Explore the vast range of titles available.

Introduction: Abdominal pain is the leading emergency department (ED) chief complaint in older (≥65 years of age) adults, accounting for 1.4 million ED visits annually. Ultrasound and computed tomography (CT) are high-yield tests that offer rapid and accurate diagnosis for the most clinically significant causes of abdominal pain. In this study we used nationally representative data to examine racial/ethnic differences in cross-sectional imaging for older adults presenting to the ED with abdominal pain. Methods: We performed a retrospective, cross-sectional analysis using data from the National Hospital Ambulatory Medical Care Survey (NHAMCS) to assess differences in the rate of imaging between White and Black older adults presenting to the ED for abdominal pain. Our primary outcome was the receipt of abdominal CT and/or ultrasound imaging. Results: Across 1,656 older adult ED visits for abdominal pain, White patients were 26.8% (relatively, 14.2% absolute) more likely to receive abdominal CT and/or ultrasound than Black patients: 802 of 1,197 (67.0%) White patients were 26.8% (relatively, 14.2% absolute) more likely to receive abdominal computed tomography and/ or ultrasound than Black patients (P=0.01). Conclusion: This study revealed that Black older adults presenting to the ED with abdominal pain receive significantly lower levels of cross-sectional imaging (CT/ultrasound) than White patients. Our findings highlight the need for further investigations into causes of disparities while initiating quality improvement processes to assess and address site- and clinician-specific patterns of care.

Universal health system-based screening for mild cognitive impairment (MCI) or impairment consistent with dementia can modify care and increase access to disease-modifying therapies and memory centers, potentially reducing disparities. Constraints on nurse and clinician time have hindered existing efforts. Large language models (LLMs) present a potential solution. However, they are subject to confabulation and poor arithmetic accuracy, and were not trained on data containing definitive assessments of cognitive impairment. We hypothesized that using an LLM to administer the Short Blessed Test (SBT) and calculating test scores from item scores would mitigate these errors. We conducted prompt engineering on OpenAI GPT4o to construct an interactive tool. A team member then simulated being patients with different levels of cognitive impairment across an orthogonal matrix spanning the 6 SBT item scores, answering each question to achieve the intended item score. We summed the LLM-reported item scores to the total score that should have been obtained based on the input answers. We evaluated 57 sets of item scores with total scores of 2-26. Of these, 54.7% had an LLM score equal to the true score. All errors except 2 were within 3 points of the true score. Using a cutoff of >=5 points for mild cognitive impairment, specificity was 100.0%, and negative predictive value was 50%. LLMs may play a role in reducing nurse and clinician screening burden in hospital settings, where the benefit of being able to conduct screening at all outweighs the possibility of missed diagnoses, but requires human confirmation.

To evaluate the prevalence and predictors of microalbuminuria in diabetics in Kumasi, Ghana. Prospective, cross-sectional study of diabetic patients. Patients with diabetes, 20 to 78 years of age. Microalbuminuria All patients (109) attending an outpatient diabetic clinic at the Komfo Anokye Teaching Hospital Diabetes Centre in Kumasi, Ghana from January to July 2005 were enrolled in the study. The mean overall age of the cohort was 54.1 +/- 10.9 years, and 28% were male. The proportion of subjects who had microalbuminuria was 43.1% (n=47). The median duration of diabetes before development of microalbuminuria was 10 years. Duration of diabetes, blood urea nitrogen, serum concentration of creatinine, and triglyceride were significantly higher in patients with microalbuminuria (P<.05). Urinary potassium concentration and fractional excretion of potassium were also significantly higher in the patients with microalbuminuria. The prevalence of microalbuminuria in patients with diabetes in this study was 43%. Significant predictors of microalbuminuria included duration of diabetes and serum concentration of creatinine. To reduce renal failure among these patients, strategies to mitigate its occurrence are needed. This includes strict glycemic control, control of hypertension, and the early blockade of the renin-angiotensin system.