This report originates from a rural group practice in Friuli-Venezia-Giulia, Italy, facing significant workforce shortages and fragmented care, tending to high patient loads and challenging diagnostic continuity. We present a case where systemic constraints and cognitive biases delayed diagnosing severe heart failure in a 61-year-old woman.A 61-year-old female smoker with hypertension and obesity presented with a six-week persistent non-productive cough and exertional dyspnea. Despite lacking fever or purulent sputum, she was diagnosed with respiratory infections by multiple physicians and prescribed three antibiotic courses (amoxicillin-clavulanate, azithromycin, levofloxacin) without improvement. Subsequently, a new GP noted bilateral basal crackles and jugular venous distension. Trusting a clinical intuition that "something smelled wrong" despite the patient's naive cardiac history, the GP ordered NT-proBNP test. The result was 1,161 pg/mL. She was hospitalized for acute decompensated heart failure. Echocardiography revealed a dilated left ventricle with an ejection fraction (LVEF) of 28%. Coronary angiography was normal. She was discharged on guideline-directed medical therapy.Persistent cough in patients with cardiovascular risk factors warrants cardiac investigation when respiratory treatments fail. Clinical intuition remains a vital safeguard against algorithmic errors. Our practice encourages testing for NT-proBNP for unexplained respiratory symptoms and has implemented structured handoffs to mitigate the risks of fragmented care.Diagnostic reasoning is vulnerable to anchoring bias, especially when systemic pressures reduce consultation time. This case demonstrates that biomarkers like NT-proBNP are essential tools to objectively validate clinical suspicion and prevent diagnostic delays.A common symptom like cough can mask life-threatening pathology. In rural settings with workforce shortages, maintaining a high index of suspicion and trusting clinical intuition are crucial for patient safety.

In 2025, the American College of Cardiology (ACC) and the American Heart Association (AHA), in partnership with eleven other academic societies, updated the guidelines for hypertension prevention, detection, and management. A major revision is the endorsement of the PREVENT risk model—which integrates cardiovascular, renal, and metabolic parameters—as a tool to support risk communication and therapeutic decision-making.To assess the predictive performance, calibration, and discrimination of the PREVENT model for cardiovascular disease (CVD) in a contemporary Chinese population.We used data from the China Health and Retirement Longitudinal Study (CHARLS) from 2011 to 2020. Primary outcomes were incident CVD and non-cardiovascular mortality. Discrimination was evaluated using Harrell’s C‑statistic under competing risk analysis; calibration was assessed via the slope of predicted versus observed risk. The PREVENT model was compared with the Framingham risk score, and the incremental value of adding glycated hemoglobin (HbA1c) was examined.Among 6,942 adults free of baseline CVD (mean age 58.1 years; 54.4% female), 1,792 CVD events occurred over a median 9‑year follow‑up. The PREVENT model showed modest discrimination for CVD events (C‑statistic = 0.609, 95% CI 0.532–0.686; AUC under competing risk = 0.596, 95% CI 0.581–0.611) and systematic underestimation of risk (calibration slope = 1.67, 95% CI 1.06–2.37). Notably, it performed better for predicting non‑CVD death (AUC = 0.760). Compared with the Framingham score, PREVENT offered slight improvements in discrimination (C‑statistic 0.598 vs. 0.583), calibration (slope 0.82 vs. 0.36), and net reclassification (overall NRI = 0.035; men = 0.139). Adding HbA1c did not meaningfully enhance performance.PREVENT showed limited discrimination and systematic CVD risk underestimation, though modestly outperforming Framingham. Competing mortality and population risk differences may reduce transportability. Recalibration and local refitting are needed before clinical risk classification; HbA1c added little.In this community-based Chinese cohort, the PREVENT model exhibited limited discriminative ability and a tendency to underestimate CVD risk. Although it outperformed the Framingham score, its accuracy remains insufficient for direct clinical use in China without prior recalibration and external validation using local population data.

Identifying patients with chest pain at risk of acute coronary syndromes (ACS) or other major adverse cardiovascular events (MACE) remains challenging in urgent out-of-hours primary care (OOH-PC). To support clinical decision making, the HEART score is used in emergency departments. The acronym HEART refers to History, Age, ECG, Risk factors, and Troponin, with each component scored from 0-2 points, for a maximum of ten. This score includes an electrocardiogram (ECG), which is often unavailable and not recommended by Dutch general practitioner guidelines.This study evaluates determinants of ECG use and its added diagnostic value within the HEART score in OOH-PC.We conducted a diagnostic accuracy study involving consecutive patients presenting with chest pain at two OOH-PC centers with ECG capabilities in the Netherlands, from March 2023 to August 2025. The primary endpoint was major adverse cardiac events (MACE), including ACS, at six weeks after the index consultation, adjudicated blindly by two independent researchers. We analyzed factors associated with ECG use and compared the added diagnostic value of the HEART score, which includes ECG points assigned for repolarization disorders (1 point) or significant ST deviation (2 points), versus the HART score, which excludes ECG.A total of 285 patients were included, 8.8% experienced MACE; 67 (23.5%) underwent ECG evaluation at the OOH-PC. ECG use was not associated with specific patient characteristics. Among patients with ECG, the c-statistics for the HEART and HART scores were 0.89 (95%CI 0.78-1.00) and 0.91 (95%CI 0.81-1.00), respectively. At a threshold of three points, the HEART score showed a sensitivity of 100% (95%CI 54.1%-100%), specificity of 24.6% (95% CI 14.5%-37.3%), negative predictive value (NPV) of 100%, and positive predictive value (PPV) of 11.5%. The three point cut-off of the HART score yielded a sensitivity of 100% (95%CI 54.1%-100%), specificity of 26.2% (95%CI 15.8%-39.1%), NPV of 100%, and PPV of 11.8%..In Dutch out-of-hours primary care, ECG use in patients with acute chest pain was limited and not guided by patient characteristics. When incorporated into structured risk scores with high-sensitivity troponin, ECG appears to offer limited added value. Funding: Dutch Heart Foundation (03-003-2022-0029)

There is a lack of knowledge about sex differences in the diagnostic trajectory of patients presenting during office hours with chest discomfort to the general practitioner (GP). We aimed to compare the diagnostic care pathways initiated by the GP for women and men presenting with chest discomfort. A sex-specific analysis was conducted within a prospective registry involving 118 GPs in XXX and XXX. All patient contacts were registered during a two-week period, detailing the diagnostic steps for individuals presenting with chest discomfort. A 30-day follow-up was used to collect the outcomes of diagnostic tests and final diagnoses. Preliminary results. Among 22,294 consultations, 1.26% (n=280) concerned chest discomfort (56.4% women; 43.6% men, p=0.030). Women were less often referred (33.5% vs. 48.4%, p=0.012) and underwent fewer diagnostic investigations (38.6% vs. 50.0%, p=0.057) compared to men. During 30-day follow-up, none of the women were diagnosed with an acute coronary syndrome (ACS) or other life-threatening events (LTE). In contrast, eight men (6.9%) had an LTE; 5 an acute myocardial infarction, 2 unstable angina pectoris, and one pulmonary embolism. There were three clear misdiagnoses in men; a pneumonia (n=1) and only suspicion of cardiac disease (n=2) had to be rectified later in the disease trajectory into an ACS. No such cases of misdiagnosis occurred in women.Women presenting with chest discomfort in primary care were less frequently referred, underwent less diagnostic testing, and had lower rates of ACS or other life-threatening diagnoses compared to men.Our results are insufficient to determine whether gender differences exist in the adequacy of diagnostic assessment of patients presenting with chest discomfort to GPs during office hours. Future studies with larger patient populations are needed to clarify this potential difference.

Drug-induced QT interval prolongation is a well-recognized risk factor for life-threatening arrhythmias such as torsades de pointes (TdP). Despite the widespread prescription of QT-prolonging drugs, especially in primary care, the incidence and prevalence of drug-induced long QT syndrome (diLQTS) remain poorly characterized, particularly in outpatient settings.The aim of this study was to evaluate the incidence and prevalence of diLQTSnin cases where screening was performed in adults taking QT-prolonging drugs.A systematic review of studies reporting incidence or prevalence of drug-induced QT prolongation in patients undergoing ECG monitoring was conducted in PubMed database. Observational and interventional studies assessing QTc >500 ms or ΔQTc ≥60 ms were included. Study quality was evaluated using Newcastle-Ottawa Scale for observational studies and RoB 2.0 tool for interventional studies.Out of 501 records, 42 studies met inclusion criteria: 5 were interventional and 37 were observational (15 prospective cohorts, 14 retrospective cohorts, 8 cross-sectional), totaling 14756 patients. Methadone was the most frequently studied drug (17 studies), with reported QTc >500 ms incidence ranging from 0% to 16.2%, and TdP incidence of up to 3.17%. Other frequently studied drugs included hydroxychloroquine (5 studies), azithromycin (4), and antipsychotics such as haloperidol and quetiapine. Incidence of QTc >500 ms varied considerably across settings and drug classes, influenced by heterogeneous QTc correction formulas. Only four studies reported TdP cases, all associated with methadone, droperidol, or tyrosine kinase inhibitors such as Cabozantinib and Dasatinib.Drug-induced QTc prolongation is not uncommon across a range of widely used medications, yet its monitoring and risk remain inconsistently assessed—particularly in outpatient settings. While torsades de pointes was rarely reported, QTc > 500 ms occurred with notable frequency. Significant heterogeneity in QTc definitions, study populations, and ECG methodologies limits cross-study comparability, underscoring the need for greater standardizationAlthough TdP was rare, several studies reported that once diQTLS was identified, clinicians responded with therapeutic modifications—such as stopping or substituting the suspected medication. For clinicians—particularly in primary care—targeted QTc monitoring and individualized risk stratification, potentially supported by portable ECG technologies, may help improve patient safety.

Despite its complexity as a symptom, severity of pain score plays a prominent in clinical practice. Pain score has long been used at out-of-hours primary care serives (OHS-PC) for triaging patients with chest discomfort in the Netherlands but its value was questioned.To examine (i) how patients with chest discomfort report pain scores during telephone triage, (ii) how triagists record these scores, and (iii) the risk of acute coronary syndrome (ACS) for these reported and recorded scores.In a cross-sectional study we used data from adult patients calling with chest discomfort to 7 OHS-PC in the Netherlands. Backed up recordings of telephone triage conversations we re-listened. Pain scores mentioned by patients were compared to pain scores recorded by triagists using descriptive statistics. Final diagnoses were retrieved from health records.We included 2,322 triage conversations with mean age 56.5 (SD 19.4) years, 54.0% were women. In 1,510 (65.0%) of all 2,322 triage conversations, the patient had rated pain score on a scale of 0 to 10. In 2,062 (88.8%) of all conversations, a pain score was registered by the triagist in 1 of 3 categories (no/light pain, moderate pain, severe pain). In the 1,250 triage conversations where the pain score was both registered and reported the scores were identical in 80.9% of the cases. A lower pain score was registered by the triagist than reported by the patient in 160 (12.8%) conversations, and a higher pain score in 48 patients (3.8%). Risk of ACS was highest for highest pain scores registered by the triagist, irrespective of the score reported by the patient.Our study shows that pain scores often are not explicitly reported by patients but estimated instead or, even when collected, up- or downscaled by triagists to tinker the outcome of triage. This reflects both the inherently subjective nature of pain and communicating about pain, and the limitations of applying a single-point score with a strict cutoff.Patient-reported pain scores offer little reliable information and have no practical value for triaging patients with chest discomfort.