Severe malaria: update on pathophysiology and treatment.

PURPOSE OF REVIEW: Malaria threatens the lives of over 200 million individuals with the disease each year. Plasmodium falciparum is the predominant cause of severe malaria which may be lethal and result in neurocognitive sequelae despite appropriate treatment. We review recent advances regarding the pathophysiology of severe malaria and treatment recommendations for severe disease in the United States. RECENT FINDINGS: Infected red blood cell (iRBC) sequestration in microvascular beds is a critical factor in the development of severe malaria syndromes. Interactions between iRBC variant adhesive peptides and the endothelial protein C receptor (EPCR) result in perturbations of coagulation and cytopreservation pathways. Alterations in the protein C/EPCR axis are implicated in cerebral malaria, respiratory distress, and anemia. Brain MRIs reveal the posterior reversible encephalopathy syndrome in cerebral malaria patients. Transcriptomic analysis reveals commonalities in disease pathogenesis in children and adults despite differences in clinical presentation. US guidelines for severe malaria treatment currently recommend intravenous artesunate including in pregnant women and children. SUMMARY: Despite advances in our understanding of malarial pathogenesis much remains unknown. Antimalarial agents eradicate parasites but no treatments are available to prevent or ameliorate severe malaria or prevent disease sequelae. Further study is needed to develop effective adjunctive therapies.

Hidden flaws behind expert-level accuracy of multimodal GPT-4 vision in medicine.

Recent studies indicate that Generative Pre-trained Transformer 4 with Vision (GPT-4V) outperforms human physicians in medical challenge tasks. However, these evaluations primarily focused on the accuracy of multi-choice questions alone. Our study extends the current scope by conducting a comprehensive analysis of GPT-4V's rationales of image comprehension, recall of medical knowledge, and step-by-step multimodal reasoning when solving New England Journal of Medicine (NEJM) Image Challenges-an imaging quiz designed to test the knowledge and diagnostic capabilities of medical professionals. Evaluation results confirmed that GPT-4V performs comparatively to human physicians regarding multi-choice accuracy (81.6% vs. 77.8%). GPT-4V also performs well in cases where physicians incorrectly answer, with over 78% accuracy. However, we discovered that GPT-4V frequently presents flawed rationales in cases where it makes the correct final choices (35.5%), most prominent in image comprehension (27.2%). Regardless of GPT-4V's high accuracy in multi-choice questions, our findings emphasize the necessity for further in-depth evaluations of its rationales before integrating such multimodal AI models into clinical workflows.

Hidden Flaws Behind Expert-Level Accuracy of Multimodal GPT-4 Vision in Medicine.

Recent studies indicate that Generative Pre-trained Transformer 4 with Vision (GPT-4V) outperforms human physicians in medical challenge tasks. However, these evaluations primarily focused on the accuracy of multi-choice questions alone. Our study extends the current scope by conducting a comprehensive analysis of GPT-4V's rationales of image comprehension, recall of medical knowledge, and step-by-step multimodal reasoning when solving New England Journal of Medicine (NEJM) Image Challenges - an imaging quiz designed to test the knowledge and diagnostic capabilities of medical professionals. Evaluation results confirmed that GPT-4V performs comparatively to human physicians regarding multi-choice accuracy (81.6% vs. 77.8%). GPT-4V also performs well in cases where physicians incorrectly answer, with over 78% accuracy. However, we discovered that GPT-4V frequently presents flawed rationales in cases where it makes the correct final choices (35.5%), most prominent in image comprehension (27.2%). Regardless of GPT-4V's high accuracy in multi-choice questions, our findings emphasize the necessity for further in-depth evaluations of its rationales before integrating such multimodal AI models into clinical workflows.

Prevalence of HTLV-I Infection and Its Association with Tuberculosis among Patients at an Urban Clinic in Haiti.

This retrospective case-control study examined the prevalence of HTLV-I and its association with tuberculosis among urban clinic patients in Haiti. Prevalence of HTLV-I among tuberculosis cases was 2.1% and among controls was 2.4%. Prevalence of HLTV-I was higher in females than males (odds ratio [OR] 2.45, P = 0.020). HTLV-I prevalence in those ≥ 50 years was 8.4% compared with 1.3% in those < 50 (OR 6.74, P < 0.001). We found no association between HTLV-I and tuberculosis in this population.

Temporal Trends in the Clinical Acuity of Patients with ST-Segment Elevation Myocardial Infarction.

BACKGROUND: Despite advances in ST-segment elevation myocardial infarction (STEMI) systems of care over the last decade, studies have shown no improvement in risk-adjusted mortality. It has been hypothesized that the population presenting to the catheterization laboratory has become sicker over time, in ways not accurately captured by current mortality models. The objective of this study was to examine changes in the clinical characteristics and in-hospital case fatality rate of the STEMI population treated with early percutaneous coronary intervention (PCI). METHODS: We conducted a retrospective analysis of a nationwide inpatient database for the period 2004-2012. All patients with a diagnosis of STEMI who underwent PCI within 24 hours of admission were identified. The primary outcome was in-hospital mortality. RESULTS: From 2004 to 2012 there was a consistent increase in unadjusted in-hospital mortality (3.9% in 2004 and 4.7% in 2012, odds ratioyear 1.03; 95% confidence interval 1.01-1.04). During this time there was an increase in the proportion of patients with ≥3 Elixhauser comorbidities (14.8% vs 29.0%, Ptrend < .001). Intubation or cardiac arrest on presentation increased from 3.2% to 7.8% (Ptrend < .001) and had a strong, independent association with mortality. After multivariable adjustment using a model that incorporated the increasing trend in intubation/cardiac arrest, mortality decreased over time (odds ratioyear 0.95; 95% confidence interval 0.94-0.97). CONCLUSIONS: During a period that corresponds to improvement in STEMI quality of care, risk-adjusted in-hospital mortality declined. An increase in comorbidities, and more importantly in the proportion of patients presenting with extreme-risk features, may explain the overall "null" effect regarding in-hospital mortality despite improvements in timely reperfusion.

Characteristics of hospitalizations for cardiogenic shock after acute myocardial infarction in the United States.

BACKGROUND: Multiple studies have reported a decline in mortality for patients with cardiogenic shock after acute myocardial infarction (CS-AMI), a finding which has been attributed to an increase in revascularization over the past decade. However, other studies that have focused on CS-AMI patients treated with early percutaneous coronary intervention (PCI) have found no improvement in risk-adjusted mortality. To reconcile these discordances, we hypothesize that the clinical complexity of the PCI-population has changed over time, in ways not precisely adjusted for in previous studies. METHODS: We conducted a retrospective analysis of the 2005-2012 Nationwide Inpatient Sample. Patients with CS-AMI who underwent PCI within 24h of hospitalization were identified. Temporal trends in clinical characteristics and in-hospital mortality were analyzed. RESULTS: There was no significant change in un-adjusted in-hospital mortality (30% in 2005-2006 and 27.8% in 2011-2012, OR: 0.90; 95% CI: 0.79-1.01, p=0.07). There was an increase in the proportion of patients with ≥3 Elixhauser comorbidities and comorbidity scores ≥5. The population of patients that suffered from cardiac arrest or needed intubation on the first hospital day increased from 27.8% to 42.6% (ptrend<0.001). In a multivariate analysis, mortality rates in 2011-2012 versus 2005-2006 decreased significantly (OR: 0.75; 95% CI: 0.65-0.85, p<0.001). CONCLUSIONS: During a period that corresponds to expanded PCI use and improved prehospital survival, risk-adjusted mortality declined. Much of the survival benefit attributable to early revascularization has been neutralized by an increase in prevalence of "extreme-risk" patients. This may contribute to the null effect on in-hospital mortality.