Performance of Artificial Intelligence Content Detectors Using Human and Artificial Intelligence-Generated Scientific Writing.

BACKGROUND: Few studies have examined the performance of artificial intelligence (AI) content detection in scientific writing. This study evaluates the performance of publicly available AI content detectors when applied to both human-written and AI-generated scientific articles. METHODS: Articles published in Annals of Surgical Oncology (ASO) during the year 2022, as well as AI-generated articles using OpenAI's ChatGPT, were analyzed by three AI content detectors to assess the probability of AI-generated content. Full manuscripts and their individual sections were evaluated. Group comparisons and trend analyses were conducted by using ANOVA and linear regression. Classification performance was determined using area under the curve (AUC). RESULTS: A total of 449 original articles met inclusion criteria and were evaluated to determine the likelihood of being generated by AI. Each detector also evaluated 47 AI-generated articles by using titles from ASO articles. Human-written articles had an average probability of being AI-generated of 9.4% with significant differences between the detectors. Only two (0.4%) human-written manuscripts were detected as having a 0% probability of being AI-generated by all three detectors. Completely AI-generated articles were evaluated to have a higher average probability of being AI-generated (43.5%) with a range from 12.0 to 99.9%. CONCLUSIONS: This study demonstrates differences in the performance of various AI content detectors with the potential to label human-written articles as AI-generated. Any effort toward implementing AI detectors must include a strategy for continuous evaluation and validation as AI models and detectors rapidly evolve.

Amino acid kinetics in the critically ill.

PURPOSE OF REVIEW: Stable isotope methods have been used for many years to assess whole body protein and amino acid kinetics in critically ill patients. In recent years, new isotope approaches and tracer insights have been developed. The tracer pulse approach has some advantages above the established primed-continuous tracer infusion approach because of the high amount of metabolic information obtained, easy applicability, and low tracer costs. Effects of disease severity and sex on amino acid kinetics in ICU patients will also be addressed. RECENT FINDINGS: Current knowledge was synthesized on specific perturbations in amino acid metabolism in critically ill patients, employing novel methodologies such as the pulse tracer approach and computational modeling. Variations were evaluated in amino acid production and linked to severity of critical illness, as measured by SOFA score, and sex. Production of the branched-chain amino acids (BCAAs), glutamine, tau-methylhistidine and hydroxyproline were elevated in critical illness, likely related to increased transamination of the individual BCAAs or increased breakdown of proteins. Citrulline production was reduced, indicative of impaired gut mucosa function. Sex and disease severity independently influenced amino acid kinetics in ICU patients. SUMMARY: Novel tracer and computational approaches have been developed to simultaneously measure postabsorptive kinetics of multiple amino acids that can be used in critical illness. The collective findings lay the groundwork for targeted individualized nutritional strategies in ICU settings aimed at enhancing patient outcomes taking into account disease severity and sex.

Firearm Ownership: A Key Factor to Consider in Firearm Suicide Deaths in Children, Despite State Gun Laws.

INTRODUCTION: Firearm-related suicides among children present a significant public health concern and a tragic loss of young lives. This study explores the relationship between firearm-related suicides, gun ownership, and state-specific gun laws. METHODS: This retrospective cohort study collected data from the Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research on children under 18 who died by firearm-related suicides between 2009 and 2016 in all 50 states and D.C. It also utilized data from the RAND State-Level Estimates of Household Firearm Ownership. The study focused on the rate of child firearm suicide deaths per 100,000 individuals. The key variable of interest was the percentage of guns owned per household in each state. Univariable analysis was conducted to examine the association between individual gun laws and child firearm suicide mortalities, while multivariable regression, adjusting for household gun ownership and significant firearm legislation, was employed to assess connection to child firearm suicide mortality. RESULTS: From 2009 to 2016, 3903 children died from firearm-related suicides in the United States. In our analysis, 15 out of 44 firearm laws were found to be associated with reducing the rates of firearm suicides among children (P < 0.05). However, multivariable regression showed that higher state gun ownership rates were the primary predictor of increased child fatalities from firearms, with children in such states being 325% more likely to die when analyzing handgun laws and 337% more likely when analyzing long gun laws, as indicated by coefficients of 4.25 and 4.37, respectively. No state laws alone notably improved death rates. CONCLUSIONS: Gun ownership has a stronger association with child suicide rates than state-specific gun laws. Given the weight of gun ownership, future research should prioritize comprehensive public health initiatives to prevent child firearm-related suicides.

Factors Associated with Discharge to Skilled Nursing Facility Following Gunshot Wounds.

INTRODUCTION: Firearm injury is a public health crisis. Most victims are minorities in underserved neighborhoods. Measuring firearm injury by mortality underestimates its impact, as most victims survive to discharge. This study was done to determine if race and insurance status are associated with discharge disposition for gunshot wound (GSW)-related trauma. METHODS: Using the 2019 Trauma Quality Improvement Program database, we identified GSW patients with Abbreviated Injury Scale (AIS) = 1-3. Exclusion criteria included patients who died in hospital and routine home discharge. We compared discharge patterns of patients based on demographics (age, gender, race, ethnicity, payor, AIS, hospital designation, and length of stay [LOS]) and injury severity. Multivariable logistic regression models identified factors associated with discharge disposition. RESULTS: Our sample included 2437 patients with GSWs. On univariable analysis, Black patients were more likely to discharge to home with home health (64.1% Black versus 34.7% White; P < 0.001). White patients were more likely to discharge to skilled nursing facility (SNF) (51.4% White versus 44.6% Black; P < 0.001). Controlling for age, race, Latin ethnicity, primary payor, LOS, AIS severity, and injury severity score factors independently associated with discharge to SNF included age (0.0462, P < 0.001), Medicaid (1.136, P < 0.0003), Medicare (1.452, P < 0.001), and LOS (0.03745, P < 0.001). CONCLUSIONS: Postacute care following traumatic injuries is essential to recovery. Black GSW victims are more likely to be discharged to home health than White patients, who are more likely to be discharged to SNF. Targeted programs to reduce barriers to appropriate aftercare are necessary to eliminate this bias and improve the care of underserved populations.

The Use of Parenteral Nutrition and Disparities in Its Allocation Following Traumatic Injury.

INTRODUCTION: Severe traumatic injury requires rapid and extensive deployment of resources to save the lives of the critically injured. The sequelae of traumatic injuries frequently require extensive intervention obligating patients to a complicated recovery process devoid of meaningful nutrition. In this setting, parenteral nutrition (PN) is key in enabling appropriate wound healing, recovery, and rehabilitation. We sought to examine the use of PN in adult trauma management and to highlight any disparities in the utilization of PN in adult trauma patients. METHODS: We queried the 2017-2019 Trauma Quality Improvement Program (TQIP) for adult patients (aged > 18 y) who sustained blunt or penetrating traumatic injuries and received PN as part of their hospitalization. We compared time to PN administration based on demographics. We then used a multivariable logistic regression model to identify factors associated with the use of PN. We hypothesized that PN would be less commonly employed in the uninsured and minority groups. RESULTS: We identified 2,449,498 patients with sufficient data for analysis. Of these, 1831 patients were treated with PN. On univariate analysis, PN patients were more commonly male (74.7% PN versus 60.2% non-PN; P < 0.001). PN use was more frequent in the Black population (24.3% PN versus 15.5% non-PN; P < 0.001) and less frequent in the White population (72.7% PN versus 81.2% non-PN; P < 0.001). PN use was also much more common among patients covered by Medicaid. Penetrating trauma was over twice as common among PN recipients relative to non-PN patients (% PN versus % non-PN). PN patients had higher injury severity scores (ISSs), more intensive care unit days, longer hospitalizations, and increased mortality compared to non-PN patients. PN patients were half as likely to discharge home and twice as likely to discharge to a long-term care facility. Multivariable analysis including age, race, trauma mechanism, primary payer, and ISS, demonstrated an association of PN use with increasing age (OR 1.01, P < 0.001), cases of penetrating trauma (odds ratio [OR], 2.47; P < 0.001), and patients with high ISS (OR, 0.1.06; P < 0.001). There was decreased use in Uninsured patient (OR, 0.54; P < 0.001). CONCLUSIONS: PN use following traumatic injury is rarely required. Patients treated with PN typically have a resource-intense hospital course. More severe injuries, penetrating trauma, and increased age are more likely to result in PN use. Variations in PN use are apparent based on insurance payer, further examination into allocation of hospital and intensive care resources, as it pertains to patient socioeconomic status, is warranted in light of these findings.

Racial and ethnic demographics in malnutrition related deaths.

BACKGROUND AND AIMS: Currently, 40 million Americans are food insecure. They are forced to skip meals and buy non-nutritious food, leading to health disparities for those of low socioeconomic status. This study aims to investigate relationships between malnutrition deaths and sociodemographic groups. METHODS: This cross-sectional study from 2009 to 2018 used aggregate data from the CDC Wide-ranging Online Data for Epidemiologic Research (CDC Wonder). Patients with known race, gender, and Hispanic origin age ≥18 who died from malnutrition (E40-E46) were included. Place of death was grouped into home, inpatient medical facility, hospice facility, nursing facility/long-term care, other (including outpatient, ED, and DOA), and unknown. Crude rates of malnutrition deaths per 100,000 persons for race, gender, and Hispanic origin were calculated using US census estimates. Gross proportions of total deaths were calculated for each place of death. RESULTS: Between 2009 and 2018, there were 46,517 malnutrition deaths in the US. Death rates for Black (1.8) and White Americans (2) were twice as high compared to Native Americans (1.1) and Asians or Pacific Islanders (0.7). Death rates among females (2.3) were higher than males (1.5). Death rates among non-Hispanics (2.1) were twice as high compared to Hispanics (0.7). Most people who died of malnutrition died in hospitals (37 %). CONCLUSION: Malnutrition deaths occur at greater rates among White, Black, non-Hispanic Americans, and females. Despite reported disparities in food access, Black and White Americans have similar malnutrition mortality rates, raising concerns that malnutrition is under-diagnosed among Black patients. Given the existing nutrition literature, this finding requires further investigation.

Child Firearm-Related Homicide and Suicide by State Legislation in the US (2009-2020).

BACKGROUND: Firearms are the leading cause of death among U.S. children and adolescents. This study evaluates whether state gun laws are associated with firearm suicides and homicides in children. STUDY DESIGN: This is a cross-sectional database study comparing childhood firearm mortality with 36 state firearm laws using data from CDC WONDER and the RAND state firearm law database. Primary outcomes were firearm-related suicide and homicide mortalities per 100,000 persons. We examined suicide deaths by all firearms, including intentional self-harm by handguns only, intentional self-harm by rifles, shotguns, or large firearms only, and intentional self-harm by other or unspecified firearms, as well as homicide deaths for the same firearm types in each state. Welch's t-tests compared mean rates of suicide and homicide mortalities between states with and without these laws. States that either enacted or rescinded firearm legislation during this period were excluded. RESULTS: From 2009-2020, there were 6,735 suicides and 10,278 homicides by firearm totaling 17,013 child deaths (<18) by firearm. States with "child access prevention- negligent storage" laws demonstrated lower suicide mortality rates across all firearm types (handguns) N=13, M (mean per 100,000)=0.68, SD=0.27, p<0.001; (long guns) N=12, M=0.65, SD=0.25, p<0.001). There were no significant differences in mean suicide death rates across all firearm types when comparing states with or without firearm laws related to "minimum age youth possession", "minimum age youth purchase and sale", or "child access prevention - intentional." Comparing homicide mortality rates for all firearm types revealed no notable distinctions between states with and without the identified laws. CONCLUSIONS: Firearm legislation is associated with decreased suicide rates for individuals under 18, but its influence on homicides is less certain. Comprehensive research and thoughtful policy formulation are essential for addressing this pressing public health concern.

Whole Blood Resuscitation for Injured Patients Requiring Transfusion: A Systematic Review, Meta-Analysis, and Practice Management Guideline from the Eastern Association for the Surgery of Trauma.

INTRODUCTION: Whole blood resuscitation has reemerged as a resuscitation strategy for injured patients. However, the effect of whole blood-based resuscitation on outcomes has not been established. The primary objective of this guideline was to develop evidence-based recommendations on whether whole blood should be considered in civilian trauma patients receiving blood transfusions. METHODS: An EAST working group performed a systematic review and meta-analysis utilizing the GRADE methodology. One PICO question was developed to analyze the effect of whole blood resuscitation in the acute phase on mortality, transfusion requirements, infectious complications, and ICU length of stay. English language studies including adult civilian trauma patients comparing in-hospital whole blood to component therapy were included. Medline, Embase, Cochrane CENTRAL, CINAHL Plus, and Web of Science were queried. GRADEpro was used to assess quality of evidence and risk of bias. The study was registered on PROSPERO (#CRD42023451143). RESULTS: A total of 21 studies were included. Most patients were severely injured and required blood transfusion, massive transfusion protocol activation, and/or a hemorrhage control procedure in the early phase of resuscitation. Mortality was assessed separately at the following intervals: early (i.e., ED, 3-, or 6-hour), 24-hour, late (i.e., 28- or 30-day), and in-hospital. On meta-analysis, whole blood was not associated with decreased mortality. Whole blood was associated with decreased 4-hour RBC (mean difference -1.82, 95% CI -3.12 to -0.52), 4-hour plasma (mean difference -1.47, 95% CI -2.94 to 0), and 24-hour RBC transfusions (mean difference -1.22, 95% CI -2.24 to -0.19) compared to component therapy. There were no differences in infectious complications or ICU length of stay between groups. CONCLUSION: We conditionally recommend WB resuscitation in adult civilian trauma patients receiving blood transfusions, recognizing that data are limited for certain populations, including women of childbearing age, and therefore this guideline may not apply to these populations. LEVEL OF EVIDENCE: Level III, Guidelines.

Mobile Application-Based Communication Facilitation Platform for Family Members of Critically Ill Patients: A Randomized Clinical Trial.

Importance: Unmet and racially disparate palliative care needs are common in intensive care unit (ICU) settings. Objective: To test the effect of a primary palliative care intervention vs usual care control both overall and by family member race. Design, Setting, and Participants: This cluster randomized clinical trial was conducted at 6 adult medical and surgical ICUs in 2 academic and community hospitals in North Carolina between April 2019 and May 2022 with physician-level randomization and sequential clusters of 2 Black patient-family member dyads and 2 White patient-family member dyads enrolled under each physician. Eligible participants included consecutive patients receiving mechanical ventilation, their family members, and their attending ICU physicians. Data analysis was conducted from June 2022 to May 2023. Intervention: A mobile application (ICUconnect) that displayed family-reported needs over time and provided ICU attending physicians with automated timeline-driven communication advice on how to address individual needs. Main Outcomes and Measures: The primary outcome was change in the family-reported Needs at the End-of-Life Screening Tool (NEST; range 0-130, with higher scores reflecting greater need) score between study days 1 and 3. Secondary outcomes included family-reported quality of communication and symptoms of depression, anxiety, and posttraumatic stress disorder at 3 months. Results: A total of 111 (51% of those approached) family members (mean [SD] age, 51 [15] years; 96 women [86%]; 15 men [14%]; 47 Black family members [42%]; 64 White family members [58%]) and 111 patients (mean [SD] age, 55 [16] years; 66 male patients [59%]; 45 Black patients [41%]; 65 White patients [59%]; 1 American Indian or Alaska Native patient [1%]) were enrolled under 37 physicians randomized to intervention (19 physicians and 55 patient-family member dyads) or control (18 physicians and 56 patient-family member dyads). Compared with control, there was greater improvement in NEST scores among intervention recipients between baseline and both day 3 (estimated mean difference, -6.6 points; 95% CI, -11.9 to -1.3 points; P = .01) and day 7 (estimated mean difference, -5.4 points; 95% CI, -10.7 to 0.0 points; P = .05). There were no treatment group differences at 3 months in psychological distress symptoms. White family members experienced a greater reduction in NEST scores compared with Black family members at day 3 (estimated mean difference, -12.5 points; 95% CI, -18.9 to -6.1 points; P < .001 vs estimated mean difference, -0.3 points; 95% CI, -9.3 to 8.8 points; P = .96) and day 7 (estimated mean difference, -9.5 points; 95% CI, -16.1 to -3.0 points; P = .005 vs estimated mean difference, -1.4 points; 95% CI, -10.7 to 7.8; P = .76). Conclusions and Relevance: In this study of ICU patients and family members, a primary palliative care intervention using a mobile application reduced unmet palliative care needs compared with usual care without an effect on psychological distress symptoms at 3 months; there was a greater intervention effect among White family members compared with Black family members. These findings suggest that a mobile application-based intervention is a promising primary palliative care intervention for ICU clinicians that directly addresses the limited supply of palliative care specialists. Trial Registration: ClinicalTrials.gov Identifier: NCT03506438.