Do large language models show decision heuristics similar to humans? A case study using GPT-3.5.

A Large Language Model (LLM) is an artificial intelligence system trained on vast amounts of natural language data, enabling it to generate human-like responses to written or spoken language input. Generative Pre-Trained Transformer (GPT)-3.5 is an example of an LLM that supports a conversational agent called ChatGPT. In this work, we used a series of novel prompts to determine whether ChatGPT shows heuristics and other context-sensitive responses. We also tested the same prompts on human participants. Across four studies, we found that ChatGPT was influenced by random anchors in making estimates (anchoring, Study 1); it judged the likelihood of two events occurring together to be higher than the likelihood of either event occurring alone, and it was influenced by anecdotal information (representativeness and availability heuristic, Study 2); it found an item to be more efficacious when its features were presented positively rather than negatively-even though both presentations contained statistically equivalent information (framing effect, Study 3); and it valued an owned item more than a newly found item even though the two items were objectively identical (endowment effect, Study 4). In each study, human participants showed similar effects. Heuristics and context-sensitive responses in humans are thought to be driven by cognitive and affective processes such as loss aversion and effort reduction. The fact that an LLM-which lacks these processes-also shows such responses invites consideration of the possibility that language is sufficiently rich to carry these effects and may play a role in generating these effects in humans. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Chemical exposures and demographic associations in cutaneous T-cell lymphoma: a large single institution physician validated cohort study.

Cutaneous T-cell lymphomas (CTCL) are a rare group of T-cell neoplasms which infiltrate the skin and can result in substantial morbidity and mortality. Risk factors for CTCL are still poorly understood though recent studies suggest chemical exposures may play a role in its development. To further characterize patient-centered risk factors for CTCL, especially compared with matched controls, we performed one of the largest prospective cohort survey studies to date to examine patient-reported exposures and health-related quality of life (HRQoL) in association with concurrent clinical disease characteristics. Patient demographics, lifestyle factors, and chemical exposures were collected via clinical data and surveys. Descriptive statistics, ANOVA, Chi-square tests and t tests were utilized to compare patient-reported exposures and HRQoL in patients with CTCL versus matched healthy controls (HC). Statistically significant differences were identified between patients and HC in terms of race (non-white race 22.4% in CTCL patients vs. 18.8% in HC, P = 0.01), and education level (high school or less 41.6% in CTCL patients vs. 14.3% in HC, P = 0.001), but not with Fitzpatrick skin type (P = 0.11) or smoking status (P = 0.28). Notably, 36.0% of the CTCL patients reported exposures to chemicals, a near threefold increased percentage when compared to HC (12.9%). Among various chemical exposures, 27.0% of the CTCL patients specifically reported industrial chemical exposure, a more than two-fold increased percentage when compared to HC (12.9%). Itch and pain were significantly associated with skin disease severity (as evaluated by CTCL-specific mSWAT score) in advanced stage disease (stages IIB-IVB) (r = 0.48 and 0.57, P < 0.05). Itch and body mass index (BMI) were weakly associated with skin disease severity in early-stage disease (stages IA-IIA) (r = 0.27 and 0.20, P < 0.05).

Lifestyle, demographic and Skindex measures associated with cutaneous T-cell lymphoma: a single institution cohort study.

Cutaneous T-cell lymphoma (CTCL) risk factors and associated quality of life are poorly understood. Previous studies of CTCL risk factors explored patient comorbidities and lifestyle exposures, but not in conjunction with disease stage, subtype, severity, or health-related quality of life (HRQoL). We investigated lifestyle exposures and demographic factors associated with advanced-stage disease, increased disease severity, and poorer HRQoL outcomes in this single-center cohort study. A cohort survey study was conducted at Northwestern's Multidisciplinary Cutaneous Lymphoma specialty clinic between April 2019 and June 2021. REDCap surveys were administered to 140 patients with CTCL, investigating patients' demographics, lifestyle and chemical exposures. QoL was evaluated using the Skindex survey; pain and itch with ten-point Likert scales. Modified Severity Weighted Assessment Tool (mSWAT), disease stage, and disease subtype were confirmed upon enrollment in the study by a single board-certified dermatologist specializing in CTCL. Factors were compared by t test or Fischer's exact test. Median age was 63 years (range 14-92) with male-to-female ratio of 1.2:1. The most common diagnosis was CD4 + MF (n = 94, 67.1%). Common lifestyle exposures included smoking (past or current) (52.3%) and chemical exposure history (all sources [53.7%]; industrial only [33.0%]). History of chemical exposures were associated with advanced stage disease (p = 0.003) and worse QoL outcomes (p = 0.001). There were significant racial differences, respectively, in early (I-IIA) vs late (IIB-IV) stage disease (p = 0.003). Obesity, hygiene, smoking, recent sun exposure, education and atopy were not significantly associated with disease stage or severity. We provide an analysis of lifestyle and demographic factors in the context of CTCL disease severity, stage, and HRQoL. We identified race as a potential risk factor for advanced stage disease and both skin phototype and chemical exposures as a risk factor for increased disease severity as measured by mSWAT. QoL outcomes were multifactorial and significantly associated with history of chemical exposure, severe pain/itch, race, disease stage and subtype. An improved understanding of these associations may lead to better individualized care. As chemical exposure and race were found to be significant factors associated with advanced-stage disease, taking exposure histories and addressing racial disparities may improve care for CTCL patients.

Retrospective analysis of sepsis in cutaneous T-cell lymphoma reveals significantly greater risk in Black patients.

BACKGROUND: Sepsis is a leading cause of morbidity, mortality, and resource utilization among patients with cutaneous T-cell lymphoma (CTCL). OBJECTIVE: To characterize the demographic, clinical, and microbial attributes distinguishing patients with CTCL sepsis from other patients with non-Hodgkin lymphoma (NHL) sepsis and patients with CTCL in general. METHODS: Two-part retrospective cohort study at an academic medical center from 2001-2019 involving patients with CTCL (n = 97) and non-CTCL NHL (n = 88) admitted with sepsis, and a same-institution CTCL patient database (n = 1094). Overall survival was estimated by Kaplan-Meier analyses. RESULTS: Patients with CTCL sepsis were more likely to be older, Black, experience more sepsis episodes, die or be readmitted within 30 days of an inpatient sepsis episode, and develop Gram-positive bacteremia than patients with non-CTCL NHL sepsis. Staphylococcus aureus and Escherichia coli were the most frequently speciated organisms in CTCL (26%) and non-CTCL NHL (14%), respectively. No between-group differences were identified regarding sex, presence of central line, chemotherapy use, or disease stage. Compared with general patients with CTCL, patients with sepsis were Black and exhibited advanced-stage disease, higher body surface area involvement, and higher lactate dehydrogenase levels. LIMITATIONS: Single institution, retrospective nature may limit generalizability. CONCLUSION: Awareness of CTCL-specific risk factors is crucial for guiding sepsis prevention and improving patient outcomes.

The use of social media platforms to discuss and educate the public on allergic contact dermatitis.

BACKGROUND: Social media platforms are increasingly used by patients to research and discuss medical problems. OBJECTIVE: The aim of this study was to identify by whom, how frequently, and in what manner allergic contact dermatitis (ACD) is discussed on social media sites. METHODS: Search terms "allergic contact dermatitis" and "contact dermatitis" were queried across Twitter, Instagram, Reddit, Facebook, YouTube, and Google search metrics. The frequency, content, and creators of the content were assessed. RESULTS: ACD content was identified on all platforms, generated by by patients, physicians, professional organizations, and companies. When comparing the volume of posts, more content was on Instagram than Twitter, particularly among patients. Patient support groups were identified on Facebook but not on Reddit. A formal analysis of YouTube videos found that the medical information presented in these videos was often of poor quality. CONCLUSIONS: Patch testing physicians should be aware that information on ACD exists across social media sites. While some content is generated by physicians, patients and industry groups also post and share material. Patch testing physicians should know that there is an opportunity to share ACD information, but they should also be aware that patients are posting and creating online support communities independent of physicians. HIGHLIGHTS: Online support communities exist for patients with ACD on Facebook. Online information is of lower quality, with a mean QUEST quality score of 7.4/28 on reviewed YouTube videos. Of the five social media sites reviewed, patients are most active on Instagram, Reddit, and Facebook. Approximately 9000 Google searches per month are conducted using contact dermatitis-related search terms. Differences in terminology exist between physicians and non-physicians. While the most popular hashtag term was "contact dermatitis", physicians disproportionately authored posts tagged with "allergic contact dermatitis."

Use of technology for the objective evaluation of scratching behavior: A systematic review.

INTRODUCTION: Pruritus is a common symptom across various dermatologic conditions, with a negative impact on quality of life. Devices to quantify itch objectively primarily use scratch as a proxy. This review compares and evaluates the performance of technologies aimed at objectively measuring scratch behavior. METHODS: Articles identified from literature searches performed in October 2020 were reviewed and those that did not report a primary statistical performance measure (eg, sensitivity, specificity) were excluded. The articles were independently reviewed by 2 authors. RESULTS: The literature search resulted in 6231 articles, of which 24 met eligibility criteria. Studies were categorized by technology, with actigraphy being the most studied (n = 21). Wrist actigraphy's performance is poorer in pruritic patients and inherently limited in finger-dominant scratch detection. It has moderate correlations with objective measures (Eczema and Area Severity Index/Investigator's Global Assessment: rs(ρ) = 0.70-0.76), but correlations with subjective measures are poor (r2 = 0.06, rs(ρ) = 0.18-0.40 for itch measured using a visual analog scale). This may be due to varied subjective perception of itch or actigraphy's underestimation of scratch. CONCLUSION: Actigraphy's large variability in performance and limited understanding of its specificity for scratch merits larger studies looking at validation of data analysis algorithms and device performance, particularly within target patient populations.

A skin-conformable wireless sensor to objectively quantify symptoms of pruritus.

Itch is a common clinical symptom and major driver of disease-related morbidity across a wide range of medical conditions. A substantial unmet need is for objective, accurate measurements of itch. In this article, we present a noninvasive technology to objectively quantify scratching behavior via a soft, flexible, and wireless sensor that captures the acousto-mechanic signatures of scratching from the dorsum of the hand. A machine learning algorithm validated on data collected from healthy subjects (n = 10) indicates excellent performance relative to smartwatch-based approaches. Clinical validation in a cohort of predominately pediatric patients (n = 11) with moderate to severe atopic dermatitis included 46 sleep-nights totaling 378.4 hours. The data indicate an accuracy of 99.0% (84.3% sensitivity, 99.3% specificity) against visual observation. This work suggests broad capabilities relevant to applications ranging from assessing the efficacy of drugs for conditions that cause itch to monitoring disease severity and treatment response.