Bias in the Assessment of Mental Health

To what extent is the diagnosis of schizophrenia influenced by cultural bias?

Table of Contents

Introduction 3

Cultural Bias 4-8

Drug Companies 9-13

Conclusion 14-15

Bibliography 16-17

Introduction

This essay aims to examine “to what extent is the diagnosis of schizophrenia influenced by cultural bias?” Schizophrenia is a long-term mental disorder that causes a range of psychological symptoms such as: hallucinations, delusions, muddled thoughts, changes in behaviour, psychosis, lack of interest and concentration. Exact causes of this illness are unknown but it has been linked to genetics, brain development, neurotransmitters, pregnancy complications, and stress and drug abuse (NHS, 2016). It affects over 23 million people worldwide and sufferers are 2-3 times more likely to die early (World Health Organisation, 2018) therefore accurate diagnosis so treatment can begin as early as possible is crucial. There isn’t a single test for schizophrenia so an assessment is usually conducted by a mental health professional. A diagnosis would usually be confirmed if the patient reports experiencing delusions, hallucinations, incoherent speech or negative symptoms (NHS, 2016).

Diagnosis can be influenced by external factors such as clinician bias. This, in a medical context, occurs when there is a deviation from the validity of a diagnosis as a result of previously held beliefs. This deviation can be a consequence of influence from drug companies or cultural bias, thus leading to an incorrect diagnosis. If a clinician is culturally biased their diagnosis may be to do more to how the clinician views people belonging to the patient’s culture in a general sense and thus ignore the symptoms they are reporting our instead interpret them differently based on the characteristics they perceive the client to have based on their background e.g. white psychiatrists may be more likely to give Afro-Caribbeans a schizophrenia diagnosis as they view that race as “strange” (Javier, 2012) or “bizarre” which is how people also view schizophrenics. In addition, pharmaceutical companies have a very close relationship with medicine, especially in the US. It is the possible that they could influence the way doctors diagnose and prescribe their drugs in order to push their product forward and reap financial reward.

In the case of schizophrenia there has been an “overdiagnosis” in which there has been an inexplicable increase in diagnoses, especially in those of African descent (Schwartz & Blankenship, 2014). The purpose of this essay is to investigate to what extent is this surge in diagnosis motivated by cultural bias on behalf of clinicians or if it is the result of toxic influence from pharmaceutical companies who, via different methods of persuasion to increase the administration of their own products, have provoked this said increase.

Cultural Bias

Cultural bias is described as the tendency to judge others based on one’s own assumptions of the other person’s culture. This can occur in the form of ethnocentrism which is when one views the world from one’s own cultural perspective and believing theirs is the standard. Therefore, any variation is regarded as strange or abnormal. It is possible that many clinicians approach patients with this ethnocentric view. This is particularly likely as 83.6% of clinicians are white (American Psychological Association, 2015) so would view patients of other ethnicities with more scrutiny than those of the same race. This was established in a study conducted by Eack in which data from the 1995 MacArthur Violence Risk Assessment Study was used to examine the impact of racial differences on perceived honesty when diagnosing African-Americans (Eack, Bahorik, Newhill, Neighbors, & Davis, 2012). They coordinated structured assessments of diagnostic, sociodemographic and clinical measures. They interviewed 215 African-Americans and 537 white people to obtain diagnosis using the DSM III, as well as reviewing their records. Interviewers then completed a questionnaire after each interview, rating their perceived honest of the patients on a 5 point scale. Interviewers consistently perceived African-Americans to be less honest than their white counterparts and researchers concluded that this was the cause behind African-Americans being 3 times more likely to be diagnosed with schizophrenia as this was the only constant mediator. This is a fairly accurate conclusion as if clinicians perceive African-American as liars they are less likely to focus on symptoms that may suggest a different diagnosis e.g. lack of concentration as a result of stress may suggest depression instead of schizophrenia but stress factor could be disregarded and schizophrenia diagnosis would be given instead. However, the participant’s perception of the interaction wasn’t assessed so it may be that African-American are less likely to trust clinicians so behave in a dishonest manner, possibly minimising the reality of their symptoms. Psychiatrists may then pick up on this and overcompensate. In addition the MacArthur study is old and thus the stud is plagued with the issue of temporal validity. The DSM-III was used to diagnose simply because it was the standard at the time of the study. It is possible that if the current DSM-V was used the likelihood of diagnosis could differ or at least the number of patients who were given a schizophrenia diagnosis would be very likely to decrease. Regardless, the study does highlight the importance of doctor-patient relations and in clinical practice it’s important that doctors establish a good relationship with their patients in order to get the most accurate diagnosis as if doctors are unable to feel they can trust patients report of symptoms (a major contributor to diagnosis) they may rely on other unreliable factors such as the patients culture to make a diagnosis.

The previously mentioned possibility that misdiagnosis is due to misperception of symptoms was supported by Barnes in which data obtained from the Management Information Services Division of the Indiana Family and Social Services Administration (IFSSA) was used to obtain a sample of 2404 clients (Barnes, 2008), 80.5% of which were African-American. By accessing such a large database containing info on the state’s psychiatric hospitals, it’s probable that the results can be fairly easy to generalise. Barnes found that African-American’s were four times more likely to receive a schizophrenia diagnosis and were severely underdiagnosed for major depressive disorder and bipolar disorder. This under-diagnosis of MDD and bipolar disorder supports the idea that misdiagnosis may be due to the misinterpretation of symptoms as MDD and bipolar disorder share symptoms with schizophrenia such as lack of concentration (NHS, 2016). Since confusion may be caused due to the similarity of symptoms it’s possible that cultural bias would then come into play and influence the clinician’s final decision. The researcher determined that the race of the client was a major predictor for whether they would receive a schizophrenia diagnosis and, therefore the conclusion that cultural bias influences the clinician’s diagnosis is quite valid, especially as Barnes controlled for other variables such as gender, age and education. However, results may differ in comparison to similar studies as semi-structured interviews were used to diagnose psychiatric disorders not the participant’s admission diagnosis. In addition, this was only conducted in the state of Indiana so the extent in which the results of this study can be generalised is questionable. Although, its hospital system is very similar to those in the rest of the country so generalisability is unlikely to be a major issue. Consequently, the implications of these findings mean that psychiatrists need to be careful when deciding a diagnosis and ensure it is not influenced by any previously held beliefs.

This pattern of “overdiagnosis” isn’t limited to African-Americans as shown by Flaskerud & Hu who aimed to determine the relationship between the ethnicity of patients and psychiatric diagnosis in white, black and Latino clients that are in the Los Angeles County mental health system (Flaskerud & Hu, 1992), using a very large sample totalling 26,400 adult clients, so as to minimalize any issue of generalisability. These clients had been seen in the county mental health facilities between January 1983 and August 1988. Multiple variables such as age, gender, socioeconomic status, and primary language were observed as well as ethnicity and it was found that black and Asian clients received a greater quantity of psychotic diagnoses than their white counterparts. Ethnic identity was found to have the strongest, consistent relationship with diagnosis. Whites and Asians received a greater number of Major Depressive Disorder than blacks and Latinos and, black and Asian clients received a greater proportion of psychotic diagnoses than white and Latino clients. These results show that there is clearly a link between the ethnicity of a patient and the diagnosis they receive as even when other variables were controlled for ethnicity was the most prominent factor. In addition, unlike other studies, this one also looked at other ethnicities apart from African-American and white; Latino’s were also found to receive psychotic symptoms. This contributes to the idea that cultural bias influences diagnosis as both Latino’s and African-American’s are stigmatised, for example they are stereotyped to be dangerous the same way someone suffering from psychosis. Therefore, predominantly white clinicians are more likely to give a psychosis diagnosis because it aligns with the stereotype they have of blacks and Latino’s. In addition the study was conducted over 5 years so one can be sure that this isn’t a random phenomenon but an established fault in the medical system. However, despite its large sample, this study was only conducted on the city of Los Angeles so its ability to be generalised is questionable but at the same time the healthcare system across America is similar so the negative impacts this carries is limited.

This phenomenon in the misdiagnosis of Latino’s was examined further by Minsky, Vega , Miskimen, Gara, & Escobar (2003). Though, they instead looked at cultural bias’ influence via a different direction of causality; is it that Latino patients present their symptoms and its severity differently? They drew data from a behavioral health service delivery system in New Jersey, and included administrative data, clinical diagnosis, and a self-reported symptoms. A clinical sample of all new admissions into the system between January 1, 2000, and August 31, 2001 (19 219 patients) was obtained and a logistic regression was performed on the sample to examine the effects of the patient’s ethnicity. It included three dependent variables: presence of major depression, a schizophrenia spectrum disorder, and bipolar disorder. In order to determine the symptoms of patients the 32-item Behavior and Symptoms Identification Scale (BASIS-32), a self report questionnaire, was used. This enabled Minsky to meausre accurately suffered symptoms and compare this to those clinicians perceieved patients to suffer. It was discovered that African-Americans were diagnosed as having a schizophrenic spectrum disorder more frequently than Latinos and European Americans despite the fact they self-reported average levels of psychotic symptoms. Researcher concluded that Latinos were exceedingly diagnosed as having major depression, despite the fact that they self-reported considerably higher levels of psychotic symptoms and depressive symptoms. From this one can deduce that it is possible Latino’s present symptoms differently as they report their symptoms in a different manner than African-Americans despite there being very few reasons why there should be a great difference in the way they experience said symptoms. Yet this same explanation can’t be used to explain why clinicians would diagnose Latinos less frequently with a schizophrenia spectrum disorder if they report higher levels of psychotic symptoms. For that reason, as well as there being cultural influence in how symptoms are presented, cultural bias still influences how these symptoms are intrpreted. In addition to that Spanish is a prominent language amongst the latino community therefore this inaccuaracy in diagnosis may be influenced by the scarcity of clinicians that speak spanish. Therefore there is cultural bias in the sense that psychiatrists approach patients with a biased view in how communication should be established so they miss out on more revealing information. This is very likely as the majority of them are white and english speaking (American Psychological Association, 2015).

Drug Companies

Drug companies develop, produce and market drugs to be administered as treatment for various ailments, either physical or mental. Some companies have financial relationships academic institutions and their influence can then permeate medical schools and teaching hospitals. Alternatively, companies can influence clinicians directly by offering financial incentives to encourage them to promote/prescribe a certain drug and therefore diagnose patients more frequently in order to sell said drug to augment profits. Such relationships were investigated by Campbell et al which surveyed department chairs in academia and their relationships with medical industries (Campbell, et al., 2007). 125 medical schools and 15 teaching hospitals were surveyed along with 459 chairs who supervise personnel that conduct research. 60% had a personal relationship with industry and 27% were even a member of a scientific advisory board. The implications of this is that chairs supervise personnel who conduct research for companies with whom they have a relationship with, therefore, there is the possibility that research may be intentionally geared towards supporting a more unconventional drug. The same goes for there being chairs on scientific advisory boards as they may be biased when reviewing and advising scientific research. The survey does have a fairly high response rate of 66.7% (459/688 chairs) which increases the generalisability of the study. However, the sample was obtained via survey so demand characteristics such as social desirability may come into play e.g. one my think they come across as ore respectable if they do have a relationship with a major pharmaceutical company. It’s also possible that the sample only consists of chairs that see no problem with their relationships with industry and therefore responded. It’s possible that the other 33.3% are aware of the problematic nature of their relationship with industry and so a darker story about academia-industry relationships may not be shown in these statistics. The results are telling nonetheless – industry’s multiple relationships with academia are strong and hence unprofessional especially as they have ties to teaching hospitals were they can influence impressionable junior doctors to push forward their antipsychotics for incentives so they then may be inclined to misdiagnose.

This relationship between academia, medical practice and drug companies has been proven to be damaging by Chris Watkins as he conducted a questionnaire to look at the personal qualities of doctors (Watkins, et al., 2003) and whether they agreed with a string of statements about their prescribing attitudes such as “When faced with a patient who expects a prescription (which is not clinically indicated) my usual response is to: Agree readily or reluctantly or Discuss but not prescribe”. The questionnaire obtained a 64% response rate (1097/1714). Watkins used a univariable logistic regression model to analyse results and then entered the most significant statements into a multivariable logistic regression model so was able to see the influence of different variables and determine their impact relevant to each other. He found that frequent contact a drug representative was connected with practitioner’s inclination to prescribe new drugs or express views that lead to unnecessary prescribing. This supports the idea that interaction with drug companies does influence how clinicians prescribe and thus how they diagnose as many clinicians would have given a diagnosis before prescribing a drug. Therefore there could be an increase in schizophrenic diagnoses in order to justify said prescription. An issue with this theory however that it’s based on the assumption that prescription isn’t given without a diagnosis, it’s plausible that a clinician may simply suggest a drug without giving a definite diagnosis. There’s also an issue in regards to direction of causality – do drug representatives influence clinicians or do they just target those more likely to prescribe their drugs? Regardless Watkins, et al. still highlights the issue with there being any contact at all.

Another reason why drug companies influencing clinician’s prescribing behaviour and diagnosis is an issue is that a meta-analysis of 52 controlled randomized trials comparing atypical (amisulpride, clozapine, olanzapine, quetiapine, risperidone, and sertindole) and conventional antipsychotics such as haloperidol (Geddes, Freemantle, Harrison, & Bebbington, 2000) concluded that there is no clear evidence that new, atypical drugs pushed forward by drug companies were more effective than conventional ones. Geddes, Freemantle, Harrison, & Bebbington measured symptoms scores and the rate of drop out during trials in order to gage tolerability, which is a good objective measure for a subjective and variable issue. The study made use of a large sample of 12,649 patients so issues with generalisability are minimal. In addition by looking at many trials various methods and results have been analysed and theire effectiveness combined so consequently this raises the reliability of the results and conclusions of this study. Although there was no evidence that showed unconventional drugs to be better they noted that they may be recommended if a patient doesn’t respond to conventional drugs, however, this is not a valid reason as to why pharmaceutical companies should promote them more than conventional drugs. Hence, drug companies are definetly influencing clinicians simply for financial gain which is at the expense of patients. On the other hand findidng no clear evidence isn’t synonymous with there being no likelihood of unconventional drugs trumping typical drugs nor does this suggest they’re strictly ineffective. However, these resuls are worrying nevertheless as Watkins, et al., 2003 showed that greater conact with drug companiy representatives lead to a stronger inclination to prescribe new drugs and thus if uncnventional drugs whose effectiveness haven’t been truly proven are being prescribed instead of typical drugs this can have adverse effects on patients who shouldn’t be eceiving a prescription in the first instance.

Besides drug representatives, clinicians can also be influenced by artificial means such as gifts, the influence of which was investigated by Wazana (2000). They investigated if gifts influence prescription of medication via a meta-analysis of 538 studies. They also investigated the extent of the relationship between physicians and the pharmaceutical industry and its impacts. It was found that interactions with the pharamaceutical industry start as early as medical school and follows on into practice. Things such as attending sponsored continuing medical education (CME) events and accepting subsides for transport or accommodation for educational conventions were associated with an increased rate of prescription of the sponsor’s medication. Attending presentations given by pharmaceutical representative speakers was also associated with irrational prescribing. On top of this physicians believe that representatives provide accurate information. Overall, interactions with company representatives was found to impact the prescribing practice of physicians in terms of prescribing cost, nonrational prescribing,  preference and rapid prescribing of new drugs, and decreased prescribing of generic drugs. Self-reporting surveys were used which is a quick way to obtain plenty of rich, firsthand data inexpensively but it can result in underestimates in terms of frequency of contact as participants may struggle to recall occurences or reported occurences my be impacted by social desirability. The studies included used cross-sectional methods which considers multiple variables, including confounding, that may influence data solididfying the study’s validity. However, the study is essentially correlational in nature so although it’s useful in establishing a relationship we cannot completely determine a cause and effect relationship. The fact that interaction starts at medical school shows that this influence is deeply ingrained into clinicians from day one. They are given industry-sponsored meals, the frequency of which decreases as they enter practice but the frequency in which grants, conference travel and research funding is received increases, most likely as such things are considered more useful as one becomes a more experienced physician. Therefore drug companies have an extremely strong hold on clinicians making it ever more likely that they are a key factor influencing clinician’s diagnosis.

Conclusion

In conclusion it can be confidently argued that cultural bias influences the diagnosis of schizophrenia as numerous studies have shown that African descendants and Latinos are disproportionally diagnosed with schizophrenia despite there being little evidence that they present their symptoms differently, though the case differs for Latino’s in which there was a difference in the presentation of symptoms as well as interpretation. Similarly, studies that compare schizophrenia diagnosis between black people and white people have shown that cultural bias on behalf of the clinician influenced their diagnosis such as their perception of the patients honesty or the fact that all of the studies that looked into cultural bias controlled for other confounding variables such as age, gender etc. and found that cultural bias was the most consistent mediator. In addition the negative stigma associated with some symptoms of schizophrenia such as psychosis (sufferers of which are viewed as aggressive) correlate with negative stigma associated with black people e.g. they’re aggressive. Therefore, it’s reasonable to infer that cultural bias influences diagnosis to a great extent.

On the other hand, it has been suggested that another factor, the influence of drug companies, can also influence the diagnosis of schizophrenia undermining the aforementioned argument. This has been suggested in studies that have established a clear relationship between drug companies and the medical sector opening up the possibility of it being influenced. In addition it has been found that frequent contact with drug company representatives impacts the prescribing behaviour of clinicians despite there being no concrete evidence newer drugs are any more effective than conventional ones. If clinicians are more likely to prescribe psychotic drugs they are more likely to give a psychotic diagnosis i.e. schizophrenia or even if a prescription is given before diagnosis it may still lead to a final psychotic diagnosis. However, it’s possible this could be due to a nocebo effect – clinician has instilled into the patients mind the possibility of a psychotic disorder and thus they exhibit those symptoms thereafter. However, research must be done in that aspect. In addition there is still the possibility that this prescribing behaviour is also influenced by culturally biased beliefs as even though they are more likely to prescribe a drug, whom they prescribe to may then be influenced by cultural bias. This suggests that other factors influence the impact of drug companies.

Subsequently, in order to confidently conclude that cultural bias has the greater influence on diagnosing schizophrenia more studies it’s essential that more research is done on the cultural differences in the presentation of symptoms as well as its interpretation in order to clarify a clear direction of causality as there is still the question on whether it’s the presentation of the symptoms or the perspective in which they are viewed that impacts diagnosis. In addition studies must be conducted to investigate closely the difference in diagnosis between Latino patients and African patients as despite the fact symptoms correlate with stereotypes associated with Latino’s and Africans diagnosis still differs between the two. Therefore, it is until this information is present that it can be assuredly concluded that cultural bias has the greatest influence on the diagnosis of schizophrenia.

Bibliography

• American Psychological Association. (2015, July). 2005-13: Demographics of the U.S. Psychology Workforce. Retrieved February 19, 2019, from APA: https://www.apa.org/workforce/publications/13-demographics/index?tab=6
• Barnes, A. (2008). Race and Hospital Diagnoses of Schizophrenia and Mood Disorders. Social Work, 77-83.
• Campbell, E. G., Weissman, J. S., Ehringhaus, S., Rao, S. R., Moy, B., Feibelmann, S., & Goold, S. D. (2007). Institutional Academic−Industry Relationships. Journal of the American Medical Association, 298(15), 1779-1786.
• Eack, S. M., Bahorik, A. L., Newhill, C. E., Neighbors, H. W., & Davis, L. E. (2012, September). Interviewer-Perceived Honesty as a Mediator of Racial Disparities in the Diagnosis of Schizophrenia. Psychiatric Services, 63(9), 875-880. Retrieved from https://ps.psychiatryonline.org/doi/pdf/10.1176/appi.ps.201100388
• Flaskerud, J. H., & Hu, L.-T. (1992, May). Relationship of Ethnicity to Psychiatric Diagnosis. The Journal of Nervous and Mental Disease, 180(5), 296-303. Retrieved from https://journals.lww.com/jonmd/pages/articleviewer.aspx?year=1992&issue=05000&article=00003&type=abstract
• Geddes, J., Freemantle, N., Harrison, P., & Bebbington, P. (2000). Atypical antipsychotics in the treatment of schizophrenia:systematic overview and meta-regression analysis. BMJ, 321, 1371-1376.
• Javier, E. I. (2012). Diagnostic Bias: Racial and Cultural Issues. Taking Issue, 9, 847. Retrieved from https://ps.psychiatryonline.org/doi/pdf/10.1176/appi.ps.20120p847
• Minsky, S., Vega , W., Miskimen, T., Gara, M., & Escobar, J. (2003). Diagnostic Patterns in Latino, African American, and European American Psychiatric Patients. Arch Gen Psychiatry, 60(6), 637-644. Retrieved February 23, 2019, from https://jamanetwork.com/journals/jamapsychiatry/article-abstract/207503
• NHS. (2016, April 26). Bipolar Disorder- Symptoms. Retrieved February 20, 2019, from NHS – Web Site: https://www.nhs.uk/conditions/bipolar-disorder/symptoms/
• NHS. (2016, October 11). Schizophrenia – Diagnosis. Retrieved February 23, 2019, from NHS Website: https://www.nhs.uk/conditions/schizophrenia/diagnosis/
• NHS. (2016, October 11). Schizophrenia – Symptoms. Retrieved December 21, 2018, from NHS: https://www.nhs.uk/conditions/schizophrenia/symptoms/
• Schwartz, R. C., & Blankenship, D. M. (2014, December 22). Racial disparities in psychotic disorder diagnosis. World J Psychiatry, 4(4), 133-140. Retrieved from https://f6publishing.blob.core.windows.net/b51eab98-36a8-44aa-ad88-f84cd9e5d71c/WJP-4-133.pdf
• Watkins, C., Moore, L., Harvey, I., Carthy, P., Robinson, E., & Brawn, R. (2003, May 31). Characteristics of general practitioners who frequently see drug industry representatives: national cross sectional study. British Medical Journal, 326, 1178-1179. Retrieved from https://www.bmj.com/content/bmj/326/7400/1178.full.pdf
• Wazana, A. (2000). Physicians and the Pharmaceutical Industry: Is a Gift Ever Just a Gift? JAMA: The Journal of the American Medical Association, 283(3), 373-380. doi:10.1001/jama.283.3.373
• World Health Organisation. (2018, April 9). Schizophrenia. Retrieved December 21, 2018, from World Health Organisation Web site: https://www.who.int/news-room/fact-sheets/detail/schizophrenia