Juan Sebastian Izquierdo-Condoy

Introduction: Coronary artery fistulas are abnormal communications between a coronary artery and a cardiac chamber or great vessel. Fistulas connecting the left circumflex coronary artery to the pulmonary artery are uncommon, and their clinical relevance requires assessment of both anatomical and physiological characteristics. Case presentation: We report the case of a 62-year-old woman with hypertension, dyslipidemia, and first-degree atrioventricular block who presented with recurrent exertional angina. Examination revealed an early diastolic murmur at the aortic area and a prominent midsystolic murmur with a palpable thrill radiating toward the right lung base. Echocardiography showed moderate aortic regurgitation with preserved left ventricular systolic function, while dobutamine stress echocardiography showed no inducible wall-motion abnormalities. She subsequently developed recurrent angina with new T-wave inversions and mildly elevated high-sensitivity cardiac troponin T concentrations that remained unchanged across serial measurements. Coronary angiography demonstrated nonobstructive epicardial coronary arteries and a fistula arising from the atrioventricular nodal branch of the distal left circumflex artery and draining into the right pulmonary artery. Coronary steal was suspected but not physiologically confirmed. Following individualized assessment, the Heart Team recommended closure; however, the patient declined intervention and remained free of recurrent angina during approximately five months of follow-up. Conclusions: This case highlights the distinction between an anatomically confirmed coronary artery fistula and its unconfirmed physiological significance. Negative stress echocardiography and alternative potential contributors to symptoms should be considered when interpreting suspected coronary steal. Management requires individualized multidisciplinary assessment and shared decision-making.

UNSTRUCTURED Artificial intelligence (AI), particularly generative AI, is rapidly transforming medical education by driving advancements in adaptive learning, clinical simulation, and comprehensive literature synthesis. In Latin America, these technologies serve as a vital incentive for educational innovation, offering a unique opportunity to give access to high-quality training materials in resource-constrained settings. However, the responsible integration of AI extends far beyond mere student adoption or technological efficiency. It demands a paradigm shift for medical educators, who must now actively master these tools to supervise their application, reimagine assessment methods, and safeguard academic integrity, all while ensuring that fundamental clinical reasoning remains at the core of medical training. To inform this Perspective, we conducted a targeted evidence search across PubMed, SciELO, and LILACS on June 30, 2026, evaluating empirical studies on AI utilization, readiness, and curricular integration among Latin American medical educators. Strikingly, the search revealed a profound gap in regional data, yielding only a single relevant study focused on medical specialty professors in Mexico. This scarcity reveals that the educators who must guide AI adoption remain critically underrepresented in the literature. We introduce a framework for AI integration in Latin American medical education. Our approach begins by identifying faculty readiness and sequentially builds competence in AI literacy, pedagogical application, and assessment redesign, all while rigorously upholding academic integrity and equity. Most importantly, this framework positions medical educators as critical mediators of human–AI interaction, preserving human agency, clinical reasoning, and academic integrity while enabling responsible AI-supported learning. The future of AI in our region depends no longer on monitoring student adoption, but rather on empowering educators and institutions to champion its responsible integration.

Background: Multivariable early-warning models can integrate physiological signals and missingness patterns, but added complexity may not provide meaningful value over simple predictors. We evaluated whether a single prespecified L2-regularized logistic-regression model using physiological and missingness predictors improved the prediction of sustained hypotension beyond recent mean arterial pressure (MAP). Methods: We conducted a retrospective pilot prognostic-model study using the MIMIC-IV Clinical Database Demo on FHIR 2.1.0. Predictors were derived from the preceding 12 h, and sustained hypotension was defined as MAP < 65 mmHg for at least two consecutive hours during the following six hours. A single prespecified L2-regularized logistic-regression model using eight physiological or derived predictors and ten missingness-related predictors was compared with six-hour mean MAP. Performance was estimated using patient-grouped nested cross-validation with five outer and three inner folds. The primary comparison was the difference in area under the precision–recall curve (AUPRC); advancement required an absolute gain of ≥0.020. Paired uncertainty was estimated using 1000 patient-cluster bootstrap replicates. A prespecified sensitivity analysis recalculated AUPRC after assigning equal total weight to each patient, irrespective of the number of eligible windows contributed. Results: Ninety-two patients contributed 8970 eligible windows, of which 1774 (19.8%) were positive. The candidate achieved an AUPRC of 0.629 versus 0.622 for the comparator and an AUROC of 0.845 versus 0.842. The AUPRC difference was +0.0075 (95% bootstrap interval, +0.0005 to +0.0171). The prespecified patient-equal sensitivity analysis yielded a larger delta AUPRC point estimate of +0.0156; both the patient-equal and window-weighted estimates remained below the +0.020 advancement margin. In a post hoc analysis, 81.6% of positive windows had observed MAP below 65 mmHg during t − 5 through t. Among windows without observed recent hypotension, the delta AUPRC was +0.0030 (95% interval, −0.0288 to +0.0238). The 8970 prediction windows were overlapping repeated observations nested within 92 patients and should not be interpreted as independent samples. Conclusions: In this pilot internal-validation study, the single prespecified L2-regularized logistic-regression model using eight physiological or derived predictors and ten missingness-related predictors produced a small improvement in discrimination over six-hour mean MAP but did not meet the prespecified development-advancement margin. These findings apply only to this model specification, dataset, and outcome and do not establish the performance of multivariable vital-sign models generally. Neither strategy was validated for clinical use.

An earthquake is an event that affects populations across multiple dimensions, extending far beyond trauma and immediate mortality. Cardiovascular events are particularly important during the acute phase and in the first days after the disaster, when physiological stress, disruption of healthcare services, medication interruption, and barriers to timely medical attention may increase cardiovascular risk. Previous evidence has documented increases in myocardial infarction, stroke, arrhythmias, cardiovascular mortality, and decompensation of chronic cardiovascular disease following major earthquakes. In Colombia, the recent earthquake affecting Chocó and neighboring regions highlights the particular vulnerability of populations facing geographic isolation and unequal access to healthcare. Disaster preparedness should therefore extend beyond trauma care and include continuity of cardiovascular medications, preservation of emergency referral pathways, capacity to manage sudden increases in cardiovascular demand, and structured follow-up after the acute phase. This article reviews these cardiovascular consequences and discusses their implications for disaster preparedness in Colombia.

Sepsis is a life-threatening, biologically heterogeneous syndrome in which immune dysregulation, endothelial injury, coagulation activation, redox imbalance, and metabolic reprogramming contribute to microvascular and organ dysfunction. This structured narrative review integrates these five domains to inform biomarker-enriched trials of host-directed interventions, emphasizing the endothelium as an interface between inflammation, immunothrombosis, and impaired cellular bioenergetics. Candidate biomarkers of immune competence, endothelial injury, coagulation, and metabolism are examined alongside investigational immunomodulatory and restorative strategies. The proposed framework distinguishes three functional tiers: screening and biological enrichment; pharmacodynamic target engagement and safety assessment; and patient-centered outcomes. Candidate host-response profiles are hypothesis-generating, may overlap, and require prospective validation, with explicit provisions for mixed and indeterminate classifications. A complementary implementation model separates routine assessments from trial-specific assays and advanced mechanistic substudies, addressing sampling, reporting, missing measurements, and assay harmonization. Evidence remains heterogeneous and largely preclinical, observational, or early-phase; predictive thresholds, therapeutic windows, and patient-centered benefits remain unestablished for most candidate strategies. This framework supports prospective trial design rather than routine bedside treatment selection. Biomarker improvement must be evaluated alongside pathogen control, organ recovery, adverse events, and longer-term outcomes and should not be assumed to indicate clinical benefit.

Severe fungal diseases are an increasingly important yet underrecognized cause of global morbidity and mortality, with the greatest impact among immunocompromised individuals and populations with limited access to timely diagnosis and treatment. Disease outcome is determined by a dynamic host–fungus interface in which innate pattern-recognition receptors sense fungal cell-wall ligands and coordinate complement activation, phagocyte responses, and cytokine signaling, while adaptive Th1- and Th17-mediated immunity supports fungal containment and clearance. Pathogenic fungi counter these defenses through cell-wall remodeling and antigen masking, morphological transitions, capsule and biofilm formation, resistance to oxidative stress, complement subversion, interference with phagosome maturation, intracellular persistence, and colonization of immune-privileged sites. These mechanisms are deployed in species-specific combinations by Candida, Aspergillus, Cryptococcus, Histoplasma, Coccidioides, Blastomyces, Mucorales, and other medically important fungi, shaping tissue tropism, persistence, dissemination, and disease severity. Therapeutic control is further undermined by intrinsic and acquired resistance to azoles, echinocandins, and polyenes through target modification, efflux-pump upregulation, compensatory cell-wall remodeling, and altered ergosterol biosynthesis. This review integrates the mechanisms of antifungal immune recognition with common and species-specific evasion strategies and discusses antifungal resistance, emerging host-directed therapies, and vaccine platforms. Defining how fungal adaptation intersects with immune dysfunction and drug resistance may reveal conserved therapeutic targets and support rational combination strategies. Translating these insights into effective antifungal therapies, safe immunomodulatory interventions, and broadly protective vaccines will be essential to improve clinical outcomes and strengthen preparedness, particularly in resource-limited settings.

Thyrotoxic periodic paralysis (TPP) is a reversible but potentially severe complication of thyrotoxicosis characterized by acute muscle weakness and hypokalemia resulting predominantly from intracellular potassium redistribution. Although classically associated with Asian populations, TPP is increasingly recognized in Latin America. We report a 38-year-old Colombian man with recurrent nocturnal proximal quadriparesis, preserved deep tendon reflexes, serum potassium of 2.3 mEq/L, and biochemical thyrotoxicosis secondary to Graves disease. In parallel, we conducted a structured review of published Colombian cases identified through searches of PubMed/MEDLINE, Scopus, Web of Science, SciELO, and Google Scholar from inception through 1 May 2026. Fourteen previously published cases were identified, yielding 15 cases including the index patient. All patients were male, with a mean age of 31.4 ± 7.4 years. Deep tendon reflexes were normal in 4/13 (30.8%) cases in which they were documented, demonstrating that preserved reflexes do not exclude TPP. TSH was suppressed below 0.05 mIU/L in 14/15 (93.3%) patients. Electrocardiographic findings were reported in only 7/15 cases, and antithyroid drug doses in 6/15, highlighting substantial heterogeneity in reporting. All reported patients recovered muscle strength following acute management. TPP should therefore be considered in patients presenting with acute proximal weakness and hypokalemia even when deep tendon reflexes are preserved or overt manifestations of thyrotoxicosis are absent. Standardized reporting of biochemical, electrocardiographic, therapeutic, and follow-up data is needed to improve characterization of TPP in underrepresented populations.