Xiaoming Zeng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Xiaoming Zeng has a total of 134 co-authors. Here are the top co-authors by number of shared publications:
- Nathaniel A. Sowa
University of North Carolina at Chapel Hill School of Medicine
7 shared publications Bambang Parmanto
University of Pittsburgh , Pittsburgh , PA , 15260 , USA Phone: + 1 412-383-6650
4 shared publications- Mei-Ying Ren
Wuyi University
4 shared publications Susie T Harris
Health Services and Information Management, East Carolina University, USA
4 shared publications- Wendy Webber Chapman
Centre for Digital Transformation of Health, University of Melbourne, Carlton, Australia
3 shared publications - Bambang Parmanto
Department of Health Information Management, University of Pittsburgh, Pittsburgh, PA 15260, USA
2 shared publications
Latest Publications
- Open Access
- 23 October 2026
- Published by JMIR Publications Inc. in JMIR Preprints
GenAI chatbots in mental health support: A thematic analysis of Reddit discussions (Preprint)
BACKGROUND Mental illness affects people worldwide and poses a significant burden to individuals and society. Digital mental health support has become an important strategy to address barriers in traditional mental health care. Advances in artificial intelligence (AI) have further expanded digital outreach. Particularly, people increasingly use large language model (LLM)-powered generative AI (GenAI) chatbots such as ChatGPT to process thoughts, emotions, and personal challenges. While most prior work focused on AI chatbots intentionally designed and clinically evaluated as digital therapeutics, less is known about general-purpose GenAI chatbots that people repurpose to support their routine mental health practices. OBJECTIVE This study aims to examine user-generated discussions shared on Reddit and investigate user experiences, positioning, and sentiment toward general-purpose GenAI chatbots through first-person narratives in naturalistic settings. METHODS We retrieved publicly available Reddit posts from four selected subreddits from 2023 to 2025, and cleaned the raw dataset using Python to remove duplicates, separate posts from comments, and convert the post file into a tabular format. The cleaned dataset was filtered using two predefined keyword dictionaries (i.e., GenAI terms and mental health terms) and further screened against inclusion and exclusion criteria. We followed the six phases of reflexive thematic analysis (RTA) to analyze sampled Reddit posts. Reddit users’ GenAI positioning and sentiment were coded and summarized descriptively. RESULTS A total of 169 Reddit posts across four subreddits met the inclusion criteria and were included for RTA. Three primary themes emerged: 1) GenAI chatbots as an emotional support system (39.6%), 2) a self-understanding tool (36.1%), and 3) limited or risky support (24.3%). Across 141 posts with sufficient contextual information for positioning coding, GenAI chatbots were primarily positioned as a supplementary tool (55.4%), although a notable portion of users relied on them as a primary (33.1%) or alternative source of support (11.5%), particularly during periods of distress or when traditional care was inaccessible. Overall, Reddit user sentiment toward using GenAI chatbots to support mental health leaned positive (38.5%) and mixed (32%). CONCLUSIONS Our findings suggest that general-purpose GenAI chatbots are most often used as a supplementary tool, helping people work through thoughts, manage emotions, and support existing coping strategies. The positioning framework identification for understanding GenAI repurposing for mental health support highlights how users actively and dynamically define or redefine its role within their mental health support systems. As GenAI use continues to grow, this understanding will help inform future research, technology design, clinical care, and even health policy.
- 23 October 2026
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessGenAI chatbots in mental health support: A thematic analysis of Reddit discussions (Preprint)
BACKGROUND Mental illness affects people worldwide and poses a significant burden to individuals and society. Digital mental health support has become an important strategy to address barriers in traditional mental health care. Advances in artificial intelligence (AI) have further expanded digital outreach. Particularly, people increasingly use large language model (LLM)-powered generative AI (GenAI) chatbots such as ChatGPT to process thoughts, emotions, and personal challenges. While most prior work focused on AI chatbots intentionally designed and clinically evaluated as digital therapeutics, less is known about general-purpose GenAI chatbots that people repurpose to support their routine mental health practices. OBJECTIVE This study aims to examine user-generated discussions shared on Reddit and investigate user experiences, positioning, and sentiment toward general-purpose GenAI chatbots through first-person narratives in naturalistic settings. METHODS We retrieved publicly available Reddit posts from four selected subreddits from 2023 to 2025, and cleaned the raw dataset using Python to remove duplicates, separate posts from comments, and convert the post file into a tabular format. The cleaned dataset was filtered using two predefined keyword dictionaries (i.e., GenAI terms and mental health terms) and further screened against inclusion and exclusion criteria. We followed the six phases of reflexive thematic analysis (RTA) to analyze sampled Reddit posts. Reddit users’ GenAI positioning and sentiment were coded and summarized descriptively. RESULTS A total of 169 Reddit posts across four subreddits met the inclusion criteria and were included for RTA. Three primary themes emerged: 1) GenAI chatbots as an emotional support system (39.6%), 2) a self-understanding tool (36.1%), and 3) limited or risky support (24.3%). Across 141 posts with sufficient contextual information for positioning coding, GenAI chatbots were primarily positioned as a supplementary tool (55.4%), although a notable portion of users relied on them as a primary (33.1%) or alternative source of support (11.5%), particularly during periods of distress or when traditional care was inaccessible. Overall, Reddit user sentiment toward using GenAI chatbots to support mental health leaned positive (38.5%) and mixed (32%). CONCLUSIONS Our findings suggest that general-purpose GenAI chatbots are most often used as a supplementary tool, helping people work through thoughts, manage emotions, and support existing coping strategies. The positioning framework identification for understanding GenAI repurposing for mental health support highlights how users actively and dynamically define or redefine its role within their mental health support systems. As GenAI use continues to grow, this understanding will help inform future research, technology design, clinical care, and even health policy.
- 31 August 2026
- Published by Wolters Kluwer Health in Pediatric Emergency Care
Factors Associated With Emergency Department Length of Stay for Pediatric Patients Evaluated by an Academic Telepsychiatry Consultation Service
Objectives: Emergency department (ED) boarding for youth in mental health crisis has increased nationally, straining capacity and delaying definitive care. Telepsychiatry may reduce ED length of stay (LOS), but limited data exist to describe factors associated with ED LOS for pediatric populations receiving telepsychiatry consultation. Methods: This retrospective study analyzed telepsychiatry consultations on patients aged below 18 years across 11 hospitals in an academic health system (July 2022 to July 2024). Demographic, clinical, and encounter-level variables were extracted from the electronic health record. ED LOS was defined as the time from ED arrival to ED discharge. Linear regression was used to evaluate independent predictors of ED LOS. A subanalysis compared the 50 longest LOS encounters with 50 controls randomly selected from the lowest 2 quartiles. Results: Among 1187 consultations, median ED LOS was 34.9 hours (IQR=14.8 to 68.5). Older age ( P =0.003) and public insurance ( P =0.01) were associated with longer LOS, while female sex was associated with shorter LOS ( P =0.007). Regression analyses identified diagnosis category as the strongest independent predictor of LOS. Extreme outlier cases (median LOS=666.5 h) differed significantly from controls in age, diagnosis, disposition, violence history, legal barriers, and antipsychotic use. Conclusions: Pediatric patients evaluated by telepsychiatry experienced prolonged ED LOS, with demographic, diagnostic, and social factors contributing to delays. Extreme outlier cases highlight the multifactorial drivers of extreme boarding, including aggression, legal constraints, and custodial issues. Early identification of high-risk patients may support targeted intensive case management and interagency coordination to reduce ED LOS.
- 31 August 2026
- Published by Wolters Kluwer Health in Pediatric Emergency Care
Factors Associated With Emergency Department Length of Stay for Pediatric Patients Evaluated by an Academic Telepsychiatry Consultation Service
Objectives: Emergency department (ED) boarding for youth in mental health crisis has increased nationally, straining capacity and delaying definitive care. Telepsychiatry may reduce ED length of stay (LOS), but limited data exist to describe factors associated with ED LOS for pediatric populations receiving telepsychiatry consultation. Methods: This retrospective study analyzed telepsychiatry consultations on patients aged below 18 years across 11 hospitals in an academic health system (July 2022 to July 2024). Demographic, clinical, and encounter-level variables were extracted from the electronic health record. ED LOS was defined as the time from ED arrival to ED discharge. Linear regression was used to evaluate independent predictors of ED LOS. A subanalysis compared the 50 longest LOS encounters with 50 controls randomly selected from the lowest 2 quartiles. Results: Among 1187 consultations, median ED LOS was 34.9 hours (IQR=14.8 to 68.5). Older age ( P =0.003) and public insurance ( P =0.01) were associated with longer LOS, while female sex was associated with shorter LOS ( P =0.007). Regression analyses identified diagnosis category as the strongest independent predictor of LOS. Extreme outlier cases (median LOS=666.5 h) differed significantly from controls in age, diagnosis, disposition, violence history, legal barriers, and antipsychotic use. Conclusions: Pediatric patients evaluated by telepsychiatry experienced prolonged ED LOS, with demographic, diagnostic, and social factors contributing to delays. Extreme outlier cases highlight the multifactorial drivers of extreme boarding, including aggression, legal constraints, and custodial issues. Early identification of high-risk patients may support targeted intensive case management and interagency coordination to reduce ED LOS.
- 23 December 2025
- Published by Springer Nature in Community Mental Health Journal
- Vol. 62 (5) , 832-837
Home Blood Pressure Monitoring in a Community Mental Health Clinic with Integrated Primary Care: A Quality Improvement Project
- 23 December 2025
- Published by Springer Nature in Community Mental Health Journal
- Vol. 62 (5) , 832-837
Home Blood Pressure Monitoring in a Community Mental Health Clinic with Integrated Primary Care: A Quality Improvement Project
- 24 October 2025
- Published by Springer Nature in Administration and Policy in Mental Health and Mental Health Services Research
- Vol. 53 (1) , 38-46
Psychiatry Consultation Clinic: A Model that Improves Access to Care and Corresponds to Decreased Anxiety and Depressive Symptoms
Access to outpatient psychiatric care is often delayed, with many patients unable to obtain timely specialty evaluation. Integrated care models such as Collaborative Care Management (CoCM) and co-located care improve outcomes but have not been widely adopted due to resource and structural barriers. Direct outpatient psychiatric consultation may offer a feasible alternative, yet its effectiveness remains understudied. We conducted a retrospective review of the Psychiatry Consultation Clinic (PCC) at a large academic medical center. The PCC provides diagnostic clarification and treatment recommendations to primary care providers (PCPs) rather than longitudinal psychiatric care. Adult patients (≥18 years) referred between October 2019 and October 2022 from family and internal medicine practices were included (N = 545). Primary outcomes were: (1) time to implementation of consult recommendations; (2) rate of psychotropic medication implementation; and (3) changes in anxiety (Gretrospective review of the Psychiatry Consultation AD-7) and depression (PHQ-9) scores. Secondary outcomes included psychiatric diagnoses and subsequent referrals for specialty care. The median time from referral to consultation was 5 weeks, substantially shorter than the average 6-month wait for traditional psychiatry appointments within the same system. Most patients (88.1%) were seen once, and 83.1% of visits were conducted via telemedicine. Pharmacotherapy was recommended for 81.8% of patients, and 81.4% had at least one recommendation implemented, typically within 1–3 weeks. Antidepressants accounted for two-thirds of recommendations. For patients with pre- and post-consultation data, mean GAD-7 and PHQ-9 scores decreased significantly (−2.9 and −5.0 points, respectively; both p < 0.0001), with clinically meaningful improvements observed in depression symptoms. Nearly one-quarter of patients were referred for longitudinal specialty psychiatric care, with 70% establishing follow-up within six months. The direct consultation model was feasible, resulted in high uptake of recommendations by PCPs, and was associated with improvements in patient symptoms, particularly depression. Compared with integrated models, this approach may be easier to implement in primary care settings with limited behavioral health infrastructure. Prospective controlled studies are warranted to establish the efficacy of this outpatient consultation model and compare it to other paradigms within the continuum of psychiatric care.
- 24 October 2025
- Published by Springer Nature in Administration and Policy in Mental Health and Mental Health Services Research
- Vol. 53 (1) , 38-46
Psychiatry Consultation Clinic: A Model that Improves Access to Care and Corresponds to Decreased Anxiety and Depressive Symptoms
Access to outpatient psychiatric care is often delayed, with many patients unable to obtain timely specialty evaluation. Integrated care models such as Collaborative Care Management (CoCM) and co-located care improve outcomes but have not been widely adopted due to resource and structural barriers. Direct outpatient psychiatric consultation may offer a feasible alternative, yet its effectiveness remains understudied. We conducted a retrospective review of the Psychiatry Consultation Clinic (PCC) at a large academic medical center. The PCC provides diagnostic clarification and treatment recommendations to primary care providers (PCPs) rather than longitudinal psychiatric care. Adult patients (≥18 years) referred between October 2019 and October 2022 from family and internal medicine practices were included (N = 545). Primary outcomes were: (1) time to implementation of consult recommendations; (2) rate of psychotropic medication implementation; and (3) changes in anxiety (Gretrospective review of the Psychiatry Consultation AD-7) and depression (PHQ-9) scores. Secondary outcomes included psychiatric diagnoses and subsequent referrals for specialty care. The median time from referral to consultation was 5 weeks, substantially shorter than the average 6-month wait for traditional psychiatry appointments within the same system. Most patients (88.1%) were seen once, and 83.1% of visits were conducted via telemedicine. Pharmacotherapy was recommended for 81.8% of patients, and 81.4% had at least one recommendation implemented, typically within 1–3 weeks. Antidepressants accounted for two-thirds of recommendations. For patients with pre- and post-consultation data, mean GAD-7 and PHQ-9 scores decreased significantly (−2.9 and −5.0 points, respectively; both p < 0.0001), with clinically meaningful improvements observed in depression symptoms. Nearly one-quarter of patients were referred for longitudinal specialty psychiatric care, with 70% establishing follow-up within six months. The direct consultation model was feasible, resulted in high uptake of recommendations by PCPs, and was associated with improvements in patient symptoms, particularly depression. Compared with integrated models, this approach may be easier to implement in primary care settings with limited behavioral health infrastructure. Prospective controlled studies are warranted to establish the efficacy of this outpatient consultation model and compare it to other paradigms within the continuum of psychiatric care.
- 19 March 2025
- Published by Springer Nature in Psychiatric Quarterly
- Vol. 96 (3) , 593-606
Factors Associated with Leaving Ambulatory Psychiatric Treatment in a Large, Academic Health System During the COVID-19 Pandemic
Detailed evaluation of changes in patient retention in psychiatric care as a result of the onset of the COVID-19 pandemic have not been studied. Here, we present a retrospective analysis of aggregate data from a large academic health system (n = 16,701 patients) to examine if there were differences in patients leaving psychiatric care from the pre-COVID-19 period to the first 12 months of the COVID-19 pandemic. Demographic, clinical, and behavioral factors were studied using logistic regressions to determine significant associations in leaving psychiatric care during COVID-19 and in the 12 months immediately prior to the pandemic. Factors identified with a higher odds of leaving psychiatric care during COVID-19 that were not associated with leaving care prior to COVID-19 included demographic (male sex, uninsured (self-pay) status), behavioral (inactive patient health portal), and diagnostic (anxiety and trauma stressor disorders, pervasive and specific developmental disorders, and disorders of childhood) factors. These results highlight that the reasons patients left care during the period immediately after the COVID-19 pandemic may have been multifactorial in nature, although certain patterns seem to have appeared. Further study is needed to elucidate why these specific factors may have driven patients to leave psychiatric treatment during the COVID-19 pandemic.
- 19 March 2025
- Published by Springer Nature in Psychiatric Quarterly
- Vol. 96 (3) , 593-606
Factors Associated with Leaving Ambulatory Psychiatric Treatment in a Large, Academic Health System During the COVID-19 Pandemic
Detailed evaluation of changes in patient retention in psychiatric care as a result of the onset of the COVID-19 pandemic have not been studied. Here, we present a retrospective analysis of aggregate data from a large academic health system (n = 16,701 patients) to examine if there were differences in patients leaving psychiatric care from the pre-COVID-19 period to the first 12 months of the COVID-19 pandemic. Demographic, clinical, and behavioral factors were studied using logistic regressions to determine significant associations in leaving psychiatric care during COVID-19 and in the 12 months immediately prior to the pandemic. Factors identified with a higher odds of leaving psychiatric care during COVID-19 that were not associated with leaving care prior to COVID-19 included demographic (male sex, uninsured (self-pay) status), behavioral (inactive patient health portal), and diagnostic (anxiety and trauma stressor disorders, pervasive and specific developmental disorders, and disorders of childhood) factors. These results highlight that the reasons patients left care during the period immediately after the COVID-19 pandemic may have been multifactorial in nature, although certain patterns seem to have appeared. Further study is needed to elucidate why these specific factors may have driven patients to leave psychiatric treatment during the COVID-19 pandemic.
- 26 October 2023
- Published by SAGE Publications in Telemedicine and e-Health
- Vol. 30 (4) , e1034-e1048
Factors Associated with Reliance on Audio-Only Technology (Telephone) for Completion of Telepsychiatry Visits
Introduction: The ability to access telepsychiatry through audio–video technology versus audio-only (telephone) technology potentially leads to inequitable outcomes. This study examines the characteristics of patients who relied on the telephone to complete outpatient telepsychiatry visits in a large health system. Methods: This is a retrospective analysis of all telepsychiatry outpatient visits conducted from May 1, 2020, to December 31, 2021. Demographic, clinical, and socioeconomic factors were extracted from the electronic health record. Two-sample t tests were used for continuous variables and χ 2 tests for categorical variables for bivariate analyses. Multiple logistic regression was used to examine the association between only telephone visits and all input variables. Results: Eight hundred ninety-four (8.9%) patients completed all visits only by telephone during the study period. In bivariate analyses, factors associated with telephone-only visits included male sex, non-English primary language, Black race, unmarried status, non-Hispanic ethnicity, older age, Medicare enrollment, uninsured status, and higher social vulnerability index (SVI). Psychiatric diagnoses associated with only telephone visits included substance use disorders and psychotic disorders. In multivariate analyses, factors associated with higher odds of only telephone visits included older age, inactive patient portal, comorbid diabetes, higher SVI, and higher broadband adoption. Psychiatric diagnoses associated with higher odds of completing only telephone visits included psychotic disorders, substance use disorders, or intellectual/developmental disabilities. Conclusions: Several patient-level factors are associated with a reliance on the telephone to complete telepsychiatry appointments.
- 26 October 2023
- Published by SAGE Publications in Telemedicine and e-Health
- Vol. 30 (4) , e1034-e1048
Factors Associated with Reliance on Audio-Only Technology (Telephone) for Completion of Telepsychiatry Visits
Introduction: The ability to access telepsychiatry through audio–video technology versus audio-only (telephone) technology potentially leads to inequitable outcomes. This study examines the characteristics of patients who relied on the telephone to complete outpatient telepsychiatry visits in a large health system. Methods: This is a retrospective analysis of all telepsychiatry outpatient visits conducted from May 1, 2020, to December 31, 2021. Demographic, clinical, and socioeconomic factors were extracted from the electronic health record. Two-sample t tests were used for continuous variables and χ 2 tests for categorical variables for bivariate analyses. Multiple logistic regression was used to examine the association between only telephone visits and all input variables. Results: Eight hundred ninety-four (8.9%) patients completed all visits only by telephone during the study period. In bivariate analyses, factors associated with telephone-only visits included male sex, non-English primary language, Black race, unmarried status, non-Hispanic ethnicity, older age, Medicare enrollment, uninsured status, and higher social vulnerability index (SVI). Psychiatric diagnoses associated with only telephone visits included substance use disorders and psychotic disorders. In multivariate analyses, factors associated with higher odds of only telephone visits included older age, inactive patient portal, comorbid diabetes, higher SVI, and higher broadband adoption. Psychiatric diagnoses associated with higher odds of completing only telephone visits included psychotic disorders, substance use disorders, or intellectual/developmental disabilities. Conclusions: Several patient-level factors are associated with a reliance on the telephone to complete telepsychiatry appointments.
- 23 May 2023
- Published by Springer Nature in Psychiatric Quarterly
- Vol. 94 (2) , 255-263
A Comprehensive Examination of Pediatric Behavioral Health Service Demand and Utilization in a Large, Academic Health System from 2019 to 2021
Studies of the effects of COVID-19 on youth suggest a worsening in mental health globally. We performed a retrospective analysis of data from January 2019-November 2021 for all outpatient referrals, as well as outpatient, inpatient, and emergency department (ED) encounters for behavioral health (BH) reasons in children aged < 18 in a large academic health system in the United States. Mean weekly rates of outpatient psychiatry referrals, outpatient psychiatry visits, ED visits, and inpatient admissions for BH reasons were compared between pre-pandemic and pandemic periods. The average weekly rate of ambulatory referrals (8.0 ± 0.33 to 9.4 ± 0.31) and completed appointments (194.2 ± 0.72 to 213.1 ± 0.71) significantly increased during the pandemic, driven largely by teenagers. The weekly average of ED pediatric encounters for BH did not increase during the pandemic, although the percentage of all pediatric ED encounters that were for BH did increase from 2.6 to 4.1% (p < 0.001). Length of stay for pediatric BH ED patients increased from 1.59 ± 0.09 days pre-pandemic to 1.91 ± 0.11 days post-pandemic (p < 0.0001). Inpatient admissions for BH reasons overall decreased during the pandemic, due to a decrease in inpatient psychiatric bed capacity. However, the weekly percentage of inpatient hospitalizations for BH reasons that occurred on medical units increased during the pandemic (15.2% ± 2.8–24.6% ± 4.1% (p = 0.0006)). Taken together, our data suggest the COVID-19 pandemic had varying degrees of impact, depending on the setting of care.
- 23 May 2023
- Published by Springer Nature in Psychiatric Quarterly
- Vol. 94 (2) , 255-263
A Comprehensive Examination of Pediatric Behavioral Health Service Demand and Utilization in a Large, Academic Health System from 2019 to 2021
Studies of the effects of COVID-19 on youth suggest a worsening in mental health globally. We performed a retrospective analysis of data from January 2019-November 2021 for all outpatient referrals, as well as outpatient, inpatient, and emergency department (ED) encounters for behavioral health (BH) reasons in children aged < 18 in a large academic health system in the United States. Mean weekly rates of outpatient psychiatry referrals, outpatient psychiatry visits, ED visits, and inpatient admissions for BH reasons were compared between pre-pandemic and pandemic periods. The average weekly rate of ambulatory referrals (8.0 ± 0.33 to 9.4 ± 0.31) and completed appointments (194.2 ± 0.72 to 213.1 ± 0.71) significantly increased during the pandemic, driven largely by teenagers. The weekly average of ED pediatric encounters for BH did not increase during the pandemic, although the percentage of all pediatric ED encounters that were for BH did increase from 2.6 to 4.1% (p < 0.001). Length of stay for pediatric BH ED patients increased from 1.59 ± 0.09 days pre-pandemic to 1.91 ± 0.11 days post-pandemic (p < 0.0001). Inpatient admissions for BH reasons overall decreased during the pandemic, due to a decrease in inpatient psychiatric bed capacity. However, the weekly percentage of inpatient hospitalizations for BH reasons that occurred on medical units increased during the pandemic (15.2% ± 2.8–24.6% ± 4.1% (p = 0.0006)). Taken together, our data suggest the COVID-19 pandemic had varying degrees of impact, depending on the setting of care.
- Open Access
Identifying Racial Disparities in Clozapine Prescriptions Among Schizophrenia Patients using Data from Electronic Health Records
Objective: This study sought to assess the demographic factors that alter the likelihood of being prescribed clozapine. The primary hypothesis is that African American patients are less likely to be prescribed Clozapine than White and other racial groups. Additionally, this study aims to expand on earlier studies of clozapine by examining how multiple variables, especially social determinants of health, gender, rurality, and care patterns affect the rates at which clozapine is prescribed. Methods: This observational study examines the racial disparities and other factors associated with receiving a clozapine prescription for patients with a schizophrenia diagnosis using structured data in the Electronic Health Records (EHR) from a multi-facility health system from 10/1/2015 - 11/30/2021. Bivariate analysis, multiple logistic regression, and sensitivity analysis tests were performed to determine which factors differed significantly between the group of patients who were prescribed clozapine and those who were not. Results: Results showed that being white and having a higher socioeconomic income was associated with an increased clozapine prescription rate, while gender, rurality, age at first diagnosis, and ethnicity did not influence prescription likelihood. Increased treatment length was also associated with a greater likelihood of being prescribed clozapine. Conclusion: African American patients are over-diagnosed with schizophrenia and under-prescribed clozapine compared to Caucasians after controlling for the variables associated with demographics, social determinants of health, and care access patterns. Future research is needed to understand and disentangle the biological, clinical, administrative, and societal causes behind the disparity in clozapine treatment.
Open AccessIdentifying Racial Disparities in Clozapine Prescriptions Among Schizophrenia Patients using Data from Electronic Health Records
Objective: This study sought to assess the demographic factors that alter the likelihood of being prescribed clozapine. The primary hypothesis is that African American patients are less likely to be prescribed Clozapine than White and other racial groups. Additionally, this study aims to expand on earlier studies of clozapine by examining how multiple variables, especially social determinants of health, gender, rurality, and care patterns affect the rates at which clozapine is prescribed. Methods: This observational study examines the racial disparities and other factors associated with receiving a clozapine prescription for patients with a schizophrenia diagnosis using structured data in the Electronic Health Records (EHR) from a multi-facility health system from 10/1/2015 - 11/30/2021. Bivariate analysis, multiple logistic regression, and sensitivity analysis tests were performed to determine which factors differed significantly between the group of patients who were prescribed clozapine and those who were not. Results: Results showed that being white and having a higher socioeconomic income was associated with an increased clozapine prescription rate, while gender, rurality, age at first diagnosis, and ethnicity did not influence prescription likelihood. Increased treatment length was also associated with a greater likelihood of being prescribed clozapine. Conclusion: African American patients are over-diagnosed with schizophrenia and under-prescribed clozapine compared to Caucasians after controlling for the variables associated with demographics, social determinants of health, and care access patterns. Future research is needed to understand and disentangle the biological, clinical, administrative, and societal causes behind the disparity in clozapine treatment.
- Open Access
Towards Efficient Architecture and Algorithms for Sensor Fusion
The safety of an automated vehicle hinges crucially upon the accuracy of perception and decision-making latency. Under these stringent requirements, future automated cars are usually equipped with multi-modal sensors such as cameras and LiDARs. The sensor fusion is adopted to provide a confident context of driving scenarios for better decision-making. A promising sensor fusion technique is middle fusion that combines the feature representations from intermediate layers that belong to different sensing modalities. However, achieving both the accuracy and latency efficiency is challenging for middle fusion, which is critical for driving automation applications. We present A3Fusion, a software-hardware system specialized for an adaptive, agile, and aligned fusion in driving automation. A3Fusion achieves a high efficiency for the middle fusion of multiple CNN-based modalities by proposing an adaptive multi-modal learning network architecture and a latency-aware, agile network architecture optimization algorithm that enhances semantic segmentation accuracy while taking the inference latency as a key trade-off. In addition, A3Fusion proposes a FPGA-based accelerator that captures unique data flow patterns of our middle fusion algorithm while reducing the overall compute overheads. We enable these contributions by co-designing the neural network, algorithm, and the accelerator architecture.
Open AccessTowards Efficient Architecture and Algorithms for Sensor Fusion
The safety of an automated vehicle hinges crucially upon the accuracy of perception and decision-making latency. Under these stringent requirements, future automated cars are usually equipped with multi-modal sensors such as cameras and LiDARs. The sensor fusion is adopted to provide a confident context of driving scenarios for better decision-making. A promising sensor fusion technique is middle fusion that combines the feature representations from intermediate layers that belong to different sensing modalities. However, achieving both the accuracy and latency efficiency is challenging for middle fusion, which is critical for driving automation applications. We present A3Fusion, a software-hardware system specialized for an adaptive, agile, and aligned fusion in driving automation. A3Fusion achieves a high efficiency for the middle fusion of multiple CNN-based modalities by proposing an adaptive multi-modal learning network architecture and a latency-aware, agile network architecture optimization algorithm that enhances semantic segmentation accuracy while taking the inference latency as a key trade-off. In addition, A3Fusion proposes a FPGA-based accelerator that captures unique data flow patterns of our middle fusion algorithm while reducing the overall compute overheads. We enable these contributions by co-designing the neural network, algorithm, and the accelerator architecture.
- 23 August 2022
- Published by Springer Nature in Academic Psychiatry
- Vol. 46 (5) , 569-573
Telehealth: Reducing Patients’ Greenhouse Gas Emissions at One Academic Psychiatry Department
Objective Academic and organizational leaders in psychiatry and all other medical fields are negatively impacted by climate change. The COVID-19 pandemic prompted a rapid shift to the use of more telehealth by behavioral health clinicians. The purpose of this study was to estimate the reduction of patients’ greenhouse gas emissions during this rapid shift at one academic psychiatry institution. Methods The authors extracted data associated with all outpatient visits to all 26 psychiatry clinics from March 16, 2020, to December 31, 2020. Once the patients’ travel miles saved by confirmed virtual visits were calculated, the authors used the standard ratio from the US Environmental Protection Agency (EPA) to calculate the total quantity of CO2 that would be emitted if the visits had occurred in person. Results During the study period, a total of 47,582 outpatient behavioral health visits with 3975 unique patients were completed. The majority of these departmental visits were telehealth (85%), with most of the telehealth visits conducted using real-time audio-video platforms (75.7%). Subtracting emissions from patient technology during telehealth visits from the estimated patient transport values produced a net savings of greenhouse gas emissions of 867,011 kg CO2. This amount is equal to the greenhouse gas emissions from 189 passenger vehicles driven for 1 year according to the EPA. Conclusions This study shows that converting in person, face-to-face behavioral health visits to telehealth has the potential to increase both energy efficiency and conservation through a reduction in greenhouse gas emissions due to reduced patient travel. If these values were extrapolated to the total adult US population who have visits for behavioral health reasons, we estimate that conversion to virtual visits could save approximately 830,000 metric tons of CO2 annually. Organizational leaders should consider these societal benefits when making decisions regarding development and support of telehealth.
- 23 August 2022
- Published by Springer Nature in Academic Psychiatry
- Vol. 46 (5) , 569-573
Telehealth: Reducing Patients’ Greenhouse Gas Emissions at One Academic Psychiatry Department
Objective Academic and organizational leaders in psychiatry and all other medical fields are negatively impacted by climate change. The COVID-19 pandemic prompted a rapid shift to the use of more telehealth by behavioral health clinicians. The purpose of this study was to estimate the reduction of patients’ greenhouse gas emissions during this rapid shift at one academic psychiatry institution. Methods The authors extracted data associated with all outpatient visits to all 26 psychiatry clinics from March 16, 2020, to December 31, 2020. Once the patients’ travel miles saved by confirmed virtual visits were calculated, the authors used the standard ratio from the US Environmental Protection Agency (EPA) to calculate the total quantity of CO2 that would be emitted if the visits had occurred in person. Results During the study period, a total of 47,582 outpatient behavioral health visits with 3975 unique patients were completed. The majority of these departmental visits were telehealth (85%), with most of the telehealth visits conducted using real-time audio-video platforms (75.7%). Subtracting emissions from patient technology during telehealth visits from the estimated patient transport values produced a net savings of greenhouse gas emissions of 867,011 kg CO2. This amount is equal to the greenhouse gas emissions from 189 passenger vehicles driven for 1 year according to the EPA. Conclusions This study shows that converting in person, face-to-face behavioral health visits to telehealth has the potential to increase both energy efficiency and conservation through a reduction in greenhouse gas emissions due to reduced patient travel. If these values were extrapolated to the total adult US population who have visits for behavioral health reasons, we estimate that conversion to virtual visits could save approximately 830,000 metric tons of CO2 annually. Organizational leaders should consider these societal benefits when making decisions regarding development and support of telehealth.