Are non-invasive estimations of plasma volume an accurate measure of congestion in patients with chronic heart failure?

AIMS

We report associations between different formulae for estimating plasma volume status (PVS) and clinical and ultrasound markers of congestion in patients with chronic heart failure (CHF) enrolled in the Hull Lifelab registry.

METHODS AND RESULTS

Cohort 1 comprised patients with data on signs and symptoms at initial evaluation (n = 3505). Cohort 2 included patients with ultrasound assessment of congestion [lung B-line count, inferior vena cava (IVC) diameter, jugular vein distensibility (JVD) ratio] (N = 341). Two formulae for PVS were used: (a) Hakim (HPVS) and (b) Duarte (DPVS). Results were compared with clinical and ultrasound markers of congestion. Outcomes assessed were mortality and the composite of heart failure (HF) hospitalisation and all-cause mortality. In cohort 1, HPVS was associated with mortality [hazard ratio (HR) per unitary increase = 1.02 (1.01-1.03); P < 0.001]. In cohort 2, HPVS was associated with B-line count (HR) = 1.05 [95% confidence interval (CI) (1.01-1.08); P = 0.02] and DPVS with the composite outcome [HR = 1.26 (1.01-1.58); P = 0.04]. HPVS and DPVS were strongly related to haemoglobin concentration and HPVS to weight. After multivariable analysis, there were no strong or consistent associations between PVS and measures of congestion, severity of symptoms, or outcome. By contrast, log[NTproBNP] was strongly associated with all three.

CONCLUSION

Amongst patients with CHF, HPVS and DPVS are not strongly or consistently associated with clinical or ultrasound evidence of congestion, nor clinical outcomes after multivariable adjustment. They appear only to be surrogates of the variables from which they are calculated with no intrinsic clinical utility.

Inferior vena cava diameter adds information to pulmonary artery systolic pressure in outpatients with heart failure

Background

Evaluation of pulmonary arterial pressure (PAP) is recommended by guidelines to diagnose and monitor congestion in patients with heart failure (HF). However, the high compliance of the venous system might buffer the effects of a large increase in intravascular volume and delay an elevation in PAP. Therefore, measuring inferior vena cava (IVC) diameter by ultrasounds might add important information in patients with HF, regardless of PAP.

Aim

To evaluate the relationship between the IVC diameter and systolic PAP (PAsP) assessed by echocardiography with mortality in outpatients with HF.

Methods

We enrolled consenting out-patients attending a community HF clinic for initial diagnosis or follow-up in Kingston upon Hull, UK between 2009–2012. HF was defined as the presence of relevant symptoms and signs and objective evidence of cardiac dysfunction: either a left ventricular systolic function (LVEF) &lt;50% or elevated amino-terminal pro-brain natriuretic peptide (NT-proBNP) ≥125 pg/ml. IVC was considered dilated when &gt;2.0 cm, PAsP elevated when &gt;35 mmHg.

Results

Amongst the 874 patients enrolled, median age was 75 years, 68% were men, median LVEF was 44% and median NT-proBNP was 1125 pg/ml. 468 patients (54%) had normal IVC and PAsP, 117 (13%) had normal IVC but elevated PAsP, 75 (8%) had dilated IVC but normal PAsP and 214 (25%) had both dilated IVC and elevated PAsP. Compared to those with normal IVC and PAsP, those with elevated PAsP but normal IVC were older, more likely to be women, and had higher LVEF and NT-proBNP, whilst those with dilated IVC but normal PAsP had similar age and LVEF, but more signs of congestion and higher NT-proBNP. Compared to those with both normal IVC and PAsP (reference), those with dilated IVC but normal PAsP (HR: 1.83; CI: 1.04–3.25; P=0.037) or elevated PAsP but normal IVC (HR: 1.88; CI: 1.25–2.85; P=0.003) had a similarly increased risk of death but those with a dilated IVC and elevated PAsP had the greatest risk (HR: 4.16; CI: 3.10–5.57; P&lt;0.001).

Conclusion

A dilated IVC is associated with mortality even when PAsP is not elevated. Tailoring treatments to reduce IVC diameter is a strategy worth exploring to improve outcomes in outpatients with heart failure.

Funding Acknowledgement

Type of funding sources: None.

The burden and pattern of comorbidities and its relation to mortality in frail vs non-frail patients with chronic heart failure

Background

Frailty is common in patients with chronic heart failure (CHF). Frail patients are at high risk of poor clinical outcomes which might be attributable to the presence of multiple comorbidities. The impact of comorbidities on mortality in frail patients with CHF is not well described.

Aim

To compare the burden and patterns of comorbidities in frail vs non-frail patients with CHF and their impact on mortality.

Methods

We studied consecutive patients attending a routine follow-up visit to a HF clinic. Frailty was assessed using the Clinical Frailty Scale (CFS); those with CFS&gt;4 were classified as frail. Patients were classified into 6 comorbidity groups including: metabolic (obesity, diabetes); respiratory; renal; cancer; neuropsychiatric (depression, dementia); and degenerative (falls, arthritis, fragility fractures). We investigated the relation between frailty, comorbidity groups and all-cause mortality in patients with CHF.

Results

Amongst 467 patients with CHF [67% male, median (IQR) age 76 (69–82) years, NTproBNP 1156 (469–2463) ng/L], 291 patients had HF with reduced ejection fraction (HFrEF, LVEF &lt;40%), and 176 had HF with preserved ejection fraction (HFpEF, LVEF ≥40%). Frailty was more common in HFpEF vs HFrEF (51 vs 40%). 64% of patients had &gt;4 comorbidities (36% 5–6, 21% 7–9 and 7% &gt;9 comorbidities).

Frail patients were more likely to have multiple comorbidities than non-frail patients (85% vs 48% with &gt;4 comorbidities, p&lt;0.001). The number of comorbidities increased with worsening frailty severity (Figure 1). Those with HFrEF were more like to suffer from cancer, whereas those with HFpEF were more likely to have neuropsychiatric, metabolic and degenerative comorbidities.

During a median follow up of 554 days, 82 (18%) patients died. Increasing number of comorbidities was associated with increasing mortality. (Figure 2) Patients who were frail with &gt;4 comorbidities had a 6-fold increased risk of mortality compared to those who were neither frail nor had multiple comorbidities [HR (95% CI) 6.6 (3.2–13.9), p&lt;0.001]. In a model adjusted for age, sex, logNTproBNP and NYHA class, amongst comorbidity groups, the presence of renal and neuropsychiatric comorbidities were independent predictors of higher mortality.

Conclusion

Frail patients with CHF have a high comorbidity burden. The co-existence of frailty and multiple comorbidities predisposes to higher risk of mortality. Future studies should investigate whether treatment focusing on comorbidities improve outcomes.

Funding Acknowledgement

Type of funding sources: None.

Hypochloraemia following admission to hospital with heart failure is common and associated with an increased risk of readmission or death: a report from OPERA-HF

AIMS

Hypochloraemia is common in patients hospitalized with heart failure (HF) and associated with a high risk of adverse outcomes during admission and following discharge. We assessed the significance of changes in serum chloride concentrations in relation to serum sodium and bicarbonate concentrations during admission in a cohort of 1002 consecutive patients admitted with HF and enrolled into an observational study based at a single tertiary centre in the UK.

METHODS AND RESULTS

Hypochloraemia (<96 mmol/L), hyponatraemia (<135 mmol/L), and metabolic alkalosis (bicarbonate >32 mmol/L) were defined by local laboratory reference ranges. Outcomes assessed were all-cause mortality, all-cause mortality or all-cause readmission, and all-cause mortality or HF readmission. Cox regression and Kaplan-Meier curves were used to investigate associations with outcome. During a median follow-up of 856 days (interquartile range 272-1416), discharge hypochloraemia, regardless of serum sodium, or bicarbonate levels was associated with greater all-cause mortality [hazard ratio (HR) 1.44, 95% confidence interval (CI) 1.15-1.79; P = 0.001], all-cause mortality or all-cause readmission (HR 1.26, 95% CI 1.04-1.53; P = 0.02), and all-cause mortality or HF readmission (HR 1.41, 95% CI 1.14-1.74; P = 0.002) after multivariable adjustment. Patients with concurrent hypochloraemia and natraemia had lower haemoglobin and haematocrit, suggesting congestion; those with hypochloraemia and normal sodium levels had more metabolic alkalosis, suggesting decongestion.

CONCLUSION

Hypochloraemia is common at discharge after a hospitalization for HF and is associated with worse outcome subsequently. It is an easily measured clinical variables that is associated with morbidity or mortality of any cause.

Inter operator variability of machine learning researchers predicting all-cause mortality in patients admitted to intensive care unit

Background

The number of publications using machine learning (ML) to predict cardiovascular outcomes and identify clusters of patients at greater risk has risen dramatically in recent years. However, research papers which use ML often fail to provide sufficient information about their algorithms to enable results to be replicated by others in the same or different datasets.

Aim

To test the reproducibility of results from ML algorithms given three different levels of information commonly found in publications: model type alone, a description of the model, and complete algorithm.

Methods

MIMIC-III is a healthcare dataset comprising detailed information from over 60,000 intensive care unit (ICU) admissions from the Beth Israel Deaconess Medical Centre between 2001 and 2012. Access is available to everyone pending approval and completion of a short training course.

Using this dataset, three models for predicting all-cause in-hospital mortality were created, two from a PhD student working in ML, and one from an existing research paper which used the same dataset and provided complete model information. A second researcher (a PhD student in ML and cardiology) was given the same dataset and was tasked with reproducing their results. Initially, this second researcher was told what type of model was created in each case, followed by a brief description of the algorithms. Finally, the complete algorithms from each participant were provided. In all three scenarios, recreated models were compared to original models using Area Under the Receiver Operating Characteristic Curve (AUC).

Results

After excluding those younger than 18 years and events with missing or invalid entries, 21,139 ICU admissions remained from 18,094 patients between 2001 and 2012, including 2,797 in-hospital deaths. Three models were produced: two Recurrent Neural Networks (RNNs) which differed significantly in internal weights and variables, and a Boosted Tree Classifier (BTC). The AUC of the first reproduced RNN matched that of the original RNN (Figure 1), however the second RNN and the BTC could not be reproduced given model type alone. As more information was provided about these algorithms, the results from the reproduced models matched the original results more closely.

Conclusions

In order to create clinically useful ML tools with results that are reproducible and consistent, it is vital that researchers share enough detail about their models. Model type alone is not enough to guarantee reproducibility. Although some models can be recreated with limited information, this is not always the case, and the best results are found when the complete algorithm is shared. These findings have huge relevance when trying to apply ML in clinical practice.

Funding Acknowledgement

Type of funding sources: None.

The efficacy of using a holistic 4-domain approach in evaluating frailty in ambulatory patient with heart failure

Background

Recently, the Heart Failure Association (HFA) of the European Society of Cardiology (ESC) proposed a four-domain approach to assess frailty in patients with heart failure (HF), to tailor treatment and potentially improve outcomes. The efficacy of such approach in detecting frailty and predicting outcome in patients with HF is unknown.

Aim

To study the prevalence and prognostic value of four different types of frailty deficits: clinical, physical, cognitive and social frailty in ambulatory patients with HF.

Methods

We assessed prospectively consecutive patients attending a routine follow-up visit. Patients with ≥5 non-HF comorbidities were classified as having a clinical deficit. Those who scored ≥3 using the Fried criteria were classified as having a physical deficit. Those who failed to complete a clock test accurately were classified as having a cognitive deficit. Those who lived alone or in a residential home were classified as having a social deficit. All patients were followed for a minimum of 1 year. The primary end point is all-cause mortality.

Results

We enrolled 467 patients (67% male, median (25th–75th centile) age 76 (69–82) years, median (25th–75th centile) NT-proBNP 1156 (469–2463) ng/L). 65% of patients had clinical deficits, 52% had a physical deficit, 39% had a social deficit and 18% had a cognitive deficit. 28% had 2, 19% had 3, 8% of patients had all 4 deficits; 16% had none. An increasing number of frailty deficits was associated with worse symptoms, higher NT-proBNP and less likelihood of being prescribed guideline-indicated HF treatment.

During a median follow-up of 554 days, 82 patients died. The presence of any frailty deficit was associated with increased risk of mortality. (Figure 1) The more frailty deficit a patient had, the higher the risk of mortality (Figure 2). A base model (including age, body mass index, NYHA class and log [NT-proBNP]) for predicting mortality at 1 year achieved a C-statistic of 0.78. Addition of all four deficits improved the performance of the base model (C-statistic = 0.82).

Conclusion

Clinical, physical, cognitive and social deficits are common in patients with HF and are associated with a poor outcome. Future studies should evaluate how a domain-based approach can be used to optimise care for frail patients with HF.

Funding Acknowledgement

Type of funding sources: None. Figure 1Figure 2

A comparison of machine learning models for predicting rehospitalisation and death after a first hospitalisation with heart failure

Background

Many machine learning models exist, including Multilayer Perceptron (MLP), Random Forest algorithm (RF), Support Vector Machine (SVM), and Gradient Boosted Machine (GBM), but their value for predicting outcome in patients with heart failure has not been compared.

Aim

To predict rehospitalisation (all-cause) and death (all-cause) at 1-, 3- and 12 months after discharge from a first hospitalisation for heart failure using four machine learning models.

Methods

The National Health Service Greater Glasgow and Clyde Health Board serves a population of ∼1.1 million. We obtained de-identified administrative data, including investigations, diagnosis and prescriptions, linked to hospital admissions and deaths for anyone with a diagnosis of vascular disease or heart failure or prescribed loop diuretics, statins or neuro-endocrine antagonists at any time between 1st January 2010 and 1st June 2018. Patients who were under 18 or had no prior hospitalisation for heart failure were excluded. Four ML algorithms using 46 variables were applied.

Results

Of 360,000 people who met the above criteria between 2010–2018, 6,372 had a hospitalisation for heart failure prior to 1st January 2010 and 8,304 had a first hospitalisation for heart failure thereafter. Between 2010 and 2018 there were 3,086 re-hospitalisations over 24 hours and 3,706 patients died, with 5,070 patients experiencing the composite outcome.

GBM and RF consistently outperformed MLP and SVM when comparing AUC, sensitivity and specificity combined, with GBM performing best in all scenarios. Since GBM and RF are both tree-based models, and with SVM and MLP regularly reporting very poor sensitivity or specificity despite a similar AUC to the others, this suggests that SVM and MLP may be suffering from overfitting and might perform better in larger data-sets.

Both GBM and RF work by ordering variables, so the final model can be used to determine the most important prediction variables. Age, number of times a blood sample was taken out of hospital, length of stay, social deprivation index and haemoglobin concentration consistently ranked amongst the most important variables. Models predicted all 1-month events better than later events.

Conclusions

Some, but not all, ML models applied to this data-set predicted rehospitalisation and death with great accuracy for up to 3 months after a first hospitalisation for heart failure. The models identified several important prognostic variables that are currently seldom collected in clinical research registries but perhaps should be.

Funding Acknowledgement

Type of funding source: Public grant(s) – National budget only. Main funding source(s): Medical Research Council

Natural history and prognostic significance of anaemia and iron deficiency in ambulatory patients with heart failure

Background

Anaemia and iron deficiency (ID) are both common and associated with adverse outcomes in patients with heart failure (HF). However, the incidence of and recovery from anaemia and ID and their impact on prognosis are not well described.

Methods

Between 2001 and 2018, patients with suspected HF referred to a clinic serving a local population of ∼500,000 were invited to be enrolled in a long-term registry. HF was defined as relevant symptoms or signs with either at least moderate systolic dysfunction on echocardiography or raised plasma concentration of amino-terminal pro-B type natriuretic peptide (NT-proBNP ≥125pg/ml). ID was defined as a transferrin saturation (TSAT) of &lt;20%, anaemia was defined as a haemoglobin (Hb) of &lt;13.0 g/dL in men and &lt;12.0 g/dL in women. At 1-year follow-up, 872 patients had repeat assessments for ID. Patients were grouped into four phenotypes according to the presence or absence of anaemia and/or ID. Those who developed or recovered from ID and anaemia were assessed separately. Survival analysis was conducted at 5 years after the one-year visit.

Results

The prevalence of ID and anaemia at baseline was 40% and 29% respectively. At baseline and at one-year, 53% of patients had either ID, anaemia or both. Compared to other groups, those with both anaemia and ID were older, had worse renal function and higher median NT-proBNP. In patients with TSAT &gt;20% without anaemia at baseline, 23% had ID at 1 year, 14% were anaemic and 6% developed both. At one year, 11% of patients with anaemia and ID at baseline had normal values for both compared to 16% of those with anaemia but not ID. In patients with ID without anaemia at baseline, 51% remained iron deficient at one year, irrespective of initial Hb.

At 5-years, survival was markedly worse for those with anaemia at baseline compared to those without anaemia, irrespective of iron status. Compared to those who had normal TSATs and Hb, those with both anaemia and ID had the worst survival (HR=2.35; 1.77–3.11; p&lt;0.001), followed by those with anaemia without ID (HR=1.93; 1.40–2.67; p&lt;0.001) and those with isolated ID (HR=1.34; 1.01–1.78; p=0.046).

Compared to patients who never had anaemia or ID, patients who developed (HR 2.01; 1.41–2.88; p&lt;0.001) or recovered from (HR 2.21; 1.45–3.39; p&lt;0.001) anaemia or ID (HR 1.61; 1.14–2.28; 0.007 and HR 1.63; 1.16–2.28; 0.005 respectively) had a worse prognosis.

Conclusions

About 30% of patients with HF who have neither anaemia nor ID will develop such problems within a year and this associated with a worse prognosis. Recovery from ID and anaemia is also common, but this is not associated with a better prognosis. Factors leading to the development of anaemia and ID may be driving prognosis rather than anaemia and ID themselves

Funding Acknowledgement

Type of funding source: None

Use of social media by health care professionals: is patients' privacy respected?

Background

The internet has a key role in sharing and expanding medical knowledge. Social networks offer health-care professionals the possibility to communicate, debate and learn from each other in real-time, thereby improving access to expertise and creating new knowledge. On Facebook, there are many groups that health-care professionals can join to discuss clinical cases. However, it is unknown if patient-privacy is respected and whether users are aware that they might be inadvertently involved in a criminal act for which they might be sued.

Methods

We identified the most popular Facebook group for cardiologists, “ECHO BOARD REVIEW forum” that has more than 47,000 users. We retrospectively evaluated all content posted by this group between February 6th and February 14th 2020. Information about the type of data posted, personal patient details, geographic location reported on images or videos, number of reactions and comments was collected. Privacy was considered fully violated when name and/or surname of the patient was identifiable. A comparison between the two cases (privacy preserved vs. privacy violated) was performed.

Results

Of 53 posts evaluated, 50 (94%) were echocardiograms; the remaining three cases were of an electrocardiogram, a cardiac magnetic resonance video and a coronary angiogram. The patient's identity was revealed in seven cases (13%). Compared to cases where identity was concealed, those revealing the patient's identify were more likely to report other important personal details including date of birth (n=0 (0%) vs 2 (29%)), age (n=14 (30%) vs 5 (71%)) and sex (n=15 (33%) vs 6 (86%)). The country (n=46, 87%), city (n=39, 74%), and hospital (n=34, 64%) where the cases were being evaluated were also frequently disclosed. Most cases were from Asia (n=23, 43%) or Africa (n=17, 32%). Of the 7 cases in which privacy was fully violated, 43% were from Africa, 27% were from South America and 14% were from Asia. In the majority of cases, the author of the post was seeking diagnostic help (n=33), less frequently a diagnosis was already made by the author (n=17). The median (interquartile range) number of comments/post was 12 (5–23), and of likes/post was 21 (12- 37), with no significant difference between cases in which privacy was violated or not.

Conclusion

Social media allows knowledge and expertise to be shared amongst health care professionals, but, alarmingly, violation of patient-confidentiality is common. In order to maintain patient-confidentiality and avoid breaking the law, strict rules should be applied to regulate the use of social media by health-care professionals.

Funding Acknowledgement

Type of funding source: None

Why do frail patients with chronic heart failure die and become hospitalised?

Background

Frailty is common in patients with heart failure (HF) and is associated with increased morbidity and mortality. A better understanding of the causes of hospitalisations and death in frail patients might help to tailor interventional strategies for these at-risk patients.

Purpose

We studied the cause of death and hospitalisations in ambulatory patients with HF and frailty.

Methods

We assessed frailty using the clinical frailty scale (CFS) in consecutive HF patients attending a routine follow-up visit. Those with CFS ≥5 were classified as frail. Mortality and hospitalisations were ascertained from medical records (updated systematically using an NHS electronic database), discharge letters, autopsy reports and death certificates. We studied the primary cause of death and hospitalisations within one year of enrolment.

Results

467 patients (67% male, median (IQR) age 76 (69–82) years, median (IQR) NT-proBNP 1156 (469–2463) ng/L) were enrolled. 206 (44%) patients were frail. Frail patients were more likely to not receive or receive suboptimal doses of ACEi/ARB and Beta-blockers; while non-frail patients were more likely to be treated with optimal doses.

At 1-year follow up, there were 56 deaths and 322 hospitalisations, of which 46 (82%) and 215 (67%) occurred in frail patients. Frailty was associated with an increased risk of all-cause mortality (HR (95% CI): 4.27 (2.60–7.01)) and combined mortality/ hospitalisation (HR (95% CI): 2.85 (2.14–3.80)), all p&lt;0.001. 57% (n=26) of frail patients died of cardiovascular causes (of which 58% were due to HF progression); although deaths due to non-cardiovascular causes (43%, n=20), especially severe infections, were also common (26%, n=12). (Figure 1)

The proportion of frail patients who had non-elective hospital admissions within 1 year was more than double that of non-frail patients (46% (n=96) vs 21% (n=54); p&lt;0.001). Compared to non-frail patients, frail patients had more recurrent (≥2) hospitalisations (28% (n=59) vs 9% (n=24); p&lt;0.001) but median (IQR) average length of hospital stay was not significantly different (frail: 6 (4–11) vs non-frail: 6 (2–12) days, p=0.50). A large proportion of hospitalisations (64%, n=137) in frail patients were due to non-cardiovascular causes (of which 34%, 30% and 20% were due to infections, falls and comorbidities respectively). Of cardiovascular hospitalisations (36%, n=78), the majority were due to decompensated HF (67%, n=46). (Figure 1)

Conclusion

Frailty is common in patients with HF and is associated with an increased risk of mortality and recurrent hospitalisations. A significant proportion suffered non-cardiovascular deaths and hospitalisations. This implies that interventions targeted at HF alone can only have limited impact on outcomes in frail patients.

Figure 1

Funding Acknowledgement

Type of funding source: None

Population prevalence of atrial fibrillation and QRS prolongation and mortality in people with suspected heart failure

Background

The 12-lead electrocardiogram (ECG) is an essential tool for the diagnosis and management of heart failure (HF). There are few population-based studies on the prevalence and prognostic implications of ECG abnormalities in patients with HF. There are also no robust diagnostic criteria for HF. We explored these issues in a large administrative database.

Methods

The National Health Service Greater Glasgow and Clyde Health Board serves a population of ∼1.1 million. We obtained de-identified administrative data, including investigations, diagnosis and prescriptions, linked to hospital admissions and deaths, for anyone with a diagnosis of vascular disease or HF or prescribed loop diuretics (LD) or neuro-endocrine antagonists between 1st January 2012 and 1st April 2018. People were classified into 5 exclusive groups: a) prevalent HF; b) incident (or latent) HF with onset during follow-up; c) people taking LDs but with no diagnosis of HF at any time; d) new prescription of LDs during follow-up but with no diagnosis of HF at any time and d) people to whom none of the above applied. ECGs were classified according to heart rhythm (sinus, AF or flutter or pacemaker/CRT/ICD) and QRS duration &lt;100ms, 100–130ms or &gt;130ms. Follow-up for each group started on 1st of January 2012 (prior to the onset of the classifying event for incident groups).

Results

During the observation period, of 316,350 people included, 158,421 had a recorded ECG (mean of 3.2 per person with an ECG), including 8,768 prevalent and 13,195 incident cases of HF. Of those who never got a diagnosis of heart failure, 11,508 were receiving and a further 14,633 were newly prescribed LD during follow-up. There were 110,317 people who did not fall into the above groups, of whom 51,089 were aged ≥60 years. A higher proportion of those who were prescribed loop diuretics without a diagnosis of heart failure were women. A similar proportion of those with heart failure and those prescribed diuretics alone had, lung disease and renal dysfunction but patients with heart failure had more ischaemic heart disease, more often had a heart rhythm other than sinus and had longer QRS duration. By three years, 8,816 people (11%) had died, of whom 2,919 (33%) had a diagnosis of heart failure and 2,694 (31%) had been prescribed LD without a diagnosis of HF, together accounting for 64% of all deaths. Patients with a rhythm other than sinus had a worse prognosis in all 5 groups of patients. QRS duration &gt;130ms was associated with a worse prognosis in patients with HF or taking LD.

Conclusions

Most people with cardiovascular disease who die will first develop HF or be prescribed a LD (indicating possible undiagnosed HF). Patient characteristics of those prescribed LD suggest that many might have HF with preserved left ventricular ejection fraction (HFpEF).

Baseline characteristics and HR

Funding Acknowledgement

Type of funding source: None

P2629Early benefits of empagliflozin in patients with type 2 diabetes with heart failure are not offset by increased adverse events: results from the EMPA-REG OUTCOME trial

Background

The cardiovascular (CV) benefits of sodium-glucose co-transporter-2 (SGLT2) inhibitors in patients with type 2 diabetes (T2D) have been demonstrated in long-term clinical trials. In the EMPA-REG OUTCOME trial, the SGLT2 inhibitor empagliflozin (EMPA), compared with placebo (PBO), significantly reduced the risk of CV death and hospitalisation for heart failure (HHF) in patients with T2D and established CV disease, with a median follow-up time of 3.1 years.

Purpose

To investigate the early benefits and safety associated with use of EMPA in patients enrolled in the EMPA-REG OUTCOME trial according to heart failure (HF) status at baseline.

Methods

We evaluated the effects of treatments on glycated haemoglobin (HbA1c) levels and on the clinical endpoints of HHF, HHF or CV death, and HHF or all-cause mortality (ACM), as well as the occurrence of adverse events (AEs), at 12 weeks, 6 months, and 1 year after randomisation. Outcomes data were explored descriptively at 12 weeks, and assessed by Cox regression models adjusting for baseline risk factors at 6 months, and 1 year, whereas safety data were explored descriptively. Effects on HbA1c were evaluated using a Mixed Model Repeated Measures (MMRM) model.

Results

A total of 7020 participants, 706 (10%) with investigator-reported HF at baseline, were randomised to PBO, or two different doses of EMPA (10 mg or 25 mg once daily). In patients with HF at baseline, the adjusted mean differences in HbA1c between pooled EMPA and PBO at 12 weeks, 6 months, and 1 year after randomisation were −0.55, −0.54 and −0.53%-point, respectively, p<0.001 vs PBO for all, with similar results in those without HF (p for interactions 0.822, 0.939 and 0.539 at 12 weeks, 6 months and 1 year, respectively). Already at 12 weeks, patients assigned to EMPA had a lower frequency of all evaluated clinical outcome events (HHF, HHF or CV death, HHF or ACM) compared with PBO, regardless of HF status. This effect was sustained and significant at 6 months and 1 year in those with and without HF (see Figure). During the same time frame, the rates of AEs were generally higher in those with HF than without HF, but were not increased by the use of EMPA. At 1 year, any AE occurred in 206 (84.4%) and 1694 (81.1%) patients with and without HF, respectively, on PBO vs 363 (78.6%) and 3246 (76.8%) patients with and without HF on EMPA; any serious AE at 1 year occurred in 79 (32.4%) and 447 (21.4%) patients with and without HF on PBO vs 105 (22.7%) and 764 (18.1%) of those with and without HF on EMPA.

Conclusions

In the EMPA-REG OUTCOME trial, EMPA led to early beneficial effects on morbidity and mortality outcomes in patients with T2D with or without HF, which were not offset by an increased risk of AEs.

P3823Body mass index or waist and hip circumference as predictors of outcome in the UK biobank

Background

Controversies exist about the relationship between body habitus and mortality, especially for patients with cardiovascular disease.

Purpose

We evaluated the relations between different anthropometric indices and mortality amongst participants with and without cardiovascular (CV) risk factors, or established CV disease (stroke, myocardial infarction and/or heart failure), enrolled in the UK Biobank.

Methods

The UK Biobank is a large prospective study which, between 2006 and 2010, enrolled 502,620 participants aged 38–73 years. Participants filled questionnaires and had a medical history recorded, physical measurements done and biological samples taken. The UK Biobank is routinely linked to national death registries and updated on a quarterly basis. Data on death were coded according to the International Classification of Diseases, 10th Revision (ICD-10). The primary end-point was all-cause mortality (ACM) across three subgroups of men and women: those with, or without, one or more CV risk factors (smoking, diabetes and/or hypertension), and those with CV disease (history of stroke, myocardial infarction and/or heart failure) at recruitment. Presence, or absence, of CV risk factors and diagnoses of CV disease were self-reported by participants at enrolment. Associations between anthropometric indices (body mass index (BMI), waist circumference (WC), waist to hip ratio (WHiR), and waist to height ratio (WHeR)) and the risk of all-cause mortality were analysed using Cox regression models.

Results

After excluding those with history of cancer at baseline (n=45,222), 453,046 participants were included (median age: 58 (interquartile range: 50 - 63) years; 53% women), of whom 150,732 had at least one CV risk factor, and 17,884 established CV disease.

During a median follow-up of 5 years, 6,319 participants died. Baseline BMI had a U-shaped relationship with ACM, with higher nadir-values for those with CV risk factors or CV disease, for both sexes (figure). WC, WHiR and WHeR (measures of central distribution of body fat) had more linear associations with ACM, regardless of CV risk factors, CV disease and sex.

Conclusions

For adults with or without CV risk factors or established CV disease, measures of central distribution of body fat are more strongly and more linearly associated with ACM than BMI. WC, or WHiR, rather than BMI, appear to be more appropriate variables for risk stratification.

P4514Agreement and prognostic significance of 6 frailty tools in patients with chronic heart failure

Background

Frailty is common in patients with chronic heart failure (CHF) and is associated with adverse outcome. Many frailty tools are available, however, there is no standard way of evaluating frailty in patients with CHF.

Purpose

To report the prevalence of frailty, agreement and prognostic significance amongst 3 frailty assessment tools and 3 screening tools in CHF patients.

Methods

We comprehensively studied frailty using 6 frailty tools. Frailty screening tools include: Clinical frailty scale (CFS); Derby frailty index & Acute frailty network frailty criteria. Frailty assessment tools include: Fried criteria; Edmonton frailty score & Deficit index. Since there is no gold standard in evaluating frailty in CHF patients, for each of the frailty tools, we used the results of the other 5 tools to produce a combined frailty index which we used as a “standard” frailty tool. Subjects were defined as frail if so identified by at least 3 out of 5 tools.

Results

467 consecutive ambulatory CHF patients (67% male, median age 76 (IQR: 69–82) years, median NTproBNP 1156 (IQR: 469–2463) ng/L) and 87 controls (79% male, median age 73 (IQR: 69–77 years) were studied.

Prevalence of frailty was much higher in CHF patients than in controls (30–52% vs 2–15%, respectively). Amongst the frailty screening tools, DFI scored the greatest proportion of patients as frail (48%) while CFS scored the lowest proportion as frail (44%). Amongst the assessment tools, Fried criteria scored the greatest proportion of patients as frail (52%) while EFS scored the lowest proportion as frail (30%). Frail patients were older, have worse symptoms, higher NTproBNP and more co-morbidities compared to non-frail patients.

Of the screening tools, CFS had the strongest agreement with assessment tools (kappa coefficient: 0.65–0.72, all p<0.001). CFS had the highest sensitivity (87%) and specificity (89%) amongst screening tools and the lowest misclassification rate (12%) amongst all 6 frailty tools in identifying frailty according to the combined frailty index.

During a median follow-up of 559 days (IQR 512–629 days), 82 (18%) patients died. 55% (N=45) of frail patients died of non-cardiovascular causes. Worsening frailty as detected by all 6 frailty tools was associated with worse outcome. A base model for mortality prediction including sex, NYHA class (III/IV vs I/II), BMI, log NTproBNP and haemoglobin had a C-statistics of 0.78. Amongst frailty tools: CFS and Fried criteria increased model performance most compared with base model (c-statistics: 0.80 for both). Patients who were frail according to CFS had a 9 times greater mortality risk than non-frail patients (Figure).

Conclusion

Frailty is common in CHF patients and is associated with worse outcome. CFS is a simple screening tool which identifies a similar group as lengthy assessment tools and has similar prognostic significance. Frailty screening should be incorporated into routine care of patients with CHF.

Acknowledgement/Funding

None

ECHO and magnetic resonance imaging in a patient with high bleeding risk and ventricular perforation: a case report and literature review

Myocardial perforation is a complication following pacemaker implantation that may cause cardiac tamponade. We present an original case of myocardial lead perforation not complicated by acute cardiac tamponade. The patient with an acute myocardial infarct had a high bleeding risk both in the acute phase of lead insertion (anticoagulant and triple platelet anti-aggregation therapy) and after few days, the percutaneous extraction lead for the double platelet antiaggregant therapy. Torrent-Guasp's theory is considered for explaining the clinical course of patient. Echocardiography and magnetic resonance imaging (MRI) evaluation showed a diffuse pericardial non-hemorrhagic fibrinous effusion and guide the clinical management.