Racial Disparity in Mental Illness: Advice for Clinicians

Hello. I am Stephen M. Strakowski, professor of psychiatry at the University of Cincinnati, where I also serve as senior vice president in the affiliated health system, University of Cincinnati Health. Today I want to talk to you about a health disparity that has been around in psychiatry for many years and then think about ways we might address it and manage it in our clinical practices. It has been recognized for 50 or maybe even 60 years now that in clinical settings, African American patients are significantly more likely than otherwise similar white patients to be diagnosed with schizophrenia. In fact, the rates of schizophrenia can be nine times higher in a comparable group of African American patients compared with white patients, suggesting that there may be misdiagnosis of schizophrenia. We have explored this for the last 20 years to try to better understand it because from epidemiologic studies, we would expect the rates of diagnosis of schizophrenia across different racial and ethnic groups in the United States to be relatively similar in clinical settings; and, in fact, that is not what is commonly observed.

Potential Contributing Factors

There are a lot of different reasons why this might be happening. From a clinician's standpoint, it may be related to not correctly eliciting symptoms or failing to understand cultural variability and differences about how distress is reported. So-called "idioms of distress" vary among groups, and if we are not sensitive to these differences, we may misinterpret information as it is provided. Unfortunately, there may also be overt or even covert stereotyping that leads to jumping to conclusions prematurely without appropriately gathering all information. This becomes relevant particularly for a diagnosis like schizophrenia, which is a diagnosis of exclusion. Other psychiatric conditions, especially mood disorders but also conditions such as posttraumatic stress disorder, have to be considered and ruled out, and this requires assessing symptoms broadly.

From a patient's standpoint, there may be differences in how symptoms are expressed for different conditions across different groups, and so we want to make sure that we understand those. There have also been reports that for minority patients in general, but particularly for African American patients, because of a negative history in mental health systems for many years, particularly in the early 1900s, there is a protective wariness that patients bring when approaching a mental health setting that can be misinterpreted as paranoia. This may lead to a decision that psychosis is present when it is not. There are also religious differences and expressions of religion; often, agnostic clinicians tend not to appreciate the role of religion in different cultural groups, and, again, it gets misinterpreted.

All of these things are potentially contributing factors. We have tried to understand this through a series of research studies over the last 20 years or so, primarily here at the University of Cincinnati, and we have found several things that may be helpful in approaching this problem.

Improving Our Understanding of the Issue

In an early study that we published in the Journal of Clinical Psychiatry in 1997,^[1] we compared clinical diagnoses made in a broad group of psychiatric patients presenting in an emergency service vs research diagnoses obtained with a structured clinical interview later in the same admission. What we found was that there was a high rate of disagreement among all patients between the research diagnosis and the emergency department diagnosis, which isn't terribly surprising because you just get more information in the research setting. Overall, there were rates between 40% and 60% of disagreement; however, these rates were much higher in African American patients, where disagreement approached 65%-70% compared with 45% or so in the white patients. When we examined why that was happening, it looked like once symptoms were acquired by clinicians, the Diagnostic and Statistical Manual of Mental Disorders (DSM) criteria were appropriately applied similarly in both groups, but in the African American group, there tended to be a lower rate of reporting the presence of affective symptoms.

This study suggested that affective symptoms were not being elicited consistently, particularly in African American patients presenting with psychotic symptoms. Because schizophrenia is a diagnosis of exclusion, harvesting of those symptoms becomes very important. Subsequently, we have done a number of different studies trying to further understand this issue, and it has led to the impression that in clinical settings, for reasons that we still don't completely understand, when African American patients present with psychosis, the affective symptoms tend to be underappreciated, and even when they are recorded, they tend to be underweighted. So, there is an overweighting of the psychotic symptoms and a tendency to prematurely conclude that schizophrenia is present before there has been a more thorough evaluation of possible exclusions.

In fact, we had questions about whether this was unique to African Americans or it was a broader problem among minorities. We conducted a very large six-site study with University of Cincinnati, University of Michigan, Rutgers University, Howard University, University of Texas San Antonio, and University of California Los Angeles to get a large, diverse sample of patients across the United States. We also came up with a way of blinding the race of the patients in the interviews through a series of recordings and then using experts to make diagnoses blind to the racial and ethnic background of the patients through a series of editing transcripts. This study can be read in detail in a publication that came out in the Archives of General Psychiatry^[2] in 2012 with the lead author Michael Gara. In this study, we found what we had seen across the previous studies, that African Americans had higher rates of schizophrenia diagnoses even after controlling for mood disorders and symptoms. In fact, we observed that the African American patients had similar rates of mood symptoms, mania symptoms, and depression symptoms, but their psychotic symptoms received higher ratings, suggesting again that they were potentially being over weighted and leading to over interpreting the psychosis even when affective symptoms were present. We also found that psychotic symptoms were associated with patients' cultural mistrust, consistent with this notion that if patients come in with cultural apprehension, they are suspicious and less open with clinicians, which potentially gets misinterpreted as psychosis.

Advice for Clinicians

So what does this mean in practice? How do we manage this idea that when you are of a different racial or ethnic background than your patient, how do you help yourself take care of them and prevent cultural misunderstanding from interfering with making good clinical decisions? We have had several ideas around that but no complete solution. Structured interviews seem to minimize these differences. A deliberate style in your clinical assessment to make sure that you always ask a broad set of questions and truly evaluate the possibility of a mood disorder in someone presenting with psychosis can really go a long way towards decreasing this potential disparity.

I encourage most clinicians to take a look at your practice and see if the ethnicity or gender of your patients is biasing your diagnostic decision-making, and then challenge yourself to try a more deliberate approach towards making diagnoses. If you can build a structured interview into your clinical practice, that is probably the best approach. But admittedly, in many clinical practices, there just is not enough time. Even in your routine clinical assessment, try to copy some of that best practice of being deliberate and broad in symptom assessment, recognizing also that schizophrenia is a diagnosis of exclusion, so it should not be the first diagnosis when first presented with a complex patient with psychosis. When you are in doubt, my recommendation is always to pick the condition with a better prognosis when given a choice of two that you are debating because it tends to keep treatment options open. When we diagnose schizophrenia, we often narrow what our expectations are, and by doing that we may not be as likely to use a mood stabilizer such as lithium, order psychotherapy, use cognitive interventions, or other things that we might be more likely to use with someone who has psychotic depression or bipolar disorder. Pick the better prognosis condition, and if your patient is not getting better (with treatment), challenge yourself to re-evaluate if the diagnosis was the correct one.

The real goal here is to try to continue to build best practices into how we deliver care so that we can really treat all patients fairly, despite potential cultural mismatches and differences between psychiatrists and their patients.

I don't believe that most psychiatrists are doing this on purpose or that they are bigots. However, I do believe that all of us are susceptible to preconceptions and cultural biases that we are often not aware of. The way to manage this is to be deliberate, systematic, and keep a very close eye on how you are making diagnoses over time. Again, if the patient is not getting better, challenge your own clinical assessment.

With that, my hope is that we will be able to continue to provide great care, and I thank you very much for taking the time to listen to this today.

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.