Bias and Inaccuracy in AI Chatbot Ophthalmologist Recommendations

Cureus, Journal Year: 2023, Volume and Issue: unknown

Published: Sept. 25, 2023

To evaluate the accuracy and bias of ophthalmologist recommendations made by three AI chatbots, namely ChatGPT 3.5 (OpenAI, San Francisco, CA, USA), Bing Chat (Microsoft Corp., Redmond, WA, Google Bard (Alphabet Inc., Mountain View, USA). This study analyzed chatbot for 20 most populous U.S. cities.Each returned 80 total when given prompt "Find me four good ophthalmologists in (city)." Characteristics physicians, including specialty, location, gender, practice type, fellowship, were collected. A one-proportion z-test was performed to compare proportion female recommended each national average (27.2% per Association American Medical Colleges (AAMC)). Pearson's chi-squared test determine differences between chatbots male versus recommendation accuracy.Female (1.61%) (8.0%) significantly less than 27.2% practicing (p<0.001, p<0.01, respectively). fewer (29.5%) (p<0.722). (73.8%), (67.5%), (62.5%) gave high rates inaccurate recommendations. Compared academic (17%), medicine or combined private greater all chatbots.This revealed substantial inaccuracy chatbots' They struggled recommend reliably accurately, with being physicians specialties other ophthalmology not near desired city. showed a significant tendency against recommending ophthalmologists, favored medicine.

Language: Английский

---

Does COVID-19 Infection Increase the Risk of Diabetes? Current Evidence

Current Diabetes Reports, Journal Year: 2023, Volume and Issue: 23(8), P. 207 - 216

Published: June 7, 2023

Abstract Purpose of Review Multiple studies report an increased incidence diabetes following SARS-CoV-2 infection. Given the potential global burden diabetes, understanding effect in epidemiology is important. Our aim was to review evidence pertaining risk incident after COVID-19 Recent Findings Incident by approximately 60% compared patients without Risk also non-COVID-19 respiratory infections, suggesting SARS-CoV-2-mediated mechanisms rather than general morbidity illness. Evidence mixed regarding association between infection and T1D. associated with elevated T2D, but it unclear whether persistent over time or differs severity time. Summary diabetes. Future should evaluate vaccination, viral variant, patient- treatment-related factors that influence risk.

Language: Английский

---

Excess natural-cause mortality in US counties and its association with reported COVID-19 deaths

Proceedings of the National Academy of Sciences, Journal Year: 2024, Volume and Issue: 121(6)

Published: Feb. 1, 2024

In the United States, estimates of excess deaths attributable to COVID-19 pandemic have consistently surpassed reported death counts. Excess non-COVID-19 natural causes may represent unrecognized deaths, caused by health care interruptions, and/or from pandemic’s socioeconomic impacts. The geographic and temporal distribution these help evaluate which explanation is most plausible. We developed a Bayesian hierarchical model produce monthly natural-cause mortality for US counties over first 30 mo pandemic. From March 2020 through August 2022, 1,194,610 occurred nationally [90% PI (Posterior Interval): 1,046,000 1,340,204]. A total 162,886 (90% PI: 14,276 308,480) were not COVID-19. Overall, 15.8 every 100 deaths. This number was greater in nonmetropolitan (36.0 deaths), West (Rocky Mountain states: 31.6 deaths; Pacific 25.5 South (East Central 26.0 Atlantic 25.0 24.2 deaths). contrast, counts metropolitan New England Middle states. Increases correlated temporally with increases same prior month. suggests that many during States

Language: Английский

---

Trends in Necrotizing Fasciitis‐Associated Mortality in the United States 2003–2020: A CDC WONDER Database Population‐Based Study

World Journal of Surgery, Journal Year: 2025, Volume and Issue: unknown

Published: March 19, 2025

ABSTRACT Background Necrotizing fasciitis (NF) is a severe and rapidly progressing soft tissue infection with high mortality rates. Despite the urgency of this condition, there limited research on long‐term NF‐related trends in United States. Objective This study aims to analyze adults aged 25 older States from 2003 2020, focusing variations by sex, race/ethnicity, geographic region. Methods deaths were identified using CDC WONDER database through ICD‐10 code M72.6. Crude age‐adjusted rates (AAMRs) calculated across demographic groups regions. Temporal assessed joinpoint regression, providing annual percent change (APC) Results From total 19,158 recorded, marking 120.6% increase, rising 824 1842 2020. The overall AAMR increased 0.44 per 100,000 0.71 Males consistently had higher than females both sexes saw sharp rise after 2015. By American Indian or Alaska Native populations exhibited highest rates, followed Black African individuals. Regional revealed that West AAMR, whereas Northeast recorded lowest. A significant was observed all regions 2014. Additionally, urban–rural analysis indicated large central metropolitan areas elevated smaller noncore experienced sharper increases. Conclusions has significantly risen since 2014, distinct disparities based race, Contributing factors may include chronic conditions, healthcare access issues, climate‐related events. Public health interventions early diagnosis, timely treatment, addressing inequities are essential for improving outcomes (highlighted shows corrections).

Language: Английский

---

Estimation of Excess Mortality in Germany During 2020-2022

Cureus, Journal Year: 2023, Volume and Issue: unknown

Published: May 23, 2023

Background This study estimates the burden of COVID-19 on mortality in Germany. It is expected that many people have died because new virus who otherwise would not died. Estimating pandemic by number officially reported COVID-19-related deaths has been proven to be difficult due several reasons. Because this, a better approach, which used studies, estimate calculating excess for years. An advantage such an approach additional negative impacts are covered as well, possible pandemic-induced strain healthcare system. Methods To calculate Germany years 2020 2022, we compare all-cause (i.e., independently underlying causes) with statistically deaths. For state-of-the-art method actuarial science, based population tables, life and longevity trends, from 2022 if there had no pandemic. Results The results show observed was close respect empirical standard deviation; approximately 4,000 occurred. By contrast, 2021, two deviations above even more than four times deviation 2022. In total, year 2021 about 34,000 66,000 deaths, yielding cumulated 100,000 both high mainly increase age groups between 15 79 started accumulate only April onward. A similar pattern stillbirths 9.4% second quarter 19.4% fourth compared previous Conclusions These findings indicate something must happened spring led sudden sustained mortality, although effects during early so far. Possible influencing factors explored discussion.

Language: Английский

---

Inability to get needed health care during the COVID-19 pandemic among a nationally representative, diverse population of U.S. adults with and without chronic conditions

BMC Public Health, Journal Year: 2023, Volume and Issue: 23(1)

Published: Sept. 26, 2023

Abstract Background Delays in health care have been observed the U.S. during COVID-19 pandemic; however, prevalence of inability to get needed and potential disparities access yet be assessed. Methods We conducted a nationally representative, online survey 5,500 American Indian/Alaska Native, Asian, Black/African American, Latino (English- Spanish-speaking), Native Hawaiian/Pacific Islander, White, multiracial adults between 12/2020–2/2021 (baseline) 8/16/2021–9/9/2021 (6-month follow-up). Participants were asked “Since start pandemic, was there any time when you did not medical that needed?” Those who responded “Yes” about type reason for receiving care. Poisson regression used estimate association sociodemographics receive care; all analyses stratified by chronic condition status. Chronic conditions included: obstructive pulmonary disease (COPD), heart conditions, 2 diabetes, kidney or on dialysis, sickle cell disease, cancer, immunocompromised state (weakened immune system). Results Overall, 20.0% participants at baseline 22.7% follow-up reported getting The most common reasons being unable included fear (baseline: 44.1%; follow-up: 47.2%) doctors canceled appointment 25.3%; 14.1%). Routine 59.9%; 62.6%) management 31.5%; 30.1%) often types delayed Fair/poor self-reported physical significantly associated with despite status (≥ 1 condition: aPR = 1.36, 95%CI 1.04–1.78); no conditions: 1.52, 95% CI 1.28–1.80). likelihood differed some instances race/ethnicity, age, insurance For example, uninsured more likely 1.76, 1.17–2.66); 1.25, 1.00–1.56). Conclusions one fifth follow-up. may exacerbate existing perpetuate among vulnerable populations received pandemic.

Language: Английский

---

COVID-19 Information, Trust, and Risk Perception Across Diverse Communities in the United States: Initial Findings from a Multistate Community Engagement Alliance (CEAL)

American Journal of Public Health, Journal Year: 2024, Volume and Issue: 114(S1), P. S112 - S123

Published: Jan. 1, 2024

Objectives. To provide initial findings from Community Engagement Alliance (CEAL), a multistate effort funded by the National Institutes of Health, to conduct urgent community-engaged research and outreach focused on COVID-19 awareness, education, evidence-based response. Methods. We collected survey data (November 2020–November 2022) 21 CEAL teams 29 state regional sites spanning 19 US states, District Columbia, Puerto Rico, which covered priority populations served trusted sources information about COVID-19, including prevention behaviors, vaccination, clinical trials. Results. A disproportionate number respondents were Latino (45%) or Black (40%). There was considerable variability between regarding information, prevention, vaccination. For example, more (70%) reported health care providers as source than any other (ranging 6% 87% site). Conclusions. rapidly developed novel infrastructure engage academic, public health, community organizations address COVID-19’s impacts underserved communities. provides an example how respond in future emergencies quickly promote trustworthy, ways that advance equity. (Am J Public Health. 2024;114(S1):S112–S123. https://doi.org/10.2105/AJPH.2023.307504 )

Language: Английский

---

The Effect of War on STEMI Incidence: Insights from Intensive Cardiovascular Care Unit Admissions

Journal of Clinical Medicine, Journal Year: 2024, Volume and Issue: 13(5), P. 1356 - 1356

Published: Feb. 27, 2024

(1) Background: The impact of armed conflicts on public health is undeniable, with psychological stress emerging as a significant risk factor for cardiovascular disease (CVD). Nevertheless, contemporary data regarding the influence war CVD, and especially acute coronary syndrome (ACS), are scarce. Hence, aim current study was to assess repercussions admission prognosis patients admitted tertiary care center intensive unit (ICCU). (2) Methods: All ICCU during first three months Israel–Hamas (2023) were included compared all same period in 2022. primary outcome in-hospital mortality. (3) Results: A total 556 (184 females [33.1%]) median age 70 (IQR 59–80) included. Of them, 295 (53%) war. Fewer Arab more ST-segment elevation myocardial infraction (STEMI) (21.8% vs. 13.2%, p < 0.001, 31.9% 24.1%, = 0.04, respectively), whereas non-STEMI (NSTEMI) frequently pre-war year (19.3% 25.7%, 0.09). In-hospital mortality similar both groups (4.4% 3.4%, 0.71; HR 1.42; 95% CI 0.6–3.32, 0.4). (4) Conclusions: During war, fewer STEMI ICCU. groups.

Language: Английский

---

2024 ACC/AHA Key Data Elements and Definitions for Social Determinants of Health in Cardiology

Journal of the American College of Cardiology, Journal Year: 2024, Volume and Issue: 84(14), P. e109 - e226

Published: Aug. 26, 2024

Language: Английский

---

Factors associated with delay or avoidance of medical care during the COVID-19 pandemic in Armenia: results from a nationwide survey

BMC Health Services Research, Journal Year: 2024, Volume and Issue: 24(1)

Published: Jan. 4, 2024

Abstract Background The coronavirus disease 2019 (COVID-19) pandemic has disrupted healthcare systems throughout the world. Many patients faced delays and cancellation of care due to scaled back services, mobility restrictions, concerns related risk infection. present study aimed assess prevalence factors associated with avoidance or delay medical COVID-19 in Armenia. Methods We conducted a cross-sectional telephone survey 3,483 adults across used stratified two-stage cluster sampling select participants from different age groups proportionate their size population. Logistic regression analysis assessed association avoidance/delay routine, urgent/emergency, any care. Results mean sample was 49.5 (SD = 14.8), ranging 18 90. About 9.9% respondents avoided/delayed type care; whereas 5.5% urgent/emergency 6.6% routine In adjusted analysis, female gender higher monthly expenditures were Factors delay/avoidance included perceived threat COVID-19. Younger age, gender, not being vaccinated against analysis. Conclusion Since avoiding delaying might increase morbidity mortality conditions COVID-19, identifying population that are more likely avoid is important. Targeting such educational interventions focusing on risks using versus times be crucial. Ensuring provision in-home services for high-risk help address important needs during pandemic.

Language: Английский

---