aureus [53]. The ISMI was used most frequently, with no less than 81 papers covering 42 completed translations [13]. Most studies used the instrument in mental health, but other studies demonstrated the usefulness of the ISMI among persons with substance abuse, leprosy, HIV, and inflammatory bowel disease [96, 99‐101]. The EMIC Affected Persons scale has been used most widely in terms of range of conditions. Originally designed to measure the impact of leprosy on the mental health of persons affected [45], it has since been used to measure experienced stigma related to mental health conditions, including depression, schizophrenia and bi-polar disorder [102‐104], onchocerciasis [105], Buruli ulcer [94], HIV [101], TB [106], and leprosy [96].

The current paper demonstrates that ‘health-related stigma’ is a viable concept with clearly identifiable characteristics that are similar across a variety of stigmatized health conditions in very diverse cultures. The etiology of stigma differs between conditions and sometimes between cultural settings. For example, persons with schizophrenia are stigmatized because people perceive them to be unpredictable or dangerous, while PLHIV may be stigmatized and discriminated against because, in certain cultures, HIV is associated with homosexuality and promiscuity, and because it is perceived to be a highly infectious, as well as fatal and incurable disease. Leprosy is often stigmatized because of the notion that the person affected has committed a sin or broken a taboo, either in this or a previous life; it may also be due to fear of the associated disfigurements. Even regarding the etiology and origins of stigma and discrimination, ‘shared dimensional features’ can be readily recognized. Pachankis et al. [120] used the six features identified by Jones et al. [1] (aesthetics, concealability, course, disruptiveness, origin, and peril) as a taxonomy for characterizing and investigating the perceived burden of stigma on health and wellbeing across no less than 93 health and other conditions.

As noted in the Background section, the expressions or manifestations and psychosocial consequences of stigma and discrimination are often remarkably similar, even across very different cultures and levels of socioeconomic development [3, 5, 6, 8]. Stigma starts when salient differences between people are recognized, labelled, and connected to stereotypes or social identities [16]. This process leads on to a separation between ‘us’ and ‘them’, resulting in status loss and discrimination. Depending on the culture and time, these differences may include a large variety of characteristics, including ethnicity, sexual orientation, skin color, body weight, religious beliefs, and a wide range of health conditions. In this paper, we limited ourselves to health conditions, though we are well aware of the intersectionality of stigma where health-related and other stigmas interacted and may compound each other [121‐123]. A substantial body of literature addresses the intersectionality of stigma related to particular conditions. For example, Lowie et al. [121] examined how gender, race, sexual orientation, and sex work intersect with HIV-related stigma. Very few studies have investigated types of stigma, stigma assessment, or stigma interventions across multiple stigmatized conditions. A notable exception are the studies that have looked jointly at HIV- and TB-related stigma [124, 125]. Mak et al. [126] compared SARS-related stigma with that of HIV and TB. However, the great majority of studies of stigma related to health conditions occurred within the specific field dealing with a specific condition or range of conditions (e.g., mental health conditions). Within these fields, authors have demonstrated the similarities and differences across cultures and languages, e.g., in leprosy [127], HIV [8], TB [106], and mental health [6]. However, very few studies have attempted in-depth analyses across different health conditions. Van Brakel [3] included mental health, epilepsy, HIV, leprosy, TB, Buruli ulcer, onchocerciasis, and physical disability in his review of measurement of health-related stigma, noting many commonalities in the approaches and tools used to measure different stigmas. A more recent review investigated stigma across 10 neglected tropical diseases and noted many similarities in the types of stigma reported, the manifestations, and the approaches used to mitigate stigma [10]. Although not limited to health-related stigma, the study of Pachankis et al. [120] included 44 health conditions. They examined similarities and differences regarding each of the six characteristics proposed by Jones et al. [1] and investigated their association with a range of different stigma-related measures, including the SDS. One of the findings was that “Visibility and course were not associated with social distance. In contrast, participants indicated a desire for greater social distance with respect to stigmatized statuses that were perceived as disruptive, aesthetically unappealing, onset controllable, and perilous” [120]; these features are shared by many stigmatized health conditions.

The above findings show that there is a scientific rationale for the concept of health-related stigma, as proposed by Weiss et al. [19] and Scambler [20, 128]. A more generic approach to the study of health-related stigma opens up important practical opportunities. This paper illustrated this with two aspects of work – stigma measurement and interventions to reduce or mitigate stigma.

If it were possible to measure stigma and discrimination using generic instruments, this would have clear advantages, especially for use in public health programs and social services. Use of measurement tools requires training. With a different tool for each condition, staff in health and social services have to learn and keep up with many different instruments, some of which they may only use infrequently, thus never acquiring a ‘feel’ for the instrument and the results it produces. In the current age of mobile data collection, one could envisage that adaptation of a given instrument to a particular condition would be done by just indicating on the opening screen which condition one wants to test; the software would automatically adapt the instrument to that condition. Tools for which this would be very easy are those indicated in Table 1 and Table 2. Instruments like the SDS, EMIC, and ISMI were shown to be highly suitable for use across conditions since the content includes manifestations and impact common to many stigmatized health conditions.

Researchers in the health-related stigma field can clearly also benefit from the use of instruments that can be adapted very easily for use across conditions; the study of Pachankis et al. [120] illustrates this point very nicely.

A disadvantage of using generic instruments is a potential lack of sensitivity and/or specificity. Where this would be required, one could envisage using an add-on module comprising a few condition-specific items. This would retain the advantage of a common core of items that can be used and compared across conditions. A very similar approach that is widely accepted is the measurement of health-related quality of life. Generic tools like the WHO Quality of Life scale, abbreviated version (WHOQOL-BREF), and the Short Form 36 items are used across a myriad of disabling and stigmatized conditions and in very culturally diverse circumstances. In certain situations, add-on modules are used, such as the WHOQOL-DIS for disability, or the WHOQOL-SRPB for the effects of spirituality, religion and personal beliefs.

Many of the same advantages that apply to cross-condition measurement tools also apply to interventions.

Table 3 and Table 4 illustrate the several interventions that have already been used successfully with multiple conditions; this is hardly surprising because of the common social and psychological processes underlying health-related stigma [5, 16, 19]. Manifestations, such as difficulties in finding and maintaining employment, broken relationships, and impacts on socioeconomic status and mental wellbeing, including shame and reduced self-esteem, are common across conditions, thus offering entry points for cross-cutting interventions. It should be noted that, although the studies included have been classified under one, or at the most two, intervention types, almost all studies used multiple interventions. Sometimes, these addressed different levels and sometimes they addressed both the sources of stigma and the persons affected by stigma. Even when used on a single level, there is evidence that using multiple interventions is more effective than using a single intervention [111].

In contrast to the use of instruments, certain interventions can even be used across multiple conditions simultaneously. This is the case for counselling services, skills-building, and economic empowerment programs and SHGs, for example.

One major problem is that funders of stigma reduction programs usually only fund condition-specific studies, measures, and interventions. Surveillance for stigma and stigma-mitigating interventions can be integrated in regular health and social services using generic tools and interventions. For example, in China, a stigma-reduction intervention focused on infection control through education and providing adequate supplies for practicing universal precautions [78, 129]. Similarly, in Vietnam, a stigma-reduction intervention allowed healthcare facility staff to develop practical skills around infection prevention and a code of practice, tailored for their own hospital’s needs, on implementing stigma-free practices and universal precautions [130]. In the field of leprosy, counselling to mitigate the effects of stigma has been integrated in a range of hospitals that offer leprosy services in Nepal and India [131, 132].

Using generic tools and interventions within the health services would help overcome the siloed approach by demonstrating the advantages of integration, while simultaneously contributing to health systems strengthening. Dr Gottfried Hirnschall, WHO HIV Director, said, “We need to ensure that frontline health workers have the information and skills required to effectively identify, address and avoid stigma and discrimination of all types, including those related to HIV”.¹ Developing generic health-related stigma assessment and monitoring tools as well as generic stigma interventions would provide essential building blocks for making this possible.

A limitation of this paper is that it is not based on a systematic literature review. We can therefore make no claim to completeness of the evidence to support the concept of health-related stigma. However, we believe that the cross-condition use of each instrument and intervention has been adequately demonstrated through our use of these selective, illustrative examples.