Stigma has repeatedly been identified as a major barrier to help seeking for mental health problems across various disorders and across the lifespan. Stigma is also an obstacle to community reintegration and rehabilitation in people suffering from severe mental illness. Moreover, people with psychiatric diagnoses suffer the effects of discrimination in health care settings. Not only do people with mental illness have diminished access to primary care, there is evidence to suggest that physicians perform fewer physical examinations and laboratory investigations, provide less preventive health care, and undertake fewer therapeutic interventions in this population. Researchers are increasingly framing the problem of stigma as a public health issue.

So begins a new paper that considers stigma and mental health.

This week’s Reading: “Explicit and Implicit Attitudes of Canadian Psychiatrists Toward People With Mental Illness” by Dr. Layla Dabby et al., which was just published in The Canadian Journal of Psychiatry.

This paper shows that members of the public demonstrated relatively negative explicit attitudes towards mental illness. In fact, Canadians reported a desire for greater social distance from the patient with schizophrenia as opposed to the patient with diabetes, even though the study describes the patient with schizophrenia as well-controlled by medication. Wow.

Except here’s the twist in the tale. The paper actually didn’t look at the public. The paper looked at psychiatrists and residents of psychiatry. In other words, the relatively negative explicit attitude wasn’t from the uninformed small businessman in Edmonton or the teacher in Halifax; it reflects the biases of people like… me.

Is this psychiatry’s dirty little secret?

You can find the paper here:

http://www.cpa-apc.org/media.php?mid=2342

The paper seeks to study bias in psychiatrists and residents of psychiatry. To that end, the authors distinguish between conscious and unconscious bias. “Social information is processed on both an explicit level (that is, consciously, controllably, and reflectively) and an implicit level (that is, subconsciously, automatically, and intuitively).”

Here’s what they did:

· Essentially, drawing on 3 measures, they considered bias in a group of psychiatrists and psychiatrists-in-training.

· Email invitations were sent out to potential respondents inviting them to participate. The sample was drawn from the McGill University Department of Psychiatry and the Canadian Psychiatric Association membership.

· Participants were asked to complete the 5-item Social Distance Scale, which considers people’s willingness to engage with certain types of patient populations. Two short descriptions of patients were presented: “John is a 40-year-old patient with schizophrenia, which is well- controlled on medication.” and “Peter is a 40-year-old patient with diabetes mellitus, which is well-controlled on medication.” Respondents were then asked questions like their willingness to see John or Peter live near them or marry into the family.

· Respondents were asked to do a 12-item Opening Minds Scale for Health Care Providers – a scale developed to assess the attitudes of health care professionals toward patients with mental illness and toward self-disclosure of a mental illness. They rate agreement or disagreement on a 5-point scale to statements like: “If I were under treatment for a mental illness I would not disclose this to any of my colleagues.”

· Respondents were finally asked to do a computerized version of the Implicit Association Test, which assesses automatic (that is, unconscious or intuitive) associations to the diagnostic labels of schizophrenia and diabetes mellitus. In this test, aspects of illness were mentioned (hallucinations for schizophrenia, for example, and hyperglycemia for diabetes) and then paired with responses (wonderful, pleasure, and happy, compared with terrible, nasty, and awful).

· Demographic information was collected on the respondents.

· Statistical analysis was done, including assessing variables for normal distribution with the Shapiro-Wilk test.

Here’s what they found:

· “A total of 198 respondents accessed the survey online; 126 consented to participate in the study, and, among these, 107 respondents completed the entire survey (41 from the McGill University Department of Psychiatry, yielding a response rate of 12.1%; 66 from the membership of the CPA, yielding a response rate of 3.3%).” (!)

· Residents were significantly younger than psychiatrists, and more likely to be from Quebec. Gender, country of medical training, and mean number of hours of direct patient contact per week did not significantly differ between residents and psychiatrists. Participating psychiatrists were disproportionately based at university-affiliated teaching hospitals (52.9%).

· There were internal consistencies on the results for the SDS, OMS-HC, and the IAP.

· In terms of results, on all three measures, “There was no significant difference between residents and psychiatrists.”

· With the SDS, there was a big difference between scores reported for the schizophrenia vignette and the diabetes one. “SDS-SCZ scores were significantly greater than SDS-DM scores, both in residents (z = –4.64, P < 0.001) and in psychiatrists (z = –6.30, P < 0.001), as determined by a Wilcoxon signed- rank test.” (!)

· Researchers did several statistics analyses to look at correlation. IAD scores were correlated negatively with age but positively with patient contact. More significantly, with a more complicated analysis: “a hierarchical multiple regression was run to determine if the addition of patient contact improved the prediction of IAT D-scores over and above age, sex, and employment type alone. The full model was statistically significant…” (!)

The authors write:

Although there is evidence supporting the existence of implicit biases against mental illness in various studies of the general population and professionals, our study does not support the hypothesis that psychiatrists and psychiatry residents harbour negative implicit associations to schizophrenia, compared with diabetes mellitus. IAT scores, while negative in an absolute sense, fell short of representing even a small effect size. Nonetheless, relatively negative attitudes were detected in explicit measures of stigma, with both residents and psychiatrists reporting greater desire for social distance from the patient with schizophrenia, even though his illness was described as well-controlled by medication.

The authors conclude:

Future research on stigma among physicians must address the complex interplay between explicit and implicit attitudes. Studying the evolution of stigmatizing attitudes during the course of training, deconstructing the relation between patient contact and attitudes, and incorporating meaningful behavioural outcome measures into the study of stigma are essential steps toward the development of evidence-based, stigma-reduction education for physicians.

A few thoughts:

1. This is a thoughtful study – it’s also a thought-provoking study.

2. The response rate is low. Does that undermine the conclusion? Two points. First, the people who did respond – as the authors note – had a good variation of attitudes and were gender balanced, suggesting that the sample size probably was representative. Second, this paper is hardly in a vacuum. Other work on bias in health care providers has shown similar problems. Consider that the “Opening Minds” interim report of the Mental Health Commission of Canada notes: “People who seek help for mental health problems report that they often experience some of the most deeply felt stigma from front–line health care personnel.” So while it would be easy to dismiss this paper, it would be a mistake to do so.

3. The paper’s conclusion – a call for further research – makes sense given the nuance and complexity of this issue. Understanding the implicit and explicit biases of providers will help us better understand a core problem that needs to be understood. But can we be more ambitious? We need more research and more action. Though things are markedly better than in years past, stigma still exists – in our society, and in our health care system as well. At the end of the day, we need to take patients with mental health problems and disorders seriously. Anything less than that is a failure – to our patients and to ourselves.

Reading of the Week. Every week I pick a reading — often an article or a paper — from the world of Psychiatry.