Artificial intelligence (AI) and virtual reality (VR) are useful tools that can improve precision medicine and can prove useful in the clinical care of patients with psychosis.

Our aim was to determine whether AI and VR have been applied to the prediction of clinical response in women with schizophrenia.

A systematic review was carried out in PubMed and Scopus from inception to September 2023 by using the PRISMA guidelines. Search terms: (“artificial intelligence” OR “intelligent support” OR “machine intelligence” OR “machine learning” OR “virtual reality” OR “intelligent agent” OR “neural networks” OR “virtual reality” OR “digital twins”) AND (“schizophrenia” OR “psychosis”) AND (“women” OR gender”). Inclusion criteria: 1)English, French, German or Spanish language, 2) reporting treatment response in schizophrenia (as long as information in women was included), and 3) including AI and VR techniques.

From a total of 320 abstracts initially screened (PubMed:182, Scopus:138), we selected 6 studies that met criteria.

• - Prediction of treatment response. (1) Clinical information, genetic risk score and proxy methylation score have been shown to improve prediction models. (2) Graph-theory-based measures have been combined with machine learning.

• - Therapeutic drug monitoring. (1) A machine learning model has been useful in predicting quetiapine blood concentrations.

• - Pharmacovigilance. (1) Machine learning has connected prolactin levels and response in olanzapine-treated patients. (Zhu et al., 2022).

• - Treatment-resistant schizophrenia (TRS). (1) Women with TRS have been found to receive clozapine less frequently than men (adjusted for sociodemographic, biological and clinical factors). (2) Statistical learning approach: Women have been found to respond better to clozapine than men.

AI, including machine learning, show promising results in the prediction of treatment response in women with schizophrenia. As of yet, digital twins have not been investigated to test specific interventions or to personalize treatment in women with schizophrenia.

None Declared

There are many theoretical reasons to implement gender-specific care for schizophrenia. For all these reasons, the Mutua Terrassa-Functional Unit for Women with Schizophrenia was inaugurated in January 2023 in the context of a community mental health service.

Our aim today is to describe the health care model applied in this newly initiated unit.

We created a healthcare model in our new unit consisting of A)Five observatories of Health (somatic morbi-mortality, hyperprolactinemia-HPRL, substance use disorders, social exclusion/discrimination, and drug safety); B)Monitoring stations or vigilance teams (reflecting the 5 observatories); and C)resulting actions (specific interventions). The observatory teams each meet monthly. In this presentation, according to the healthcare model we implemented, we first describe data about the original patient recruitment and then focus on the observatories of somatic morbi-mortality and hyperprolactinemia.

From 265 potentially eligible women, 42 were included in the 5 observatories. (A) of the 11 women in the observatory of somatic morbi-mortality, 10 women had died within the last 24 months. Causes of Death: (1)respiratory tract disease (n=5,45.4%), (2)cancer (n=3;27.3%): lung cancer (n=1), pancreatic cancer (n=1), kidney cancer (n=1), (3)ischemic colitis (n=1;9%), (4)Alzheimer disease (n=1;9%). 2) Morbidity. One woman had an ongoing glioblastoma. (B)Observatory of HPRL. Eight women with moderate/severe HPRL were included. Strategies for lowering prolactin levels were discussed with neuroendocrinologists. Interventions:adjunctive aripiprazole (n=3), switch to aripiprazole (n=2), lowering antipsychotic doses (n=2), and adjunctive cabergoline (n=1).

Designating special teams to focus on specific problems of women with schizophrenia will reduce morbidity and improve outcomes in this vulnerable population.

None Declared

The Cordoba Naming Test (CNT) is a 30-item lexical retrieval task that was developed to be administered in multiple languages. Research shows that self-identifying Mexicans residing in Mexico outperform self-identifying Mexicans that reside in the United States on the CNT. Studies indicate that the process of acculturation can influence cognitive performance. Previous studies demonstrated that Generation Z individuals (i.e., people born between 1997 and 2012) have underperformed on the CNT compared to Generation Y individuals (i.e., people born between 1981 and 1996). To our knowledge, no study has examined the influence of acculturation on Generation Z Mexicans’ CNT performance. We expected Mexicans residing in Mexico (MRM) to outperform Mexicans residing in the United States on the CNT and to report higher acculturation traits. We also predicted that acculturation would correlate with CNT performance.

The present study sample consisted of 285 Generation Z psychologically and neurologically healthy Mexicans with a mean age of 20.32 (SD = 1.60). Participants were divided into three groups: MRM, Mexicans residing in the United States, and Mexican-Americans residing in the United States (MARUS). All participants completed the CNT and acculturation measure in Spanish. Acculturation traits were measured by the Abbreviated Multidimensional Acculturation Scale (AMAS). ANCOVAs were used to evaluate differences in the CNT and AMAS (i.e., Spanish language, Latino competency, Latino identity). Pearson’s correlation coefficient was used to evaluate the relationship between acculturation on CNT performance.

MRM outperformed the Mexicans residing in the United States and MARUS on the CNT, p = .000, np2 = .49. The MRM group reported better Spanish language abilities compared to Mexicans residing in the United States and the MARUS groups, p = .000, np2 = .10. Additionally, MRM reported better Latino competency than the MARUS group, p = .000, np2 = .08. Pearson’s correlation coefficient revealed that the MARUS’s Spanish language abilities impacted CNT performance, p = .000, r = .48. In addition, we found that Latino competency correlated with CNT performance, p’s < .05, r’s = .20-.47, in both the MRM and Mexicans residing in the United States groups. Latino identity did not significantly correlate with CNT performance in any group.

Results confirmed that MRM individuals perform better on the CNT than Mexicans residing in the United States and MARUS. Additionally, we found that several acculturation traits correlated with Mexican groups’ CNT performance. Our research indicates that while all Generation Z individuals of Mexican heritage feel strongly connected to their Latino identity regardless of where they live, MARUS feel less competent in Spanish and Latinx culture than MRM and Mexicans residing in the United States. Future work should further explore these differences for better insight into how acculturative factors influence lexical retrieval performance. Future work with bigger sample sizes can additionally examine CNT performance and acculturation in Generation Z first-generation and non-first-generation Mexicans (e.g., second-generation, third-generation) residing in the United States.

A 30-item confrontation naming test was developed in Argentina for Spanish speakers, The Cordoba Naming Test (CNT). The Boston Naming Test is an established confrontation naming task in the United States. Researchers have used the Boston Naming Test to identify individuals with different clinical pathologies (e.g., Alzheimer’s disease). The current literature on how Spanish speakers across various countries perform on confrontational naming tasks is limited. To our knowledge, one study investigated CNT performance across three Spanish-speaking countries (i.e., Argentina, Mexico, and Guatemala). Investigators found that the Guatemalan group underperformed on the CNT compared to the Argentine and Mexican groups. The purpose of this study was to extend the current literature and investigate CNT performance across five Spanish-speaking countries (i.e., Argentina, Mexico, Guatemala, Colombia, United States). We predicted that the Argentine group would outperform the other Spanish-speaking countries.

The present study sample consisted of 502 neurologically and psychologically healthy participants with a mean age of 29.06 (SD = 13.41) with 14.75 years of education completed (SD = 3.01). Participants were divided into five different groups based on their country of birth and current country residency (i.e., United States, Mexico, Guatemala, Argentina, & Colombia). All participants consented to voluntary participation and completed the CNT and a comprehensive background questionnaire in Spanish. The CNT consisted of 30 black and white line drawings, ranging from easy to hard in difficulty. An ANCOVA, controlling for gender, education, and age, was used to evaluate CNT performance between the five Spanish-speaking country groups. Meanwhile, a Bonferroni post-hoc test was utilized to evaluate the significant differences between Spanish-speaking groups. We used a threshold of p < .05 for statistical significance.

Results revealed significant group differences between the five Spanish speaking groups on the CNT, p = .000, np2 = .48. Bonferroni post-hoc test revealed that the United States group significantly underperformed on the CNT compared to all the Spanish-speaking groups. Next, we found the Guatemalan group underperformed on the CNT compared to the Argentinian, Mexican, and Colombian groups. Additionally, we found the Argentinian group outperformed the Mexican, Guatemalan, and United States groups on the CNT. No significant differences were found between the Argentinian group and Colombian group or the Mexican group and Colombian group on the CNT.

As predicted, the Argentinian group outperformed all the Spanish-speaking groups on the CNT except the Colombian group. Additionally, we found that the United States group underperformed on the CNT compared to all the Spanish-speaking groups. A possible explanation is that Spanish is not the official language in the United States compared to the rest of the Spanish-speaking groups. Meanwhile, a possible reason why the Argentinian and Colombian groups demonstrated better CNT performances might have been that it was less culturally sensitive than the United States, Mexican, and Guatemalan groups. Further analysis is needed with bigger sample sizes across other Spanish-speaking countries (e.g., Costa Rica, Chile) to evaluate what variables, if any, are influencing CNT performance.