Stigma is defined as a discrediting characteristic that occurs when someone’s attributes are inconsistent with our stereotype, making them appear different and inferior.1,2 If stigma is related to negative beliefs about having mental disorders (MDs) and seeking professional help, it can have two forms: public stigma, which is typical of social groups, and self-stigma, which is the internalisation of these beliefs and prejudices by the person suffering from the disorder.3 Self-stigma of seeking help is the self-perception that affects self-esteem and self-efficacy if mental health care is requested.4 This type of stigma means that treatment is not sought or is delayed, which has negative effects on the prognosis of MDs.5,6 In Colombia, only 38% of those affected by MDs seek health care and the rest do not do so, mainly due to attitudinal barriers, including self-stigma.7–9

Mental health problems are common in university student populations in which self-stigma can act as an access barrier to mental health services that are offered to them.10–13 Some studies have pointed out its role as a predictor of attitudes towards psychological treatment and its relationship with a lower likelihood of seeking services.4,13,14 In particular, medical students have reported a high prevalence of depression and anxiety and exposure to sources of stress such as the acquisition of clinical skills, social expectations and the suffering and death of patients, among others.15–18 In a study on the use of mental health services, 33.7% of medical students reported not seeking psychological help due to the possible impact that this fact would have on their professional career.19 Moreover, 25.3% of the students associated the use of psychological services with weakness and a similar percentage showed concern about what their fellow students might think of them.19

It is important to take into account this construct and its relationship with help-seeking behaviour in the design and evaluation of mental health and wellbeing programmes, because it would allow for a more complete model that would increase the chances for success of interventions to reduce the duration of untreated MDs, which would have great value for public health, since timely care improves the prognosis.20 Therefore, in Colombia an instrument is required to measure self-stigma related to seeking professional help, since its research would provide new and relevant information for understanding the reasons and attitudes why people limit themselves in seeking help and deprive themselves of the positive effects of available mental health care. In addition, a measurement tool is required to identify the manifestations of self-stigma, its distribution in the population and its effects on people's quality of life.

Most of the available tools for measuring stigma associated with seeking help were developed and validated in samples of university students and focus on measuring public stigma, not self-stigma.21–26 The only self-stigma scale related to seeking help that was identified was the one developed by Vogel in 2006: the Self-Stigma of Seeking Help (SSOSH) scale.4 It is a self-applied scale composed of 10 items that measure concerns about the loss of self-esteem that a person would feel if they decided to seek help from a mental health professional. The SSOSH was built under the assumptions of the Classical Test Theory (CTT) and a reflective model. It has been translated, validated and used in several countries and, although in Spain there is a version in Spanish, it lacks published evidence of evaluation of psychometric properties.27

For clinicians and researchers to be able to use the SSOSH scale in Colombia, it is necessary to carry out an adaptation process and an evaluation of its psychometric properties to ensure its validity, reliability and usefulness.28 This study aims to translate it into Spanish, adapt the SSOSH scale to Colombian culture, and evaluate the psychometric properties of the new version in a sample of university students from a medical school.

The study had two phases: a) translation and adaptation, and b) evaluation of psychometric properties. It was approved by the Ethics Committee of the Faculty of Medicine of the University of Antioquia.

This study evaluated the validity and reliability of the SSOSH scale for its use in Colombia. The SSOSH scale measures the psychological impact of internalising a stigma associated with seeking professional mental health treatment in university populations. It is a short and quick-to-complete scale that, when validated, will contribute to research and the generation of evidence-based interventions.41

In general, the results of the evaluation of psychometric properties were satisfactory. The description of the scores in the items showed a percentage of losses <3% in all items, which is acceptable and indicates that they were understandable and relevant for almost the entire sample. The distribution of scores in the response categories showed that all response options were informative and had discriminative qualities. Furthermore, no floor or ceiling effect was found in the distribution of scores.

The internal consistency of the SSOSH scale indicates a high degree of interrelation between the items and the homogeneity of the measurement. The Cronbach's alpha found was close to that reported by previous studies.4,33,36,42,43 However, it differs from the Cronbach's alpha of 0.71 reported by Acun.37 This is probably due to the use of a smaller number of items by this author, which is known to decrease the value of Cronbach's alpha.28 In this case, Acun used only 9 of the 10 definitive items of the SSOSH in English, arguing a low factor loading of item 10.

The results of the test-retest reproducibility analysis indicate that the scale scores in the sample are stable over time. We can trust that 77% of the total variance between measurements corresponds to true differences in the scores obtained from the participants, and the remaining variability can be attributed to measurement errors. The ICC calculated was similar to that found in the validation study of the scale in the United States (0.72).4 Moreover, a standard error of measurement of 2.7 points is acceptable, showing a very favourable degree of precision when compared with the range of scores that goes from 10 to 50 points. In the Blant-Altman limits of agreement, there were no systematic measurement errors between the applications, which provides more evidence on the reliability of the scores.

The CFA made it possible to establish that the data collected in the empirical sample in Colombia fit well the one-dimensional theoretical model proposed by the developers of the scale. Unlike previous studies that performed a CFA with the maximum likelihood method4,36, the present study used the diagonally weighted least squares (WLSMV) method to estimate the goodness-of-fit statistics, as it is considered more robust and appropriate for the SSOSH scale, which has items with an ordinal measurement level.35 Meanwhile, the CFA modification indices led to the examination of Model 5, which presented the best fit, showing covariation between items 5 and 8, 1 and 8, and 6 and 10. These covariances may be due to the similarity in the wording of the contents of the items. It should be added that our result contrasts with what was found in a validation study of the scale in Turkey, where they observed that the model was two-dimensional.44 However, it was not the same scale because they used the version with 23 items that came from the original scale construction study.

Regarding the convergent construct validity, it was expected that those who are more susceptible to perceptions of public stigma for receiving psychological help would tend to internalise the stigma.45,46 The moderate positive correlation with a public stigma scale favours the validity of the SSOSH. In the validation study of the scale in the United States, a moderate and positive correlation was also reported between these two constructs in two different samples (ρ = 0.48 and ρ = 0.46).4 The moderate negative correlation of the SSOSH with the measure of attitudes towards seeking help―that is, the greater the experience of self-stigma, the more negative attitudes towards seeking help in mental health would be produced―is also in favour of validity. This finding is consistent with the existing theory that the search for psychological treatment requires that individuals feel sympathy for themselves and feel worthy of receiving help.47 The internalisation of stigma erodes the individual's self-esteem, which leads to less sympathy for themselves and more negative attitudes about seeking help.33 This is in line with the original validation of the scale, in which moderate negative correlations were observed between the SSOSH and the ATSPPH-SF in two samples (ρ = –0.53 and ρ = –0.63).4

The lack of significant association with the measure of social desirability supports the notion that the self-stigma of seeking help scale measures the internalisation of the stigma of seeking psychological help, and not the social benefit of the individual's reactions to possible scenarios of mental health problems and help-seeking behaviour. In the study, the self-stigma measure was not correlated with the social desirability scale, which supports the divergent-type construct validity of the scale. Validation of the original scale reported a non-significant negative correlation of –0.13 between these two measures.4

The differences found with a high effect size between those who view the intention to seek help in case of suffering a mental health problem positively and those who do not support the discriminative validity of the scale scores, that is, the ability to distinguish through the scores between relevant groups with respect to the construct. Those who had a negative disposition towards seeking help obtained the highest scores on the measure of self-stigma. However, these results should be interpreted with caution, as it is most likely that at the time of the evaluation the majority of the participants did not have significant mental health problems. Therefore, the reported willingness to seek help may not be a reliable predictor of actually consulting a professional when needed.

Some of the limitations of this study have to do with the self-selection of the participants in the sample, since the variability of the sample could be different than in the target population and this would affect the reliability observed.29

Another limitation is that the ATSPPH-SF and MC-SDS scales, used for the evaluation of convergent and divergent validity, are not validated in Colombia. However, the Spanish versions of these instruments were used, which were validated in Spanish-speaking populations of university students in the United States and Argentina, where they showed adequate psychometric properties.34,48 The decision to use the aforementioned instruments is due to the need to obtain new information about the consistency between the SSOSH scale scores and the conceptual model of the construct. Although this information was not obtained under ideal conditions, it is a starting point for further refinements of the evidence provided, taking into account the continuous nature of the validation processes.

It is important to bear in mind that this validation of the SSOSH was performed in students of health sciences programmes and it might be thought that the psychometric properties cannot be generalised to other university student populations. This has been considered because it can be assumed that they would have greater homogeneity in the self-stigma for seeking help construct as they are sensitised to mental health problems and therefore have less tendency to stigmatise both the condition and the search for treatment. However, the opposite could also be expected: people who are being trained to help others may self-stigmatise if they themselves seek care because they believe, for example, that they should be able to solve their problems on their own, and that their aptitude would be called into question, etc.17,49 In addition, there is evidence that medical students with MDs have little access to mental health services offered at universities and that approximately one third do not seek help, because they believe it would signify weakness and are concerned about the impact this fact would have on their professional career.19 Therefore, we are not certain that this is a population whose homogeneity is so high that the psychometric properties are significantly altered. Moreover, the generalisation of the measure for other non-university populations is a limitation, since the psychometric properties of the scale scores are unknown in populations with different levels of education, income or severity of the mental health problem.

Future research needs to carry out the process of evaluating the sensitivity to change of the measure's scores. The discriminative capacity of the scale should also be examined between groups that have actually sought professional help in the past and those who have not, or better still, through prospective monitoring of this behaviour. In addition, in the future, studies will be carried out that evaluate differences in the functioning of the items and other psychometric properties between populations from different faculties and university programmes.

The SSOSH scale is a useful, valid and reliable measure to investigate help-seeking self-stigma in university populations in the area of health in Colombia. Although greater knowledge of its sensitivity to change and its psychometric properties in different contexts is required, it can be used to measure this construct as an initial step in the study of the search for and access to mental health services in university students.

The authors have no conflicts of interest to declare in relation to the research topic.