Advancing Diagnostic Accuracy and Treatment With the Cultural Formulation Interview: Assessment of Psychosis in a Jamaican-American Woman

"Mrs. A" is a 45-year-old married, Pentecostal, Jamaican-born woman with a history of psychosis and anxiety. She presented to the emergency department in June 2018 with 4 months of progressive dysphagia, which she believed was related to her loxapine and lurasidone, which she had discontinued 4 months prior. After a nonrevealing medical workup and moderate improvement with 1 mg lorazepam, the psychiatry consult service was asked to evaluate Mrs. A and consider whether anxiety was causing her dysphagia.

Through interview with the patient and chart review, the consult service collected a thorough psychiatric and social history. Mrs. A was born in Jamaica, where she was raised by an emotionally abusive aunt before immigrating to the United States at age 12. Her mother was neglectful and struggled with alcohol and cannabis use, but there was no other known family history of psychiatric conditions. Mrs. A entered college to study nursing but withdrew as a sophomore when she became "disturbed by spirits." She moved in with family, worked various jobs, and eventually took courses to become a health aide. She worked and had three children before joining the United States military in the late 1990s; she was honorably discharged after 3 years of service.

In 2008, she suddenly started experiencing derogatory command hallucinations. She saw a psychiatrist for the first time and reported that she believed Lucifer was spiritually attacking her body. As a result, she was diagnosed as having schizophrenia and panic disorder. She was briefly given risperidone and then aripiprazole, but these were discontinued after less than 2 weeks due to decreased energy and sedation, respectively. In 2009, her diagnosis was changed to unspecified psychosis because hallucinations disappeared when she was not taking antipsychotic medications and symptoms did not appear to affect her daily functioning.

The patient remained asymptomatic until a visit with her mother and brother in 2017, when she experienced auditory hallucinations and the delusion that "the devil [is] speaking through me." She was initially given aripiprazole again for 2–3 weeks, but it was discontinued due to side effects and lack of efficacy. She was then started on loxapine up to 10 mg every night for several months. This low dose was also ineffective, and a 3-week cross-taper was started with lurasidone (up to 40 mg nightly). During the cross-taper in February 2018, she discontinued both medications due to "difficulty swallowing, drooling, muscle cramps, and pain in her shoulders and back." She ultimately presented to the emergency department in June 2018 for ongoing dysphagia, and the psychiatry team was consulted to evaluate the role of anxiety in her presentation.

Upon initial interview by the psychiatry consult team, Mrs. A was organized and insightful, providing a plausible cultural context for some symptoms. She endorsed daily auditory hallucinations with no change in pattern while off medications over the past 4 months, variable sleep, and anxiety due to difficulty swallowing. She denied all other psychiatric symptoms, including depression, mania, visual hallucinations, or nightmares. She also denied substance use. To gain a more complete cultural context, the CFI was employed. During the CFI, the team asked several questions and elicited notable cultural themes that ultimately altered the patient's care (Table 1). For example, in response to "What have you done on your own to cope with your problem?" she responded, "Both religion and the medical field; I pray daily, and pastors at church are aware of my condition."

For her dysphagia, esophagogastroduodenoscopy, barium esophagram, bedside scope, and general labs were unrevealing. Her dysphagia improved with a low-dose benzodiazepine, indicating that anxiety may have contributed to her symptoms. Because the dysphagia persisted several months after she discontinued the antipsychotics, these medications were not thought to be the cause of her dysphagia. The primary team started omeprazole for a discharge diagnosis of laryngopharyngeal reflux. After the nonrevealing medical workup, the differential diagnoses for Mrs. A's psychotic symptoms included delusional disorder, schizophrenia, and posttraumatic stress disorder (PTSD).

Although she met some criteria for schizophrenia, her level of functioning for most of her course was not significantly affected. In addition, it would be unusual for hallucinations to disappear when a patient was off antipsychotic medications so early in the disease course. We considered a diagnosis of PTSD, especially as her psychotic symptoms recurred after a visit with her mother, which could have been traumatic given her mother's history of neglect. However, during the CFI she did not identify her mother as a trigger and instead reported that her symptoms were more broadly due to her cultural beliefs.

From the CFI, we learned that Mrs. A viewed her symptoms as a possible manifestation of past conflicts. The symptoms responded to prayer, which she had effectively used in the past. Although her Jamaican heritage and Pentecostal beliefs were contributing to her presentation, we determined that her symptoms were unusual and caused some degree of distress. As such, we felt that the diagnosis of delusional disorder was most appropriate. We discussed the benefit of an adequate dose of an antipsychotic for a sufficient trial. However, the "demonic possession" was short lived, not significantly distressing, and likely delusional in nature and had an uncertain response to medications. Thus we respected her wish to abstain from restarting an antipsychotic medication. She did not request chaplain services in the hospital but planned to continue working with the pastors at her church. A recent chart review showed that her dysphagia had resolved, and she had resumed work.

Overall, the CFI helped facilitate a dialogue that ultimately influenced Mrs. A's diagnosis and treatment. Although she was initially diagnosed as having schizophrenia, she had not been made aware of this until recently gaining access to her medical record. Using the CFI, we discussed her perceptions of her illness more fully. Her insight into the etiology of her distress and her primary use of prayer as a form of healing helped us appreciate that her condition may not improve with antipsychotics, and thus we altered her care. Furthermore, use of the CFI ultimately spared her potential lifelong exposure to adverse effects of antipsychotics. We suspect her dysphagia improved due to a combination of treatment with omeprazole and reduction in her anxiety through support from her treatment team and faith community.

Notably, as a black person, Mrs. A was more likely than a non-Hispanic white person to receive a diagnosis of schizophrenia (5). It has been found that black individuals are diagnosed as having psychotic disorders at a rate of three to four times higher than the rate among white individuals. Given that misinterpretation of symptomatology and clinician racial bias may play a role in this trend, caution should be used when making inferences about psychosis among individuals from ethnic and racial minority groups (6).

The CFI can help reveal multiple facets of an individual's diversity. For example, although Mrs. A identified as Pentecostal, it is not necessarily representative of the Jamaican population, which consists of diverse groups and of which only 11% identify as Pentecostal (7). This supports the use of the CFI to identify multiple potential cultural influences in a patient.

Multiple studies have found the CFI to be useful for patients and clinicians in practice (2–4). In one study of 30 monolingual Spanish-speaking patients, use of the CFI improved trust among patients and clinicians (2). To address the concern of practicality in a busy clinical setting, a study of 75 clinicians found that their ratings of the CFI's feasibility improved significantly after administering it just once (3). Another concern is that certain CFI questions are limited by their wording and do not directly address significant cultural domains, such as gender or education (4). To mitigate these concerns, in our interview we asked CFI questions by following the patient's lead and not in the order they appear in the DSM. We also minimally rephrased some questions to match the flow of the interview.

Further investigation is needed on how to best utilize the CFI. It may need to be adjusted depending on time limitations and cannot be used as a stand-alone assessment during an initial consult intake. In addition, one must be cautious to avoid generalizing the experience of one patient to a whole culture. Despite limitations, the CFI creates the opportunity to better understand a patient's experience of distress and can help provide treatment consistent with the patient's values.

The authors thank Grace Chang, M.D., Professor of Psychiatry, Harvard Medical School, for her counsel.

The authors have confirmed that details of the case have been disguised to protect patient privacy.