Explore the vast range of titles available.

ObjectivePatients with congenital heart disease (CHD) have an increased cancer risk. The aim of this study was to determine cancer-related mortality in CHD patients compared with non-CHD controls, compare ages at cancer diagnosis and death, and explore the most fatal cancer diagnoses.DesignRegistry-based cohort study.Setting and participantsCHD patients born between 1970 and 2017 were identified using Swedish Health Registers. Each was matched by birth year and sex with 10 non-CHD controls. Included were those born in Sweden with a cancer diagnosis.ResultsCancer developed in 758 out of 67814 CHD patients (1.1%), with 139 deaths (18.3%)—of which 41 deaths occurred in patients with genetic syndromes. Cancer was the cause of death in 71.9% of cases. Across all CHD patients, cancer accounted for 1.8% of deaths. Excluding patients with genetic syndromes and transplant recipients, mortality risk between CHD patients with cancer and controls showed no significant difference (adjusted HR 1.17; 95% CI 0.93 to 1.49). CHD patients had a lower median age at cancer diagnosis—13.0 years (IQR 2.9–30.0) in CHD versus 24.6 years (IQR 8.6–35.1) in controls. Median age at death was 15.1 years (IQR 3.6–30.7) in CHD patients versus 18.5 years (IQR 6.1–32.7) in controls. The top three fatal cancer diagnoses were ill-defined, secondary and unspecified, eye and central nervous system tumours and haematological malignancies.ConclusionsCancer-related deaths constituted 1.8% of all mortalities across all CHD patients. Among CHD patients with cancer, 18.3% died, with cancer being the cause in 71.9% of cases. Although CHD patients have an increased cancer risk, their mortality risk post-diagnosis does not significantly differ from non-CHD patients after adjustements and exclusion of patients with genetic syndromes and transplant recipients. However, CHD patients with genetic syndromes and concurrent cancer appear to be a vulnerable group.

Survival in left-sided valvular heart disease (VHD; aortic stenosis [AS], aortic regurgitation [AR], mitral stenosis [MS], mitral regurgitation [MR]) in out-of-hospital cardiac arrest (OHCA) is unknown. We studied all cases of OHCA in the Swedish Registry for Cardiopulmonary Resuscitation. All degrees of VHD, diagnosed prior to OHCA, were included. Association between VHD and survival was studied using logistic regression, gradient boosting and Cox regression. We studied time to cardiac arrest, comorbidities, survival, and cerebral performance category (CPC) score. We included 55,615 patients; 1948 with AS (3,5%), 384 AR (0,7%), 17 MS (0,03%), and 704 with MR (1,3%). Patients with MS were not described due to low case number. Time from VHD diagnosis to cardiac arrest was 3.7 years in AS, 4.5 years in AR and 4.1 years in MR. ROSC occurred in 28% with AS, 33% with AR, 36% with MR and 35% without VHD. Survival at 30 days was 5.2%, 10.4%, 9.2%, 11.4% in AS, AR, MR and without VHD, respectively. There were no survivors in people with AS presenting with asystole or PEA. CPC scores did not differ in those with VHD compared with no VHD. Odds ratio (OR) for MR and AR showed no difference in survival, while AS displayed OR 0.58 (95% CI 0.46–0.72), vs no VHD. AS is associated with halved survival in OHCA, while AR and MR do not affect survival. Survivors with AS have neurological outcomes comparable to patients without VHD.

BackgroundBoth Takotsubo syndrome (TS) and ST-elevation myocardial infarction (STEMI) are conditions characterised by the acute onset of left ventricular (LV) dysfunction. While LV thrombus is a known complication of LV dysfunction, its epidemiology in these two patient groups remains poorly understood.MethodsWe used data from the Stunning in Takotsubo versus Acute Myocardial Infarction (STAMI) study, which prospectively enrolled patients with TS and STEMI at Sahlgrenska University Hospital. Serial echocardiography was performed on admission and on days 1, 2, 3, 7, 14 and 30. Predictors of LV thrombus were identified using Cox regression analyses.Results314 patients were included; 68 with TS, 148 with anterior STEMI and 98 with non-anterior STEMI. Mean LV ejection fraction (LVEF) at admission was 39% (95% CI 35.8 to 42.2) in TS, 46.7% (95% CI 43.3 to 50.1) in anterior STEMI and 52.8% (95% CI 48.9 to 56.7) in non-anterior STEMI. LV thrombus occurred in 20 of 246 (8.1%) STEMI patients but in none of the TS patients. All but one LV thrombus was found in anterior STEMI. All LV thrombi in anterior STEMI were detected within 7 days, while the single non-anterior LV thrombus was found on day 30. All patients with LV thrombi received anticoagulation. Predictors of LV thrombus included lower LVEF and higher troponin levels.ConclusionsDespite more severe LV dysfunction in TS compared with STEMI, LV thrombus was exclusively found in STEMI patients. Almost all LV thrombi were found in anterior STEMI within the first week and showed a high-resolution rate at 30 days. Our findings highlight pathophysiological differences between these two conditions, warranting further investigation and implications for differing surveillance needs after TS and STEMI.

ObjectiveIn-hospital cardiac arrest (IHCA) presents a critical challenge with low survival rates and limited prediction tools. Despite advances in resuscitation, predicting 30-day survival remains difficult, and current methods lack interpretability for timely decision-making. This study developed a machine learning (ML) model to predict 30-day survival after IHCA, using peri-arrest variables available on the rescue team’s arrival, while ensuring a balance between predictive accuracy and clinical interpretability through Shapley Additive Explanations (SHAP).DesignA nationwide, registry-based observational study.SettingData were sourced from the Swedish Cardiopulmonary Resuscitation Registry (2010–2020), merged with the Patient Registry.ParticipantsWe analysed 25 905 IHCA cases with attempted resuscitation, of which 8166 patients survived for 30 days.Outcome measure and analysis30-day survival after IHCA was the outcome measure. An ML model was developed using fivefold cross-validation. Key predictors were identified through in-built variable importance and validated using SHAP. Model performance was evaluated with metrics such as area under the receiver operating characteristics (AUROC), calibration, sensitivity, specificity, false negative rate (FNR) and F-score.ResultsThe CatBoost model achieved an AUROC of 0.9136 (95% CI 0.9075 to 0.9191) with all features, and 0.9034 (95% CI 0.8955 to 0.9037) with the top 15 features, along with Brier scores of 0.1028 and 0.1103, respectively. Performance plateaued after including the top 15 predictors, with few key variables, such as epinephrine administration, age, initial rhythm, ROSC within 15 min, breathing on rescue team arrival and witnessed cardiac arrest, being most influential. The model showed strong calibration for patients with low predicted survival probabilities and demonstrated high sensitivity with a low FNR across relevant survival thresholds.ConclusionThe CatBoost model provides an effective and interpretable tool for predicting 30-day survival after IHCA. Key predictors such as epinephrine administration, age and initial rhythm inform clinical decision-making. This model has strong clinical utility and can be externally validated via the open-access Application Programming Interface (API) at www.gocares.se

ObjectiveThis study aimed to examine the association between exercise workload, resting heart rate (RHR), maximum heart rate and the risk of developing ST-segment elevation myocardial infarction (STEMI).MethodsThe study included all participants from the UK Biobank who had undergone submaximal exercise stress testing. Patients with a history of STEMI were excluded. The allowed exercise load for each participant was calculated based on clinical characteristics and risk categories. We studied the participants who exercised to reach 50% or 35% of their expected maximum exercise tolerance. STEMI was adjudicated by the UK Biobank. We used Cox regression analysis to study how exercise tolerance and RHR were related to the risk of STEMI.ResultsA total of 66 949 participants were studied, of whom 274 developed STEMI during a median follow-up of 7.7 years. After adjusting for age, sex, blood pressure, smoking, forced vital capacity, forced expiratory volume in 1 s, peak expiratory flow and diabetes, we noted a significant association between RHR and the risk of STEMI (p=0.015). The HR for STEMI in the highest RHR quartile (>90 beats/min) compared with that in the lowest quartile was 2.92 (95% CI 1.26 to 6.77). Neither the maximum achieved exercise load nor the ratio of the maximum heart rate to the maximum load was significantly associated with the risk of STEMI. However, a non-significant but stepwise inverse association was noted between the maximum load and the risk of STEMI.ConclusionRHR is an independent predictor of future STEMI. An RHR of >90 beats/min is associated with an almost threefold increase in the risk of STEMI.

Background Takotsubo syndrome (TTS) is an acute heart failure syndrome with symptoms similar to acute myocardial infarction. TTS is often triggered by acute emotional or physical stress and is a significant cause of morbidity and mortality. Predictors of mortality in patients with TS are not well understood, and there is a need to identify high-risk patients and tailor treatment accordingly. This study aimed to assess the importance of various clinical factors in predicting 30-day mortality in TTS patients using a machine learning algorithm. Methods We analyzed data from the nationwide Swedish Coronary Angiography and Angioplasty Registry (SCAAR) for all patients with TTS in Sweden between 2015 and 2022. Gradient boosting was used to assess the relative importance of variables in predicting 30-day mortality in TTS patients. Results Of 3,180 patients hospitalized with TTS, 76.0% were women. The median age was 71.0 years (interquartile range 62–77). The crude all-cause mortality rate was 3.2% at 30 days. Machine learning algorithms by gradient boosting identified treating hospitals as the most important predictor of 30-day mortality. This factor was followed in significance by the clinical indication for angiography, creatinine level, Killip class, and age. Other less important factors included weight, height, and certain medical conditions such as hyperlipidemia and smoking status. Conclusions Using machine learning with gradient boosting, we analyzed all Swedish patients diagnosed with TTS over seven years and found that the treating hospital was the most significant predictor of 30-day mortality.

Aims Individuals with congenital heart disease (CHD) are at an increased risk for cancer. As cancer survival rates improve, the prevalence of late side effects, such as heart failure (HF), is becoming more evident. This study aims to evaluate the risk of developing HF following a cancer diagnosis in patients with CHD, compared with those without CHD and with CHD patients who do not have cancer. Methods CHD patients (n = 69 799) and randomly selected non‐CHD controls (n = 650 406), born in Sweden between 1952 and 2017, were identified from the Swedish National Health Registers and Total Population Register (excluding those with syndromes and transplant recipients). CHD patients who developed cancer (n = 1309) were propensity score‐matched with non‐CHD patients who developed cancer (n = 9425), resulting in a cohort of 1232 CHD patients with cancer and 2602 non‐CHD controls with cancer (after exclusion of individuals with HF prior to cancer diagnosis). In a separate analysis, CHD patients with cancer were propensity score‐matched with CHD patients without cancer (n = 68 490). A total of 1233 CHD patients with cancer and 2257 CHD patients without cancer were included in the study. Results Among CHD patients with cancer, 73 (5.9%) developed HF during a mean follow‐up time of 8.5 ± 8.7. Comparatively, in the propensity‐matched control population, 29 (1.1%) non‐CHD cancer patients (mean follow‐up time of 7.3 ± 7.5) and 101 (4.5%) CHD patients without cancer (mean follow‐up time of 9.9 ± 9.2) developed HF. CHD patients exhibited a significantly higher risk of HF post‐cancer diagnosis compared with the non‐CHD control group [hazard ratio (HR) 4.39, 95% confidence interval (CI) 2.83–6.81], after adjusting for age at cancer diagnosis and comorbidities. In the analysis between CHD patients with cancer and those without cancer, the results indicated a significantly higher risk of developing HF in CHD patients with cancer (HR 1.53, 95% CI 1.13–2.07). Conclusions CHD patients face a more than four‐fold increased risk of developing HF after a cancer diagnosis compared with cancer patients without CHD. Among CHD patients, the risk of HF is only modestly higher for those with cancer than for those without cancer. This suggests that the increased HF risk in CHD patients with cancer, relative to non‐CHD cancer patients, may be more attributable to CHD itself than to cancer treatment‐related side effects.

BackgroundElevated troponin levels are a sensitive biomarker for cardiac injury. The quick and reliable prediction of troponin elevation for patients with chest pain from readily available ECGs may pose a valuable time-saving diagnostic tool during decision-making concerning this patient population.Methods and resultsThe data used included 15 856 ECGs from patients presenting to the emergency rooms with chest pain or dyspnoea at two centres in Sweden from 2015 to June 2023. All patients had high-sensitivity troponin test results within 6 hours after 12-lead ECG. Both troponin I (TnI) and TnT were used, with biomarker-specific cut-offs and sex-specific cut-offs for TnI. On this dataset, a residual convolutional neural network (ResNet) was trained 10 times, each on a unique split of the data. The final model achieved an average area under the curve for the receiver operating characteristic curve of 0.7717 (95% CI±0.0052), calibration curve analysis revealed a mean slope of 1.243 (95% CI±0.075) and intercept of −0.073 (95% CI±0.034), indicating a good correlation between prediction and ground truth. Post-classification, tuned for F1 score, accuracy was 71.43% (95% CI±1.28), with an F1 score of 0.5642 (95% CI±0.0052) and a negative predictive value of 0.8660 (95% CI±0.0048), respectively. The ResNet displayed comparable or surpassing metrics to prior presented models.ConclusionThe model exhibited clinically meaningful performance, notably its high negative predictive accuracy. Therefore, clinical use of comparable neural networks in first-line, quick-response triage of patients with chest pain or dyspnoea appears as a valuable option in future medical practice.

Background and objectivesDuring the last decades, the survival rates in patients with congenital heart disease have increased dramatically, particularly in patients with complex heart malformations. However, the survival in patients with simple defects is still unknown. We aimed to determine the characteristics and the risk of mortality in patients with isolated pulmonary valve stenosis (PS).MethodsSwedish inpatient, outpatient and cause of death registries were used to identify patients born between 1970 and 2017 with a diagnosis of PS, without any other concomitant congenital heart lesion. For each patient with PS, 10 control individuals without congenital heart disease were matched by birth year and sex from the total population registry. We used median-unbiased method and Kaplan-Meier survival analysis to examine the risk of mortality.ResultsWe included 3910 patients with PS and 38 770 matched controls. The median age of diagnosis of PS was 0.7 years (IQR 0.3–7.0). During a median follow-up of 13.5 years (IQR 6.5–23.5), 88 patients with PS and 192 controls died; 500 patients with PS (12%) underwent at least one transcatheter or surgical valve intervention. The overall mortality rate was significantly higher in patients with PS compared with matched controls (HR 4.67, 95% CI 3.61 to 5.99, p=0.001). Patients with an early diagnosis of PS (0–1 year) had the highest risk of mortality (HR 10.99, 95% CI 7.84 to 15.45).ConclusionsIn this nationwide, register-based cohort study, we found that the risk of mortality in patients with PS is almost five times higher compared with matched controls. Patients with an early diagnosis of PS appears to be the most vulnerable group and the regular follow-up in tertiary congenital heart units may be the key to prevention.

PurposeWe examined whether end-to-end deep-learning models could detect moderate (≥50%) or severe (≥70%) stenosis in the left anterior descending artery (LAD), right coronary artery (RCA) or left circumflex artery (LCX) in iodine contrast-enhanced ECG-gated coronary CT angiography (CCTA) scans.MethodsFrom a database of 6293 CCTA scans, we used pre-existing curved multiplanar reformations (CMR) images of the LAD, RCA and LCX arteries to create end-to-end deep-learning models for the detection of moderate or severe stenoses. We preprocessed the images by exploiting domain knowledge and employed a transfer learning approach using EfficientNet, ResNet, DenseNet and Inception-ResNet, with a class-weighted strategy optimised through cross-validation. Heatmaps were generated to indicate critical areas identified by the models, aiding clinicians in understanding the model’s decision-making process.ResultsAmong the 900 CMR cases, 279 involved the LAD artery, 259 the RCA artery and 253 the LCX artery. EfficientNet models outperformed others, with EfficientNetB3 and EfficientNetB0 demonstrating the highest accuracy for LAD, EfficientNetB2 for RCA and EfficientNetB0 for LCX. The area under the curve for receiver operating characteristic (AUROC) reached 0.95 for moderate and 0.94 for severe stenosis in the LAD. For the RCA, the AUROC was 0.92 for both moderate and severe stenosis detection. The LCX achieved an AUROC of 0.88 for the detection of moderate stenoses, though the calibration curve exhibited significant overestimation. Calibration curves matched probabilities for the LAD but showed discrepancies for the RCA. Heatmap visualisations confirmed the models’ precision in delineating stenotic lesions. Decision curve analysis and net reclassification index assessments reinforced the efficacy of EfficientNet models, confirming their superior diagnostic capabilities.ConclusionOur end-to-end deep-learning model demonstrates, for the LAD artery, excellent discriminatory ability and calibration during internal validation, despite a small dataset used to train the network. The model reliably produces precise, highly interpretable images.