Risk of acute myocardial infarction, stroke and thromboembolism following COVID-19 vaccination compared to testing positive for COVID-19 infection: a nationwide cohort study of 4.6 mio individuals

Background

Large randomized controlled trials (RCT) have shown that COVID-19 vaccines are effective at preventing severe COVID-19. However, the RCT's are not powered to detect rare adverse events. It has been reported that the new mRNA based COVID-19 vaccines may increase the risk of thromboembolic and ischemic events. Likewise, thromboembolic and ischemic events are also known complications to infection with SARS-CoV-19. Currently, less is known about the risk-reward relationship of receiving an mRNA-based COVID-19 vaccine versus contracting COVID-19 infection with respect to thromboembolic and ischemic outcomes.

Purpose

To compare the risk of thromboembolic and ischemic events following COVID-19 vaccination to the risk following infection with SARS-CoV-19.

Methods

The study period was from March 2020 to August 2021. All individuals were >18 years old. The population was stratified into two different groups. The vaccinated group consisted of recipients of the first dose of either Moderna (mRNA-1273, n=488,220) or Pfizer-BioNTech (BNT162b2 mRNA, n=3,186,164) vaccines. Individuals who had previously tested positive for SARS-CoV-19 were excluded. The other group consisted of individuals who had tested positive for SARS-CoV-19 in the same period who had not yet received their first vaccination dose (n=233,926). The exposure period for both groups was set to 28 days following vaccination/testing positive for SARS-CoV-19 (Figure 1). Patient level data were obtained on all included individuals using nationwide registries. Primary outcomes were acute myocardial infarction (AMI), ischemic stroke, pulmonary embolism (PE), and deep venous thrombosis (DVT). Odds ratios were obtained from logistic regression models with the vaccinated group acting as reference. Multivariable models were adjusted for demographics and comorbidities.

Results

In the vaccinated group, mean age was 53±19 years and 50.3% were female. In the group of participants testing positive for SARS-CoV-19, mean age was 42.1±17.4 years and 50.2% were female. In total, 773 suffered a stroke, 472 suffered a PE, 500 suffered an AMI, and 484 suffered a DVT during the 28-day exposure period. We observed an increased absolute risk of all outcomes for participants testing positive for SARS-CoV-19 as compared to participants being vaccinated (stroke: 0.049% vs 0.019%, p<0.001), (PE: 0.91% vs 0.0072%, p<0.001), (AMI: 0.021 vs 0.013, p=0.0004), and (DVT: 0.037% vs 0.011%, p<0.001). In multivariable models, participants testing positive for SARS-CoV-19 had a significantly increased risk of all outcomes compared to participants being vaccinated: (stroke: OR: 4.0, 95% CI: [2.9–5.6], p<0.001), (PE: OR: 38.6 95% CI: [30.3–48.5], p<0.001), (AMI: OR: 3.3, 95% CI: [2.1–5.00], p<0.001), and (DVT: OR: 5.3, 95% CI: [3.8–7.5], p<0.001) (Figure 2).

Conclusion

The risks of thromboembolic and ischemic events were substantially higher after SARS-CoV-19 infection than after vaccination in the Danish population.

Funding Acknowledgement

Type of funding sources: Public hospital(s). Main funding source(s): Gentofte University Hospital

Association between hemodialysis and patient characteristics, microbiological etiology, cardiac surgery, and mortality in patients with infective endocarditis: a nationwide study

Background

Hemodialysis and infective endocarditis are both associated with poor patient outcome. However, despite high mortality rates for each disease entity, little attention is given to patients on hemodialysis who develop infective endocarditis.

Purpose

To examine patient characteristics, microbiological etiology, cardiac surgery, and outcome among patients on hemodialysis with infective endocarditis compared with patients with infective endocarditis without hemodialysis treatment.

Methods

With Danish nationwide registries, we identified patients with infective endocarditis between 2010–2018 and linked them to microbiological data from a nationwide microbiological registry with complete blood culture data. We included patients in the hemodialysis group if they received hemodialysis treatment within 6 months prior to their first-time infective endocarditis admission. Patients not meeting this criteria were put in the non-hemodialysis group. We used Kaplan-Meier estimates for difference in mortality and Cox regression for adjusted analysis.

Results

We included 4,106 patients with infective endocarditis of which 265 (6.5%) patients were also in hemodialysis treatment (66.8% men). Patients on hemodialysis were younger (median age 66 years [IQR=54.2–74.9] vs. 72.3 years [IQR=62.3–80.4]) and had a higher burden of comorbidities including hypertension (68.7 vs. 56.9%), diabetes (47.2% vs. 18.8%), and ischemic heart disease (41.1% vs. 32.2%) compared to patients without hemodialysis treatment, all p-values <0.01. Cardiac surgery was less frequently performed in patients in the hemodialysis group than in the non-hemodialysis group (11.9% vs. 19.4%, respectively, p<0.001) and Staphylococcus aureus was more frequently the microbiological etiology of infective endocarditis in the hemodialysis group than in the non-hemodialysis group (57.0% vs. 25.3%, respectively, p<0.0001). No statistically significant difference for in-hospital mortality was found. Figure 1 shows difference in mortality between the two groups. 1- and 5-year mortality were significantly higher in the hemodialysis group than in the non-hemodialysis group (34.3% vs. 17.2% and 50.5% vs. 33.9%, respectively, p<0.00001) and in adjusted analysis hemodialysis was associated with higher 1- and 5-year mortality (hazard ratio of 2.41, 95% CI 1.85–3.13 and 2.50, 95% CI 2.05–3.05, respectively), as compared with patients in the non-hemodialysis group.

Conclusion

Patients on hemodialysis with infective endocarditis are younger, sicker and have Staphylococcus aureus as causing agent more than twice as often as patients with infective endocarditis without hemodialysis treatment. This patient group have a higher mortality and by 5 years, 75% of patients in our hemodialysis group were dead.

Funding Acknowledgement

Type of funding sources: None.

Oral fluoroquinolones and risk of aortic dissection and aortic aneurysm: a nationwide nested case-control study

Background

Oral fluoroquinolones are commonly prescribed antibiotics. Observational studies have shown an association between fluoroquinolone-use and subsequent risk of aortic aneurysm (AA) and aortic dissection (AD) due to a potential collagen degrading effect of fluoroquinolones.

Purpose

To investigate if fluoroquinolone-use was associated with increased rates of AA or AD in patients without known aortic disease. Secondly, to investigate if fluoroquinolone-use was associated with increased all-cause mortality and aortic interventions in high-risk patients with known aortic disease.

Methods

We used a nested case-control study design in which individuals aged 30–100 years from 2003 to 2018 were included from Danish nationwide registers. Exclusion criteria were bicuspid aortic valve, coarctation of the aorta, and connective tissue disease. A main cohort and a secondary high-risk cohort were defined. The main cohort comprised patients without history of AA/AD in which two case definitions were used: 1) A broad case definition of first-time AA/AD. 2) A severe case definition of ruptured AA/AD. The high-risk cohort comprised patients surviving index AA/AD admission in which cases were defined as all-cause mortality and aortic interventions.

Cases were matched on age, sex, and year of inclusion in a 1:30 ratio with controls. For the main cohort, a potential dose-response effect was investigated using groups of cumulative defined daily doses (cDDD) of fluoroquinolones. Hazard ratios (HR) with 95% confidence intervals (CI) for fluoroquinolone-use compared with amoxicillin as an active comparator were obtained from time-dependent Cox regression models using multiple exposure windows.

Results

The main cohort comprised 4.81 million individuals with 43,280 cases. Short-term 30-day, intermediate-term 90-day, and long-term 1-year fluoroquinolone use were all not associated with AA/AD (30-day HR 1.18 [95% CI: 0.84 to 1.66]; 90-day HR 1.12 [95% CI 0.96 to 1.30]; 1-year HR 1.00 [95% CI 0.93 to 1.07]). Using a severe case definition of ruptured AA/AD yielded comparable results. For the dose-response analysis, increasing cDDD did not confer increased rates of AA/AD (1–5 cDDD: Reference group; 6–10 cDDD: HR 1.03 [95% CI: 0.87 to 1.23]; >10 cDDD: HR 1.00 [95% CI 0.83 to 1.29]) (Figure 1).

The secondary high-risk cohort included 20,195 patients surviving index admission with 9,183 cases of all-cause mortality and 1,768 cases of aortic interventions. The 30-day HR for all-cause mortality was 1.21 (95% CI 0.92 to 1.60) and the 60-day HR 1.06 (95% CI 0.89 to 1.26). No association with aortic interventions was found either (Figure 2).

Conclusion

Fluroquinolone-use was not associated with AA/AD. Furthermore, fluoroquinolone-use was not associated with all-cause mortality or aortic interventions in potentially susceptible patients with known aortic disease. These findings do not support an increased risk of AA/AD with fluoroquinolone-use.

Funding Acknowledgement

Type of funding sources: None.

Incidence of atrial fibrillation/flutter, one-year re-admission rates, and practice patterns among patients <65 years of age: a Danish nationwide study

Background/Introduction

The general atrial fibrillation/flutter (AF) population is well explored and described, but there is sparse data on temporal changes in the incidence, AF-readmission rates, and practice patterns in patients with AF under 65 years of age from unselected cohorts.

Purpose

To investigate temporal changes, AF readmission rates, and practice patterns in patients under 65 years of age with first-time AF diagnosed between 2000–2018.

Methods

Using Danish nationwide registries, we identified patients >18 years and <65 years with a first-time AF-diagnosis from 2000–2018. The cohort was categorized according to calendar periods; 2000–2002, 2003–2006, 2007–2010, 2011–2014 and 2015–2018. Incidence rate (IR) of AF per 100,000 person years (PY), AF-readmission, and practice patterns of medical treatment, electrical cardioversion, and catheter ablation was investigated in the first year following AF-diagnosis.

Results

In this study 60,917 patients were included; 8,150 patients (13.4%) in 2000–2002, 11,898 (19.5%) in 2003–2006, 13,560 (22.3%) in 2007–2010, 14,167 (23.3%) in 2011–2014 and 13,142 (21.6%) in 2015–2018. No major differences were seen in patient characteristics according to calendar period. A stepwise increase, as seen in the Table, in the crude IR of AF per 100,000 PY was observed across calendar periods, except for 2015–2018 (Crude IR [95% CI]: 2000–2002: 78.7 [77.0; 80.4], 2003–2006: 86.3 [84.7; 87.8], 2007–2010: 97.9 [96.3; 99.6], 2011–2014: 102.3 [100.7; 104.0], 2015–2018: 93.6 [92.0; 95.2], while no difference in AF readmission was identified (AF-readmissions: 2000–2002: 32.7%, 2003–2006: 31.1%, 2007–2010: 32.2%, 2011–2014: 32.1% and 2015–2018: 31.7%), as seen in the Figure, right panel. In the first year following AF-diagnosis, the cumulative incidence of catheter ablation increased stepwise from 1.2% in 2000–2002 to 7.6% in 2015–2018 and electrical cardioversion from 2.0% in 2000–2002 to 8.7% in 2015–2018 (Figure, left panel). Treatment with oral anticoagulant therapy (OAC) increased from 28.5% in 2000–2002 to 47.8% in 2015–2018, while there was no change in treatment with rhythm or rate medication therapy.

Conclusion

From 2000–2018, we found an increase in the incidence of atrial fibrillation/flutter (AF) in patients <65 years from 78.7/100,000 person years (PY) to 93.6/100,000 PY and an increase in the use of catheter ablation, electrical cardioversion and OAC in the first year following first-time AF-diagnosis. AF readmission rates were stable over calendar periods.

Funding Acknowledgement

Type of funding sources: None.

PMHnet-alpha: development and validation of a neural network based discrete-time survival model for mortality prediction in ischemic heart disease

Background

Current risk prediction models in ischemic heart disease (IHD) use a small set of well-known risk factors, have limited predictive capabilities, and are largely the same as they were twenty years ago.

We developed and externally validated PMHnet-alpha, a neural-network based survival model for risk-stratification in ischemic heart disease that leverages the multitude of clinical features available in modern electronical health records.

Methods

We included 39,746 IHD patients from the regional Heart Registry that had been subjected to a coronary angiography between 2006 and 2017 with confirmed coronary artery disease.

Clinical data was extracted from the Danish National Patient Registry, and electronic health records.

595 different features, consisting of diagnosis codes, procedure codes, biochemical test results, and clinical measurements were used as model inputs.

Prior to model development, patients were randomly divided into a training set (n=34,746) and a tesing set (n=5,000).

The testing set was not used for model development.

Model performance was evaluated at six months, one years, three-, and five years of follow-up using time-dependent ROC curve analysis and Harrels' C-index.

Lastly, we also assessed the calibration of the model.

We benchmarked the performance of PMHnet-alpha against the GRACE Risk Score 2.0, which is widely considered to the best-performing model in current clinical use.

We explored the importance of individual features using SHAP values on the trained models.

Findings

PMHnet-alpha had very high model discrimination on the testing data with time-dependent AUCs of 0.88 (95% CI 0.86–0.90) at six months, 0.88 (95% CI 0.86–0.90) at one year, 0.84 (95% CI 0.82–0.86) at three years, and 0.82 (95% CI 0.80–0.84) at five years.

The discrimination of the benchmark model GRACE2.0 on the same data was considerably lower, 0.77 (95% CI 0.73–0.80) at six months, 0.77 (95% CI 0.74–0.80) at one year, and 0.73 (95% CI 0.70–0.75) at three years.

PMHnet-alpha is undergoing external validation in other nordic countries.

We identified that on-average, age, coronary pathology and smoking status were the most impactful features.

Interpretation

Here we present a significant improvement of the state of the art in cardiac risk prediction.

PMHnet-alpha supports better and optimized use of available healthcare data, signified by the vast improvement compared to GRACE2.0.

This also signifies an important paradigm shift in which data-driven strategies are necessary to transform the increasing amount of data generated in the modern healthcare system into evidence-based clinical decision making.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): The Novo Nordisk Foundation, NordForsk

Mobitz type I 2nd degree atrioventricular (Wenckebach) block and cardiovascular death using 978,901 12 lead ECGs recordings

Background

Mobitz type I 2nd degree atrioventricular (AV) block (Wenckebach) is usually considered benign. Guidelines recommend permanent cardiac pacing for patients with Mobitz type II second degree AV block (Mobitz II), but for patients with Wenckebach, permanent pacing is only indicated if the AV block causes symptoms or if the conductions delay occurs below the bundle of His. However, these guidelines are based on evidence of modest quality and a consensus amongst experts (1).

Purpose

This study aims to investigate if Wenckebach really is benign by comparing the risk of cardiovascular death for patients with Wenckebach to patients with normal ECGs.

Methods

This retrospective cohort study included 978,901 ECGs obtained from general practitioners in Denmark from 01/02/2001 to 31/10/2014. Index date was the day of the ECG recording and the patients were followed until death or end of follow up at December 2019.

The association between Wenckebach and cardiovascular death was analyzed using: 1) multivariate Cox models adjusted for age and comorbidities, 2) cause-specific Cox models and 3) cumulative risk and cause-specific hazard function plots, compared to matched controls. Information about comorbidities, pacemaker, indications, and death was retrieved from Danish nationwide registries.

Results

From the 978,901 ECG recordings, we found 262 patients with Wenckebach, 131 patients with Mobitz II, and 229,056 patients with normal ECGs. In Wenckebach, Mobitz II, and normal ECG the median age was 76, 80, and 50 years, 76%, 63%, and 41% were male, 25%, 16%, and 3% had diabetes, 35%, 30%, and 8% had hypertension, respectively.

During a mean follow-up of 11.2 years, cardiovascular death occurred in a total of 11,301 patients: 77 (29%) patients with Wenckebach, 40 (31%) patients with Mobitz II, and 11,184 (5%) patients with normal ECGs. In a matched cohort 262 Wenckebach patients were matched with 520 controls with normal ECGs. In the multivariate Cox model, Wenckebach was associated with cardiovascular death (HR: 2.14 [95% CI: 1.46–3.13], P<0.001). Furthermore, in multivariate cause-specific Cox analysis with non-cardiovascular death and pacemaker as competing risk, Wenckebach was still associated with cardiovascular death (HR: 2.27 [95% CI: 1.37–3.75], P=0.001).

Furthermore, the results showed that 43% of the Wenckebach patients received pacemaker with a median time to pacemaker from ECG recording being 252 days. The vast majority of the Wenckebach patients who received pacemaker had a higher degree AV block than Wenckebach as indication for the implantation.

Conclusion

Wenckebach on routine ECG was associated with a significant higher hazard rate of cardiovascular death compared to matched controls with normal ECGs.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): The Danish Heart Association

Prognosis and antithrombotic practice patterns in recurrent and transient atrial fibrillation following acute coronary syndrome: a nationwide study

Background

First-time detected atrial fibrillation (AF) during acute coronary syndrome (ACS) aggravates the prognosis and increases the risk of ischemic stroke. In this setting, AF may present as brief and transient or with recurrent episodes after discharge. However, data on the association between transient or recurrent AF and ischemic stroke in patients with ACS are sparse. Further, despite being associated with ischemic stroke, first-time detected AF patients have been reported with low oral anticoagulation (OAC) rates.

Purpose

To examine the associated rate of ischemic stroke and mortality in ACS survivors with transient or recurrent AF and to assess the antithrombotic practice patterns one year after ACS.

Methods

Using data from Danish nationwide registries, we identified all patients with first-time ACS, without known AF prior to ACS, from 2000–2017 who were alive one year after ACS discharge (index date). According to a grace period between ACS discharge and one year after ACS discharge, patients were categorized into: i) no AF; ii) first-time detected AF during ACS admission without AF recurrence (transient AF); and iii) first-time detected AF during ACS admission with a subsequent recurrent AF episode (recurrent AF). Patients who developed AF during the grace period were excluded. Patients were followed from one year post ACS discharge, and two-year rates of ischemic stroke and mortality were compared using multivariable adjusted Cox proportional hazards analysis. Further, we assessed the prescribed OAC rates in a three-month period following the index date.

Results

We included 116,793 patients surviving one year post ACS discharge: 111,708 (95.6%) without AF (64.9% male, median age 64 years), 2,671 (2.3%) with transient AF (58.0% male, median age 74 years), and 2,414 (2.1%) with recurrent AF (55.2% male, median age 76 years). The cumulative two-year incidence of ischemic stroke was 0.9%, 1.5%, and 2.3% for patients without AF, transient AF, and recurrent AF, respectively (Figure 1). The cumulative two-year incidence of mortality was 7.4%, 12.1%, and 20.3% for patients without AF, transient AF, and recurrent AF, respectively (Figure 1). Compared to those without AF, the adjusted two-year rates of outcomes were as follows: ischemic stroke: HR 1.15 (95% CI: 0.81–1.61) for patients with transient AF and HR 1.50 (95% CI: 1.14–1.98) for patients with recurrent AF; mortality: HR 0.98 (95% CI: 0.87–1.10) for patients with transient AF and HR 1.35 (95% CI: 1.23–1.49) for patients with recurrent AF (Figure). We identified that 20.9% for transient AF and 42.2% for recurrent AF were prescribed OAC therapy in the three-month period after one year.

Conclusion

In patients surviving one year after ACS with first-time detected AF, a recurrent AF episode was associated with an increased long-term rate of ischemic stroke and mortality, while transient AF yielded no statistically difference as compared with patients without AF.

Funding Acknowledgement

Type of funding sources: None.

High bleeding risk in all-comers with ST-segment elevation myocardial infarction and use of P2Y12-inhibitiors

Background

Bleeding has an important prognostic impact in patients with ST-segment elevation myocardial infarction (STEMI), yet stratification of bleeding risk to guide dual antiplatelet therapy (DAPT) is not routinely performed in clinical practice.

Purpose

To describe high bleeding risk (HBR) patients according to the PRECISE-DAPT (predicting bleeding complications in patients undergoing stent implantation and subsequent DAPT) score and use of P2Y12-inhibitors.

Methods

This single-centre observational study included consecutive patients with STEMI who were treated with percutaneous coronary intervention (PCI) from 2009–2016. Individual linkage to Danish nationwide registries was conducted to obtain information on diagnoses, claimed drugs, and vital status. Age, prior bleeding diagnosis, and blood samples before PCI (maximum 30 days before hospitalisation) were used to calculate the PRECISE-DAPT score. A score ≥25 was considered as HBR. Due to 26.7% missing on blood parameters (mainly leucocytes), the maximum and minimum values of the missing parameters and respective imputed PRECISE-DAPT scores were calculated. If both the maximum and minimum score were ≥25 or <25, patients were categorised accordingly, and a maximum score of ≥25 and minimum score of <25 as missing. Differences between continuous (median [interquartile range, IQR]) and categorical variables (frequency [percentage]) were assessed using Wilcoxon rank-sum and χ2-test for patients with vs. without HBR. Cumulative incidence of major bleeding (composite of bleedings leading to hospitalisation) and major adverse cardiovascular events (MACE) (composite of all-cause mortality, recurrent MI, and ischemic stroke) 1 year after PCI were plotted for patients with and without HBR. Number of HBR patients alive and collecting a P2Y12-inhibitor prescription within 30 days from discharge was reported.

Results

We identified 6179 PCI-treated patients with STEMI, of whom 5530 (89.5%) had imputed PRECISE-DAPT scores (Figure 1). A total of 1821 (32.9%) were at HBR, and these were more often female (38.3 vs. 18.2%, p-value<0.001), elderly (median age 75 [IQR 67, 81] vs. 57 years [IQR 51, 64], p-value<0.001), and had more comorbidities (diabetes [16.7 vs. 12.1%], heart failure [16.2 vs. 7.6%], cardiac arrhythmia [24.9 vs. 12.3%], cancer [17.5 vs. 5.7%], and ischemic stroke [8.1 vs. 2.6%], all p-values<0.001) compared with patients not at HBR. One-year cumulative incidence of major bleeding and MACE for patients with and without HBR were plotted (Figure 2). Of the 1431 (78.6%) HBR patients who were alive and claimed a P2Y12-inhibitior prescription 30 days from discharge, 459 (32.1%) were treated with clopidogrel, 672 (46.9%) with ticagrelor, and 300 (21.0%) with prasugrel (Figure 1).

Conclusion

Every third PCI-treated all-comer with STEMI was at HBR according to the PRECISE-DAPT score. HBR patients were more often treated with potent P2Y12-inhibitors (prasugrel or ticagrelor) instead of clopidogrel.

Funding Acknowledgement

Type of funding sources: Private grant(s) and/or Sponsorship.

First-in-man trial of b3-adrenoreceptor agonist treatment in chronic heart failure – impact on diastolic function

Background

Diastolic dysfunction (DD) in heart failure (HF) is associated with increased myocardial cytosolic calcium, and calcium-efflux via the sodium-calcium-exchanger depends on the sodium gradient. Beta-3-adrenoceptor (β3-AR)-agonist lowers cytosolic sodium and has been shown to reverse organ congestion.

Purpose

To assess whether β3-AR-agonist treatment improves DD.

Methods

In a first-in-man randomized controlled, double-blind trial, we assigned 70 patients with HF with reduced ejection fraction (HFrEF) (NYHA II–III) and LVEF <40% to receive mirabegron (300 mg/day) or placebo for 6 months, in addition to recommended HF-therapy. Patients were assessed with echocardiography and cardiac computed tomography (CCT) at baseline and follow-up. DD was graded according to the current American/European guidelines.

Results

Baseline and follow-up echocardiographic data were available in 57 patients (59±11 years, 88% male, 49% ischemic heart disease). Baseline LVEF was 34%±8%. No significant change in DD grade was found between the groups at follow-up, p=0.72. Neither was there any clinical differences in any singular diastolic parameters within or between groups by echocardiography (E/e' placebo: 13.3±6.9 to 12.6±5.1, p=0.19 vs. mirabegron: 12.0±5.7 to 12.8±7.9, p=0.67, mean difference 1.12 [95% CI −1.68 to 4.3], p=0.37), or CCT (left atrial max volume index: between group mean difference 0.2 [95% CI −6.2 to 5.6] ml/m2, p=0.91).

Conclusions

In patients with HFrEF, no improvement nor worsening in DD gradings or singular diastolic parameters after β3-AR stimulation compared to placebo were identified. The findings add to previous literature questioning the role of impaired Na+-Ca2+ mediated Ca2+ export as a major culprit in DD.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): The Heart Centre Research Foundation, RigshospitaletThe Novo Nordic Foundation

Temporal trends in end-stage renal disease in patients with heart failure with or without diabetes: a nationwide study from 2002 to 2017

Background

Advances in treatment of heart failure (HF) have increased survival rates. However, whether the improved life expectancy for HF patients has resulted in an increased risk of a significant comorbidity like end-stage renal disease (ESRD) is less clear. Renal dysfunction is associated with increased morbidity and mortality in HF and constitutes an important prognostic factor for HF. Further, diabetes (DM) is closely related to both HF and ESRD, but it is unknown how DM affects the risk of ESRD in patients with HF.

Purpose

To investigate temporal trends in ESRD in patients with HF and the subsequent risk of mortality stratified by DM.

Methods

Using Danish nationwide registies, we identified patients, aged 18 to 100 years, with incident HF between 2002 and 2017. The outcomes were ESRD (defined as dialysis treatment), worsening of HF (wHF, defined as rehospitalization for HF) and all-cause mortality. Three study periods were investigated 2002–2006, 2007–2011 and 2012–2017. We estimated crude 5-year incidence rates (per 1000/person-years) of the outcomes stratified by DM. Multivariate Cox regression models were performed for all outcomes stratified by DM. Further, we computed the 1-year all-cause mortality risk after diagnosis with ESRD.

Results

Of 124,141 patients with HF, 50,690 (41%) were women and the median age was 74.5 years [95% confidence interval (CI) 64.5–82.8]. At baseline DM was present in 20% of the patients. These patients were older, more often men and more comorbid than HF patients without DM. Over time (2002–2006 to 2012–2017) the incidence rates of ESRD (9.0 to 7.9 and 2.1 to 1.9 per 1000/person-years for DM and no-DM, respectively) and wHF (124.0 to 124.8 and 84.3 to 81.9 per 1000/person-years for DM and no-DM) remained stable, while all-cause mortality rates decreased (217.0 to 170.3 and 172.9 to 127.8 per 1000/person-years for DM and no-DM). The incidence of ESRD was lower compared with the incidence of wHF and all-cause mortality [Figure 1]. HF patients with DM had significantly higher associated rates of all three outcomes (in 2012–2017 the rates for DM vs no-DM of ESRD: 3.99 [3.27–4.86], wHF: 1.42 [1.36–1.49], all-cause mortality: 1.36 [1.31–1.41]) compared with patients without DM. We found no significant interaction between time period and DM on the rates of outcomes (p>0.05 for all) [Figure 2]. One-year all-cause mortality risk after diagnosis with ESRD was high both for HF patients with and without DM through all time periods (identical risks and 95% CI in 2012–2017: 32% [0.25–0.39]).

Conclusions

We did not observe a change over time in the 5-year risk of ESRD for HF patients. The incidence of ESRD remained low compared to wHF and all-cause mortality. DM was associated with increased rates of all three events, not changed over time. Conversely, all-cause mortality after diagnosis with ESRD was markedly high, irrespectively of DM. Our analyses suggest that ESRD is a less common, but fatal event in HF patients.

Funding Acknowledgement

Type of funding sources: None.

Risk of worsening heart failure and all-cause mortality following mRNA COVID-19 vaccination in patients with heart failure: a Danish nationwide real-world safety study

Background

The studies investigating the safety and efficacy of the SARS-COV2 mRNA vaccines only included a limited number of heart failure patients and no separate analyses were performed regarding the safety of the vaccines in this patient population.

Purpose

The aims of this study were to investigate the risk of worsening heart failure and all-cause mortality associated with the SARS-COV-2 mRNA vaccines in a nationwide cohort of patients with heart failure.

Methods

Using the Danish nationwide registries, two cohorts were constructed; 1) all prevalent heart failure patients in 2019 and 2) all prevalent heart failure patients in 2021 who were vaccinated with either of the two mRNA vaccines (BNT162B2 or mRNA-1273). The patients in the two cohorts were matched 1:1 using exact exposure matching on age, sex, and duration of heart failure (intervals). For patients in the 2021 cohort, the index date was defined as the date of the patients' second vaccination. Patients in the 2019 cohort were assigned the index day and month of their 1:1 match in the 2021 cohort, but used the pre-vaccination index year 2019. The primary outcomes were worsening heart failure and all-cause mortality and secondary outcomes were myocarditis and venous thromboembolism. Standardized risks were estimated based on outcome-specific Cox regression analyses, and all models were standardized to age, sex, duration of heart failure, use of SGLT2 inhibitors or Entresto, ischemic heart disease, cancer, diabetes, atrial fibrillation, and admission with heart failure <90 days before index.

Results

The total study population comprised 101,786 patients, with 50,893 patients in each cohort. The median age of the study population was 74 (interquartile range (IQR); 66,81), and duration of heart failure was 4.1 (IQR: 2.0,6.7) years. The standardized risk of all-cause mortality within 90 days was 2.2% (95% CI: 2.1% to 2.4%) in the 2021 cohort and 2.6% (95% CI: 2.4% to 2.7%) in the 2019 cohort, showing a significantly lower risk difference for all-cause mortality in 2021 versus 2019 (risk difference: −0.3% (95% CI: −0.5% to −0.1%)) Figure 1)). The standardized risk of worsening heart failure within 90 days was 1.1% (95% CI: −1.0% to 1.2%) in the 2021 cohort and 1.1% (95% CI: 1.0% to 1.2%) in the 2019 cohort showing no significant difference in the risk of worsening heart failure between the two cohorts (risk difference: 0% (95% CI: −0.1% to 0.1%)). No significant differences were found for venous thromboembolism or myocarditis.

Conclusion

This study showed that the SARS-COV2 mRNA vaccines were not associated with an increased risk of worsening heart failure, venous thromboembolism or myocarditis, but was associated with a decreased risk of all-cause mortality. Our study may suggest that these vaccines are safe in heart failure patients.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): The Danish Heart FoundationLæge Sofus Carl Emil Friis og hustrus legat

Adherence and discontinuation of optimal heart failure therapies according to age

Background

Guideline-recommended disease-modifying pharmacological therapies for heart failure (HF) with reduced ejection fraction are underutilized, particularly among elderly patients.

Purpose

We examined adherence with and discontinuation of evidence-based HF pharmacotherapy, comprising of angiotensin-converting enzyme inhibitors (ACEi)/angiotensin-II receptor blockers (ARB), beta-blockers (BB) and mineralocorticoid receptor antagonists (MRA), according to age.

Methods

Using Danish nationwide registries, we included patients with a first HF diagnosis between 2011 and 2018. Patients were stratified into three age groups, <65 years (reference group), 65–79 years, and ≥80 years. The average daily drug dose was calculated as median proportions of target doses one year after inclusion. Adherence was estimated by the proportion of days covered (PDC), i.e., the total number of days with the drug available for a patient alive for the whole first year of the follow-up period. Discontinuation was defined as a break of >90 days, and the 5-year risk of discontinuation according to age groups was estimated with the Aalen-Johansen estimator. Multivariable Cox regression models were used to evaluate the treatment discontinuation rate according to age groups.

Results

We included a total of 29,482 patients (<65 9,449 (25.4% female), 65–79 13,746 (33.1%), ≥80 6,287 (46.3%)). Advancing age was associated with lower median proportions of daily target doses (ACEi 100%, 88%, 63%; ARB 75%, 67%, 50%; BB 75%, 56%, 44%), and lower adherence (ACEi/ARB 79.1%, 77.5%, 69.4%; BB 79.1%, 78.6%, 73.8%), in the <65, 65–79 and ≥80 age groups respectively, one year after inclusion. Age ≥80 was associated with a higher 5-year risk of discontinuation; cumulative incidence, ACEi/ARB 41%, 44%, 51%; BB 38%, 35%, 39%, in the same age group order as above (adjusted hazard ratio: ACEi/ARB 1.60 [95% CI, 1.51–1.69]; BB 1.33 [95% CI, 1.25–1.41]). Conversely, the risk of discontinuation of MRAs differed little with age (<65 50%, 65–79 54%, ≥80 56%), although MRA initiation in the most elderly was less frequent (<65 33%, 65–79 33%, ≥80 22%).

Conclusion

Among a nationwide cohort of HF patients, advanced age was associated with lower proportions of daily target doses, lower adherence, and a higher rate of discontinuation of ACEi/ARB and BBs. Focus on treatment adherence and optimal dosages among elderly HF patients could improve outcomes.

Funding Acknowledgement

Type of funding sources: None.

Temporal trends in infection-related hospitalizations in patients with heart failure: a nationwide study from 1997 to 2017

Background

Over the last 20 years mortality has decreased for patients with heart failure (HF). However, re-hospitalization for HF is still a challenge. Further, whether the improved survival has resulted in increased rates of non HF hospitalization is unknown.

Purpose

This study examined the temporal trends in infection-related hospitalizations among new-onset HF patients and compared it to temporal trends in risk of worsening HF and death.

Methods

The study population included all Danish patients aged between 18 and 100 years old, with new-onset HF (defined according to the ICD10-code system) diagnosed between 1st January 1997 and 31st December 2017. Patients who were diagnosed with any type of cancer up to five years before their HF diagnosis were excluded to avoid cancer related infections.

The outcomes of interest were infections (defined according to the ICD10-code system) and worsening of heart failure (defined as a hospital admission with HF covering at least to dates).

The Aalen Johansen's estimator was used to estimate unadjusted 5-year absolute risk for all outcomes. Furthermore, a multivariate Cox analysis was made, and hazard ratios were estimated for the four time periods presented in a forest plot with the period 1997–2001 being the reference group. Adjustments for sex, age and history of comorbidities were conducted. Additionally, we stratified the infection outcome on different types of infections illustrated in 5-year cumulative incidence curves.

Results

The total population consisted of 147,737 patients. Over time there was a slight decrease in median age (1997–2001: 76.8 years, 2011–2017: 73.1 years) and the patients were more likely to be male (1997–2001: 53.5%, 2011–2017: 60%).

Figure 1 illustrates overall absolute risk of death decreased over time 1997–2001 (62.7% [95% CI 62.2–63.2]) vs. 2011–2017 (57.9% [95% CI 41.5–42.7]). Unadjusted curves for absolute risk showed that patients with HF had a higher risk of infection over time 1997–2001 (16.4% [95% CI 16.0–16.8] vs. 2011–2017 (24.5% [95% CI 24.0–24.9]). In contrast, they have a lower risk of worsening HF 1997–2011 (26.5% [95% CI 26.1–27.0] vs. 2011–2017 (23.2% [95% CI 22.8–23.7]). Adjusted analyses provided the same result for all outcomes illustrated in figure 2.

The risk of infection stratified by infection type, mark the risk of pneumonia infection as the most significant in all subintervals 1997–2001 (11.4% [95% CI 11.1–11.7]) vs. 2011–2017 (16.1% [95% CI 15.7–16.5]). The second most important was the risk of urogenital infection 1997–2001 (3.5% [95% CI 3.31–3.69]) vs. 2011–2017 (7.8% [95% CI 7.52–8.12]).

Conclusion

In this nationwide study, we observed that overall mortality risk and risk of hospitalization for worsening HF decreased from 1997 to 2017. In contrast, an increase in the risk of hospitalization for infection, especially pneumonia infections, increased during the same period. Future HF management programs should include strategies to prevent infections.

Funding Acknowledgement

Type of funding sources: None.

The impact of statins and RAS inhibitors on the association between delayed antidiabetic treatment and the risk of cardiovascular event in patients with a first HbA1c between 48–57 mmol/mol

Background

In addition to lifestyle intervention, guidelines recommend initiation of antidiabetic (AD) treatment within 3 months of diagnosing type 2 diabetes (T2D). Yet, patients with an initial HbA1c level between 48 and 57 mmol/mol may await effects of lifestyle intervention up to 6 months. Omitting initial AD treatment and any lifestyle-induced remission, may affect initiation of statins and renin-angiotensin system inhibitors (RASi) and, thus, cardiovascular risk.

Purpose

To examine whether omission of initial AD treatment is associated with an increased 5-year risk of first-time major cardiovascular event (MACE: myocardial infarction/stroke/all-cause death) compared with well-controlled patients on AD. Further, whether lower initial use of statins and RASi could explain this excess risk of MACE.

Methods

We used Danish registers to identify patients with a first-measured HbA1c of 48–57 mmol/mol between 2014 and 2020. We included patients aged 40–80 years without prior atherosclerotic disease that were alive the following 180 days (the index date). At date of index, we divided patients into four groups according to AD treatment and achieved HbA1c (mmol/mol): well-controlled (HbA1c ≤47) on AD; poorly controlled (HbA1c ≥48) on AD; remission (HbA1c ≤47) not on AD; poorly controlled (HbA1c ≥48) not on AD. Based on a Cox-regression model and imputations of treatment values of statins and RASi from two logistic regression models, we examined to what extent the observed standardised 5-year risk of MACE within each group could be reduced if each group had the same probability of treatment initiation with statin and RASi as well-controlled patients on AD.

Results

We included 14,206 patients (median age 59 [IQR 51–68] years; 52.0% men) with the following distribution according to AD group: well-controlled on AD: 22.3%; poorly controlled on AD: 14.7%; remission not on AD: 38.3%; poorly controlled not on AD: 24.6%. Patients not on AD had lower probabilities of initiation of statins and RASi compared with patients on AD (Figure 1). Compared with well-controlled on AD, the absolute 5-year risk of MACE was increased with 3.7% (95% CI 1.6–6.1) in poorly controlled on AD; 2.1% (95% CI 0.3–3.8) in remission not on AD; 3.4% (95% CI 1.6–5.3) in poorly controlled not on AD (Figure 1 and 2). If initiation of statins and RASi were the same as in the well-controlled group on AD, patients not on AD could reduce their risk of MACE with 1.0% (95% CI 0.2–1.8) in the remission group and with 2.2% (95% CI 1.2–3.2) in the poorly controlled group (Figure 2).

Conclusions

Patients not on initial AD treatment had an increased 5-year risk of MACE, even among those who experienced remission of T2D. Lower initial use of statin and RASi seem to explain some of the excess risk of MACE in patients not on initial AD treatment. This study emphasizes the need for greater focus on primary prevention with statins and RASi in T2D, especially among patients not on AD treatment.

Funding Acknowledgement

Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): Research Grant from Steno Diabetes Center Sjaelland

Temporal changes in aortic valve replacement according to age in Denmark: nationwide data from 2008 to 2020

Background

Since the introduction of transcatheter aortic valve implantation (TAVI), the management of symptomatic severe aortic stenosis has changed. Recent published European guidelines (2021) favours TAVI over surgical aortic valve replacements (SAVR) in patients with older age (≥75 years of age) or patients with high surgical risk. The study of nationwide practice patterns for AVR is important and renders the possibility to evaluate whether clinical practice differs from current guidelines.

Purpose

To evaluate temporal changes in use of isolated aortic valve replacement (AVR) procedures according to age in the era of TAVI in Denmark.

Methods

We identified all first-time aortic valve replacement procedures (TAVI or SAVR) from 2008 until the end of 2020 through administrative registries in Denmark. Patients with no prior diagnosis of aortic stenosis at time of AVR were excluded. Patients with prior AVR or valve repair were excluded. SAVR was divided according to type of prostheses: surgical bioprostheses and mechanical prostheses. To evaluate changes according to age the study cohort was divided into two age groups: <75 and ≥75 years of age.

Results

Between 2008 and 2020, 12,313 first-time isolated AVR procedures were performed in Denmark. Volume of isolated AVR increased from 621 to 1256 procedures per year (ptrend <0.001). Isolated SAVR was performed in 6,548 patients (53.2%) and TAVI in 5,765 patients (46.8%). Median age of TAVI patients was 81.4 [76.9–85.2] years of age compared to 73.1 [68.0-≥77.7] in patients receiving surgical bioprostheses and TAVI patients had a higher degree of comorbidity (TAVI: 70% of patients with Charlson comorbidity score ≥1, surgical bioprostheses: 50% of patients with Charlson comorbidty score ≥1). TAVI increased during study period compared to isolated SAVR, where a decreasing trend was observed from 2014 and onwards. In <75-year-old patients, volume of TAVI significantly increased during study period (ptrend<0.001), whereas volume of surgical bioprostheses remained stable. Volume of mechanical prostheses decreased over time (ptrend <0.001) TAVI increased in ≥75-year-old patients (ptrend <0.001) and TAVI accounted for 91.5% of all isolated AVR procedures in 2020. In contrast, volume of isolated SAVR declined driven by a decreasing use of surgical bioprostheses (ptrend=0.001). (Figure 1).

Conclusions

Volume of isolated aortic valve replacement (AVR) doubled from 2008 and 2020. The increase in isolated AVR was driven by transcatheter aortic valve implantation (TAVI). TAVI has become the predominant choice of isolated AVR in management of aortic stenosis and our results suggest that real-world practise patterns are in line with current guideline recommendations.

Funding Acknowledgement

Type of funding sources: None.

History of betablocker treatment breaks and risk of ventricular tachyarrhythmias among patients with heart failure and implantable cardioverter defibrillator: a nationwide cohort study

Background

Beta-blockers have in randomized clinical trials been shown to reduce the risk of life-threatening arrhythmias and sudden cardiac death (SCD) in patients with heart failure (HF), and treatment is a class 1A recommendation in current guidelines. Thus, beta-blocker treatment breaks (i.e. planned break, beta-blocker related side-effects, or poor adherence) may increase risk of life-threatening arrhythmias and SCD. Whether patients with HF and a history of beta-blocker treatment breaks before implantable cardioverter defibrillator (ICD) is associated with increased risk of device related therapy and mortality is largely unknown.

Aims

In patients with HF and an ICD alone or combined with cardiac resynchronization therapy (CRT-D), we examined the association between a history of a beta-blocker treatment breaks prior to device implantation and the risk of appropriate and inappropriate device related therapy (i.e., anti-tachycardia pacing [ATP] or DC shock [DC]), and all-cause mortality.

Methods

Using the Danish Pacemaker and ICD Registry, we identified all patients with HF receiving a first-time ICD (2000–2018). Beta-blocker treatment breaks >60 consecutive days up to 3 years prior to device implantation were identified using the National Prescription Registry. Patients were able to switch between beta-blockers and were required to be in treatment at the time of implantation. We used multivariable Cox regressions to compare the 1-year risks of device-related therapy and all-cause mortality between patients with and without a history of a beta-blocker treatment break.

Results

We identified 9,239 patients with HF and an ICD (82.6% male; median age 67 years). A total of 82.5% had ischemic heart disease, 33.9% atrial fibrillation, and 33.1% of ICDs were secondary prophylaxis. During one-year follow-up, 5.7% of all patients died and appropriate DC and appropriate ATP was identified for 3.9% and 6.7% of patients, respectively. Overall, 14.6% of all HF patients had one or more beta-blocker treatment break >60 days. Compared with HF patients with no history of treatment breaks, a history of treatment breaks >60 days were associated with increased risk of appropriate DC (hazard ratio (HR)=1.33; 95% confidence interval [CI], 1.02–1.73) and appropriate ATP (HR 1.30; CI, 1.06–1.59), but also inappropriate DC and ATP therapy (Figure 1). There was no difference between groups with respect to all-cause mortality (HR=0.96; CI: 0.76–1.22). Treatment breaks of >30 or >90 days were also evaluated and yielded similar results as the main analysis.

Conclusion

Patients with heart failure who had a history of treatment breaks with beta-blockers prior to ICD implantation was associated with a higher 1-year risk of appropriate and inappropriate shocks and anti-tachycardia pacing, but not all-cause mortality.

Funding Acknowledgement

Type of funding sources: None.

Rehospitalizations, repeated aortic surgery, and death in initial survivors of surgery for Stanford type A aortic dissection and the significance of age – a nationwide registry-based cohort study

Background

All patients with type A aortic dissections, regardless of age, are recommended urgent aortic surgery. However, studies exploring long term outcomes in survivors are sparse, and especially, the significance of age on long-term outcomes remain unclear.

Purpose

We described and compared incidences across age groups of post-discharge readmission, repeated aortic surgery, and death in patients who survived surgery and hospitalization for type A aortic dissection.

Methods

Using data from Danish nationwide registries, we identified patients hospitalized with Stanford type A aortic dissections from 2006–2018. Survivors of hospitalization and surgery on the ascending aorta and/or aortic arch comprised the study population (Figure 1). Using cumulative incidence plots taking death into account as a competing risk and Cox regression analysis, we described long-term outcomes (rehospitalizations, repeated aortic surgery, and death) and compared different age groups. The diagnosis of type A aortic dissection in the registries used, was validated from 191 clinical records to have a positive predictive value of 94.8%.

Results

Of 606 initial survivors of surgery and hospitalization with type A aortic dissection, 236 (38.9%) were <60 years old (group I), 194 (32.0%) were 60–69 years old (group II), and 176 (29.1%) were >69 years old (group III). Figure 2 shows cumulative incidences of outcomes according to age. During the first year, 62.5% were re-hospitalized (median number of days hospitalized was 2 days (IQR 1–8 days) and 1.4% underwent repeated aortic surgery with no significant differences across age groups (P=0.68 and P=0.39, respectively). Further, 5.9% died (group I: 3.0%, group II: 8.3%, group III: 7.4%, P=0.04). After 10 years of follow up, 8.0% had undergone repeated aortic surgery (group I: 11.5%, group II: 8.5%, group III: 1.6%, P=0.04) and 10.2% (group I), 17.0% (group II), and 22.2% (group III) had died (P=0.01). In adjusted analyses, no age differences were found in one-year outcomes, while age >69 years (group III) compared with age <60 years (group I) was associated with a lower rate of repeated aortic surgery (hazard ratio 0.17, 95% confidence interval 0.04–0.78) and a higher rate of all-cause mortality (hazard ratio 2.44, 95% confidence interval 1.37–4.34) in the 10-years analysis.

Conclusion

Among survivors of type A aortic dissections, rehospitalizations the first year after discharge were common among all age groups, but survival was high. Repeated aortic surgery was rare, and significantly more common among younger than older patients. Evaluations of quality of life in survivors of type A aortic dissections are needed.

Funding Acknowledgement

Type of funding sources: None.

Long-term cardiovascular outcomes in five-year cancer survivors: a nationwide cohort study

Background

Patients with cancer have higher short-term rates of cardiovascular events than the general population. However, little is known about rates of long-term cardiovascular outcomes in 5-year cancer survivors, especially in older patients.

Objective

We investigated the long-term rates of cardiovascular outcomes, including heart failure, atrial fibrillation, venous thromboembolism, ischemic stroke and myocardial infarction in five-year cancer survivors, overall and according to age.

Methods

Using Danish nationwide registries, five-year survivors of 20 of the most common cancers (diagnosed between 1994 and 2013; 15 years of age or older at the time of the diagnosis) were matched with four non-cancer controls from the background population by age and sex. Study participants with a history of any the outcomes of interest prior to index date were excluded. Rates of outcomes in the cancer and non-cancer group were compared with Cox regression models, overall and according to age (i.e., 15–39, 40–59, and >60 years).

Results

In total, 167,215 five-year cancer survivors were age- and sex-matched with 668,860 non-cancer controls (median age 66 years; 34.4% men, median follow-up of 6.8 years). Five-year survivors had higher associated rates of cardiovascular outcomes, irrespective of age, and the incidence rates per 1,000 person-years of cardiovascular outcomes for cancer survivors and non-cancer controls were: HF: 6.2 (95% CI: 6.1–6.4) and 5.2 (5.1–5.3), respectively; atrial fibrillation: 11.1 (10.9–11.3) and 9.3 (9.3–9.4), respectively; venous thromboembolism: 5.1 (5.0–5.2) and 2.8 (2.8–2.9), respectively; ischemic stroke: 5.8 (5.6–5.9) and 5.4 (5.4–5.5), respectively; and myocardial infarction: 3.6 (3.5–3.7) and 3.4 (3.3–3.4), respectively. The absolute rates of cardiovascular outcomes were highest in the oldest group, whereas the relative rates were more pronounced in the youngest cancer group compared with matched controls (Figure 1).

Conclusions

Compared with the general population, five-year cancer survivors had higher associated rates of cardiovascular outcomes across the spectrum of age. The increased rates of cardiovascular outcomes were more pronounced in the youngest group. These data underline the importance of risk assessment and prevention of cardiovascular diseases in five-year cancer survivors.

Funding Acknowledgement

Type of funding sources: None.

Perioperative atrial fibrillation in major emergency abdominal surgery: does it affect postoperative outcome?

Background

Atrial fibrillation (AF) in relation to surgery remains a clinical challenge. Major emergency abdominal surgery (e.g. ileus, perforation) is associated with postoperative complications and mortality. However, the prevalence and impact of perioperative AF in this setting is not well examined.

Purpose

We compared 30-days and 1-year outcomes (i.e. hospitalization of any causes, AF-related hospitalization, thromboembolic events and all-cause mortality) in patients who did and did not develop perioperative AF (POAF) in relation to their major emergency abdominal surgery.

Methods

We crosslinked data from Danish nationwide registries and identified all patients who underwent major emergency abdominal surgery (2000–2018) and discharged alive. Patients who developed POAF during hospitalization were matched in a 1:3 ratio on age, sex, year of surgery and category of surgery with those without POAF. Starting follow up at discharge, we examined the rates of outcomes at 30-days and 1-year post-discharge. The cumulative incidences and ratios of outcomes were assessed with the Aalen Johanson estimator together with Kaplan-Meier estimator and multivariable Cox regression analysis, respectively.

Results

We identified 891 patients with POAF and 64,914 patients without POAF. The matched cohort were composed of 889 patients with POAF and 2667 patients without POAF with a median age of 79 years [25th-75th percentile; 72–84 years] and 45.2% males. In general, patients with POAF had higher comorbid burden compared with patients without POAF. The cumulative incidences of a hospitalization of any cause after 30-days post-discharge were 31.2% and 22.3% in patients with and without POAF, respectively. The corresponding numbers for AF-related hospitalization were 20.8% and 1.2%, respectively. In adjusted analyses, POAF was associated with a significantly higher risk of hospitalization of any causes together with AF-related hospitalization (Figure 1 and 2).

The cumulative incidences of a thromboembolic event after 30-days post-discharge were 2.2% and 0.9% in patients with and without POAF, respectively. The corresponding numbers for all-cause mortality were 9.7% and 3.2%, respectively. In adjusted analyses, POAF was associated with a significantly higher risk of a thromboembolic event together with all-cause mortality within 30-days of follow up as well as 1-year of follow up. However, the results regarding thromboembolic events did not reach statistical significance after 1-year of follow up (Figure 1 and 2).

Conclusions

Perioperative atrial fibrillation in relation to major emergency abdominal surgery was associated with higher 30-days and 1-year rates of hospitalizations of any causes, atrial fibrillation related hospitalization, a thromboembolic event and all-cause mortality. These findings suggest that perioperative atrial fibrillation is a strong prognostic marker of increased morbidity following major emergency abdominal surgery.

Funding Acknowledgement

Type of funding sources: None.

Cardiac resynchronisation therapy and implantable cardioverter-defibrillator in non-ischaemic systolic heart failure: extended follow-up of the DANISH trial

Background

In the Danish Study to Assess the Efficacy of Implantable Cardioverter-Defibrillators [ICDs] in Patients with Non-ischaemic Systolic Heart Failure on Mortality (DANISH) trial, ICD implantation did not provide an overall survival benefit in patients with non-ischaemic systolic heart failure. A high proportion of patients in the DANISH trial received a cardiac resynchronisation therapy (CRT) device, which improves the prognosis in patients with heart failure. Therefore, it is of interest to examine whether the effect of ICD implantation in patients with non-ischaemic systolic heart failure is modified by CRT.

Purpose

Adding 4 years of additional follow-up to the DANISH trial, we examined the effect of ICD implantation according to status with respect to CRT implantation at baseline.

Methods

In the DANISH trial, 556 patients with non-ischaemic systolic heart failure were randomised to receive an ICD and 560 to receive usual clinical care (control). Patients fulfilling indications for a CRT device received a CRT-defibrillator (if randomised to ICD arm) or CRT-pacemaker (if randomised to control arm). In the ICD group, 322 patients (57.9%) received a CRT device; in the control group, 323 patients (57.7%) received a CRT device. In this extended follow-up study, patients were followed until May 18, 2020. The primary outcome was death from any cause; secondary outcomes were cardiovascular death and sudden cardiovascular death.

Results

During a median follow-up of 9.5 years, the ICD group did not have significantly lower all-cause mortality compared with the control group (hazard ratio [HR] 0.89 [95% CI, 0.74–1.08]). The results were independent of whether the patient received a CRT device at randomisation (patients with a CRT device: HR 0.92 [95% CI, 0.72–1.18]; patients without a CRT device: HR 0.86 [95% CI, 0.64–1.14]; P for interaction, 0.72). Similarly, ICD implantation did not reduce rates of cardiovascular death overall (HR 0.87 [95% CI, 0.70–1.09]), and this association was not modified by CRT (patients with a CRT device: HR 0.89 [95% CI, 0.66–1.19]; patients without a CRT device: HR 0.85 [95% CI, 0.60–1.20]; P for interaction, 0.86). The ICD group had significantly lower rates of sudden cardiovascular death in the overall population (HR, 0.60 [95% CI, 0.40–0.92]), and this association was not modified by CRT (patients with a CRT device: HR 0.69 [95% CI, 0.40–1.21]; patients without a CRT device: HR 0.51 [95% CI, 0.26–0.97]; P for interaction, 0.47). See Figure 1 for all results.

Conclusions

In this extended follow-up study of the DANISH trial, the effect of ICD implantation in patients with non-ischaemic systolic heart failure was not modified by CRT.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): The DANISH trial was supported by unrestricted grants from Medtronic, St Jude Medical, Tryg Fonden, and the Danish Heart Foundation. No further funding was obtained for this follow-up study.

Temporal trends in the incidence of endocarditis among patients with a prosthetic heart valve: a nationwide study

Background

The incidence of infective endocarditis (IE) is increasing in the adult population, as is the insertion of prosthetic heart valves. Patients with prosthetic heart valves are considered at high risk of IE – a complication with a high mortality. However, data on temporal changes in the incidence of IE among patients with prosthetic heart valves from unselected cohorts are sparse

Purpose

We aimed to examine nationwide temporal trends in the incidence of IE in patients with an implanted prosthetic heart valve in Denmark from 1999 to 2018.

Methods

Using Danish nationwide health-care registries we identified all patients, who underwent heart valve implantation between 1996–2018. Crude annual incidence rates per 1,000 person years (PY) of IE were computed and presented in two year intervals. Analyses were stratified by sex and age groups (<50, 50–59, 60–69, 70–79, >80 years).

Results

We identified 26,604 patients with first time prosthetic valve implantation with a median age of 72.7 years at the time of implantation, 63.1% were men with a median follow-up of 6.5 years. We found 1,442 cases of first time IE. The IE incidence rate ranged from 5.4 /1,000 PY (95% CI 3.9–7.4) in calendar period 2001–2002 to 10.0/1,000 PY (95% 8.84–11.11) in calendar period 2017–2018 with an unadjusted increasing trend during the study period (ptrend<0.0001), (Figure 1). Overall, men had a higher crude incidence rate compared with women, however no significant temporal changes were seen in the incidence rate during the study period. For age groups, a trend of stepwise increase in the incidence rate of IE was observed for increasing age groups, however no temporal changes were observed (Figure 2).

Conclusion

The incidence of IE following prosthetic heart valve implantation has increased slightly over the last 20 years in Denmark.

Funding Acknowledgement

Type of funding sources: None.

Adherence and discontinuation of sglt2-inhibitors and glp1-r agonists in patients with type 2 diabetes with and without cardiovascular disease

Background

Sodium-glucose cotransporter-2 (SGLT2) inhibitors and glucacon-like-peptide-1 receptor (GLP-1R) agonists are two novel therapies in type 2 diabetes (T2D) that are recommended in patients with manifestations of cardiovascular disease or high risk of cardiovascular disease due to their cardioprotective benefits. Despite the increasing use of these drugs, there is still limited knowledge on the adherence patterns and the risk of discontinuation, according to the presence or absence of different cardiovascular conditions.

Purpose

To investigate the adherence and estimate the risk of discontinuation of SGLT2 inhibitors and GLP1-R agonists in patients with T2D with and without cardiovascular disease.

Methods

From Danish nationwide registers, we included all individuals >40 years with T2D who redeemed the first prescription of a SGLT2 inhibitor or GLP1-R agonist, between December, 2012 and December, 2018. Patients were divided into subgroups according to the presence or absence of different cardiovascular conditions (heart failure, ischemic heart disease, peripheral artery disease, atrial fibrillation), at the time of inclusion. Adherence over the course of one year after initiation of treatment was estimated by the proportion of days covered (PDC) for each subgroup, and adherence was defined as PDC 80%. The risk of discontinuation was estimated using multivariable cause-specific Cox regression models, and was defined as a break in treatment of 90 days. The Aalen-Johansen estimator was used to account for censoring and competing risks. Patients were followed until date of emigration, death or study end (December 31, 2019).

Results

We included 24,061 patients with T2D who initiated treatment with a SGLT2 inhibitor, and 13,899 patients with T2D who initiated treatment with a GLP1-R agonist, for the first time between December 10, 2012 and December 31, 2018. Median age at inclusion was 62 years (IQR 54–70) and 40% were female. In the analyses, 67% of the patients treated with SGLT2 inhibitors were adherent to therapy throughout the first year, whereas the same proportion of patients, 67%, were adherent to therapy with GLP-1R agonists. Mean PDC was 77% (SD 33) and 79% (31) for patients in therapy with SGLT2 inhibitors and GLP1-R agonists, respectively. No significant difference was observed related to the presence or absence of cardiovascular disease (SGLT2 inhibitor: mean PDC 77 (33) vs. 77 (33), GLP1-R agonist: 78 (32) vs. 79 (31)).

Conclusions

In this nationwide cohort study, we found that patients with T2D who initiated therapy with a SGLT2 inhibitor and those who initiated therapy with a GLP1-R agonist were almost equally adherent to therapy throughout the first year. Adherence in both treatment groups did not differ according to the presence or absence of cardiovascular disease.

Funding Acknowledgement

Type of funding sources: None.

Readmission after transcatheter aortic valve implantation according to frailty

Background

Readmissions and time spent hospitalized following transcatheter aortic valve implantation (TAVI) are important parameters of patient autonomy, particularly for frail patients with limited life-expectancy. Yet, such data remain scarce.

Purpose

To investigate actual time spent hospitalized the first year after TAVI. Secondly, to investigate time spent hospitalized according to frailty risk.

Methods

Through Danish, nationwide registries, we included all patients undergoing TAVI and alive at discharge between January 2008 and June 2020. From discharge, patients were followed until death, emigration, end of study period, or one year of follow-up, whichever came first. During follow-up, all in-patient hospital admissions were identified according to ICD-10 diagnosis codes. Length of stay was calculated, and cumulative numbers of days hospitalized was presented. Further, the proportion of patients dying within one year of follow-up was calculated.

Using The Hospital Frailty Risk Score, a validated frailty risk assessment tool, patients were categorized as low, intermediate, and high frailty risk. We then evaluated the time spent hospitalized stratified by frailty risk group.

Results

The study population comprised 5,464 patients undergoing first-time TAVI with a median age of 81 years among whom 55.2% were males. After one year, 445 (8.1%) patients had died. In total, 2,452 (44.9%) of TAVI patients survived one year and were never admitted, whereas 3,012 (55.1%) patients were admitted at least once or died within one year of TAVI. Of these, 1,200 (21.9%) patients were admitted for more than two weeks or died within one year of TAVI (Figure 1).

Regarding frailty, 3,296 (60.3%), 1,965 (36.0%), and 203 (3.7%) patients were classified as low, intermediate, and high frailty risk, respectively. In the low frailty risk group, 6.2% of patients died within one year and 50.4% survived one year without a hospital admission. By contrast, 16.7% of patients in the high frailty risk group died within one year and only 24.6% survived one year without a hospital admission. Further, 17.1% of patients in the low frailty risk group were admitted for more than two weeks or died within one year of TAVI compared with 47.3% in the high frailty risk group (Figure 2).

Conclusion

Readmissions in the first year after transcatheter aortic valve implantation were common and time spent hospitalized after transcatheter aortic valve implantation was significant. Our results were clearly related to frailty, which should be considered for future prevention strategies.

Funding Acknowledgement

Type of funding sources: None.

Accidental diagnosis of bradyarrhythmia in patients monitored for atrial fibrillation

Funding Acknowledgements

Type of funding sources: Foundation. Main funding source(s): The study was supported by The Innovation Fund Denmark [12-135225], The Research Foundation for the Capital Region of Denmark [no grant number], The Danish Heart Foundation [11-04-R83-A3363-22625], Aalborg University Talent Management Programme [no grant number], Arvid Nilssons Fond [no grant number], Skibsreder Per Henriksen, R. og Hustrus Fond [no grant number], Medtronic [no grant number], and the AFFECT-EU consortium which has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No 847770.

Background

The interest in heart rhythm monitoring and technologies to detect arrhythmia is increasing. The prevalence and prognostic significance of subclinical bradyarrhythmias is unknown.

Objectives

To assess the accidental diagnosis of bradyarrhythmia and its subsequent treatment and prognostic impact in persons screened for atrial fibrillation compared to unscreened persons.

Methods

We utilized a randomized trial of ≥70-year-olds with cardiovascular risk factors recruited outside the hospital setting to receive implantable loop recorder screening for atrial fibrillation (ILR group) vs. usual care (Control group). Time-to-event analyses were performed for bradyarrhythmia, pacemaker implantation, syncope, and sudden cardiovascular death.

Results

A total of 6004 participants were randomized (mean age 75 years, 47% women, 91% with hypertension, 20% with prior syncope), 4503 to Control and 1501 to ILR. The median follow-up period was 64.5 [59.3, 69.8] months. A total of 675 deaths occurred with an overall rate of 2.16 (2.00-2.33) per 100 person-years, and 67 sudden cardiovascular deaths occurred with a rate of 0.21 (0.15-0.28) for the Control group and 0.23 (0.14-0.37) for the ILR group (hazard ratio (HR) 1.11 (0.64-1.90), p=0.71)).

The overall rate of incident bradyarrhythmia was 1.63 (1.49-1.79) per 100 person-years, and bradyarrhythmia was diagnosed in 172 (3.82%) and 312 (20.8%) participants in the Control and ILR group, respectively (HR 6.21 (5.15-7.48), p<0.0001) (Figure 1). The most common bradyarrhythmia was sinus node dysfunction (SND) which was diagnosed in 68 participants in the Control group (1.51%) and 214 in the ILR group (14.26%). In the Control group, 57.35% of diagnoses of sinus node dysfunction resulted in pacemaker implantation, compared to 12.15% in the ILR group where the majority was treated conservatively (Figure 2). The second-most common type of bradyarrhythmia was high-grade atrioventricular block (AVB) which was diagnosed in 86 participants in the Control group (1.91%) and 54 in the ILR group (3.60%). In both groups, the majority of high-grade AVB was treated with pacemaker, although 29.63% in the ILR group were treated conservatively. Risk factors for bradyarrhythmia included higher age, male sex, and prior syncope.

Overall, a pacemaker was implanted in 132 (2.93%) and 66 (4.40%) participants (HR 1.53 (1.14-2.06), p<0.0001), syncope occurred in 120 (2.66%) and 33 (2.20%) participants (HR 0.83 (0.56-1.22), p=0.34), and sudden cardiovascular death occurred in 49 (1.09%) and 18 (1.20%) participants (HR 1.11 (0.64-1.90), p=0.71) in the Control and ILR group, respectively.

Conclusions

Bradyarrhythmias are highly common in ≥70-year-olds with cardiovascular risk factors. Compared to Control, ILR monitoring led to a six-fold increase in diagnosis of bradyarrhythmia and a significant increase in pacemaker implantations, but no change in the risk of syncope or sudden death.

Use of non-recommended drugs in Brugada Syndrome: a Danish nationwide cohort study

Funding Acknowledgements

Type of funding sources: None.

Background

To lower the risk of sudden cardiac death, patients with Brugada Syndrome (BrS) are recommended to avoid intake of drugs known to increase the risk of arrhythmias or the development of type-1 BrS ECG. However, information on adherence to these recommendations among patients with BrS is limited.

Purpose

To examine treatment with non-recommended drugs before and after diagnosis with BrS, risk factors of treatment with these drugs, and whether treatment was associated with a higher risk of hospitalization with ventricular arrhythmias or death.

Methods

All patients diagnosed with BrS in Denmark (1995-2018) with >12 months of follow-up were identified through nationwide registries using the ICD-10 diagnosis code DI472M (PPV 95.8%). Relevant BrS risk drugs were identified and grouped as drugs to "avoid" or "preferably avoid" in agreement with the likelihood of promoting arrhythmias and type-1 BrS ECG according to brugadadrugs.org(1) (accessed August 2021). Multiple logistic regression (adjusted for sex, age, year of diagnosis, and relevant comorbidities and drugs) was performed to identify factors associated with risk drug use during follow-up.

Results

We identified 270 patients with BrS. Median age at the time of diagnosis was 46.2 years [IQR 32.6-57.6], 70.4% were male. Before the time of diagnosis, 16 patients (5.9%) were treated with a drug to "avoid" or "preferably avoid" (n=5 and n=12, respectively). During a median follow-up of 79 months, 89 patients (33%) were treated with at least one BrS risk drug after the time of diagnosis (table). A total of 22 patients with BrS (8.1%) received ≥2 different drugs at any time during follow-up. There was no significant difference in proportions of patients receiving a risk drug the year prior to diagnosis (12.2%) compared to each of the five years following diagnosis (year 1-5, respectively: 12.2%; 9.7%; 12.3%; 13.6%; 13.5% (p>0.05 for all)).

Females had an odds ratio (OR) of 2.21 [95% CI 1.21-4.03] for use of risk drugs. Also associated with a greater likelihood of risk drug use after diagnosis were having a psychiatric disease at baseline (OR=4.80 [1.72-13.41]) and any use of a risk drug within 90 days prior to diagnosis (OR=8.54 [2.13-34.31]) (figure).

During follow-up, six patients were hospitalized for ventricular arrhythmias; none had redeemed a prescription of a risk drug. In total, 12 patients died, of which five (41.7%) had redeemed a prescription of one or more risk drugs within 50 days of death.

Conclusions

1/3 patients with BrS received a risk drug at any time point after diagnosis. No change in proportions of patients treated with risk drugs was identified after time of diagnosis. 5/12 patients that died during follow-up had redeemed a prescription of one or more risk drugs within 50 days of death. Female sex, any psychiatric diagnosis, and use of a non-recommended drug before diagnosis with BrS were associated with a greater likelihood of risk drug use after diagnosis.

Risk of pericardiac effusion after cardiac implantable electronic device implantation a nationwide study

Funding Acknowledgements

Type of funding sources: None.

Background

Procedural pericardiac effusion (PE) is considered a major complication to implantation of cardiac implantable electronic devices (CIED), including permanent pacemakers (PM), cardiac resynchronization therapy devices with defibrillators (CRT-D) or without (CRT-P), and implantable cardioverter defibrillators (ICD), as it can cause life threatening cardiac tamponade. Very little is known about risk factors for procedural PE.

Aim

To identify the patient- and procedure related risk factors associated with clinically relevant procedural PE.

Methods & Results

This is a nationwide retrospective observational cohort study based on data on 51.599 patients from the Danish Pacemaker Register. Included were all Danish patients who received their first PM, CRT or ICD from 2000 – 2018. Procedural PE was defined related to the invasive procedure if it occurred within 1 months after the invasive procedure and no cancer was diagnosed before the procedure. Pre-specified risk factors, including sex, age, year, implantation center-type and device type were analyzed by multivariable logistic regression models to estimate the association with PE. A total of 78 (0.2%) patients were diagnosed with procedural PE, with a median age of 73 years and 43% were females. In adjusted logistic regression analysis age > 70, heart failure [aOR 1.64 (1.01;2.67)], ischemic heart disease [aOR 1.84 (1.13;2.99)], direct oral anticoagulation [aOR 1.77 (1.13–2.77.)], amiodarone use [aOR 3.03 (1.75–5.22)], beta blocking agent [aOR 2.26 (1.23 –4.14)], university hospitals [aOR 2.59 (1.18 –5.67)] and PM implantation [aOR 3.38 (1.77;6.45)], were associated with PE.

Conclusion

Procedural PE is a rare complication after CIED implantation in Denmark. Importantly most of the risk factors for PE are modifiable. Optimizing the modifiable risk factors may reduce the risk of complication.

Familial clustering of unexplained heart failure - A Danish nationwide cohort study

Funding Acknowledgements

Type of funding sources: Public grant(s) – EU funding. Main funding source(s): This project has received funding from the European Union’s Horizon 2020 research and innovation programme under acronym ESCAPE-NET, registered under grant agreement No 733381, and the European Union’s COST programme under acronym PARQ, registered under grant agreement No CA19137.

Background

Although family history of heart failure (HF) is associated with increased risk of HF, the extent to which a family history contributes to the risk of HF needs further investigation.

Purpose

To determine whether a family history of unexplained HF in first-degree relatives (children or sibling) increases the rate of unexplained HF.

Methods

Using Danish nationwide registry data (1978-2017), we identified patients (probands) diagnosed with first unexplained HF (HF without any known comorbidities) in Denmark, and their first-degree relatives. All first-degree relatives were followed from the HF date of the proband and until an event of unexplained HF, exclusion diagnosis, death, emigration, or study end, whichever occurred first. Using the general population as a reference, we calculated adjusted standardized incidence ratios (SIR) of unexplained HF in the three groups of relatives using Poisson regression models.

Results

We identified 57,845 first-degree relatives to individuals previously diagnosed with unexplained HF. Having a family history was associated with a significantly increased unexplained HF rate of 2.08 (95% CI 1.82-2.38) (Figure 1). The estimate was higher among siblings (SIR 4.82 [95% CI 3.17-7.32]). Noteworthy, the rate of HF increased for all first-degree relatives when the proband was diagnosed with HF in a young age (≤50 years, SIR of 3.60 [95% CI 2.37-5.47]) and having >1 proband (SIR of 2.73 [95% CI 1.14-6.56]). The highest estimate of HF was observed if the proband was ≤40 years at diagnosis (6.12 [95% CI 3.39-11.05]) (Figure 2).

Conclusion

A family history of unexplained HF was associated with a two-fold increased rate of unexplained HF among first-degree relatives. If the proband age was ≤40 years, the risk was six-folded. These findings suggest that screening families of unexplained HF with onset below 50 years is indicated.

Systolic blood pressure and effects of screening for atrial fibrillation with long-term continuous monitoring

Funding Acknowledgements

Type of funding sources: Other. Main funding source(s): The LOOP Study was supported by Innovation Fund Denmark [grant number 12-1352259], The Research Foundation for the Capital Region of Denmark, The Danish Heart Foundation [grant number 11-04-R83-A3363-22625], Aalborg University Talent Management Program, Arvid Nilssons Fond, Skibsreder Per Henriksen, R og Hustrus Fond, the European Union’s Horizon 2020 program [grant number 847770 to the AFFECT-EU consortium], Læge Sophus Carl Emil Friis og hustru Olga Doris Friis’ Legat, and an unrestricted grant from Medtronic.

Background

The recently published LOOP Study was a randomized controlled clinical trial to evaluate systematic atrial fibrillation (AF) screening with long-term continuous monitoring in an elderly population at risk and found no significant reduction in stroke. However, the screening effects seemed to differ across levels of systolic blood pressure (SBP). It is well-known that hypertension constitutes a prominent risk factor for clinical AF and stroke alike, but data on the impacts of SBP on subclinical AF and hereby AF screening efficacy are lacking.

Purpose

With this post hoc analysis of the LOOP Study, we aimed to provide insights into the interaction between SBP and benefits of systematic AF screening.

Methods

The LOOP Study randomized individuals aged 70-90 years with ≥1 stroke risk factor (hypertension, diabetes, heart failure, or previous stroke) and without prior AF to either monitoring with implantable loop recorder (ILR) and initiation of oral anticoagulation upon detection of new-onset AF episodes lasting ≥6 minutes, or usual care (control group). In total, 5997 participants with available SBP measurements at enrolment were included in the present analysis. The interaction between SBP and ILR screening efficacy on stroke or systemic arterial embolism (SAE), as indicated by hazard ratio (HR) for ILR versus control, was assessed with polynomial moving-average regression. The lowest SBP threshold with significant screening benefits was further determined and used to examine clinical outcomes and the occurrence of AF with respect to dichotomized SBP. Additionally, penalized spline models were employed to assess AF occurrence by SBP as a continuous variable.

Results

HR of stroke/SAE for ILR versus control decreased with increasing SBP and the lowest threshold for significant screening benefits was at SBP ≥150 mmHg. ILR screening of participants with SBP ≥150 mmHg yielded a 45% risk reduction of stroke/SAE (HR 0.55 [0.37-0.82]). Within the ILR group, SBP ≥150 mmHg was associated with an increased risk of AF episodes ≥24 hours as compared to lower SBP (HR 1.57 [1.01-2.45]), but not with the overall occurrence of AF (HR 1.14 [0.95-1.36]). No significant association between SBP and AF occurrence in the ILR group was reported in penalized spline models either (p-value: 0.73).

Conclusions

The benefits of ILR screening for AF on stroke/SAE increased with increasing blood pressure. SBP ≥150 mmHg was associated with a 1.5-fold increased risk of AF episodes ≥24 hours, along with an almost 50% risk reduction of stroke/SAE by ILR screening.

Socioeconomic position and initiation of SGLT-2 inhibitors or GLP-1 receptor agonists in patients with type 2 diabetes – a Danish nationwide observational study

Background

Between 2015 and 2017, Sodium-glucose cotransporter-2 (SGLT-2) inhibitors and glucacon-like-peptide-1 receptor agonists (GLP-1 RA) were shown to reduce cardiovascular events in patients with type 2 diabetes and cardiovascular disease. Thus, in 2018, guidelines were updated to favor these drugs in patients with cardiovascular disease and type 2 diabetes. Lower socioeconomic position may adversely affect use of SGLT-2 inhibitors and GLP-1 RA.

Purpose

We aimed to examine socioeconomic differences in initiation of SGLT-2 inhibitors and GLP-1 RA in a contemporary population of patients with type 2 diabetes.

Methods

Through the Danish nationwide registers, we identified all patients with type 2 diabetes who initiated second-line add-on therapy after metformin monotherapy between December 10, 2012, and December 31, 2018. Patients aged 40–79 years and without a history of end-stage renal disease were included. We measured socioeconomic position according to level of income: Low = 1st quartile; Middle = 2nd and 3rd quartile; High = 4th quartile. Based on multivariable logistic regression models adjusted for age, sex, cohabitation status, duration of type 2 diabetes, comorbidities, and cardiovascular medications, we reported the standardised probabilities of initiating each drug class at time of first intensification according to income group and time period: 2012–2014, 2015–2017, and 2018.

Results

The 33,201 patients had a median age of 63 years (interquartile range 53–69). The probability of initiating a SGLT-2 inhibitor or a GLP-1 RA increased over time in all income-groups. In each time period, the standardised probability of initiating a SGLT-2 inhibitor or a GLP-1 RA at time of first intensification increased with increasing income (Figure): in 2012–2014, from 9.6% (95% confidence interval (CI) 8.4–10.9) in the lowest income group to 14.4% (CI 12.9–15.9) in the highest income group; in 2015–2017, from 19.5% (CI 18.3–20.7) to 24.6% (CI 23.3–25.9); in 2018, from 39.9% (CI 37.5–42.3) to 50.7% (CI 48.2–53.1). The absolute difference between high and low income groups increased over time, reaching 10.8% (CI 7.3–14.3) in 2018. A similar trend was observed in both subgroups of patients with and without established cardiovascular disease (data not shown). Initiation of a dipeptidyl peptidase-4 (DPP-4) inhibitor increased with income in the early time periods, but this trend reversed in 2018 (Figure). Initiation of sulfonylureas (SU) showed a consistent inverse association with income in each time period.

Conclusions

Low socioeconomic position was consistently associated with a lower probability of initiation of a GLP-1 RA or a SLGT-2 inhibitor at time of first intensification of antidiabetic treatment, even after guidelines recommended these drugs to patients with established cardiovascular disease. These disparities may adversely affect cardiovascular outcomes in patients with low socioeconomic position.

Funding Acknowledgement

Type of funding sources: None.

Prodromal symptoms of out-of-hospital cardiac arrest among patients calling emergency and non-emergency medical help services

Background

Early identification of individuals at risk of out-of-hospital cardiac arrest (OHCA) remains difficult. Little is known about symptoms presented when contacting a medical helpline prior to OHCA.

Aim

To examine the registered prodromal symptoms when patients phoned to seek medical help prior to OHCA.

Methods

OHCA patients (≥18 years) were identified from the Danish Cardiac Arrest Registry (2014–2018) and linked with calls to the non-emergency (1813-Medical Helpline) and Emergency Medical Services 1–1-2 (112). We examined (1) symptoms registered within 30 days before OHCA, categorized into eight groups and stratified by time-period and call-type; (2) hospital diagnoses and medical prescriptions according to symptom groups within 180 days before these calls.

Results

Among 974 OHCA patients who called in total within 30 days before OHCA, 816 OHCA patients (males 57%, median age 76 years [Q1-Q3: 65–84]) had a registered symptom and some of them called more than once (1,145 calls by 816 patients). Overall, the most reported group of symptoms was “Other” (29%), containing a diverse group of prodromal symptoms registered by the caregiver that did not fit into the other categories (Figure), followed by breathing problems (15%). When stratified by time-period (Figure) the most common symptom group remained “Other”. This was followed by symptoms related to the Central Nervous System (CNS)/Unconsciousness (17%) for the time-period within 0–7 days before OHCA, and by breathing problems (19%) and trauma/exposure (17%) for the time-period within 8–30 days before OHCA (Figure). When stratified by call-type, most patients (60.8%) called the 1813-Medical Helpline, where “Other” (35%) and abdominal/back/urinary (14%) symptom groups were the most common. While breathing problems (24%) and CNS/Unconsciousness (21%) were highly reported among calls to 112. Within 180-days before contact with the medical helpline, independently of the symptom group presented, respiratory-related hospital diagnoses and antibiotic medications were common.

Conclusions

Patients with OHCA who called emergency and non-emergency medical helpline 30 days before OHCA had diverse prodromal symptoms; the largest category were “Other” symptoms, followed by breathing-related symptoms.

Funding Acknowledgement

Type of funding sources: Public grant(s) – EU funding. Main funding source(s): European Union's Horizon 2020 under the ESCAPE-NET program;Helsefonden Figure 1. Classification of the prodromal symptoms among patients that called for medical assistance. Stratified by the time-period within 0–7 days and 8–30 days before OHCA. The number of calls within 0–7 days before OHCA = 471 (399 patients), and the number of calls within 8–30 days before OHCA = 674 (500 patients).

Effect of empagliflozin in hfref patients treated with angiotensin receptor neprilysin inhibitor an analysis of EMPIRE HF

Introduction

Inhibition of neprilysin/valsartan (ARNi) or sodium glucose cotransporter 2 (SGLT2) in patients with heart failure (HF) and reduced ejection fraction (HFrEF) has been shown to reduce the risk of Cardiovascular death and hospitalization for HF. Recent trails suggested that SGLT2 reduces the risk for cardiovascular death or hospitalization for HF, regardless of underlying ARNi treatment and that the effect may even be greater in those receiving the combination. Whether there exist an interaction between effect of ARNi and SGLT2 on functional endpoints related to mechanism of action is unknown.

Purpose

This post-hoc analysis of the randomized double-blinded Empire HF trial evaluated the influence of ARNi on the effect of the SGLT2 Empagliflozin on N-terminal prohormone B-type natriuretic peptide (NT-proBNP), pulmonary capillary wedge pressure (PCWP), Left ventricular end-systolic and end-diastolic volumes index; (LVESVI) (LVEDVI), left atrial volume index (LAVI), Left ventricular ejection fraction (LVEF), and Kansas City Cardiomyopathy Questionnaire (KCCQ) HFrEF patients.

Methods

Empire HF trial randomized 190 patients with HFrEF (LVEF ≤40%) to placebo or empagliflozin (10 mg/day), on top of recommended treatment for HFrEF, for 12 weeks of treatment. A total of 58 (31%) received ARNi at baseline and no patients initiated ARNi during study period.

Results

Patients on ARNi were well-treated with a similar baseline characteristic as those who were not treated with ARNi (Table 1). Patients with ARNi had a lower systolic blood pressure (P=0.01), with a higher NT-proBNP (P<0.001) when compared with those not receiving ARNi. When compared to placebo, empagliflozin did not reduce the ratio of change of NT-proBNP with or without ARNi (0.94 [95% CI, 0.75 to 1.19] pg/ml; P=0.62) and (1.02 [95% CI, 0.86 to 1.22] pg/ml; P=0.78), respectively, adjusted (age, atrial fibrillation) interaction P=0.57. Empagliflozin reduced PCWP regardless of ARNi treatment (with ARNi; −4.9 [95% CI, −9.1 to −0.6] mmHg; P=0.02) and (without ARNi; −2.1 [95% CI, −3.8 to −0.4] mmHg; P=0.01), adjusted interaction P=0.20. Overall, empagliflozin was associated with a reduction in LVESVI, LVEDVI, and LAVI volumes, but no effect on LVEF. However, Empagliflozin combined with ARNi at baseline, significantly reduced LVEDVI (−11.2 [95% CI, −21.2 to −1.2] ml/m2; P=0.03), but not without ARNI (−2.9 [95% CI, −8.7 to 2.9] ml/m2; P=0.32), adjusted interaction P=0.13. Treatment-by-subgroup interaction P-values for LVESVI, LAVI, and LVEF analysis were >0.05 (Figure 1). KCCQ total symptom score were significantly increased in those not receiving ARNi (5.4 [95% CI, 1.1 to 9.6]; P=0.013), but not with ARNi (−4.0 [95% CI, −10.3 to 2.3]; P=0.22), adjusted P=0.02.

Conclusion

In this post hoc analysis the effects on empagliflozin to reduce PCWP and LV volumes were not diminished in patients receiving ARNi, however KCCQ change were diminished in patients receiving ARNi.

Funding Acknowledgement

Type of funding sources: Public hospital(s). Main funding source(s): This work was supported by the Danish Heart Foundation [grant numbers 17-R116-A7714-22076, 18-R124-A8573-22107]; Steno Diabetes Center Odense, Denmark [grant number 3363] and A.P. Møller Foundation for the Advancement of Medical Science [grant number 17-L-0339]. Table 1. Baseline characteristicsFigure 1. Change in echo variables +/− ARNi

Elevated markers of liver function are associated with poorer outcomes in HFREF: an analysis of DAPA-HF

Background

Abnormalities of liver tests in patients with heart failure with reduced ejection fraction (HFrEF) is a well-recognised phenomenon. We examined the prognostic value of measures of liver function in a large contemporary cohort of patients with HFrEF enrolled in the Dapagliflozin and Prevention of Adverse Outcomes in Heart Failure (DAPA-HF) trial

Methods

In this post-hoc analysis of the DAPA-HF trial we studied 4625 patients with liver function tests available at baseline. Cox proportional hazards models were used to assess the association between liver tests (total bilirubin, alkaline phosphatase [ALP], alanine transaminase [ALT], aspartate transaminase [AST]) and the Model for End-stage Liver Disease excluding INR (MELD-XI) score (calculated as 5.11 Ln [total bilirubin as mg/dL] + 11.76 Ln [creatinine as mg/dL] + 9.44), and the risk of the primary composite endpoint (hospitalisation or urgent visit for heart failure or cardiovascular death). Models were adjusted for age, sex, race, region, systolic blood pressure, heart rate, LVEF, eGFR, log-transformed NT-proBNP, NYHA class, history of hypertension, stroke, myocardial infarction, atrial fibrillation, heart failure aetiology and randomized treatment to dapagliflozin and stratified by diabetic status at baseline. An interaction term between each measure and the effect of treatment on the primary composite outcome was tested as a fractional polynomial.

Results

Total bilirubin, ALP, and MELD-XI score were associated with a higher risk of all the primary outcome (Figure 1) but not ALT or AST. These relationships persisted after adjustment: total bilirubin per log unit increase (HR=1.46; 95% CI 1.28 – 1.67, p<0.001), ALP per log unit increase (HR=1.39; 95% CI 1.15 – 1.66, p<0.001), MELD-XI per 1 SD increase (HR 1.27; 95% CI 1.13 – 1.42, p<0.001). The effect of dapagliflozin on the primary outcome was not modified by the baseline levels of either total bilirubin, ALP or MELD-XI score (Figure 2)

Conclusions

Higher total bilirubin, ALP and MELD-XI score were independently associated with a higher risk of cardiovascular death or worsening HF and may be useful routinely available biomarkers to assess prognosis. The efficacy of dapagliflozin was the not modified by baseline levels of any of these markers.

Funding Acknowledgement

Type of funding sources: None. Figure 2

Age- and Sex-Specific Rates of Heart Failure and other Adverse Cardiovascular Outcomes in Systemic Sclerosis

Background

Age at disease onset and sex appear to modify the disease course in patients with systemic sclerosis (SSc). Although patients with SSc have a higher risk of adverse cardiovascular outcomes than people without SSc, there are few data on age- and sex-specific risks of heart failure (HF) and other adverse cardiovascular outcomes in patients with SSc.

Objectives

To investigate the long-term rates of HF and other adverse cardiovascular outcomes (including arrhythmias, myocardial infarction, ischemic stroke, venous thromboembolism, and pulmonary hypertension) in a nationwide cohort of patients with SSc compared with the background population according to age and sex, separately.

Methods

Using Danish nationwide registries, all patients >18 years with newly diagnosed SSc (1996–2018) were identified. SSc patients were matched at a 1:4 ratio by age, sex, and comorbidities with controls from the background population without SSc. Rates of outcomes according to age (above/below median age) and sex were compared between cases and controls using Cox regression.

Results

Of the 2,019 patients diagnosed with SSc, 1,569 patients were matched with 6,276 controls from the background population (median age 55 years, 80.4% women). SSc was associated with a higher rate of HF in both women (HR 2.99 [95% CI, 2.18–4.09]) and men (HR 3.01 [1.83–4.95]) (Pfor interaction=0.88), with similar findings for other cardiovascular outcomes.For age interaction, SSc was associated with an increased rate of HF in patients <55 years (HR 4.14 [2.54–6.74]) and ≥55 years (HR 2.74 [1.98–3.78]), with similar effect of younger and older groups on HF (P for interaction=0.21), and other cardiovascular outcomes.

Conclusions

SSc was associated with an increased long-term rate of cardiovascular outcomes compared with a matched background population, with similar extent in different gender and age groups.

Funding Acknowledgement

Type of funding sources: None. Adjusted hazard ratios according to sexAdjusted hazard ratios according to age

Increased prevalence of premature coronary atherosclerosis after preeclampsia

Background

Women with preeclampsia have increased risk of manifest coronary artery disease later in life. However, it remains unknown if and when premature coronary atherosclerotic stages can be identified. This knowledge could enable early intervention in women considered at high risk for future manifest coronary artery disease.

Purpose

Using cardiac computed tomography, we aimed to investigate the prevalence of premature coronary atherosclerosis in women with previous preeclampsia in comparison with women from the general population.

Methods

Women, aged 40–55 years, with previous preeclampsia were recruited in the CPH-PRECIOUS study and compared 1:1 with age- and parity-matched women from the CGPS. Both groups underwent a cardiac computed tomography, including a contrast-enhanced coronary computed tomography angiography and a non-contrast coronary artery calcium scoring, as well as an overall assessment of cardiovascular risk factors imbedded in an extensive questionnaire. Cardiac computed tomography examinations were analysed blindly. The main outcome of the study was the prevalence of any coronary atherosclerosis defined as any plaque at coronary computed tomography angiography or a calcium score >0 in case of a non-diagnostic coronary computed tomography angiography.

Results

A total of 1,424 women were included (715 women with previous preeclampsia and 709 controls from the general population). Women with previous preeclampsia were more likely to have cardiovascular risk factors (hypertension, dyslipidaemia, diabetes mellitus). The prevalence of any coronary atherosclerosis was significantly higher in the preeclampsia group (27.4% vs. 20.0%) (P=0.001). A calcium score >0 was also more prevalent in the preeclampsia group (16.6% vs. 11.8%) (P=0.009). Preeclampsia remained an independent risk factor for the presence of any coronary atherosclerosis after adjusting for cardiovascular risk factors (age, hypertension, dyslipidaemia, diabetes, smoking, body mass index, parity) (OR=1.37, 95% CI (1.05–1.79), P=0.021).

Conclusion

Women with previous preeclampsia are more likely show premature signs of coronary atherosclerosis compared with an age- and parity matched control group from the general population. Preeclampsia is an independent risk factor for premature coronary atherosclerosis.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): The Danish Heart FoundationSkibsreder Per Henriksen, R og Hustrus FondKong Christian den Tiendes FondBrødrene Hartmanns FondHans og Nora Buchards FondArvid Nilssons FondAnita og Tage Therkelsens FondLægefondenAase og Ejnar Danielsens FondHjertecentrets Forskningsudvalg (Rigshospitalet)Direktør Kurt Bønnelycke og Hustru Fru Grethe Bønnelyckes FondLægeforeningens ForskningsfondTorben & Alice Frimodt FondHenry og Astrid Møllers Fond

Left atrial remodeling and cerebrovascular disease assessed by magnetic resonance imaging in patients undergoing continuous heart rhythm monitoring

Background

Atrial remodeling and atrial fibrillation (AF) have both been associated with cerebrovascular lesions. We wished to investigate the possible direct association between atrial remodeling and cerebrovascular disease including white matter lesions and lacunar infarcts in patients with and without atrial fibrillation (AF) as documented by implantable loop recorder (ILR).

Methods

Cardiac and cerebral MRI scans were acquired in a cross-sectional study including participants ≥70 years of age with stroke risk factors (history of hypertension, diabetes mellitus, congestive heart failure and/or previous stroke) but without known AF. Cerebrovascular disease was visually rated using the Fazekas scale and number of lacunar strokes. Left atrial (LA) (see figure) and ventricular volumes and function were analyzed, and associations between atrial remodeling and cerebrovascular disease were assessed with logistic regression models. Multivariable models were adjusted for sex, age, diabetes, hypertension, heart failure and history of stroke/transient ischemic attack. The analyses were stratified according to sinus rhythm or any AF during three months of continuous ILR monitoring to account for subclinical AF.

Results

Of 200 participants investigated, 87% had a Fazekas score≥1 and 45% had ≥1 lacunar infarct. Within three months of ILR implantation, AF was detected in 28 (14%) participants, while 172 (86%) had sinus rhythm only. Results are summarized in table. For participants with sinus rhythm, lower LA passive emptying fraction was associated with Fazekas score after multivariable adjustment, while LA total emptying fraction was borderline significant, and increased LA maximum and minimum volumes were associated with lacunar infarcts. There were no significant associations in patients with AF.

Sensitivity analyses showed similar results with longer screening periods for AF.

Conclusions

In patients free from AF as documented by ILR monitoring, we found an independent association between LA passive emptying and Fazekas score, and between atrial volumes and lacunar infarcts. This supports that atrial remodeling alone without AF is associated with an increased risk of cerebrovascular lesions.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): The Innovation Fund, DenmarkThe Research Foundation for the Capital Region of Denmark

Infarct size following loading with ticagrelor/prasugrel versus clopidogrel in ST-segment elevation myocardial infarction

Background

Treatment with newer direct-acting anti-platelet drugs (Ticagrelor and Prasugrel) prior to primary percutaneous coronary intervention (PCI) is associated with improved clinical outcome in patients with ST-segment elevation myocardial infarction (STEMI) when compared with Clopidogrel.

Purpose

We retrospectively compared infarct size following non-randomized treatment with

Ticagrelor/Prasugrel versus Clopidogrel in a population of STEMI patients treated with primary PCI.

Methods

Patients were loaded with Clopidogrel, Ticagrelor or Prasugrel in the ambulance before primary PCI. Infarct size and myocardial salvage index were calculated using cardiac magnetic resonance (CMR) during index admission and at three-month follow-up.

Results

693 patients were included in this analysis. Clopidogrel was given to 351 patients and Ticagrelor/Prasugrel to 342 patients. The groups were generally comparable in terms of baseline and procedural characteristics. Median infarct size at three-month follow-up was 12.9% vs 10.0%, in patients treated with Clopidogrel and Ticagrelor/ Prasugrel respectively (p<0.001), and myocardial salvage index was 66% vs 71% (p<0.001). Results remained significant in a multiple regression model (p<0.001).

Conclusion

Pre-hospital loading with Ticagrelor or Prasugrel compared to Clopidogrel, was associated with smaller infarct size and larger myocardial salvage index at three-month follow-up in patients with STEMI treated with primary PCI.

Infarct size at three month follow-up

Funding Acknowledgement

Type of funding source: Foundation. Main funding source(s): The Novo Nordisk Foundation. The Alfred Benzon Foundatioun.

Efficacy and safety of clopidogrel, ticagrelor, and prasugrel in an all-comers population of patients with ST-segment elevation myocardial infarction

Background

Until 2009, aspirin and clopidogrel were recommended for most patients with acute coronary syndrome (ACS). After 2009, this recommendation was changed to aspirin combined with prasugrel or ticagrelor since randomised trials had demonstrated reduced cardiovascular mortality and ischemic events, however with a slight increase in bleeding risk. Randomised clinical trials often include selected patients and the results may not be generalisable to an all-comers population of high-risk ACS patients.

Purpose

To compare efficacy and safety of clopidogrel, ticagrelor, and prasugrel in all-comers patients with ST-segment elevation myocardial infarction (STEMI).

Methods

The Eastern Danish Heart Registry was utilised to identify all consecutive STEMI patients admitted to the capital region from 2009–2016. By individual linkage to Danish nationwide registries, claimed drug prescriptions and end points were obtained. Patients alive a week after discharge were included and stratified according to clopidogrel, ticagrelor, or prasugrel treatment, and followed for 18 months. The risk of the primary efficacy end point (a composite of all-cause mortality, recurrent myocardial infarction, and ischemic stroke) and the safety end point (bleeding events leading to hospital admission) were assessed by multivariate Cox proportional-hazard models.

Results

In total, 4841 STEMI patients were included (clopidogrel [n=1222], ticagrelor [n=1820], prasugrel [n=1799]). The median age was 66 (IQR 57–76) for clopidogrel, 64 (IQR 54–73) for ticagrelor, and 59 (IQR 51–67) for prasugrel, and only 19% were women of the prasugrel treated patients (29% for clopidogrel, 25% for ticagrelor). Treatment with anticoagulant therapy was 21% for clopidogrel treated patients (4% for ticagrelor, 5% for prasugrel). Number of events and incidence rates/100 years (IR) for the primary efficacy end point were 165 (IR 9.7) for clopidogrel, 134 (IR 5.1) for ticagrelor, and 107 (IR 4.1) for prasugrel, and for bleeding events 57 (IR 3.3) for clopidogrel, 60 (IR 2.3) for ticagrelor, and 55 (IR 2.1) for prasugrel treatment. Compared with clopidogrel, a reduction in the primary efficacy end point was found in patients treated with both ticagrelor (HR 0.47, 95% CI 0.36–0.63, p<0.001) and prasugrel (HR 0.49, 95% CI 0.36–0.67, p<0.001) with no difference in bleeding events (HR 0.71, 95% CI 0.45–1.13, p=0.15 and HR 0.72, 95% CI 0.44–1.17, p=0.18, respectively). No differences were found between prasugrel and ticagrelor treated patients for the primary efficacy end point (HR 0.83, 95% CI 0.60–1.16, p=0.28) or safety end point (HR 0.97, 95% CI 0.61–1.54, p=0.90).

Conclusion

Ticagrelor and prasugrel treatment in all-comers STEMI patients were associated with reduced rates of all-cause mortality and ischemic events without an increase in bleeding events when compared with clopidogrel treatment. No differences in efficacy or safety were found between prasugrel and ticagrelor treated patients.

Efficacy+safety end points at 18 months

Funding Acknowledgement

Type of funding source: None

Surgical treatment for infective endocarditis over three decades: a nationwide study

Background

An increasing number of patients with infective endocarditis (IE) are treated surgically over time. It is important to know how this affects patient outcome. Current studies are mainly from tertiary centres which may bias estimations of outcomes. We have therefore conducted a nationwide study of surgical outcomes during admission for IE over three decades.

Purpose

We set out to examine temporal trends in use of valve surgery for IE and these patients' characteristics and related outcomes in Denmark in the period 1998–2017.

Methods

Using Danish nationwide registries, we included patients with first-time IE (1998–2017). The study population was categorized into four groups of five-year intervals (1998–2002, 2003–2007, 2008–2012, 2012–2017). Annual number of patients with IE and the proportion who underwent valve surgery during admission were reported. Kaplan-Meier estimates and multivariable logistic regression analyses were used to compare the associated 30-day mortality risk between calendar periods. Kaplan-Meier estimates and multivariable adjusted Cox proportional hazard analyses were used compare the associated 1-year mortality risk between calendar periods.

Results

A total of 8,455 patients with first-time IE were identified in the period of 1998–2017 of which 1,906 (22.5%) underwent valve surgery (1998–2002; N=320, 2003–2007; N=468, 2008–2012; N=528, 2013–2017; N=595). The proportion of patients who underwent surgery was 21.5% in 1998 and 19.4% in 2017 (P=0.02 for trend). See figure.

For patients undergoing surgery, the median age and proportion of males increased from 58.3 years (P25-P75: 48.2–67.4) and 69.1% to 66.7 years (P25-P75: 55.2–73.0) and 73.1% in 1998–2002 and 2013–2017, respectively. Patients had an increasing burden of comorbidities including diabetes (10.3% to 14.3%), hypertension (16.9% to 37.5%) and renal disease (9.1% to 9.6%) across calendar periods. The 30-day mortality risk for patients with IE who underwent valve surgery was 10.0% (1998–2002), 10.8% (2003–2007), 6.4% (2008–2012) and 8.5% (2013–2017), respectively (P=0.09). One-year mortality risk for patients with IE who underwent valve surgery was 16.7% (1998–2002), 21.2% (2003–2007), 15.2% (2008–2012) and 16.6% (2013–2017), respectively (P=0.08). The declining 30-day and 1-year mortality was statistically significant over time when adjusting for patient characteristics (P=0.01 and P≤0.0001, respectively).

Conclusion

From a nationwide, unselected cohort of patients with first-time IE, around 1/5 undergo surgery during admission. Surgical IE-cases are older and sicker now compared to 10–20 years ago. In spite of this, there was a trend towards a decreased associated 30-day and 1-year mortality over time. Our data show a lower rate of surgery in IE than in most prior studies and we believe that this is due to the nationwide, unselected nature of our study.

Infective endocarditis and surgery

Funding Acknowledgement

Type of funding source: None

Severity of congenital Long QT Syndrome disease onset and risk of depression, anxiety, and mortality: a nationwide study

Introduction

The congenital Long QT Syndrome (cLQTS) is associated with an increased risk of sudden cardiac death (SCD). Thus, cLQTS patients are susceptible to develop depression or anxiety, both of which have been associated with poor outcomes including risk of mortality.

Aim

We examined if a cLQTS diagnosis and the severity of cLQTS disease onset was associated with an increased risk of depression, anxiety, and all-cause mortality compared with a matched control population.

Methods

Using Danish nationwide registries and inherited cardiac disease clinics, we identified all patients with known cLQTS (1994–2016) who were ≥18 years at the time of diagnosis. The disease onset for cLQTS was identified as asymptomatic, ventricular tachycardia [VT]/ syncope, aborted SCD [aSCD], or unknown (i.e. no available information). After cLQTS diagnosis, we determined the risk of depression (i.e. depression diagnosis or prescription of antidepressants), anxiety (i.e. anxiety diagnosis or prescription of anxiolytics), and mortality using multivariable Cox proportional hazards regression. Patients were followed for three years. An age and gender matched control population was identified (matching 1:4). Competing risk analysis with death as competing risk was used to generate cumulative incidence plots.

Results

Overall, 428 cLQTS patients were identified of which 107/428 (25%) developed depression or anxiety after being diagnosed with cLQTS compared with 285/1712 (16.6%) from the control population (p<0.001). The severity of disease onset was identified for 229/428 (55%) cLQTS patients; 104 (24%) were asymptomatic, 89 (21%) had VT/ syncope, and 36 (8.4%) had aSCD. A graded relationship between the severity of cLQTS disease onset and risk of depression or anxiety was identified (Figure 1). In multivariable models, patients with aSCD as disease onset had a higher risk of developing depression or anxiety compared with asymptomatic cLQTS patients (HR=2.34, CI: 1.03–5.32). Furthermore, previous depression (HR=6.38, CI: 4.80–8.48) and anxiety (HR=4.20, CI: 3.15–5.59) was found as associated risk factors. However, no risk was associated with concurrent treatment with beta-blockers (HR=1.23, CI: 0.90–1.69). During follow-up, 8 cLQTS patients died of which 4 had developed depression or anxiety (50%). No significant association between all-cause mortality and depression or anxiety was found, although numbers were low (P=0.22).

Conclusion

The prevalence of depression and anxiety was high among cLQTS patients after diagnosis. Moreover, a graded relationship between severity of disease onset and risk of depression or anxiety was identified. These findings highlight a need for increased awareness following a cLQTS diagnosis in order to reduce the risk of adverse outcomes.

Cumulative incidence curve

Funding Acknowledgement

Type of funding source: Public hospital(s). Main funding source(s): Fund of Rigshospitalet

Risk of heart failure associated with thiazide diuretics compared with calcium channel blockers in patients with type 2 diabetes: a nationwide nested case-control study

Background

Thiazide diuretics and calcium channel blockers (CCB's) are two important and widely used antihypertensive drugs classes among patients with type 2 diabetes (T2D). The risk of developing heart failure (HF) is increased in patients with T2D but whether use of these two drugs are associated with changes in HF risk is unknown.

Purpose

To examine and compare the association of two different classes of antihypertensive drugs, thiazide diuretics and CCB's, with the development of new onset HF in patients with T2D.

Methods

The study cohort comprised T2D patients >40 years on metformin and renin-angiotensin system inhibitor (RAS-i) without a history of HF or use of loop diuretics identified in Danish health care registers (period 1995 to 2015). A nested case-control study was conducted by matching all HF cases on sex, age and duration of T2D with 10 controls from the T2D population. Exposure was defined as three redeemed prescriptions of either a thiazide diuretic or a CCB up to 365 days before index, which corresponds to one year of antihypertensive therapy. Conditional logistic regression adjusted for comorbidities (atrial fibrillation, chronic obstructive pulmonary disease and anemia) was used to estimate and compare the treatment effect of thiazide diuretics and CCB's, with patients receiving neither of the two drugs as reference.

Results

The study population consisted of 170,514 T2D patients using metformin and RAS-i, comprising 13,814 HF cases each matched on sex, age and duration of T2D with 10 controls. The median age was 62 years and 55% were men. T2D patients, who had received antihypertensive treatment with only thiazide diuretics one year prior to index had a significantly lower risk of HF compared to the reference group who did not receive treatment with neither thiazide diuretics or CCB's: Hazard ratio (HR) 0.79 [95% confidence interval (CI) 0.74–0.85]. Patients who had received treatment with only CCB's had a comparable risk of HF: HR 0.98 [95% CI 0.94–1.02]. Patients who had received treatment with both thiazide diuretics and CCB's were not associated with a lower risk of HF: HR 1.01 [95% CI 0.96–1.08].

Conclusion

Patients with T2D who received antihypertensive therapy with thiazide diuretics for at least one year had a significantly lower risk of HF compared to those who were not treated with either thiazide diuretics or CCB's. No association between use of CCB's and HF was observed. Use of thiazide diuretics may prevent development of HF in T2D and a randomized clinical trial evaluating diuretics is patients with T2D is warranted.

Risk of new onset heart failure

Funding Acknowledgement

Type of funding source: None

Temporal trends in utilization of transcatheter aortic valve implantation and patient characteristics – a nationwide cohort study

Introduction

In recent years, there has been a surge in the utilization of transcatheter aortic valve implantation (TAVI) for the treatment of severe symptomatic aortic stenosis. Randomized controlled trials have compared TAVI to surgical aortic valve replacement (SAVR) in patients at high-, intermediate-, and low perioperative risk. As TAVI continues to be utilized in patients with lower risk profiles, it is important to investigate the temporal trends in “real-world” patients undergoing TAVI.

Purpose

To investigate temporal trends in the utilization of TAVI and examine changes in patient characteristics of patients undergoing first-time TAVI.

Methods

Using complete Danish nationwide registries, we included all patients undergoing first-time TAVI between 2008 and 2017. To compare patient characteristics, the study population was stratified according to calendar year in the following groups: 2008–2009, 2010–2011, 2012–2013, 2014–2015, and 2016–2017.

Results

We identified 3,534 patients undergoing first-time TAVI. In 2008–2009, 180 patients underwent first-time TAVI compared with 1,417 patients in 2016–2017, resulting in a 687% increase in TAVI procedures performed. During the study period, the median age remained stable (2008–2009: Median age 82 year [25th–75th percentile: 78–85] vs. 2016–2017: Median age 81 years [25th–75th percentile: 76–85]; P-value: 0.06). The proportion of men undergoing first-time TAVI increased over the years (2008–2009: 49.4% vs 2016–2017: 54.9%; P-value for trend: <0.05), also the proportion with diabetes increased (2008–2009: 12.2% vs. 2016–2017: 19.3%; P-value for trend: <0.05). The proportion of patients with a history of stroke decreased over the years (2008–2009: 13.9% vs. 2016–2017: 12.1%; P-value for trend: <0.05). The same trend was seen in patients with a history of myocardial infarction (2008–2009: 24.4% vs. 2016–2017: 11.9%; P-value for trend: <0.05), ischaemic heart disease (2008–2009: 71.7% vs. 2016–2017: 29.4%; P-value for trend: <0.05), and heart failure (2008–2009: 45.6% vs. 2016–2017: 29.4%; P-value for trend: <0.05).

Conclusions

In this nationwide study, there was a marked increase in the utilization of TAVI in the years 2008–2017. Patients undergoing first-time TAVI had a decreasing comorbidity burden, while the age of the patients at first-time TAVI remained stable.

Funding Acknowledgement

Type of funding source: None

Increasing time to thrombolysis is associated with worse long-term outcomes in patients with ischaemic stroke: a nationwide study

Background

It is well-established that the short-term benefits of intravenous thrombolytic therapy are reduced with increasing treatment delay in patients with acute ischaemic stroke. However, there is a paucity of contemporary data on the association between time from symptom onset to initiation of thrombolysis and long-term outcomes. With improving post-stroke survival in the Western world, data on time to thrombolysis and subsequent long-term outcomes are warranted in order to provide further insight into the importance of time to treatment.

Purpose

To examine the long-term risk of adverse outcomes according to time from symptom onset to intravenous thrombolytic therapy in patients with acute ischaemic stroke.

Methods

In this observational cohort study, we identified all patients with first-time ischaemic stroke treated with intravenous thrombolysis between 2011–2015 and alive at discharge through the Danish National Stroke Registry. Patients who received thrombolysis after >270 min were excluded. Using multivariable Cox regression, we examined associations between time from symptom onset to thrombolysis and risks of the composite of death, recurrent ischaemic stroke, and dementia, as well as each of these components separately. Patients were followed until the outcome of interest, emigration, or December 31, 2017.

Results

Of the 4,313 patients with first-time ischaemic stroke treated with intravenous thrombolysis, 4,119 were alive at discharge (median age 69 years [25th-75th percentile 59–78 years], 60% males). The median follow-up was 3.3 years (25th-75th percentile 2.3–4.7 years). The median time from symptom onset to initiation of thrombolytic therapy was 140 min (25th-75th percentile 106–187 min), and the median National Institutes of Health Stroke Scale score at presentation was 5 (25th-75th percentile 3–10). The unadjusted absolute 3-year risks of the composite outcome, death, recurrent ischaemic stroke, and dementia according to time to thrombolysis are displayed in the figure. Compared with thrombolysis within 90 min, time to thrombolysis >90 min was associated with a higher relative risk of the composite outcome (91–180 min: adjusted hazard ratio [HR] 1.37 [95% confidence interval [CI], 1.13–1.68]; 181–270 min: adjusted HR 1.42 [95% CI 1.15–1.76]). The risks of each component of the composite outcome according to time to thrombolysis were similar to results for the composite endpoint, as illustrated in the figure.

Conclusions

In this nationwide cohort of patients with acute ischaemic stroke treated with thrombolysis, increasing time from symptom onset to initiation of intravenous thrombolytic therapy was associated with higher long-term risks of the composite of death, recurrent ischaemic stroke, and dementia, as well as all three outcomes separately. These data indicate that long-term outcomes of patients with ischaemic stroke treated with intravenous thrombolysis can be greatly improved by reducing treatment delay.

Time to thrombolysis and outcomes

Funding Acknowledgement

Type of funding source: None

Long-term mortality in patients with infective endocarditis who undergo aortic root replacement versus isolated aortic valve replacement

Background

Infective endocarditis (IE) with involvement of the aortic root (root abscess or aortic prosthetic valve endocarditis (PVE)) is associated with high mortality and morbidity, and it often requires root replacement. IE-guidelines recommend surgery in patients with aortic root involvement (class B recommendation), but the surgical technique is challenging, and the perioperative risk is high. Long-term data are sparse for these high-risk patients and may help better select appropriate patients for surgery.

Purpose

We set out to investigate the short- and long-term mortality for patients with destructive aortic valve endocarditis who underwent aortic root replacement. For comparison, we included those who underwent isolated aortic valve replacement for IE (isolated AVR).

Methods

We included patients with first-time IE from 2000 to 2016 who underwent aortic valve surgery identified from The Eastern Danish Thoracic surgery database. Patient characteristics were identified by cross-linking Danish nationwide databases. Patients who underwent aortic root replacement were compared with those who underwent isolated AVR. Kaplan-Meier plots and multivariable Cox regression analyses were used to estimate and compare the associated 30-days and 10-year mortality risks between groups.

Results

We included 368 patients with aortic valve IE who underwent AVR surgery; 126 patients underwent aortic root replacement and 242 underwent isolated AVR. Median age for root replacement patients was 65.4 years (interquartile range [IQR] 56.2–73.0) compared with 62.1 years (IQR 52.3–71.6) for isolated AVR patients. In the root replacement group, 40.5% had prosthetic valve endocarditis (PVE), whereas 6.6% had PVE in the isolated AVR group. 30-day mortality was 12.7% (CI95: 7.6%-19.2%) in the root replacement group and 7.0% (CI95: 4.3%-10.7%) in the isolated AVR group (P=0.06). Estimated 10-year mortality was 54.4% (CI95: 40.3%-67.6%) in the root replacement group and 45.3% (CI95: 35.7%-54.5%) (P=0.07) after isolated AVR (figure 1). At up to 10 years follow-up, there was no significant difference in adjusted mortality between the groups, adjusted HR=1.34 (CI 95: 0.90–2.00).

Conclusion

Patients with IE who underwent aortic root replacement surgery more often had a prosthetic heart valve, were older, and were more often male. There was no significant difference in long-term mortality between the groups. Nonetheless, long-term mortality was high – 50% of patients died by 10 years, and our results underline the need for stringent patient selection.

Figure 1

Funding Acknowledgement

Type of funding source: None

Heart failure in survivors of out-of-hospital cardiac arrest – factors associated with reduced systolic ventricular function at follow-up

Background

Survival after out-of-hospital cardiac arrest (OHCA) has increased in recent years but is still only 10%. Little is known about the association between post-resuscitation comorbidity and heart failure after discharge from the initial OHCA-admission.

Purpose

In OHCA-survivors we aimed to describe predictors of left ventricular (LV) dysfunction, defined as LV ejection fraction (LVEF) <40%, at follow-up.

Methods

A consecutive cohort of OHCA-patients with cardiac cause from 2007 to 2011 without a pre-OHCA congestive heart failure diagnosis (according to the Danish National Patient Registry, which holds data on all Danish citizens) were retrospectively examined. Logistic regression analyses were used to assess factors associated with LV dysfunction (LVEF <40%) at follow-up after a median of 6 months. Follow-up was not performed systematically in the OHCA-survivors and data from follow-up was assessed by reading of patient charts.

Results

A total of 365 OHCA-survivors with a mean age of 61 years were discharged alive from hospital. LVEF <40% at hospital discharge was seen in 54% (n=184, 7% missing), and at follow-up after a median of 6 months 19% (n=69) of the total OHCA-cohort of survivors still had LV dysfunction. Factors associated with LV dysfunction at follow-up were chronic ischemic heart disease (IHD) prior to OHCA (odds ratio (OR) = 2.9 (95% CI: 1.2 – 7.1)) and ST-elevation myocardial infarction (STEMI) as cause of OHCA (OR = 2.9 (1.4–6.0)), whereas age, gender, high comorbidity burden prior to OHCA or pre-hospital circumstances (including shockable cardiac arrest rhythm) were not.

Conclusion

More than half of OHCA-survivors with LVEF <40% at hospital discharge improved LV function and LV dysfunction at follow-up after a median of 6 months after discharge was present in 1 in 5 (19%) of the cohort. Chronic IHD and STEMI were the only factors significantly associated with LV dysfunction at follow-up. A systematic follow-up including echocardiography in the outpatient clinic for OHCA-survivors is recommended especially in patients with reduced LV function at discharge and in STEMI-patients in order to assess the appropriateness of heart failure medication and an implantable cardiac defibrillator.

Funding Acknowledgement

Type of funding source: Private grant(s) and/or Sponsorship. Main funding source(s): Danish Foundation Trygfonden

Declining incidence and mortality of ischaemic stroke between 1996–2016: a nationwide study

Background

The incidence and mortality of ischaemic heart disease have been declining over many years. The development with ischaemic stroke is less well studied, and with an increasing elderly population, there is a need for large-scale studies. Recent changes in stroke prevention and treatments may have affected the incidence and mortality of ischaemic stroke.

Purpose

To examine time trends and sex and age differences in the incidence and mortality of first-time ischaemic stroke in Denmark between 1996–2016.

Methods

In this observational cohort study, we used Danish nationwide registries to identify all individuals >18 years of age admitted with a first-time diagnosis of ischaemic stroke between 1996–2016. We calculated age- and sex-stratified annual incidence rates and absolute 30-day and 1-year mortality risks. Further, we calculated annual incidence rate ratios using multivariable Poisson regression, odds ratios for 30-day mortality using multivariable logistic regression, and hazard ratios for 1-year mortality using multivariable Cox regression.

Results

The study population consisted of 224,617 individuals >18 years of age with first-time ischaemic stroke between 1996–2016. The figure displays the unadjusted incidence rates and 1-year mortality risks of ischaemic stroke by calendar year. The overall unadjusted incidence rates of ischaemic stroke per 1,000 person-years increased from 1996 (2.43 [95% confidence interval [CI], 2.38–2.47]) to 2002 (2.91 [95% CI, 2.86–2.96]) and then gradually decreased to below the initial level until 2016 (1.99 [95% CI, 1.95–2.03]). Men had higher incidence rates than women in all age groups except in patients between 18–30 years and >85 years. The absolute mortality risk decreased between 1996–2016 (30-day mortality from 17.1% to 7.6% and 1-year mortality from 30.9% to 17.3%). Women had higher mortality than men in the age groups 55–64 years and >85 years. Similar trends were observed for all analyses after multivariable adjustment.

Conclusions

The overall incidence of first-time hospitalization for ischaemic stroke increased from 1996–2002 and then gradually decreased to below the initial level until 2016. The absolute 30-day and 1-year mortality risk decreased between 1996–2016. These findings correspond to the increased awareness of stroke prevention and introduction of new treatment options during the study period.

Trends in stroke incidence and mortality

Funding Acknowledgement

Type of funding source: None

Risk of infective endocarditis among patients with tetralogy of fallot

Introduction

Patients with tetralogy of Fallot (ToF) are considered at high risk of infective endocarditis (IE) as a result of altered hemodynamics and multiple surgical and interventional procedures including pulmonary valve replacement (PVR). The overall survival of patients with ToF has increased in recent years. However, data on the risk of adverse outcomes including IE are sparse.

Purpose

To investigate the risk of IE in patients with ToF compared with controls from the background population.

Methods

In this nationwide observational cohort study, all patients with ToF born in 1977–2017 were identified using Danish nationwide registries and followed from date of birth until occurrence of an outcome of interest (i.e. first-time IE), death, or end of study (July 31, 2017). The comparative risk of IE among ToF patients versus age- and sex-matched controls from the background population was assessed.

Results

A total of 1,156 patients with ToF were identified and matched with 4,624 controls from the background population. Among patients with ToF, 266 (23.0%) underwent PVR during follow-up. During a median follow-up time of 20.4 years, 38 (3.3%) patients and 1 (0.03%) control were admitted with IE. The median time from date of birth to IE was 10.8 years (25th-75th percentile 2.8–20.9 years). The incidence rates of IE per 1,000 person-years were 2.2 (95% confidence interval (CI) 1.6–3.0) and 0.01 (95% CI 0.0001–0.1) among patients and controls, respectively. In multivariable Cox regression models, in which age, sex, pulmonary valve replacement, and relevant comorbidities (i.e. chronic renal failure, diabetes mellitus, presence of cardiac implantable electronic devices, other valve surgeries), were included as time-varying coefficients, the risk of IE was significantly higher among patients compared with controls (HR 171.5, 95% CI 23.2–1266.7). Moreover, PVR was associated with an increased risk of IE (HR 3.4, 95% CI 1.4–8.2).

Conclusions

Patients with ToF have a substantial risk of IE and the risk is significantly higher compared with the background population. In particular, PVR was associated with an increased risk of IE. With an increasing life-expectancy of these patients, intensified awareness, preventive measures, and surveillance of this patient group are advisable.

Figure 1. Cumulative incidence of IE

Funding Acknowledgement

Type of funding source: None

Hemoglobin level at stabilization is associated with long-term all-cause mortality in patients with left-sided endocarditis, a POET substudy

Background

Left-sided infectious endocarditis (IE) has a high 1-year mortality. Anemia is a common finding in patients with IE, yet little is known about frequency, severity, and associated outcomes in this setting.

Purpose

To examine the relationship between Hemoglobin (Hgb) level measured at IE stabilization (time of randomization) in the Partial Oral versus intravenous Antibiotic Treatment of Endocarditis (POET) trial - and long-term all-cause mortality.

Methods

In the POET trial, 400 patients with left-sided IE were randomized, after medical and/or surgical stabilization, to conventional antibiotic treatment or partial oral treatment. Only non-surgically treated patients were considered in this study. Patients were divided by quartiles into four groups based on Hgb level at randomization.

Results

We examined 248 patients with non-surgically treated IE. Median time from diagnosis of IE to randomization was 14 days (IQ 12–19). At long-term follow-up (median 3.2 years, IQ 2.18–4.60), 71 patients had died (28.6%). Patients in the lowest quantile (Hgb ≤6.0 mmol) had a HR of 4.17 (95% CI 1.81–9.61, p<0.001) for death compared to patients in the highest quantile (Hgb >7.5 mmol/L). This association remained significant after multivariable adjustment for age, sex, renal disease, C-Reactive Protein, and Prosthetic heart valve (HR 2.69, 95% CI 1.11–6.50); p=0.028).

Conclusion

Low Hemoglobin level at stabilization in patients with IE was associated with an increased risk of long-term mortality. Whether intensified treatment of anemia in patients with IE could improve long-term outcome requires investigation.

Funding Acknowledgement

Type of funding source: Foundation. Main funding source(s): The Danish Heart Foundation, The Capital Regions Research Council

1347The flu vaccine and mortality in hypertension. A Danish nationwide cohort study

Background

Influenza infection is associated with an increased risk of acute myocardial infarction (AMI) and stroke. It is currently unknown whether influenza vaccination may reduce mortality in patients with hypertension.

Purpose

To determine whether influenza vaccination is associated with lower risks of death in hypertensive patients without significant cardiovascular or other chronic disease.

Methods

Using nationwide registers, we identified all patients with hypertension in Denmark during 9 consecutive influenza seasons in the period 2007–2016 who were treated with at least 2 different classes of antihypertensive medication (beta-blockers, diuretics, calcium antagonists or renin-angiotensin system inhibitors). Patients who were not 18–100 years old or had ischemic heart disease, heart failure, chronic obstructive lung disease, cancer or cerebrovascular disease were excluded. Prior to each influenza season we assessed the exposure to influenza vaccination. End-points were death from all causes, from AMI or stroke, or cardiovascular death. For each season, patients were followed from December 1 until April 1 the next year, spanning the period of high influenza activity in Denmark.

Results

A total of 608,452 Patients were followed for a median of 5 seasons (interquartile-range: 2–8 seasons), with total follow-up time of 975,902 person-years. The vaccine coverage during study seasons ranged from 26% to 36%. During follow-up, 21,571 patients died of all-causes (3.5%), 12,270 patients died of cardiovascular causes (2.0%) and 3,846 patients died of AMI/stroke (0.6%). Vaccination was associated with older age, Diabetes Mellitus, atrial fibrillation, lower educational level, lower income and higher medication use. In unadjusted analysis considering all seasons, vaccination was significantly associated with increased risk of all-cause death, cardiovascular death and death from AMI/stroke. However, following adjustment for season, age, sex, comorbidities, medications, income, education, and more, vaccination was significantly associated with reduced risks of all-cause death, cardiovascular death and death from AMI/stroke (Figure).

PY, person-years.

Conclusion

In a nationwide study spanning 9 consecutive influenza seasons including more than 600,000 hypertensive patients without significant cardiovascular disease identified through medication use, influenza vaccination was significantly associated with a reduced risk of death from all-causes, cardiovascular causes and AMI/stroke. Influenza vaccination may improve patient outcome in hypertension.

Acknowledgement/Funding

Daniel Modin was supported by the Herlev & Gentofte University Hospital Internal Research Fund and by the Novo Nordisk Foundation.

P3661Two-fold increase in incidence of infective endocarditis in the period 1997–2016: a Danish nationwide study

Background

Infective Endocarditis (IE) is a disease with high mortality. Previous studies have shown considerable differences and contradicting trends in overall incidence and mortality why data from an unselected nationwide cohort is needed.

Purpose

We investigated temporal trends in the incidence rate and in-hospital mortality of IE in Denmark in the period of 1997–2016.

Methods

We included cases of first-time IE (1997–2016) using Danish nationwide registries. Crude incidence rates were given for each calendar year. Further, incidence rates were reported for subgroups of age and sex. For the analysis of patient characteristics and in-hospital mortality, the study cohort was grouped into four 5-year intervals (1997–2001, 2002–2006, 2007–2011, 2012–2016). Multivariable adjusted Cox proportional hazard model was used to compare in-hospital mortality between groups.

Results

A total of 8,147 patients with IE were identified in the period of 1997–2016. The median age and proportion of males increased from 64.3 years (P25-P75: 48–75.5) and 59.1% to 71.8 years (P25-P75: 62.1–79.9) and 67.1% in 1997–2001 and 2012–2016, respectively. The overall incidence rate (Figure 1) increased from 4.68/100.000-person-years (PY) (CI95: 4.17–5.26) to 8.23/100.000 PY (CI95: 7.53–8.99) in 1997 and 2016, respectively. Male incidence increased from 5.35/100.000 PY (CI95: 4.59–6.23) to 11.03/100.000 PY (CI95: 9.9–12.29) and female incidence increased from 4.03/100.000 PY (CI95: 3.38–4.8) to 5.44/100.000 PY (CI95: 4.67–6.35) in 1997 and 2016 respectively. Incidence rates increased more than seven-fold for the oldest age group (≥80 years) from 1997 to 2016 (6.95/100.000 PY [CI95: 5.32–9.08] to 51.19/100.000 PY [CI95: 43.41–60.38], respectively). In-hospital mortality was significantly lower for patients with IE in the period of 2011–2016 compared with 1997–2001 HR: 0.8 (CI95: 0.69–0.92).

Figure 1

Conclusion

Infective endocarditis incidences are increasing mostly among men and elderly patients. In order to prevent this disease as best as possible, we need more knowledge on causes for this increasing incidence.

P765One-year mortality risk after intensification of outpatient diuretics

Background

Mortality is increased following a hospitalization for heart failure (HF). It is not clear whether outpatient intensification of diuretic confers the same increased risk in the general population with heart failure

Purpose

This study sought to assess 1-year mortality risk after worsening HF, defined either as hospitalization due to HF or as intensified diuretic therapy in an outpatient setting, in a complete nationwide cohort of patients with HF on angiotensin converting enzyme inhibitors/ angiotensin receptor blocker and beta blockers.

Methods

From nationwide administrative registers, we identified all patients in Denmark diagnosed with HF in 2001–2016 and prescribed angiotensin converting enzyme inhibitor/ angiotensin receptor blocker and beta blocker within 120 days. During follow-up we defined worsening HF by the following events: Inpatient worsening (HF readmission) and outpatient worsening (intensified diuretic therapy, defined as the first event of new addition or doubled dosage of loop diuretic therapy or new onset addition of thiazide to loop diuretic therapy). Patients with a worsening event were risk set matched to two HF controls each at time of the event – based on age, sex and calendar year. One-year mortality risk was estimated with Kaplan-Meier and multivariable Cox regression models.

Results

We included 74,990 patients, median age 71 years (interquartile range: 62–79), 36% women. During five years of follow up, 8,727 patients had an inpatient worsening event, and 12,290 had an outpatient worsening event as first event. Absolute risk of 1-year mortality was 22.6% (95%-confidence interval (95%-CI): 21.7%-23.5%) after inpatient worsening, 18.0% (95%-CI: 17.3%-18.7%) after outpatient worsening compared to 9.8% (95%-CI: 9.5%-10.1%) for the matched controls. In a multivariable Cox model adjusted ischemic heart disease, atrial fibrillation, chronic obstructive pulmonary disease and diabetes, the hazard ratio for mortality among patients experiencing inpatient worsening was 2.46 (95%-CI: 2.33–2.60) and for outpatient worsening was 1.87 (95%-CI: 1.77–1.97), compared with the matched HF controls as reference (figure 1). Among patients who had an outpatient worsening as first event, 1,245 (10.1%) had a subsequent HF readmission within one year.

Conclusion

In a nationwide cohort of patients with HF, outpatient worsening defined by a diuretic intensification was associated with almost 2-fold risk of mortality during the next year. Although HF hospitalization is associated with a higher risk, the need to intensify diuretics in the outpatient setting is a signal to review and intensify efforts to improve HF outcomes.

Acknowledgement/Funding

The Danish Heart Foundation, (grant number 17-R116-A7610-22048)