Rushabh H. Doshi

Colon cancer stage at diagnosis strongly predicts survival, yet racial disparities remain underexplored for Asian American, Native Hawaiian, and Pacific Islander (AANHPI) subpopulations. In light of COVID-19 and rising anti-Asian hate, we examined temporal changes in colon cancer stage at diagnosis among disaggregated AANHPI populations before and during the pandemic. Using the National Cancer Database, we examined colon cancer stage at presentation (0-IV) among AANHPI subgroups in pre-pandemic (2004–2019) and pandemic (2020–2021) periods. Ordinal logistic regressions with race-year interaction terms evaluated odds of higher-stage diagnosis, adjusting for demographic, socioeconomic, and clinical factors. Among 913,119 patients, 28,845 (3.2%) were AANHPI. The yearly increase in odds of advanced-stage diagnosis was greater for Black than White patients (AOR, 1.008; 95% CI, 1.005–1.010; P < .001). Within-AANHPI subgroups, Thai patients had lower yearly odds of than the Chinese reference (AOR, 0.943, 95% CI, 0.903–0.986; P = .01). During the pandemic, disparities persisted or widened: Chinese (AOR 1.21, 95% CI, 1.05–1.38), Korean (AOR 1.35, 95% CI, 1.11–1.65), Vietnamese (AOR 1.26, 95% CI, 1.04–1.54), and Hmong (AOR 2.15, 95% CI, 1.19–2.36) patients exhibited higher odds of advanced-stage presentation than White patients. Significant heterogeneity exists in colon cancer stage at presentation among AANHPI subgroups. Disparities worsened during the pandemic, showing the importance of addressing subgroup-specific inequities in post-pandemic cancer care.

Introduction: Patients undergoing total shoulder arthroplasty (TSA) may experience postoperative new-onset depression (NOD), which is gaining attention because of its detrimental effect on patient outcomes. Methods: The PearlDiver Mariner database from 2010 through October 2022 was queried for adult patients undergoing first-time TSA. Exclusion criteria included patients with diagnoses of tumors, trauma, and infections within 30 days before surgery and patients with previous depression or psychiatric risk factors for depression. The incidence of NOD diagnoses within 1 year after surgery was identified. Those with and without NOD were exact matched 1:4 based on age, sex, and Elixhauser Comorbidity Index and compared on multivariate logistic regression to determine whether various 90-day postoperative complications were associated with the development of NOD. After Bonferroni correction, significance was established at P < 0.003. Five-year revision rates were also compared with the log-rank test. Results: Overall, 144,773 TSA patients meeting inclusion criteria were identified. NOD diagnoses in the year after surgery were identified for 4,321 (3.0%). Those with NOD were more likely to be female and have greater comorbidity burden ( P < 0.003 for both). Among matched comparisons, those with NOD had higher odds of experiencing various 90-day complications, visits to the emergency department, hospital readmission, and were more likely to undergo revision TSA within 5 years ( P < 0.003). Conclusions: Postoperative NOD was diagnosed in 3% of patients after TSA. Having NOD was associated with various postoperative complications and higher revision rates. Recognizing predisposing factors for NOD may help target risk mitigation strategies.

Background: Chemotherapy at the end of life may not extend survival and can diminish quality of life. Understanding patterns of use is critical to ensuring equitable, patient-centered care. Objective: To examine racial and ethnic disparities in chemotherapy receipt within 30 days of death among patients with colon cancer, and whether these disparities changed over time. Design: We conducted a retrospective cohort study using the National Cancer Database, identifying patients with colon cancer who died within 30 days between 2004 and 2021. The final sample included 43 490 patients. The primary outcome was chemotherapy receipt within 30 days of death. Multivariable logistic regression adjusted for demographics, insurance, comorbidities, facility characteristics, and socioeconomic factors. A difference-in-differences approach compared 2 periods: 2004-2012 and 2013-2021. Results: Among 43 490 patients (mean age, 75.6 years; 51.1% female), 1275 (2.9%) received chemotherapy near death. Adjusted analyses found no statistically significant racial or ethnic differences in use. However, Asian American patients had a marginally increased likelihood of receiving chemotherapy over time compared with non-Hispanic White patients (interaction OR, 2.25; 95% CI, 0.95-5.32; P = .065). Older age (≥80 years: OR, 0.12), higher comorbidity burden (Charlson-Deyo score ≥2: OR, 0.61), and uninsured status were associated with lower odds of treatment. Private insurance (OR, 1.67) and treatment at comprehensive cancer centers were associated with higher odds. Conclusion: While overt racial disparities were not observed, disparities by insurance and facility type point to structural inequities that should be addressed in future quality improvement efforts.

Radiology reports are an integral part of patient medical records; however, these reports often contain complex medical terminology that are difficult for patients to comprehend, potentially leading to anxiety, misunderstanding, and misinterpretation. The development of user-friendly instruments to improve understanding is thus critically important to enhance health literacy and empower patients. In this study, we introduce a novel artificial intelligence (AI) interface, the Rads-Lit Tool, which can simplify radiology reports for patients using natural language processing (NLP) techniques. This manuscript presents the development process, methodology, and results of the Rads-Lit Tool, demonstrating its potential to simplify radiology reports across various examination types and complexity levels. Our findings highlight that patient-facing AI-driven tools can enhance patient health literacy and foster improved patient-provider communication in radiology.

Background: As digital health resources become increasingly prevalent, assessing the quality of information provided by publicly available AI tools is vital for evidence-based patient education. Objective: This study evaluates the accuracy and readability of responses from four large language models—ChatGPT 4.0, ChatGPT 3.5, Google Bard, and Microsoft Bing—in providing contraceptive counseling. Methods: A cross-sectional analysis was conducted using standardized contraception questions, established readability indices, and a panel of blinded OB/GYN physician reviewers comparing model responses to an AAFP benchmark. Results: The models varied in readability and evidence adherence; notably, ChatGPT 3.5 provided more evidence-based responses than GPT-4.0, although all outputs exceeded the recommended 6th-grade reading level. Conclusions: Our findings underscore the need for the further refinement of LLMs to balance clinical accuracy with patient-friendly language, supporting their role as a supplement to clinician counseling.