Tag: nevin manimala

JMIR Med Educ. 2026 Mar 30;12:e82988. doi: 10.2196/82988.

ABSTRACT

BACKGROUND: Generative artificial intelligence (AI) tools, such as ChatGPT, are increasingly used in higher education and have raised significant concerns about assessment validity and academic integrity. In Digital Health and Health Information Management (DIGHIM) programs, assessments are designed to evaluate a mix of technical skills, contextual reasoning, and professional judgment that underpin medical and health practice. Understanding how generative AI performs across different assessment types is, therefore, critical to identifying which formats are most susceptible to AI-generated content and how assessments may be redesigned to remain authentic and educationally meaningful.

OBJECTIVE: This study aimed to evaluate ChatGPT’s performance across diverse assessment types in DIGHIM education by examining how task complexity influences AI-generated output quality, and develop recommendations for ethical and effective AI integration in assessments.

METHODS: A pilot quasi-experimental design compared ChatGPT-generated responses with deidentified student submissions across 5 assessment types: digital health solution design, business case analysis, reflective assessment, SQL health database programming, and a health classification quiz. For each task, multiple AI submissions were produced using different prompting strategies, including rubric integration and the use of ChatGPT (GPT-4 and o1 Preview model). Blinded academic markers evaluated all AI-generated submissions and previously submitted deidentified student assessments against standard rubrics, and descriptive statistics were used to compare performance.

RESULTS: ChatGPT’s performance varied considerably across assessment types. It achieved its highest accuracy scores in objective, rule-based tasks such as multiple-choice quiz items in health classification (mean 88%, SD 0%) and produced well-structured, coherent responses for reflective assessments (mean 69%, SD 12.8%), though these often lacked personalization and nuanced industry context. In descriptive analytical tasks, such as digital health business cases and solution designs, ChatGPT produced logically structured work with reasonable use of evidence but failed to provide deep contextualization, domain-specific insights, or visual elements expected in DIGHIM practice. Technical assessments revealed the greatest limitations: SQL programming tasks averaged 42% (SD 17.2%) with persistent schema errors, incomplete queries, and weak interpretation of health data outputs, while scenario-based clinical coding scored just 7% (SD 0%), reflecting a lack of precision in applying ICD-10-AM (International Classification of Diseases, Tenth Revision, Australian Modification) rules and coding conventions. Structured prompting and rubric integration improved results, particularly in descriptive and reflective tasks (up to 80%), but the advanced o1 Preview model did not consistently outperform earlier versions.

CONCLUSIONS: While ChatGPT performs well in structured, rule-based, and reflective tasks, it remains limited in technical accuracy, contextual reasoning, and applied DIGHIM competencies. To support academic integrity and workforce readiness, assessment design should prioritize critical thinking, ethical reasoning, and scenario-based problem-solving aligned with health care practice. Using AI as a tool for critique and refinement, rather than a substitute for student work, may help educators prepare learners for responsible AI use in medical and health professional education.

PMID:41911020 | DOI:10.2196/82988

JMIR Ment Health. 2026 Mar 30;13:e93040. doi: 10.2196/93040.

ABSTRACT

BACKGROUND: Generative artificial intelligence (AI) chatbots have rapidly entered public use, including in contexts involving emotional support and mental health-related interactions. Although these systems are increasingly accessible, concerns have emerged regarding potential adverse psychiatric outcomes reported in public discourse, including psychosis, suicidal ideation, self-harm, and suicide. However, these reports largely originate from journalistic accounts rather than systematically verified clinical data.

OBJECTIVE: This rapid scoping review aimed to systematically map and characterize mass media narratives describing alleged adverse psychiatric outcomes temporally associated with generative AI chatbot interactions.

METHODS: A rapid scoping review methodology was applied to publicly accessible news articles identified primarily through Google News searches. Articles published from November 2022 onward were screened for eligibility if they described a specific case in which psychiatric deterioration or crisis was temporally linked to generative AI use. Data were extracted using a structured coding template capturing article characteristics, demographic information, AI platform features, interaction intensity, outcome type and severity, type of evidence reported, and causal attribution language. Descriptive statistics and cross-tabulations were performed.

RESULTS: A total of 71 news articles representing 36 unique cases were included. Suicide death was the most frequently reported outcome (35/61, 57.4% cases with complete severity coding), followed by psychiatric hospitalization (12/61, 19.7%). Fatal outcomes were disproportionately represented among minors (19/21, 90.5%) compared with adults (17/35, 48.6%). ChatGPT was the most frequently cited platform (51/71, 71.8%), followed by Character AI (10/71, 14.1%). Causal attribution most commonly referenced AI system behavior (45/61, 73.8%), and the term “alleged” was the predominant causal descriptor (33/61, 54.1%). Evidence sources were primarily chat logs or screenshots (34/61, 55.7%), while police or medical documentation was rare (1/61, 1.6%). Regulatory calls were present in 51 of 60 (85%) articles with nonmissing data.

CONCLUSIONS: Mass media reporting of generative AI-related psychiatric harms is concentrated around severe outcomes, particularly suicide deaths among youth, and is frequently framed within regulatory and corporate accountability narratives. While causality cannot be established from media reports, consistent patterns of high-intensity interactions, user vulnerability, and limited safeguard reporting highlight the need for structured safety surveillance, transparent AI risk auditing, and clearer governance frameworks. As generative AI becomes increasingly integrated into everyday psychosocial contexts, systematic research and formal safety monitoring will be necessary to determine whether media-reported harms correspond to verifiable clinical risk patterns.

PMID:41911018 | DOI:10.2196/93040

Psychol Sci. 2026 Mar 30:9567976261432207. doi: 10.1177/09567976261432207. Online ahead of print.

ABSTRACT

Income inequality is frequently cited as a forceful determinant of mental health and as a possible contributor to the rising trend in adolescent depressive symptoms. However, research findings often rely on low-powered cross-sectional designs. We conducted a preregistered study of the within-municipality effect of income inequality on adolescent depressive symptoms in Norway, covering ≈550,000 respondents nested within 863 municipality years and 340 municipalities. Using multilevel modeling and equivalence testing, the overall within-municipality effect of income inequality was neither statistically significant nor practically meaningful and did not significantly interact with family financial situation. A significant gender interaction showed that rising inequality predicted slightly higher depressive symptoms among females and slightly lower among males; however, the main gender effects were also probably too small to be meaningful. We conclude that changes in income inequality likely do not meaningfully predict nor help explain changes in adolescent depressive symptoms in Norway from 2010 to 2019.

PMID:41911005 | DOI:10.1177/09567976261432207

J Clin Psychol. 2026 Mar 30. doi: 10.1002/jclp.70133. Online ahead of print.

ABSTRACT

OBJECTIVE: Emotional labour has long been associated with personal and organizational outcomes such as burnout. However, theoretically dichotomising regulation into surface and deep acting may constrain the ecological validity of research as iterative and person-centered approaches to emotion regulation are not considered. Furthermore, recent research suggests self-compassion and experience may predict emotional labour regulation in psychologists, but specific mechanisms accounting for this relationship are unknown. We addressed these concerns by examining how self-compassion and career experience predict latent profiles of emotional labour regulation strategies in psychologists and subsequent burnout.

METHOD: We performed latent profile analysis, multinomial logistic regression, and a one-way between-groups ANOVA on data from 232 international psychologists across two time points.

RESULTS: We found a similar but not identical pattern of latent profiles when compared to previous studies in different occupations. Self-compassion and career experience significantly predicted subsequent profile membership and profiles characterized by less surface acting and more authentic and genuine emotional displays had statistically significantly lower levels of emotional exhaustion.

CONCLUSIONS: Our findings suggest that self-compassion promotes adaptive emotional labour regulation strategies in psychologists, that experienced clinicians express emotion more authentically, and that regulation that involves authentic and genuine expression is linked with lower emotional exhaustion.

PMID:41910994 | DOI:10.1002/jclp.70133

JAMA Netw Open. 2026 Mar 2;9(3):e262420. doi: 10.1001/jamanetworkopen.2026.2420.

ABSTRACT

IMPORTANCE: While the clinical benefits of timely percutaneous coronary intervention (PCI) are well established, it remains unclear whether the expansion of PCI-capable facilities enhances access to critical care or contributes to overuse.

OBJECTIVE: To assess whether new PCI-capable hospital openings are associated with changes in the overall procedural volume at the community level and the health characteristics of patients undergoing PCI.

DESIGN, SETTING, AND PARTICIPANTS: This cohort study used California all-payer data from January 1, 2011, to December 31, 2022, including 651 585 patients across 2348 communities (defined by zip code). Using a difference-in-differences framework, changes in PCI volume and patient characteristics in communities exposed to a PCI facility opening were compared with those without, stratified by baseline PCI access. Statistical analysis was completed between January and July 2025.

EXPOSURES: Opening of a PCI-capable facility within a 30-minute driving time of a zip code community.

MAIN OUTCOMES AND MEASURES: Community-level PCI volume and patient-level indicators, including primary diagnosis of stable angina, prior acute myocardial infarction (AMI) or coronary artery bypass grafting (CABG), and procedure complexity (number of vessels treated).

RESULTS: The final sample included 651 585 patients (463 526 male [71%]; 128 469 Hispanic [20%], 370 672 White [57%]); 47 003 patients (7%) lived in rural communities. At baseline, 84 349 patients (13%) had no access to PCI within 30 minutes. Community PCI volume increased by 7.5% (95% CI, 6.4%-8.6%) after a local PCI facility opened, with a 19.9% increase (95% CI, 15.7%-24.1%) in communities without prior 30-minute access. Among patients, the proportion with stable angina increased by 2.5 percentage points (95% CI, 2.0 to 3.1 percentage points), and by 3.5 percentage points (95% CI, 1.3 to 5.7 percentage points) in communities with no PCI at baseline. In communities with prior access, there was a 0.7 percentage point increase (95% CI, 0.3 to 1.1 percentage points) in patients without prior AMI or CABG and a 0.6 percentage point increase (95% CI, 0.4 to 0.9 percentage points) in those receiving PCI on 3 or more vessels. In contrast, communities with no baseline access to PCI experienced a 2.2 percentage point increase (95% CI, 0.3 to 4.1 percentage points) in single-vessel PCI and a 2.1 percentage point decrease (95% CI, -3.0 to -1.2 percentage points) in patients receiving PCI on 3 or more vessels.

CONCLUSIONS AND RELEVANCE: In this retrospective cohort study of 651 585 patients, the opening of PCI-capable hospitals was associated with increased community PCI volumes, particularly in underserved areas. Changes in patient profiles suggested potential supply-induced demand in areas that had existing access, and a release of unmet need in previously underserved areas; these findings highlighted the dual implications of service expansion.

PMID:41910975 | DOI:10.1001/jamanetworkopen.2026.2420

JAMA Netw Open. 2026 Mar 2;9(3):e263392. doi: 10.1001/jamanetworkopen.2026.3392.

ABSTRACT

IMPORTANCE: Despite the growing number of residents with disabilities, barriers to equitable access persist in medical training. Program access to accommodation has been linked to improved training and mental health outcomes, but little is known about possible resident and program characteristics associated with access to and requests for needed accommodations.

OBJECTIVE: To examine demographic, training, and disability-related factors associated with program access and accommodation requests among internal medicine (IM) residents with disabilities.

DESIGN, SETTING, AND PARTICIPANTS: This national cross-sectional study looked at accredited IM residency programs in mainland US and Puerto Rico. Participants were IM residents who took the 2023 Internal Medicine In-Training Examination and reported having at least 1 type of disability.

MAIN OUTCOMES AND MEASURES: The primary outcomes were program access, defined as receiving accommodations or not needing them, and requesting needed accommodations. Multivariable logistic regression models were conducted for each of the outcomes.

RESULTS: Of 19 205 respondents, 1824 (9.5%) reported a disability; participants were predominantly men (979 men [53.7%]), US medical graduates (1398 participants [76.6%]), and enrolled in categorical IM programs (1532 participants [84.0%]). With regard to race, 340 participants (18.6%) were Asian, 415 (22.8%) were from groups underrepresented in medicine (including self-reported Black or African American or Afro-Caribbean; Latinx or Latino or Hispanic; Native American or American Indian or Indigenous or Alaskan Native; Native Hawaiian or Pacific Islander), and 823 (45.1%) were White. Among 1052 with complete accommodation information, 811 (77.1%) had program access and 241 (22.9%) did not. In multivariable regression models, having cognitive disabilities (adjusted odds ratio [aOR], 0.27; 95% CI, 0.15-0.49) and identifying as women (aOR, 0.55; 95% CI, 0.40-0.75), Asian (aOR, 0.53; 95% CI, 0.34-0.82) and underrepresented racial or ethnic groups (aOR, 0.58; 95% CI, 0.38-0.87) were associated with lower odds of program access. Among 699 residents coded as needing disability accommodations with classifiable responses, 200 (28.6%) did not request them. Fear of stigma (164 respondents [82.0%]) and unclear institutional processes (60 respondents [30.0%]) were the most cited reasons for nonrequest for needed accommodations. Requesting accommodations was less likely among residents with cognitive disabilities (aOR, 0.16; 95% CI, 0.08-0.31) and who identify as women (aOR, 0.37; 95% CI, 0.25-0.54), genderqueer or nonbinary (aOR, 0.11; 95% CI, 0.02-0.68), Asian (aOR, 0.50; 95% CI, 0.30-0.85), or underrepresented in medicine (aOR, 0.60; 95% CI, 0.37-0.97).

CONCLUSIONS AND RELEVANCE: These findings suggest that despite growing disability representation, substantial inequities in access to and requests for accommodations persisted for IM residents with disabilities, particularly those with cognitive disabilities and marginalized identities. Institutions should implement inclusive, transparent policies to foster psychological safety and disability inclusion.

PMID:41910974 | DOI:10.1001/jamanetworkopen.2026.3392

JAMA Netw Open. 2026 Mar 2;9(3):e263885. doi: 10.1001/jamanetworkopen.2026.3885.

ABSTRACT

IMPORTANCE: The use of robotic-assisted surgery has increased rapidly despite limited evidence of superior outcomes over more established surgical approaches such as laparoscopy.

OBJECTIVE: To evaluate whether and to what extent surgeons’ financial relationships with industry are associated with the use of robotic-assisted surgery.

DESIGN, SETTING, AND PARTICIPANTS: In this cohort study, fee-for-service Medicare claims (January 1, 2011, to December 31, 2021) for patients undergoing 1 of 4 common surgical operations (bariatric surgery, cholecystectomy, colectomy, or ventral hernia repair) were linked to surgeon-level Open Payments data on receipt of industry payments from a large robotic surgical device company. Data were analyzed from April to August 2025.

EXPOSURE: Receipt of a direct industry payment from a robotic-assisted surgical device company.

MAIN OUTCOME AND MEASURES: Each surgeon’s use of robotic-assisted surgery as a proportion of all surgeries performed by that surgeon. A staggered difference-in-differences (DID) approach was used to isolate the association of industry payments with the proportional use of robotic-assisted surgery among surgeons who received payment compared with control surgeons who never received a payment.

RESULTS: Among 20 313 surgeons (mean [SD] age, 50.7 [10.2] years; 86.2% male) performing 886 385 surgeries, 5933 (29.2%) received at least 1 industry payment. Receipt of an industry payment was associated with a significant increase in the proportional use of robotic-assisted surgery, with a DID estimate of 9.9 percentage points (pp) (95% CI, 9.30-10.6 pp). Results were consistent across discrete procedures (eg, DID estimate of 11.7 pp [95% CI, 9.4-13.9 pp] for bariatric surgery and 10.3 pp [95% CI, 9.4-11.4 pp] for ventral hernia repair). There was a significant dose-dependent response. For example, surgeons receiving less than $500 increased use of robotic-assisted surgery after payment from a mean of 1.5% (95% CI, 1.4%-1.6%) to 3.7% (95% CI, 3.5%-3.9%), compared with 0.4% (95% CI, 0.4%-0.5%) to 17.0% (95% CI, 16.7%-17.3%) among surgeons receiving more than $10 000.

CONCLUSIONS AND RELEVANCE: In this cohort study, receipt of industry payments by surgeons was associated with increased use of robotic-assisted surgery compared with no receipt of payment, with a significant dose-dependent response. These results suggest that surgeon-industry financial relationships may be an important contributor to greater use of robotic-assisted surgery across the US.

PMID:41910972 | DOI:10.1001/jamanetworkopen.2026.3885

JAMA Netw Open. 2026 Mar 2;9(3):e264017. doi: 10.1001/jamanetworkopen.2026.4017.

ABSTRACT

IMPORTANCE: Limited research exists on sex, racial, and ethnic disparities in required graduate medical education (GME) resident competency ratings across specialties during sensitive periods when career decision-making occurs. Rating disparities using an antideficit-based approach measured by exceptionally high ratings are underexplored in GME.

OBJECTIVE: To assess the association of exceptionally high ratings in the Accreditation Council for Graduate Medical Education (ACGME) Milestones during time-sensitive training periods across specialties with differences among residents’ characteristics, including sex, race, and ethnicity.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional analysis was conducted between March 15 and December 31, 2025, using 2018 to 2021 Association of American Medical Colleges and ACGME data. Postgraduate year (PGY) 2 residents training at US ACGME-accredited emergency medicine, family medicine, internal medicine, obstetrics and gynecology, pediatrics, and surgery residency programs between 2018 and 2021 who self-reported sex, race, or ethnicity were studied.

EXPOSURE: Required Milestones ratings at the end of PGY-2 training associated with resident sex and race or ethnicity (underrepresented in medicine [URiM] and Asian), while controlling for preresidency Step 2 Clinical Knowledge examination scores.

MAIN OUTCOMES AND MEASURES: Proportion and adjusted odds ratios (AORs) for exceptionally high resident-level ratings (80th percentile level) across competencies in interpersonal and communication skills, medical knowledge, patient care, practice-based learning and improvement, professionalism, and systems-based practice.

RESULTS: Among 19 492 PGY-2 residents across 1754 programs, 10 384 (53.3%) were female, 28 (0.14%) American Indian or Alaskan Native, 4327 (22.2%) Asian, 1106 (5.7%) Black, 1008 (5.2%) Hispanic or Latinx, 3 (0.02%) Native Hawaiian or Pacific Islander, 12 269 (62.9%) White, 751 (3.9%) reporting 2 or more races, and 3423 (17.6%) classified as URiM. Exceptional rating differences were identified by sex, race, and ethnicity. Across all specialties, female residents had greater odds for 80th percentile ratings (AOR, 1.12; 95% CI, 1.05-1.21; P < .001); whereas when compared with White residents, URiM residents (AOR, 0.68; 95% CI, 0.62-0.76; P < .001) and Asian residents (AOR, 0.67; 95% CI, 0.60-0.74; P < .001) were less likely to have 80th percentile ratings than White residents. Within specialties, URiM residents in emergency medicine, family medicine, internal medicine, obstetrics and gynecology, and surgery were less likely to have 80th percentile ratings, whereas Asian residents in family medicine, internal medicine, pediatrics, and surgery were also less likely than White residents.

CONCLUSION AND RELEVANCE: In this cross-sectional national study of residents, exceptionally higher ratings were associated with differing resident characteristics during crucial career planning phases. These results suggest the need for more studies to explore factors of resident success during GME training.

PMID:41910971 | DOI:10.1001/jamanetworkopen.2026.4017

JAMA Netw Open. 2026 Mar 2;9(3):e264024. doi: 10.1001/jamanetworkopen.2026.4024.

ABSTRACT

IMPORTANCE: Mental health morbidity is a proximal factor in suicide. Using the research domain criteria (RDoC) framework to investigate sex and age differences in psychopathology at the time of death may provide better characterization than focusing on clinical diagnosis alone.

OBJECTIVES: To (1) evaluate whether token-based and large language model scoring of RDoC can be successfully applied to law enforcement and coroner or medical examiner death narratives in the US National Violent Death Reporting System (NVDRS), and (2) investigate sex and age differences in clinically relevant scores to illuminate underidentified dimensions of mental health dysfunction proximal to suicide.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study drew on death records in the restricted-access 2020 to 2021 NVDRS. Participants were limited to suicide decedents aged 12 years and older from all 50 states whose death record included a law enforcement and coroner or medical examiner death narrative of 20 words or more. Analyses were conducted between May 2024 and September 2025.

EXPOSURES: Sex and age of decedents.

MAIN OUTCOMES AND MEASURES: RDoC symptom scores; mental health status measures in death records (mental health diagnosis, current depressed mood, and alcohol or drug misuse).

RESULTS: Using both a token-based system and a large language model approach, law enforcement and coroner or medical examiner narratives of 72 585 suicide decedents were scored (mean [SD] age, 46.3 [19.3] years; 57 770 [80.6%] male). Both methods were previously validated with psychiatric electronic health records. To validate this approach, token density and large language model scores were compared with current NVDRS mental health status measures. Both scoring methods correlated with precoded measures and demonstrated levels of neurobehavioral dysfunction at the time of death similar to psychiatric inpatients on admission. Sex- and age-related differences in clinically relevant dysfunction showed the highest levels among female vs male and younger vs older decedents after adjusting for demographic confounding.

CONCLUSIONS AND RELEVANCE: This cross-sectional study of suicide decedents found that information relevant to RDoC domains is encoded in NVDRS death narratives and can be extracted using large language models. The approach used here observed more pervasive neurobehavioral dysfunction among suicide decedents than that captured by currently employed NVDRS measures of mental health.

PMID:41910970 | DOI:10.1001/jamanetworkopen.2026.4024

JAMA Netw Open. 2026 Mar 2;9(3):e264171. doi: 10.1001/jamanetworkopen.2026.4171.

ABSTRACT

IMPORTANCE: A sense of belonging is a fundamental need that predicts mental, physical, social, economic, and behavioral outcomes. Experiencing a sense of belonging at work may be influenced by organizational culture and whether a person feels supported by teammates. There is little evidence of what proportion of physicians feel a strong sense of belonging and teammate support. There exists a gap in literature showing whether these factors are associated with burnout or work intentions, which leaves health care leaders without a valuable tool to improve professional well-being and workforce retention.

OBJECTIVE: To provide health care leaders with a better understanding of how fostering belonging and teammate support can decrease burnout and potentially prevent costly turnover.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used data from a survey administered to physicians by their organization leaders as part of the organization’s quality improvement efforts through participation in the American Medical Association Organizational Biopsy program. The survey was administered within health care organizations and health systems across the US with more than 50 physicians. The study population comprised physicians working in US-based health care organizations and health systems. Survey responses were collected between November 7, 2023, and November 12, 2024. Analysis was performed from June to August 2025.

MAIN OUTCOMES AND MEASURES: Belonging and teammate support perceptions were measured by asking participants to indicate how much they agree with 2 statements. Burnout was assessed using the single-item Mini Z assessment, and work intentions were indicated by likeliness to reduce clinical hours in the next 12 months or leave practice within the next 2 years. Odds ratios (ORs) and 95% CIs were used.

RESULTS: Of 14 051 physicians, 7315 were male (52.1%), 1978 were Asian (14.1%), 406 were Black or African American (2.9%), 425 were Latinx or Hispanic (3.0%), and 8681 were White (61.8%). Additionally, 4733 (33.7%) had 20 or more posttraining years of clinical practice and 4475 (31.8%) practiced in primary care. A strong sense of belonging was endorsed by 8425 (60.0%) and 11 293 (80.4%) perceived teammate support. Females (odds ratio [OR], 0.91, 95% CI, 0.84-0.98) and physicians with more than 5 years of clinical practice (eg, 6-10 years: OR, 0.76; 95% CI, 0.67-0.85; ≥20 years: OR, 0.87; 95% CI, 0.78-0.96) had lower odds of having strong sense of belonging. A strong sense of belonging was associated with lower odds of burnout (OR, 0.22; 95% CI, 0.21-0.24), lower odds of intent to reduce clinical hours (OR, 0.48; 95% CI, 0.44-0.53), and lower odds of intent to leave the organization (OR, 0.23; 95% CI, 0.21-0.26). Similar patterns were observed for factors associated with teammate support and for the association of teammate support with burnout, intent to reduce hours, and intent to leave the organization.

CONCLUSIONS AND RELEVANCE: In this cross-sectional study of physicians in the US, a strong sense of belonging was associated with lower odds of burnout, intent to reduce clinical hours, and intent to leave the organization. These findings suggest that fostering belonging and teammate support can benefit physicians, care teams, patients, and organizations by potentially avoiding physician turnover. Organizations should prioritize belonging and teammate support in their efforts to improve culture, work environment, and physician well-being.

PMID:41910969 | DOI:10.1001/jamanetworkopen.2026.4171