The Use of Person-Centered Language in Scientific Research Articles Focused on Psychosis

Abstract

Purpose

Societal stigma of psychosis leads to delayed treatment for individuals experiencing early symptoms, for fear of being labeled “crazy” or “psychotic.” Delayed treatment can worsen the prognosis of individuals with psychosis. Proper adherence to person-centered language (PCL) guidelines can curb the widely held stigmatized preconceptions about psychotic conditions.

Methods

Our meta-epidemiological analysis began with a systematic search of literature in PubMed discussing psychosis from 2020–2022. From the 14,522 results, we randomly selected 500 articles in journals with at least 20 articles on psychosis (n=9,775), of which 241 articles met the inclusion criteria. These papers were screened for the usage of stigmatized language; article characteristics were also noted for further analysis.

Results

Our analysis revealed that 53.1% of articles screened included stigmatized language. The most common stigmatized phrase identified was “schizophrenia patient,” found in 35.3% of papers. “Schizophrenic [patient/subject/etc]” and “psychosis patient” were each found in 12.4% of papers. The type of intervention (ie, pharmacologic, therapeutic, or observational) was the only variable that was found to be significant for PCL usage (P = 0.006).

Conclusions

The majority of medical articles screened did not adhere to PCL guidelines. The usage of stigmatized language in literature reinforces the use of these labels in medical education, doctor’s offices, and popular culture. Given the impact of stigma on treatment delay – and the poor prognosis that may result – proper care should be taken to enforce PCL in medical literature. Reducing the fear of labeling is a necessary step in encouraging treatment for people experiencing early symptoms of psychosis.

Patient-Friendly Recap

• Person-centered language emphasizes the individual and their experience, not their diagnosis or other labels.

• Previous studies have found that the use of stigmatizing language, which labels people in stereotypical ways, is associated with negative attitudes toward those people, even from healthcare providers.

• Our study found that over 53% of the academic articles on psychosis that we looked at used stigmatizing language, like “schizophrenia patient” or “schizophrenic.”

• We suggest that academic journals and authors writing about psychosis should screen their articles for stigmatizing language and adhere to person-centered language guidelines.

Psychosis refers to a set of symptoms frequently involving hallucinations and delusions, with or without disorganized thoughts, speech, or behaviors. ^1,2 While these symptoms can exist secondary to a medical condition – such as toxin exposure, dementia, epilepsy, and various autoimmune diseases – they are classically associated with psychotic conditions. Psychosis is a DSM-5 category that includes schizophrenia, schizophreniform, brief psychotic disorder, delusional disorder, and various catatonic or “other/unspecified” disorders with psychosis. ³ It is estimated that 1.5–3.5% of individuals experience a diagnosable psychotic disorder in their lifetime, ⁴ with schizophrenia being the most common. ²

People at risk for psychosis reported experiencing internalized stigma and having greater perceived discrimination compared to people without psychosis and reported higher levels of distress, shame, and fear. ⁵ Stigmas around psychotic conditions have been shown to delay the initiation of therapies and other treatments for individuals who are experiencing early symptoms, ⁶ which can lead to a worsened prognosis, ^7,8 such as higher rates of suicidality and transition to chronic psychosis. ⁵ This delay is secondary to fears of the individual being labeled with derogatory terms and the associated social rejection that comes with those labels. ^9,10 Because of the heavily entrenched societal stigma associated with these conditions, even clinical adjectives such as “schizophrenic” or “psychotic” are stigmatized. ¹¹

In 1992, the American Psychiatric Association (APA) led the movement to use person-centered language (PCL) in healthcare. ^12,13 PCL is a way of speaking or writing that emphasizes the individual and their experiences, rather than their medical diagnosis or label. ¹⁴ It is exemplified by the Well Being Campaign’s Mental Health Language Guide, which recommends that individuals avoid language such as “schizophrenic person” and instead use person-first phrases such as “person who has schizophrenia.” ¹⁵ Similarly, the Australian mental health institute, Everymind, recommends avoiding stigmatizing words such as “crazy,” “psychotic,” or “deranged,” when describing behavior that is unusual or erratic. ¹¹

The usage of PCL is essential for healthcare providers. A survey of physicians speaking about a “substance abuser,” compared to a person having substance use disorder, found that this stigmatizing phrasing was associated with providers feeling the person should be “punished” and that the person did not need treatment. ¹⁶ Stigmatizing language used by providers, such as using the term “sickler” for someone with sickle cell disease, is associated with negative attitudes toward the patient and lower adherence to pain management guidelines. ¹⁷ Continued prejudice against patients has been associated with social isolation, decreased treatment initiation, and reduced likelihood of staying in treatment. ¹⁸ Strict adherence to PCL in scientific and medical literature reinforces this necessity, which from the early APA movement led to the adoption of PCL guidelines in the Institute of Medicine in 2001; ¹³ these guidelines were later adopted within the American Medical Association Manual of Style (AMAMS) in 2007. ¹⁹ Generally, AMAMS has three guidelines for using PCL: 1) avoid labeling people with their disabilities, 2) avoid describing people as victims or using emotional terms that suggest helplessness, and 3) avoid euphemistic descriptions. ¹⁹

Although using PCL is the standard practice for many organizations, scientific associations, and governmental agencies, there are no published investigations on the regular adherence to these guidelines for conditions of psychosis within the medical literature. Therefore, this study’s primary objective was to explore PCL usage in articles focused on schizophrenia and other psychotic conditions. Secondarily, we investigated article characteristics that may be associated with the use of PCL that may elucidate reasons for PCL adherence.

METHODS

This study was conducted as meta-epidemiological research and followed reporting guidelines proposed by Murad and Wang. ²⁰

Journal Selection and Article Inclusion Criteria

A systematic search via PubMed was conducted on November 28th, 2023, to identify publications that focused on schizophrenia or psychosis (full search string in Supplemental Table 1) between January 1st, 2021, to December 31st, 2022. To determine which journals would be included in the search process, journals must have met the requirement of having at least 20 search returns that included psychotic conditions. Searches were further limited to articles published in English and on human subjects. Searches that met these criteria were then randomized, and the first 500 articles were selected for the screening process. Randomization was conducted in Stats v16.1 (StataCorp, LLC, College Station, TX). Publications were screened in a masked, duplicative fashion and were limited to the following article types: original research, research letters, brief reports, and case reports. As we were investigating articles that further medical evidence, commentaries and editorials were excluded from our study. After completion, conflicts regarding the interpretation of analysis results were resolved through discussion and revisiting the manuscripts until 100% interrater agreement was achieved.

Data Extraction

After article screening, data were extracted from articles to assess adherence to, or deviance from, two of the three guidelines presented in the AMAMS – the use of labels and emotional language. ¹⁹ Our criteria for labeling, in adherence to recommendations made by the AMAMS, included the use of an adjective before the individual or the use of a descriptor as the noun, in opposition to placing the person first, or person-first language. Labels assessed through our investigation included “schizophrenic,” “psychotic,” “mentally ill,” and “crazy” (full list of screened labels in Figure 2). If the article used shorthand for terms (eg, “SCZ” or “SZ” for “schizophrenia”), those terms were also screened and bucketed in our analysis under the full term that they represent. Furthermore, our investigation included the categorization of phrases utilizing emotive terminology evoking implications of an inherent flaw or weakness of a person or group. Our search included the following phrases: “afflicted with” and “suffers from.” Articles were only tagged as having “emotional [stigmatized] language” if they stated that the person “suffers from” psychosis or schizophrenia – or equivalents such as shorthand “SCZ” or “PS” – not if they “suffered from” another, unrelated medical symptom (eg, “suffered from heart disease,” even if the article was primarily about psychosis). These stigmatized words-and-phrases to avoid coincided with the recommendation, regarding both labeling and emotional language, from the Language Basics guide created by Well Beings in conjunction with APA, ¹⁵ as well as from the Australian mental health institute, Everymind. ¹¹ Articles’ full text, including titles and abstracts, were searched as they pertained to participants. References were not included in the search (eg, if the title of a referenced work used stigmatized phrasing). Other elements that were extracted from articles included the type of article, study method and interventions involved, institution of origin, funding source, and mention of adherence to reporting guidelines for their methodology.

Figure 2.

Prevalence of non-PCL terminology found in scientific publications focused on disorders of psychosis.

Data Analysis

First, we reported frequencies of the listed terms and phrases that did not adhere to PCL recommendations from the manuscripts, that is, from the total number of articles. Articles were considered to be PCL-adherent if no stigmatizing terms (labels or emotional language) were found and non-adherent if any such terms were present. We then calculated the ratio of PCL-adherent and non-adherent articles by the total number of articles overall. We reported the rates of adherence by the extracted article characteristics and conducted logistic regression analyses to determine if associations existed. Alpha was set at 0.05 and statistical analyses were conducted in Stata v16.1 (StataCorp, LLC., College Station, TX). The results reported in this study are in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines, and the methodology employed here is similar to that used in previous publications. ²¹

RESULTS

Our search returned 14,522 articles, of which 9,775 papers were from journals with more than 20 publications on the topic, across 127 journals. Of the 500 articles randomly selected for our analysis, 241 (48.2%) met our inclusion criteria – articles focused primarily on disorders of psychosis (Figure 1) – across 79 journals.

Figure 1.

Flowchart of the article selection process and the rate of stigmatized language within screened articles.

Of the 241 articles screened, 124 (53.1%) contained stigmatized language (48.1% labeling and 9.5% emotionally charged language; Figure 2). The most common stigmatized phrase was “schizophrenia patient,” found in 35.3% of articles, followed by “schizophrenic [person]” and “psychosis patient,” in 12.4% of articles each. Of the two emotive phrases screened, only “suffered from” was found – appearing in 9.5% of articles. A breakdown of stigmatized language found in papers by journal is available in Supplemental Table 2.

Our analyses showed a statistically significant relationship between an article’s adherence to PCL guidelines and the type of intervention (P = 0.009; Table 1). Within this category, 34.4% of studies with pharmacologic interventions, 38.7% with non-pharmacologic or surgical interventions, and 58.3% with no interventions (ie, observational) included stigmatizing language. The type of article, type of research, source of funding, and type of institution of the first and last author were not significantly associated with PCL guideline adherence.

Table 1.

Article Characteristics and Associations of Adherence to PCL Language

Article Characteristics	Articles With Non-PCL (128) No. (%)	Articles Adhering to PCL (113) No. (%)	Total (241) No. (%)	Fisher’s Exact
Type of Article
Original research, brief, letter	126	103	229	0.668
Case report	2	10	12
Type of Research
Systematic review/Meta-analysis/Literature review	20	11	31	0.265
Clinical trial	6	9	15
Observational/Cross-sectional	102	93	195
Type of Intervention
Drug/Pharmacologic treatment	7	16	23	0.009
Non-pharmacologic or surgical treatment (psychosocial/behavioral intervention)	12	19	31
No treatment of any kind (observational/cross-sectional study)	109	78	187
Article Funding
No funding	39	44	83	0.177
Funding received	89	69	158
First Author Employment
Government	8	9	17	0.374
Public	96	91	187
Private	23	13	36
Last Author Employment
Government	7	7	14	0.877
Public	98	86	184
Private	19	14	33

DISCUSSION

Our study showed that more than half of the articles that focused on psychosis in our sample were not aligned with PCL guidelines, with most utilizing an identity-first approach to labeling, such as “schizophrenic patient,” “psychosis patient,” or “schizophrenia patient.” In a 2013 survey by Anglin et al,²² people were significantly more likely to elicit a fear response when presented with a character vignette that used stigmatized labeling of people with psychosis (eg, “schizophrenic”). Interestingly, this study also found that survey participants were not likely to spontaneously assign labels if the vignette merely described symptoms of psychosis, and stigmatized responses were less likely to be elicited. ²² This result emphasizes the importance of avoiding labeling for conditions that are significantly stigmatized, such as psychosis. It is important to note that “person-centered language” describes the way a group of individuals prefers to be referenced and can be either “person-first” (eg, person with schizophrenia) or “identity-first” (eg, deaf person). Identity-first style is preferred among some disabilities, such as the Blind and Deaf communities. ^23,24 Little research has been done into the PCL preferences of people with psychosis; however, the aforementioned impact of labeling and stigmatizing phrases should encourage researchers to use person-first language, whereas providers work with their patients to determine individual preferences.

The type of intervention was the only statistically significant article characteristic for the adherence to PCL guidelines, with trials of pharmacologic or therapeutic interventions having the highest adherence to PCL guidelines and observational reviews having the lowest. This finding is consistent among other analyses of PCL in medical literature, such as with alcohol use disorder. ^21,25,26 Past research has shown that researchers are less likely to use stigmatized language if they personally interact with the people who are being researched, ²⁷ such as in clinical trials, in contrast to systematic reviews. Furthermore, a study conducted with healthcare providers in Nepal found that an educational program, which included contact with people with mental illness, significantly increased providers’ willingness to treat people with mental disorders. ²⁸ Pharmaceutical companies that fund clinical trials may also be encouraged to use PCL to market their product to government agencies and the general public, although more research is needed on this association.

Implications and Recommendations

A survey found that 98% of medical practitioners considered medical literature to be “important” or “very important” to their practice. ²⁹ When stigmatized language is used in medical literature, this vocabulary can become normalized for use in charts and verbally in the office. Studies have found that the use of stigmatized language in medical records can negatively impact providers’ attitudes toward those patients. ^17,30 Further, particularly with the advent of Open Notes, the use of non-PCL can damage the patient-provider relationship, ³¹ perpetuate self-stigma, and precipitate avoidance of treatment due to label-avoidance. ⁶ Stigmatizing phrasing has also been found to elicit negative attitudes among providers, decreased treatment recommendations for people with substance use disorder, ¹⁶ and decreased adherence to pain management guidelines in patients with sickle cell disorder. ¹⁷ Although the AMAMS has adopted PCL guidelines, ¹⁹ these data show that the majority of literature published on psychosis still contains stigmatized labels and emotional phrases. Therefore, we recommend that journal editors and reviewers screen for adherence to these guidelines and that stigmatized language found in submitted articles be changed. Suggested alternatives for common stigmatizing words and phrases can be found in Table 2.

Table 2.

Recommendations for Avoidance of Stigmatized Language in Discussions on Psychosis

Instead of this...	Use this...
Schizophrenic, schizophrenia patient	Patient with schizophrenia
Psychotic, psychosis patient	Patient with psychosis
Is bizarre, neurotic, odd	Use medically accurate symptomology
Suffer from psychosis/hallucinations/etc.	Has psychosis, experiences hallucinations, etc.

Limitations and Future Research

Screening of each manuscript for the use of stigmatizing language was conducted manually, which opened up the possibility of human errors; however, errors were minimized by conducting our procedure in a masked, duplicative fashion following best practices for meta-research. ²⁰ Additionally, our analysis did not track the number of stigmatized phrases (ie, one use of “schizophrenia patient” or 10), nor did we track the location of these stigmatized phrases (ie, results, discussion, etc). Our analysis found no significant relationship between article funding and adherence to PCL guidelines, although we did not differentiate sources of funding. Investigation into these data may elucidate patterns of non-PCL usage. Country of origin was not included in our analysis, although it may provide insights into the cultural bases of stigmatizing language. While societal stigma toward people with severe mental illness has been described globally, ³² research into PCL across cultures is limited. Research into the impact of stigmatized language in medical literature through qualitative interviews with providers or with individuals diagnosed with psychosis may be conducted to further investigate the impact of this stigma.

CONCLUSIONS

Our analysis of recent literature published on disorders of psychosis found that most of the screened articles contained stigmatized language, such as labeling or emotional terms in describing the people diagnosed with the conditions. Previous research has found associations between stigmatizing language and negative attitudes toward those individuals among healthcare providers. Therefore, journal editors and reviewers should screen submitted articles for stigmatized language and recommend adherence to AMAMS PCL guidelines, as this language is passed on through medical education and to the general public via media outlets.