BACKGROUND:Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, with significant associated morbidity and mortality. The pathophysiology underlying AF is complex, and this has limited the development of effective therapies. ADAMs (a disintegrin and metalloproteinase) are transmembrane proteinases with diverse functions involving cell-cell communication and extracellular matrix homeostasis. ADAM15 specifically has been implicated in AF pathogenesis in large-scale genetic association studies. Given the importance of atrial remodelling in AF pathogenesis, we sought to ascertain the effect of ADAM15 deficiency in murine atrial biology. METHODS:We performed extensive cardiac phenotyping in Adam15 knockout (Adam15-/-) mice; wildtype mice were used as controls. These studies included invasive electrophysiology, echocardiography, optical mapping and biochemical assays. RESULTS:ADAM15 deficiency resulted in increased AF susceptibility without overt abnormalities in cardiac structure or function by echocardiography. These findings were associated with shortened atrial action potential duration, increased atrial fibrosis, increased TGFβ signalling, as well as reduced connexin expression. CONCLUSIONS:Here, we provide evidence that ADAM15 deficiency is associated with increased AF susceptibility and atrial remodelling in mice. Further work will be needed to define the molecular mechanisms underlying these findings and determine their relevance to human AF.
BACKGROUND:Andes virus (ANDV) has recently emerged as a significant public health concern following a multi-country outbreak aboard the MV Hondius cruise ship in April 2026, with 13 identified cases and three deaths among passengers and crew. METHODS:We reviewed available literature from PubMed and Google Scholar databases and searched international health agencies, including the World Health Organisation and national public health authorities, for current outbreak information and up-to-date epidemiological data. RESULTS:ANDV is a zoonotic orthohantavirus endemic in southern Chile and Argentina, with Oligoryzomys longicaudatus as its principal reservoir. Approximately 300 cases of hantavirus cardiopulmonary syndrome occur annually in the Americas. The virus causes severe disease through endothelial dysfunction and vascular leakage, resulting in rapidly progressive pulmonary edema and shock. Human-to-human transmission is characteristic of ANDV and occurs through respiratory and salivary routes during prolonged close contact. Clinical presentation includes nonspecific prodromal symptoms followed by rapid cardiopulmonary deterioration. Diagnosis relies on serological assays and RT-qPCR. Management is predominantly supportive, emphasising hemodynamic monitoring and early consideration of extracorporeal membrane oxygenation in severe cases. CONCLUSIONS:Despite ANDV's severity and high mortality, its epidemic potential seems to be low. However, the recent outbreak underscores the importance of international surveillance coordination and rapid diagnostic capabilities for emerging zoonotic infections.
BACKGROUND:Artificial intelligence (AI) is transforming global health care through innovations in deep learning, generative models and agentic AI systems. Traditional reductionist approaches to complex pathophysiological pathways fail to capture the true complexity of disease, motivating the adoption of network medicine (NM), which models biological systems as dynamic, interconnected networks. When combined with AI, NM enables integration of multiomic data and better characterizes disease mechanisms to guide precision therapies. Nevertheless, the rapid diffusion of AI in health care also raises profound ethical, regulatory and social challenges, since only a small fraction of AI tools have achieved routine clinical use, often due to limited generalizability, opaque algorithms and workflow incompatibility. METHODS:Key strategies in this scenario are explainable AI (XAI) and counterfactual reasoning, which enhance transparency, accountability and fairness. These methods allow clinicians and regulators to interpret model decisions, identify biases and ensure human oversight in high-stakes contexts. Ethical frameworks such as the European Commission's Assessment List for Trustworthy AI (ALTAI) operationalize these goals through auditable requirements spanning transparency, fairness, privacy and societal well-being. RESULTS:Yet, challenges persist, including algorithmic bias, data inequity and variable regulatory standards across regions. Machine learning and generative AI hold significant promise for improving diagnostics, drug discovery and population health, but their deployment must be guided by fairness, transparency and human rights principles. Indeed, inadequate governance risks can impact the already existing health inequalities. CONCLUSIONS:For these reasons, the convergence of AI, NM and telemedicine requires a co-evolutionary model which must be rooted in ethical design, rigorous validation and equitable global implementation. Only by aligning technical innovation with sound ethical frameworks and explainability standards will AI become a highly transformative yet trustworthy force in the field of precision and public health.
BACKGROUND:People living with HIV (PWH) are at increased risk for metabolic and cardiovascular diseases. However, there is no structured prevention strategy for these subjects. The triglyceride-glucose (TyG) index has shown predictive value for Type 2 diabetes mellitus (T2DM) and cardiovascular events (CVEs) in the general population, but its utility in PWH remains unclear. METHODS:We conducted a retrospective study including 477 PWH without baseline T2DM. TyG index was calculated as follows: ln[fasting triglycerides (mg/dL) × fasting glucose (mg/dL)]/2. ROC analysis was used to identify the optimal TyG threshold for T2DM, and logistic regression analysis was used to evaluate the association with T2DM. The association between TyG index and CVEs was investigated using Kaplan-Meier survival analysis and Cox proportional hazards regression models. RESULTS:477 PWH were included. Median age was 38 years and 26% were women. During a median follow-up of 14.2 years, 38 participants developed T2DM and 45 experienced CVEs. The TyG index demonstrated fair discrimination for incident T2DM (AUC: 0.675, 95% CI 0.591-0.759), with an optimal threshold of 4.78. Participants with TyG ≥ 4.78 had a higher risk of developing T2DM (aOR 2.86, 95% CI 1.35-6.12). PWH in the highest TyG tertile had also higher risk for CVEs compared with those in the lowest tertile (aHR 3.31, 95% CI 1.10-9.97). CONCLUSION:In PWH, a higher TyG index was independently associated with an increased risk of both T2DM and CVEs over long-term follow-up. The TyG index represents a practical, low-cost tool for early metabolic risk stratification in this population.
BACKGROUND:Early neuroprognostication in Out-of-Hospital Cardiac Arrest (OHCA) remains challenging due to the lack of reliable biomarkers. Circulating microRNAs have emerged as potential indicators of inflammation and brain injury (HIBI). This study aimed to evaluate whether in silico-predicted microRNAs could discriminate neurological outcomes early in OHCA patients. METHODS:Consecutive OHCA patients admitted to the Acute Cardiac Care Unit with available cryopreserved blood samples were analysed. Neurological outcome was classified into favourable (Cerebral Performance Category (CPC) 1-2) and unfavourable (CPC 3-5). Blood samples at 0, 24, 72 h after hospital admission were collected. An in silico-derived network of inflammation-HIBI-related microRNA was generated and validated in all patients at the three time-points. RESULTS:A total of 100 patients were included (59.7 ± 12.5 years, 19% women). 53% had unfavourable neurological outcomes. Eleven candidate microRNAs were identified, showing differential time-dependent expression within 72 h. The earlier and most predictive microRNAs for unfavourable neurological outcome were miR-1-3p (AUC 0.637) and miR-124-3p (AUC 0.606) at 0 h, miR-21-5p (AUC 0.614) and miR-499a-5p (AUC 0.645) at 24 h. The miR-499-5p maintained the best AUC across all time-points. A multiparametric model achieved an AUC of 0.690. A pro-inflammatory environment, characterised by C-Reactive Protein (CRP) and neutrophil-to-lymphocyte ratio, showed distinct temporal expression patterns regarding neurological outcome. Neuroinflammatory miR-let-7a-5p, miR-23a-3p and miR-146a-5p were significantly overexpressed in the unfavourable group and correlated positively with CRP. CONCLUSIONS:In silico derived inflammatory and HIBI-related microRNAs showed time-dependent expression patterns that may distinguish neurological outcomes within 72 h after OHCA.
BACKGROUND:Evaluative measures, helping to decide whether a marginal graft can be used for transplantation or rather be discarded are of utmost importance but yet basically underdeveloped. Recent technical developments allow for the isolated detection of 13C-labelled carbon dioxide in the oxygenator exhaust of an isolated perfused organ. By adding a 13C-labelled tracer, its metabolism can be monitored by measuring the increase in 13C carbon in the exhaust air. Here we evaluated the potential of the technique to evaluate pig livers during machine perfusion. METHODS:Porcine livers (six per group) were subjected to 5, 20 or 45 min of warm ischemia and cold stored overnight. Evaluative machine perfusion was done with MPS at 20°C for 2 h. After injection of 13C-methacetin as a tracer, 13CO2 as metabolic end-product can be quantified by high precision laser-based spectroscopy in the gas outflow of the oxygenator. RESULTS:The 13CO2 test showed clear differences between all three groups, and provided superior discriminative potential when compared to conventional parameters (e.g., flow, transaminases, lactate). ROC curve analysis for discrimination between 5 and 20 min of warm ischemia revealed an area under the curve of 0.916 while comparison of 20 and 45 min of warm ischemia results in an area under the curve of 1.0 (p < 0.01). CONCLUSION:Thus, the 13C turnover test appears to have the potential of an objective, quantitative and valuable adjunct in the toolbox for pre-transplant evaluation of questionable liver grafts.
BACKGROUND:Calcific aortic stenosis (AS) is the most relevant valvular heart disease in aging populations and remains both underdiagnosed and undertreated. Screening studies reveal a high prevalence of unrecognized disease, while up to 40% of patients with symptomatic severe AS do not undergo valve intervention, resulting in poor survival. No pharmacologic therapy has yet altered disease progression, although targeted approaches such as lipoprotein(a) inhibition are under investigation. METHODS:This paper provides a comprehensive perspective review of the evolving epidemiology, timing of intervention, and long-term management of aortic stenosis, drawing on randomized trials and contemporary clinical data. RESULTS:Randomized trials have challenged the traditional watchful waiting strategy in asymptomatic severe AS, showing that early aortic valve replacement (AVR) reduces cardiovascular events, though survival benefit depends on patient selection and timing. Moderate AS is increasingly recognized as a high-risk condition, prompting ongoing trials evaluating earlier intervention. As transcatheter aortic valve implantation (TAVI) expands to younger, lower-risk patients, long-term issues-including conduction disturbances, paravalvular leak, cerebrovascular events, valve hemodynamics and durability, and redo feasibility-become central to care. Contemporary data suggest comparable mid-term durability between TAVI and surgery, but lifetime management planning is essential. CONCLUSIONS:Optimal care of aortic stenosis requires earlier diagnosis, optimized procedural strategies, and a lifetime-oriented approach to management, supported by continued investigation into pharmacologic therapy and the timing of intervention.
BACKGROUND:Early oral step-down has been proposed as a strategy to optimize antimicrobial use in Gram-negative bloodstream infections (BSIs). METHODS:We performed a systematic review and meta-analysis of full-text studies published in English and indexed in MEDLINE and EMBASE up to April 2025. Randomized controlled trials (RCTs) and observational studies including patients with Gram-negative BSIs managed either with oral step-down within 5 days of IV therapy or with continued IV therapy were eligible. Risk of bias was assessed using the Cochrane Risk of Bias Tool for RCTs and the Newcastle-Ottawa Scale for observational studies. RESULTS:Of 4120 screened records, 13 studies (3 RCTs and 10 cohort studies) including 10,000 patients met the inclusion criteria; 5016 of enrolled patients (52%) switched to oral therapy, mainly to fluoroquinolones (2992, 59.6%). A lower all-cause mortality was observed in patients receiving oral step-down (RR: 0.39, 95% CI: 0.16-0.93; p = 0.034). When stratifying studies according to timing of the switch to oral therapy, no significant difference in mortality was observed in studies with a median switch time ≤ 3 days (p = 0.29). Conversely, studies with a median switch time > 3 days showed a lower mortality rate in the oral step-down group, approaching statistical significance (p = 0.051). No significant differences were found between treatment groups in terms of clinical failure (p = 0.28). Microbiological failure was significantly less frequent among patients receiving oral step-down therapy (RR = 0.69, 95% CI: 0.49-0.97; p = 0.032), while adverse event rates were comparable (p = 0.52). However, sensitivity analyses limited to propensity score-adjusted data demostrated no significant differences in mortality (p = 0.32) or microbiological failure (p = 0.076). CONCLUSIONS:Oral step-down therapy for Gram-negative BSIs appears to be a valuable alternative to prolonged IV treatment, potentially reducing IV exposure, complications and healthcare costs without compromising clinical or microbiological outcomes. TRIAL REGISTRATION:PROSPERO registration number: CRD420251056335.
BACKGROUND:Tafamidis is the only approved disease-modifying therapy for transthyretin amyloid cardiomyopathy (ATTR-CM) in Australia. Although its clinical efficacy is established, its effects on cardiac remodelling assessed by contemporary multimodality imaging remain incompletely defined. METHODS:MEDLINE, Embase and Cochrane databases were systematically searched from 2000 to 2025. Observational studies evaluating tafamidis-treated versus untreated/control patients that reported serial TTE, CMR or bone-tracer scintigraphy measurements were eligible. Change-from-baseline values were used to derive measures. Outcomes were pooled as standardised mean differences (SMDs) with 95% confidence intervals using random-effects modelling. A two-sided p < 0.05 was considered statistically significant. RESULTS:Over a mean follow-up of 16 months, tafamidis was associated with significant improvements in left atrial reservoir strain (LASr) (SMD +0.58, 95% CI +0.27 to +0.89; p < 0.001; I2 = 0) on TTE and right ventricular ejection fraction (RVEF) (SMD +0.51, 95% CI +0.16 to +0.86; p = 0.004; I2 = 0) on CMR. A reduction in left atrial volume index (LAVi) was observed (SMD -0.24, 95% CI -0.46 to -0.02; p = 0.034; I2 = 0) on TTE. No consistent effects were seen for LV parameters including LVEF, LV-GLS, LV mass index, wall thickness, diastolic indices, pulmonary pressures, tissue characterisation markers (native T1, ECV), or H/CL ratio, despite trends toward benefit in some. CONCLUSIONS:Tafamidis demonstrates selective remodelling effects in ATTR-CM, with the most benefits noted in left atrial function and RV systolic performance, rather than LV structural or functional parameters. Imaging parameters such as LASr, LAVi, and RVEF may serve as sensitive endpoints of therapeutic response in future clinical trials, offering mechanistic insights beyond conventional LV metrics. STUDY REGISTRATION:PROSPERO CRD420251145891.
BACKGROUND:Acute myocarditis is an inflammatory disease of the myocardium with an annual incidence of 4-14 per 100,000 individuals, predominantly affecting young adults. Its clinical features are frequently nonspecific, mimicking acute coronary syndrome, which makes early recognition and management challenging. The disease results from infectious (predominantly viral) and noninfectious triggers, including autoimmune disorders, immune checkpoint inhibitors and mRNA vaccines. Pathophysiology involves a dysregulated interplay between innate immunity and adaptive immunity. CLINICAL FEATURES AND DIAGNOSIS:Presentation is typically dominated by chest pain, with dyspnea and syncope reported less frequently. Cardiac magnetic resonance (CMR), with the 2018 updated Lake Louise criteria, has become the cornerstone of noninvasive diagnosis, whereas endomyocardial biopsy (EMB), in experienced centres, remains the gold standard for histological characterization and guiding immunosuppressive therapy. OUTCOMES AND MANAGEMENT:Uncomplicated myocarditis usually resolves spontaneously. However, approximately 25% of patients with myocarditis have left ventricular systolic dysfunction, ventricular arrhythmias or acute heart failure. Mortality ranges from 1% to 7%, depending on presentation, aetiology and specific populations. Treatment centers on guideline-directed heart failure therapy, with immunosuppression reserved for complicated presentations and virus-negative, autoimmune or histologically specific subtypes. Mechanical circulatory support is critical in fulminant cases, where mortality is high. CONCLUSIONS:This clinical review synthesizes recent guideline updates and emerging trial data, primarily from literature published in the past 10 years, to support a phenotype-driven approach to acute myocarditis, in which management is guided by clinical severity, suspected aetiology, selective use of CMR and EMB and targeted therapy. Ongoing trials investigating corticosteroids, targeted biologics and novel therapies may further refine personalized immunomodulatory strategies.
BACKGROUND:Pulsed field ablation (PFA) is an effective alternative to radiofrequency ablation (RFA) for pulmonary vein isolation (PVI) in atrial fibrillation (AF), offering procedural advantages and a favourable safety profile. However, single-shot PFA (ssPFA) has been associated with biochemical evidence of haemolysis and myocardial injury, raising concerns regarding potential renal impairment. METHODS:In this prospective observational study, 40 patients undergoing first-time PVI were analysed (20 RFA and 20 pentaspline ssPFA). Blood samples were obtained pre-procedure and 1, 4 and 24 h post-ablation. Markers of myolysis (myoglobin), haemolysis (haptoglobin and bilirubin), renal function (creatinine, eGFR and Cystatin C) and tubular stress (NGAL and KIM-1) were assessed. Biomarker trajectories were analysed using two-way repeated-measures ANOVA and correlation analyses were performed. RESULTS:Significantly higher postprocedural myoglobin levels, greater reductions in haptoglobin and greater elevations in bilirubin were observed after ssPFA than after RFA (all interaction effects p < 0.001). Changes in renal function and tubular stress biomarkers were modest and largely comparable between modalities. Exploratory correlation analyses demonstrated inverse associations between changes in myoglobin and eGFR, with the strongest correlation observed at 24 h (ρ = -0.55). The association between early myoglobin release and subsequent eGFR changes was confined to the ssPFA cohort. CONCLUSIONS:ssPFA induces substantially greater acute myolysis and haemolysis than RFA following PVI. No patient met serum creatinine-defined criteria of acute kidney injury within the observation period. Myoglobin release was associated with transient renal functional changes and may represent a mechanistically plausible, non-causal marker of early renal stress following ssPFA.
BACKGROUND:Heart rate variability (HRV) has remained a relatively finite and niche tool in cardiology despite decades of research supporting its physiological and clinical relevance. This limited adoption may resemble the early history of electrocardiography (ECG), which was initially regarded by many physicians as a laboratory instrument rather than a routine clinical tool. The delayed acceptance of ECG reflected technological limitations, cultural resistance and the need for clinicians to master unfamiliar concepts derived from physics and electrophysiology. HRV faces comparable barriers today. Although derived from ECG RR intervals, HRV requires interpretation of time-domain, frequency-domain, geometric and nonlinear indices that may appear mathematically complex and distant from conventional bedside reasoning. AIMS:We argue that HRV should not be viewed as a replacement for ECG, but as an extension of ECG from electrical morphology to physiological dynamics. MATERIALS AND METHODS:Lessons from ECG history were examined and compared with the current state of HRV adoption in clinical practice. The complementary diagnostic roles of ECG morphology and HRV analysis were considered, together with the potential contribution of wearable sensors, remote monitoring, artificial intelligence and large language models to facilitate HRV interpretation, education and clinical integration. RESULTS:Whereas conventional ECG morphology identifies arrhythmias, conduction disturbances, ischemic alterations and overt electrical abnormalities, HRV provides insight into autonomic modulation, cardiovascular adaptability and systemic physiological regulation. The emergence of wearable sensors, remote monitoring, artificial intelligence and large language models creates an opportunity to overcome barriers that have limited HRV adoption. Artificial intelligence may serve as an educational and interpretive bridge, translating complex HRV metrics into clinically meaningful concepts while supporting medical training, artefact awareness, case-based learning and workflow integration. DISCUSSION:HRV faces barriers comparable to those encountered during the early adoption of ECG, including technological limitations, educational challenges and resistance to incorporating unfamiliar physiological concepts into routine clinical practice. CONCLUSION:Lessons from ECG history suggest that HRV adoption will depend not only on evidence but also on standardization, education, clinical interpretation and cultural acceptance within cardiology.
BACKGROUND:Immune checkpoint inhibitors (ICI) are widely used in cancer therapy, but biomarkers for thromboembolic risk, treatment response, and survival remain limited. We evaluated the association of systemic inflammatory indices with these clinical outcomes. METHODS:In this retrospective cohort study, 580 ICI-treated patients at the Medical University of Vienna, Austria, were included. Inflammatory indices including neutrophil-lymphocyte-ratio (NLR), platelet-lymphocyte-ratio (PLR), lymphocyte-monocyte-ratio (LMR), systemic immune-inflammation-index (SII) and C-reactive-protein-albumin-ratio (CAR) were calculated at ICI-start and longitudinally within 3 months. Co-primary outcomes were risk of venous thromboembolism (VTE), assessed using competing risk analysis, and overall- and progression-free-survival (OS/PFS), evaluated with Cox regression, whereas longitudinal biomarker dynamics were analysed using time-dependent analyses. RESULTS:Higher baseline NLR (HR 1.65, 95% CI: 1.28-2.12), CAR (HR 1.99, 95% CI: 1.67-2.38), and SII (HR 1.18, 95% CI: 1.00-1.39) were independently associated with shorter OS, while higher LMR (HR 0.44, 95% CI: 0.30-0.67) was associated with longer survival. For PFS, elevated NLR (HR 1.29, 95% CI: 1.05-1.57), higher PLR (HR 1.18, 95% CI: 1.00-1.39), lower LMR (HR 0.57, 95% CI: 0.43-0.70), and higher CAR (HR 1.42, 95% CI: 1.28-1.62) predicted poorer outcomes. None of the indices at treatment start were associated with VTE, yet doubling of levels within 3 months indicated higher VTE risk for PLR (sub-distribution hazard ratio [SHR]: 3.03 95% confidence interval [CI]: 1.05-8.08) and a borderline-significant association for CAR (SHR: 2.02, 95% CI: 0.89-4.60). CONCLUSION:We found that selected inflammatory indices were associated with poor OS and PFS in ICI-treated patients, and longitudinal dynamics might identify patients at higher risk for VTE.
BACKGROUND:Growth differentiation factor-15 (GDF-15) is upregulated in coronary artery disease (CAD). The relationship of intensive statin therapy with circulating GDF-15 levels has not been explored yet. In the current secondary analysis of an observational cohort, we investigated whether high-intensity statin treatment is associated with GDF-15 in patients with advanced CAD. METHODS:In 102 CAD patients we measured GDF-15 levels before and after median 7 [6-8] months from initiation atorvastatin 80 mg/day (n = 63, 61.8%) or rosuvastatin 40 mg/day (n = 39, 38.2%). At the two time points we also determined C-reactive protein (CRP), thrombin generation and fibrinolysis inhibitors. There was no concurrent control group. RESULTS:At baseline GDF-15 levels (median 942; interquartile range 639-1284 pg/mL) were associated solely with age (R = 0.567, p < 0.001) and CRP (R = 0.464, p < 0.001), but not with lipid profile. On high-intensity statin therapy GDF-15 reduction by 17.8% was observed (p = 0.006), which was associated with CRP lowering by 46.4% (R = 0.788, p < 0.001), but not with changes in total cholesterol (by 17.6%) and low-density lipoprotein cholesterol (LDL-C) (by 28.1%). However, GDF-15 levels were lower by 28.5% (p < 0.001) and 18.9% (p = 0.013) in 43 patients (42.2%) who achieved the target LDL-C < 1.8 mmol/L and 19 (18.6%) with LDL-C < 1.4 mmol/L, respectively, as compared to the remainder. The type of statin was not associated with follow-up GDF-15. Age (p = 0.029), baseline GDF-15 (p < 0.001) and CRP (p < 0.001), but not LDL-C, were independently associated with follow-up GDF-15. CONCLUSIONS:We demonstrated that high-dose statin therapy is associated with lower GDF-15 in CAD patients and this effect is largely related to the anti-inflammatory properties of statins.
BACKGROUND:Sex-specific differences regarding the prognostic impact of permanent pacemaker (PM) implantation after transcatheter aortic valve implantation (TAVI) remain largely unexplored. OBJECTIVES:To investigate sex-specific associations between pre-existing and new post-procedural PM implantation and short- and long-term mortality after TAVI. METHODS:This post hoc analysis of a prospective registry included consecutive patients who underwent TAVI between 2016 and 2020. Patients were stratified by sex and assigned to three groups: no PM, pre-existing PM (pre-PM) and PM implantation within 30 days of TAVI (post-PM). The primary endpoint was 5-year all-cause mortality; the secondary endpoint was 30-day mortality. Sex-stratified multivariable proportional hazards models and Kaplan-Meier analyses were performed. RESULTS:Among 1114 patients (81 ± 5 years, 49.8% female), women more frequently had no PM (80.2% vs. 70.8%, p < 0.001) and less often had a pre-existing PM (7.0% vs. 14.5%, p < 0.001). Post-PM rates were similar between sexes (12.8% vs. 14.7%, p = 0.172). Thirty-day survival differed by PM status only in women, with worse outcomes in the post-PM group (log-rank p = 0.033), whereas no difference was observed in men (p = 0.542). Five-year survival did not differ by PM status among women (p = 0.572) but differed significantly among men (p < 0.001), with the lowest survival in men with pre-PM. CONCLUSION:Post-procedural PM implantation was associated with higher 30-day mortality in women, whereas pre-existing PM predicted reduced 5-year survival in men, supporting sex-specific risk assessment before and after TAVI.
The g formula is a cornerstone method for estimating causal effects of time-varying treatments using longitudinal observational data in the presence of time-varying confounders that are affected by prior treatment. Standard regression techniques often fail in this setting because adjustment for such covariates can distort the very effects under study. The g-formula addresses this problem by expressing the mean potential outcome under specified static or dynamic treatment regimes as a function of the joint distribution of covariates, treatments and outcomes, which can be approximated via parametric or semi-parametric models and simulation. This review presents the g-formula, emphasizing intuitive explanations. After outlining the causal framework and the core identification assumptions-consistency, sequential exchangeability and positivity-the article describes practical parametric g-formula: model specification for covariate and outcome processes, implementation via forward simulation, and the interpretation of marginal causal contrasts between clinically relevant regimes. The g-formula is then situated within the family of g-methods alongside inverse probability weighting and targeted maximum likelihood estimation, highlighting complementary strengths and limitations. A dedicated section discusses concrete applications, including analyses of highly active antiretroviral therapy and AIDS or death, dynamic 'when to start' antiretroviral strategies in HIV and electronic health record-based evaluations of blood pressure treatment targets. Practical guidance on modelling choices, diagnostics and transparent reporting is provided to support applied researchers considering g-formula in clinical and epidemiological investigations.
Science is a self-correcting process, and scientific research aims to improve adequacy, accuracy, and utility. However, improving scientific research is a demanding task, especially when currently some key principles of science are challenged and renegotiated. Key challenges in the current environment include the increasing loss of trust in scientists; the production of most of the published scientific literature in countries without full democracy and/or in countries without fundamental freedoms, e.g., freedom of the press; limited public availability and transparency as most research is funded by non-public sponsors that do not prioritize or even seek publication of results; and rapid developments on the frontier of artificial intelligence where non-human agents can supplement and/or replace human researchers. Concurrently, there have been many proposals on how to improve research. Among a plethora of suggestions and guidance, some may not be useful or may even be harmful, and most lack evidence. Revisiting some key proposals shows mixed track records of failures and successes. Examples are provided from efforts to enhance collaboration, team science, and large studies; replication culture; registration and open science; containment of conflicts of interest; and statistical, computational, and informatics improvements. The optimal stages to improve research (early and/or late in the scientific process) may be debated, and the role, function, and mode of optimal peer review are also under scrutiny. Eventually, science and scientific research are demanding, hard enterprises. Genuine progress requires openness, honesty, and selflessness.
