Measuring congestion with a non-invasive monitoring device in heart failure and haemodialysis: CONGEST-HF
使用非侵入性监测设备测量心力衰竭和血液透析中的充血：CONGEST-HF
首次发表：2024 年 5 月 13 日 https://doi.org/10.1002/ejhf.3290 IF：16.9 Q1
Abstract 摘要
Aims 目标
We examined the effectiveness of a novel cardiopulmonary management wearable sensor (worn for less than 5 mins) at measuring congestion and correlated the device findings with established clinical measures of congestion.
我们检查了一种新型心肺管理可穿戴传感器（佩戴时间少于 5 分钟）在测量充血方面的有效性，并将该设备的发现与已建立的充血临床指标进行了相关性分析。
Methods and results 方法与结果
We enrolled three cohorts of patients: (1) patients with heart failure (HF) receiving intravenous diuretics in hospital; (2) patients established on haemodialysis, and (3) HF patients undergoing right heart catheterization (RHC). The primary outcomes in the respective cohorts were a Spearman correlation between (1) change in weight and change in thoracic impedance (TI) (from enrolment, 24 h after admission to discharge) in patients hospitalized for HF; (2) lung ultrasound B-lines and volume removed during dialysis with device measured TI, and (3) pulmonary capillary wedge pressure (PCWP) and sub-acoustic diastolic, third heart sound (S3) in the patients undergoing RHC. A total of 66 patients were enrolled. In HF patients (n = 25), change in weight was correlated with both change in device TI (Spearman correlation [rsp] = −0.64, p = 0.002) and change in device S3 (rsp = −0.53, p = 0.014). In the haemodialysis cohort (n = 21), B-lines and TI were strongly correlated before (rsp = −0.71, p < 0.001) and after (rsp = −0.77, p < 0.001) dialysis. Volume of fluid removed by dialysis was correlated with change in device TI (rsp = 0.49, p = 0.024). In the RHC cohort (n = 20), PCWP measured at one time point and device S3 were not significantly correlated (rsp = 0.230, p = 0.204). There were no device-related adverse events.
我们招募了三个患者队列：（1）在医院接受静脉利尿剂治疗的心力衰竭（HF）患者；（2）已建立血液透析的患者；（3）接受右心导管检查（RHC）的 HF 患者。各个队列的主要结果是：（1）在因 HF 住院的患者中，体重变化与胸部阻抗（TI）变化之间的 Spearman 相关性（从入组、入院后 24 小时到出院）；（2）肺部超声 B 线与透析期间去除的液体量与设备测量的 TI 之间的关系；（3）接受 RHC 的患者的肺毛细血管楔压（PCWP）与亚声学舒张期第三心音（S3）之间的关系。共招募了 66 名患者。在 HF 患者中（n = 25），体重变化与设备 TI 变化（Spearman 相关性[rsp] = −0.64，p = 0.002）和设备 S3 变化（rsp = −0.53，p = 0.014）相关。在血液透析队列中（n = 21），B 线与 TI 在透析前（rsp = −0.71，p < 0.001）和透析后（rsp = −0.77，p < 0.001）均呈强相关。透析去除的液体量与设备 TI 变化相关（rsp = 0.49，p = 0.024）。 在 RHC 队列中（n = 20），在一个时间点测量的 PCWP 与设备 S3 之间没有显著相关性（rsp = 0.230，p = 0.204）。没有与设备相关的不良事件。
Conclusions 结论
A non-invasive device was able to detect changes in congestion in patients with HF receiving decongestion therapy and patients having fluid removed at haemodialysis. The cardiopulmonary management device, which measures multiple parameters, is a potentially useful tool to monitor patients with HF to prevent hospitalizations.
一种非侵入性设备能够检测接受去充血治疗的心力衰竭患者和在血液透析中排除液体的患者的充血变化。该心肺管理设备测量多个参数，是监测心力衰竭患者以防止住院的潜在有用工具。
Graphical Abstract 图形摘要
CONGEST-HF: correlation of the cardiopulmonary monitoring (CPM) wearable device with measures of congestion. IV, intravenous; PCWP, pulmonary capillary wedge pressure.
CONGEST-HF：心肺监测（CPM）可穿戴设备与充血指标的相关性。IV，静脉注射；PCWP，肺毛细血管楔压。
Introduction 介绍
Patients with heart failure (HF) are admitted to hospital many times during their disease1, 2 for treatment with intravenous (IV) diuretics for the relief of congestion. Although there are protocols to provide IV diuretic therapy in the outpatient setting,3 preventing hospitalizations through the early detection of deterioration and identification of subclinical congestion is preferable, allowing alteration of therapy and preventing hospitalization.
心力衰竭（HF）患者在疾病期间多次入院 1, 2，接受静脉（IV）利尿剂治疗以缓解充血。尽管有在门诊环境中提供 IV 利尿剂治疗的方案，3 但通过早期发现病情恶化和识别亚临床充血来预防住院更为理想，这样可以调整治疗并防止住院。
Methods to detect congestion rely on expensive invasive monitoring through functions in specially designed pacemakers4, 5 or monitoring of pulmonary artery pressure through a sensor implanted directly in the lung. Pulmonary artery pressure monitoring has been shown to reduce hospitalizations6 but the therapy remains expensive. Simpler measures such as patient-recorded daily weight are recommended by guidelines7 but are insensitive.8 Wearable devices represent a potential new approach to monitor patients with HF.
检测充血的方法依赖于通过专门设计的起搏器进行昂贵的侵入性监测 4, 5，或通过直接植入肺部的传感器监测肺动脉压力。研究表明，肺动脉压力监测可以减少住院 6，但该疗法仍然昂贵。指南推荐的更简单的措施，如患者记录的每日体重 7，但其敏感性较差 8。可穿戴设备代表了一种潜在的新方法来监测心力衰竭患者。
We tested a non-invasive cardiopulmonary monitoring (CPM) device designed to monitor patients with HF. The wearable device is applied to the skin of the patient's chest and connects via Bluetooth to a mobile App to initiate a reading and to record physiological data. The device is worn for less than 5 min twice a day removing the need for continuous wear or invasive procedures. The CPM device uses several sensors to detect changes in thoracic impedance, a sub-acoustic diastolic, third heart sound (S3), as well as respiratory rate and tidal volume. Changes in these parameters might identify developing congestion allowing action to be taken before a patient presents with decompensation and avoiding a hospitalization. We aimed to determine if the device could detect changes in fluid status in patients with decompensated HF receiving decongestion therapy in hospital, in patients before and after a session of haemodialysis and the correlation of the measures obtained from the device with invasive measures derived from a right heart catheterization (RHC).
我们测试了一种非侵入性的心肺监测（CPM）设备，旨在监测心力衰竭（HF）患者。该可穿戴设备贴在患者胸部的皮肤上，通过蓝牙连接到移动应用程序以启动读数并记录生理数据。该设备每天佩戴不到 5 分钟，分为两次，消除了持续佩戴或侵入性程序的需要。CPM 设备使用多个传感器检测胸部阻抗的变化、亚声学舒张期第三心音（S3）、呼吸频率和潮气量。这些参数的变化可能识别出正在发展的充血，从而在患者出现失代偿之前采取措施，避免住院。我们的目标是确定该设备是否能够检测接受去充血治疗的失代偿 HF 患者的液体状态变化、透析前后患者的变化，以及从设备获得的测量值与右心导管检查（RHC）获得的侵入性测量值之间的相关性。
Methods 方法
Study design and population
研究设计和人群
The COrrelation of the Non-invasive CardioPulmonary Management wearable device with measures of conGESTion in Heart Failure (CONGEST-HF, NCT05026034) study was a prospective, observational study designed to examine the effectiveness of the CPM wearable device (Analog Devices Inc. [ADI], Wilmington, MA, USA) at detecting congestion and changes in congestion in three cohorts of patients. The first cohort were patients admitted to the Queen Elizabeth University Hospital (QEUH), Glasgow, UK with a diagnosis of HF and receiving IV diuretics to relieve congestion. Patients were identified early into their admission by their responsible medical teams as being volume overloaded with a primary diagnosis of decompensated HF and were then approached by the investigators. Left ventricular ejection fraction at baseline was not used to determine study eligibility. They were assessed at four time points; on the day of enrolment, the following day while still on IV diuretics, on the day of first dose of oral diuretics, and the day of discharge. The decisions to switch from IV to oral diuretics or discharge patients were made by the usual care teams independently of any information collected for the study. The second cohort were patients who had chronic end-stage renal failure and were established on haemodialysis at the QEUH for at least 90 days with a target volume removal of at least 1.5 L in a dialysis session. Fluid was removed via an indwelling catheter or arteriovenous fistula. They underwent study assessment immediately before and after a single dialysis session. The final cohort were patients undergoing a clinically indicated RHC as part of their cardiac transplant evaluation at the Scottish National Advanced Heart Failure Unit. This cohort underwent study assessment at one time point, on the day of the RHC. Full inclusion and exclusion criteria are detailed in online supplementary Table S1 and a flow chart detailing the screening and recruitment process is given in online supplementary Figure S1.
非侵入性心肺管理可穿戴设备与心力衰竭中充血指标的相关性研究（CONGEST-HF，NCT05026034）是一项前瞻性观察研究，旨在评估 CPM 可穿戴设备（Analog Devices Inc. [ADI]，美国马萨诸塞州威尔明顿）在检测充血及充血变化方面的有效性，研究对象为三组患者。第一组为因心力衰竭（HF）入院于英国格拉斯哥的伊丽莎白女王大学医院（QEUH）并接受静脉利尿剂以缓解充血的患者。患者在入院早期由负责的医疗团队识别为体液过载，主要诊断为失代偿性心力衰竭，随后由研究者接触。基线时左心室射血分数未用于确定研究资格。患者在四个时间点进行评估；入组当天、次日仍在接受静脉利尿剂时、首次口服利尿剂的当天以及出院当天。 从静脉注射利尿剂转为口服利尿剂或出院的决定由常规护理团队独立于为研究收集的任何信息作出。第二组是慢性终末期肾衰竭患者，他们在 QEUH 接受血液透析至少 90 天，透析会话的目标液体去除量至少为 1.5 升。液体通过留置导管或动静脉瘘去除。他们在单次透析会话前后立即接受研究评估。最后一组是作为心脏移植评估的一部分，接受临床指征下的右心导管检查（RHC）的患者。该组在一个时间点进行研究评估，即 RHC 当天。完整的纳入和排除标准详见在线补充表 S1，筛选和招募过程的流程图见在线补充图 S1。
Investigational medical device
研究性医疗器械
The CPM device is a wearable, battery-operated, class 2a medical device. The CPM device is applied to the skin of the patient's chest, with adhesive islands, one to the left of the sternum, one at the apex of the heart, and one in the mid-axillary line (Figure 1). Each participant was uniquely matched to a single device for the duration of their study participation. The device communicated via Bluetooth with a mobile App which guided the investigator through a device reading. A single reading consisted of two measurements, performed whilst sitting up, then the most supine position tolerated. Each measurement lasted approximately 60 s. The CPM device measured raw data signals that were transferred by the site investigator at the time of the assessment to a secure portal managed by the Glasgow Clinical Trials Unit at the Robertson Centre for Biostatistics (RCB). ADI then downloaded the raw data and processed them into derived measurements which were returned securely to the RCB's portal. The investigators (and clinical team) were blinded to the derived device data and ADI were blinded to the investigator-collected clinical study data. The derived CPM device measurements included thoracic impedance, S3 energy and tidal volume.
CPM 设备是一种可穿戴的、由电池驱动的 2a 类医疗设备。CPM 设备贴在患者胸部的皮肤上，使用粘合岛，一个在胸骨左侧，一个在心脏顶端，一个在腋中线（图 1）。每位参与者在研究期间与单一设备唯一匹配。该设备通过蓝牙与移动应用程序通信，指导研究者进行设备读数。一次读数包括两个测量，分别在坐起和能够忍受的最平卧位置进行。每个测量持续约 60 秒。CPM 设备测量的原始数据信号由现场研究者在评估时传输到由格拉斯哥临床试验单位在罗伯逊生物统计中心（RCB）管理的安全门户。ADI 随后下载原始数据并将其处理为派生测量，这些测量安全地返回到 RCB 的门户。研究者（和临床团队）对派生设备数据保持盲态，而 ADI 对研究者收集的临床研究数据保持盲态。 衍生的 CPM 设备测量包括胸部阻抗、S3 能量和潮气量。
设备的图片和示意图，显示传感器的位置和相关解剖标志。示意图显示设备在女性和男性患者左胸的位置，以及设备放置的侧视图。
Clinical study procedures
临床研究程序
For all cohorts, at each study assessment, before device readings, an 8-zone lung ultrasound (LUS), tidal volume by bedside spirometry (Pneumotrac, Vitalograph, UK), echocardiogram (Vivid E90 for the RHC cohort and a Canon Aplio i900 for the patients hospitalized for HF or receiving dialysis; performed by an investigator [JPC] with European Association for Cardiovascular Imaging accreditation), physical symptoms and signs (including the EVEREST clinical congestion score [ECCS] comprising an assessment of dyspnoea, orthopnoea, fatigue, jugular venous distension, oedema, and pulmonary rales) and observations were recorded. Whole blood was collected at each visit for analysis of N-terminal pro-B-type natriuretic peptide (NT-proBNP) (e411, Roche Diagnostics, UK) and mid-regional pro-adrenomedullin (BRAHMS Kryptor, Thermofisher ThermoFisher Scientific, UK). All other blood samples, including haematocrit, were collected and analysed as standard care in the local hospital laboratory.
对于所有队列，在每次研究评估时，在设备读数之前，进行了 8 区肺超声（LUS）、床边肺活量计测量潮气量（Pneumotrac，Vitalograph，英国）、心脏超声（RHC 队列使用 Vivid E90，住院心衰或接受透析的患者使用 Canon Aplio i900；由具有欧洲心血管影像学协会认证的研究者[JPC]进行）、身体症状和体征（包括 EVEREST 临床充血评分[ECCS]，该评分包括对呼吸困难、平卧呼吸困难、疲劳、颈静脉扩张、水肿和肺啰音的评估）和观察记录。每次就诊时收集全血以分析 N 末端前 B 型利钠肽（NT-proBNP）（e411，罗氏诊断，英国）和中段前肾上腺髓质素（BRAHMS Kryptor，Thermofisher ThermoFisher Scientific，英国）。所有其他血样，包括血细胞比容，均按照当地医院实验室的标准护理进行收集和分析。
Study outcomes 研究结果
The primary outcomes by cohort were: in hospitalized patients with HF, the correlation between CPM device thoracic impedance and CPM device S3 and LUS B-lines and change in weight; in the haemodialysis cohort, the correlation between CPM device thoracic impedance and both LUS B-lines and volume of fluid removed by dialysis; in the RHC cohort, the correlation between CPM device S3 and pulmonary capillary wedge pressure (PCWP) on RHC.
各队列的主要结果为：在住院心力衰竭患者中，CPM 设备胸部阻抗与 CPM 设备 S3 和 LUS B 线及体重变化之间的相关性；在血液透析队列中，CPM 设备胸部阻抗与 LUS B 线及透析去除的液体体积之间的相关性；在 RHC 队列中，CPM 设备 S3 与 RHC 上的肺毛细血管楔压（PCWP）之间的相关性。
Secondary outcomes included the correlation of CPM device measured tidal volume with tidal volume measured by spirometry; the correlation of device measures with echocardiography, physical signs and symptoms of congestion and, in the RHC cohort only, RHC parameters additional to PCWP.
次要结果包括 CPM 设备测量的潮气量与肺活量测定法测量的潮气量之间的相关性；设备测量与超声心动图、充血的体征和症状之间的相关性，以及仅在 RHC 队列中，RHC 参数与 PCWP 的额外相关性。
Statistical analysis 统计分析
Analyses were performed by a statistician at the RCB (AMcI) and verified by a statistician at the BHF Glasgow Cardiovascular Research Centre (AT) after the database was locked according to a pre-specified statistical analysis plan. The RHC cohort required assessment at a single time point and we estimated a sample size of 20 patients would be recruited, giving 80% power to detect a correlation of 0.6 at 5% statistical significance. The hospitalized HF cohort and dialysis cohort required multiple assessments per patient, so we aimed to recruit 40 patients into each cohort, giving protection against missing outcome data, while ensuring at least the same level of statistical power as for the RHC cohort. The study was delayed and interrupted by the COVID-19 pandemic and the total number of patients recruited into the inpatient and dialysis cohort was revised to 20 to match the RHC which preserved power. Continuous variables are summarized using the mean and standard deviation (SD) or median with interquartile ranges (IQR). Categorical variables are summarized with frequencies and percentages. Patient-individual Spearman correlations (rsp) were used to determine the correlations between device measurements and clinical parameters that were continuous variables. Scatter plots graphically report the correlations and heat maps are used to depict multiple correlations where darker colour indicates correlations closer to +1 and lighter colour correlation closer to −1. Relationships between continuous and categorical variables were analysed using Mann–Whitney–Wilcoxon tests or Kruskal–Wallis tests as appropriate. Wilcoxon signed rank and McNemar's tests were used to analyse the relationship between change in continuous or categorical variables, respectively, on serial assessments. A p-value <0.05 was considered statistically significant. Analyses were performed using R (R Foundation for Statistical Computing, Vienna, Austria) and Stata version 18 (College Station, TX, USA).
分析由 RCB（AMcI）的统计学家进行，并在数据库根据预先指定的统计分析计划锁定后，由 BHF 格拉斯哥心血管研究中心（AT）的统计学家进行验证。RHC 队列需要在单一时间点进行评估，我们估计将招募 20 名患者，提供 80%的能力以在 5%的统计显著性下检测到 0.6 的相关性。住院 HF 队列和透析队列需要对每位患者进行多次评估，因此我们计划在每个队列中招募 40 名患者，以防止缺失结果数据，同时确保至少与 RHC 队列相同的统计能力。研究因 COVID-19 大流行而延迟和中断，住院和透析队列招募的患者总数修订为 20，以匹配 RHC，从而保持能力。连续变量使用均值和标准差（SD）或中位数与四分位数范围（IQR）进行总结。分类变量以频率和百分比进行总结。 患者个体的斯皮尔曼相关系数（rsp）用于确定设备测量与连续变量的临床参数之间的相关性。散点图以图形方式报告相关性，热图用于描绘多个相关性，其中较深的颜色表示相关性接近+1，较浅的颜色表示相关性接近−1。连续变量和分类变量之间的关系使用 Mann–Whitney–Wilcoxon 检验或 Kruskal–Wallis 检验进行分析，具体取决于情况。Wilcoxon 符号秩检验和 McNemar 检验分别用于分析连续或分类变量在连续评估中的变化关系。p 值<0.05 被认为具有统计学意义。分析使用 R（维也纳，奥地利，R 统计计算基金会）和 Stata 版本 18（美国德克萨斯州大学城）进行。
Safety monitoring 安全监测
Patients were reviewed within 48 h of a device reading to assess for device-related adverse effects. Patients with pre-existing cardiac implantable electronic devices (CIED) had real-time monitoring of the CIED while wearable device readings were performed for surveillance of potential device–device interactions.
患者在设备读数后的 48 小时内进行了评估，以检查与设备相关的不良反应。具有既往心脏植入电子设备（CIED）的患者在进行可穿戴设备读数以监测潜在的设备间相互作用时，CIED 进行了实时监测。
Ethics and consent 伦理与同意
Ethical approval was granted by the London-Dulwich Research Ethics Committee. All patients provided written informed consent prior to enrolment. The study was registered (ClinicalTrials.gov identifier NCT05026034).
伦理批准已由伦敦-达尔维奇研究伦理委员会授予。所有患者在入组前提供了书面知情同意。该研究已注册（ClinicalTrials.gov 标识符 NCT05026034）。
Study funding 研究资金
CONGEST-HF was sponsored by NHS Greater Glasgow and Clyde Research and Innovations. ADI provided funding to the University of Glasgow for the design and conduct of the study but played no role in the collection of clinical data, the statistical analyses or the decision to publish the results.
CONGEST-HF 由 NHS 大格拉斯哥和克莱德研究与创新赞助。ADI 为格拉斯哥大学提供了研究设计和实施的资金，但在临床数据的收集、统计分析或决定发布结果方面没有发挥任何作用。
Results 结果
Performance of the device in patients hospitalized with heart failure
心力衰竭住院患者设备的性能
A total of 25 patients with a primary diagnosis of HF on admission to hospital were recruited. The median time to switch from IV to oral diuretics was 4 days (IQR 3–5) and the median time from enrolment to discharge was 7 days (IQR 4–14). Patients enrolled were representative of hospitalized patients with HF with a mean age of 72 (SD 12) years, mean ejection fraction of 31% (SD 15%), 72% in New York Heart Association (NYHA) class III or IV and NT-proBNP was elevated at 3899 pg/ml (IQR 1567-9151) (Table 1). The mean change in weight from day 1 on IV diuretics to the day of discharge was –3.7 ± 5.5 kg (p < 0.001) and the number of B-lines on LUS fell from a median of 80 (IQR 49–124) on day 1 of IV diuretics to 44 (IQR 24–59) on day of discharge (p-value for difference with visit 1 < 0.001).
共招募了 25 名入院时主要诊断为心力衰竭（HF）的患者。由静脉（IV）利尿剂转为口服利尿剂的中位时间为 4 天（四分位数范围 3–5），从入组到出院的中位时间为 7 天（四分位数范围 4–14）。入组患者代表了住院心力衰竭患者，平均年龄为 72 岁（标准差 12），平均射血分数为 31%（标准差 15%），72%的患者属于纽约心脏协会（NYHA）III 或 IV 级，NT-proBNP 水平升高至 3899 pg/ml（四分位数范围 1567-9151）（表 1）。从 IV 利尿剂治疗的第 1 天到出院当天，体重的平均变化为–3.7 ± 5.5 kg（p < 0.001），LUS 上 B 线的数量从 IV 利尿剂治疗第 1 天的中位数 80（四分位数范围 49–124）降至出院当天的 44（四分位数范围 24–59）（与第 1 次访视的差异 p 值 < 0.001）。
表 1. 每个队列中入组患者的基线特征和治疗情况
| Hospitalized patients with HF (n = 25) 住院心力衰竭患者（n = 25） |
Dialysis cohort (n = 21) 透析队列（n = 21） |
RHC cohort (n = 20) RHC 队列 (n = 20) |
|
|---|---|---|---|
| Male sex, n (%) 男性性别，n (%) | 18 (72) | 15 (71) | 14 (70) |
| Age (years) 年龄（岁） | 72 ± 12 | 60 ± 14 | 55 ± 9 |
| Race, n (%) 种族，n (%) | |||
| White 白色 | 23 (92) | 19 (91) | 20 (100) |
| Asian 亚洲 | 2 (8) | 2 (9) | - |
| BMI (kg/m2) BMI（千克/平方米） | 29 ± 5 | 25 ± 5 | 27 ± 4 |
| Medical history, n (%) 病史，n (%) | |||
| Hypertension 高血压 | 12 (48) | 15 (71) | 1 (5) |
| Diabetes 糖尿病 | 7 (28) | 6 (29) | 3 (15) |
| AF on baseline ECG 基线心电图上的房颤 |
7 (30) | 5 (24) | 4 (20) |
| History of any AF 任何房颤的历史 |
12 (48) | 6 (29) | 9 (45) |
| MI | 7 (28) | 4 (19) | 4 (20) |
| Stroke 中风 | 4 (16.0) | 2 (10) | 1 (5) |
| COPD | 3 (12) | 1 (5) | 1 (5) |
| LVEF (%) 左室射血分数（%） | 31 ± 15 | 48 ± 17 | 30 ± 11 |
| NYHA class, n (%) | |||
| I | 0 (0) | 4 (19) | 0 (0) |
| II | 7 (28) | 5 (24) | 7 (35) |
| III | 10 (40) | 9 (43) | 12 (60) |
| IV | 8 (32) | 3 (14) | 1 (5) |
| Symptoms/signs, n (%) | |||
| Dyspnoea | 9 (36) | 16 (76) | 19 (95) |
| Orthopnoea | 13 (52) | 8 (38) | 9 (45) |
| PND | 4 (16) | 5 (24) | 3 (15) |
| Fatigue | 21 (84) | 19 (91) | 20 (100) |
| Bendopnoea | 11 (44) | 9 (43) | 13 (65) |
| Peripheral oedema | 22 (88) | 6 (29) | 4 (20) |
| Systolic BP (mmHg) | 121 ± 25 | 138 ± 34 | 103 ± 13 |
|
Heart rate (bpm) |
79 ± 17 | 80 ± 15 | 75 ± 19 |
|
Respiratory rate (/min) |
17 ± 2 | 16 ± 1 | 16 ± 2 |
|
Chest circumference (cm) |
108 ± 13 | 102 ± 13 | 108 ± 10 |
| Duration of HF, n (%) | |||
| <1 year | 9 (36) | 3 (14) | 3 (15) |
| 1–5 years | 13 (52) | 3 (14) | 7 (35) |
| >5 years | 3 (12) | 1 (5) | 10 (50) |
| Ischaemic aetiology, n (%) | 11 (44) | 3 (14) | 6 (30) |
| Prior HF hospitalization, n (%) | 13 (52) | 1 (5) | 11 (55) |
| HF hospitalization within previous 6 months, n (%) | 6 (24) | 1 (5) | 3 (15) |
| NT-proBNP (pg/ml) | 3899 (1567–9151) | 43 624 (4955–88 720) | 2116 (784–4341) |
| eGFR (ml/min/1.73 m2) | 49 (33–69) | 7 (5–10) | 59 (45–77) |
| Baseline treatment, n (%) | |||
| Intravenous diuretics | 25 (100) | 5 (24) | 1 (5) |
| Loop diuretic | 0 (0) | 0 (0) | 18 (90) |
| Thiazide/thiazide-like diuretic | 8 (32) | 5 (24) | 1 (5) |
| ACE inhibitor or ARB | 6 (24) | 1 (5) | 4 (20) |
| Sacubitril/valsartan | 17 (68) | 16 (76) | 12 (60) |
| Beta-blocker | 9 (36) | 0 (0) | 17 (85) |
| MRA | 11 (44) | 0 (0) | 15 (75) |
| SGLT2 inhibitor | 1 (4) | 1 (5) | 13 (65) |
| Digoxin | 4 (16) | 1 (5) | 2 (10) |
| Any implanted device | 2 (8) | 0 (0) | 12 (60) |
| ICD | 0 (0) | 0 (0) | 11 (55) |
| CRT-D | 0 (0) | 0 (0) | 1 (5) |
| CRT-P | 2 (8) | 1 (5) | 0 (0) |
| Pacemaker | 0 (0) | 0 (0) | 0 (0) |
-
ACE, angiotensin-converting enzyme; AF, atrial fibrillation; ARB, angiotensin receptor blocker; BMI, body mass index; BP, blood pressure; COPD, chronic obstructive pulmonary disease; CRT-D, cardiac resynchronization therapy-defibrillator; CRT-P, cardiac resynchronization therapy-pacemaker; ECG, electrocardiogram; eGFR, estimated glomerular filtration rate; HF, heart failure; ICD, implantable cardioverter-defibrillator; LVEF, left ventricular ejection fraction; MI, myocardial infarction; MRA, mineralocorticoid receptor antagonist; NT-proBNP, N-terminal pro-B type natriuretic peptide; NYHA, New York Heart Association; PND, paroxysmal nocturnal dyspnoea; RHC, right heart catheterization; SGLT2, sodium–glucose cotransporter 2.
ACE，血管紧张素转化酶；AF，心房颤动；ARB，血管紧张素受体拮抗剂；BMI，身体质量指数；BP，血压；COPD，慢性阻塞性肺病；CRT-D，心脏再同步治疗-除颤器；CRT-P，心脏再同步治疗-起搏器；ECG，心电图；eGFR，估计肾小球滤过率；HF，心力衰竭；ICD，植入式心脏复律除颤器；LVEF，左心室射血分数；MI，心肌梗死；MRA，矿物皮质激素受体拮抗剂；NT-proBNP，N 端前 B 型利钠肽；NYHA，纽约心脏协会；PND，阵发性夜间呼吸困难；RHC，右心导管检查；SGLT2，钠-葡萄糖共转运蛋白 2。
Change in CPM measured thoracic impedance and change in weight between day 1 on IV diuretics and the day of discharge were significantly correlated (rsp = −0.64, p = 0.002) (Figure 2). Correlation of the change in device S3 and change in weight from visit 1 to day of discharge was statistically significant (rsp = −0.53, p = 0.014) (Figure 2 and online supplementary Figure S2). The correlations between thoracic impedance and S3 with LUS B-lines at each visit or between visits did not reach statistical significance (online supplementary Tables S2, S3 and Figures S3–S6).
在 IV 利尿剂治疗的第 1 天与出院当天之间，CPM 测量的胸部阻抗变化与体重变化显著相关（rsp = −0.64，p = 0.002）（图 2）。设备 S3 的变化与从第一次就诊到出院当天的体重变化之间的相关性具有统计学意义（rsp = −0.53，p = 0.014）（图 2 和在线补充图 S2）。每次就诊或就诊之间胸部阻抗与 S3 与 LUS B 线之间的相关性未达到统计学意义（在线补充表 S2、S3 和图 S3–S6）。
设备测量的胸部阻抗变化与肺部超声（LUS）中 B 线数量的相关性（A），以及体重变化（B）和设备测量的第三心音（S3）变化与体重变化的相关性（C）。CPM，心肺监测；Rsp，斯皮尔曼相关性。
The CPM device tidal volume correlated with spirometry tidal volume (online supplementary Table S4) most clearly on the first visit (online supplementary Figures S3–S6). Median values (IQR) are reported for the measured echocardiographic parameters per study visit in online supplementary Table S5 and correlations (online supplementary Figures S3–S6 and Tables S6, S7).
CPM 设备的潮气量与肺活量测定的潮气量相关性在第一次访问时最为明显（在线补充表 S4，在线补充图 S3–S6）。每次研究访问中测量的超声心动图参数的中位值（IQR）在在线补充表 S5 中报告，相关性（在线补充图 S3–S6 和表 S6、S7）。
The proportion of patients with symptoms and the ECCS are shown in online supplementary Table S8 at each visit. Change in total ECCS was correlated with change in device thoracic impedance between day of first IV diuretics and day of discharge (rsp = −0.44, p = 0.045) (online supplementary Figure S7) but not tidal volume or S3 (online supplementary Figure S2).
患者有症状的比例和 ECCS 在每次就诊时显示在在线补充表 S8 中。总 ECCS 的变化与首次使用静脉利尿剂当天和出院当天的设备胸部阻抗变化相关（rsp = −0.44，p = 0.045）（在线补充图 S7），但与潮气量或 S3 无关（在线补充图 S2）。
Performance of the device in patients undergoing haemodialysis
透析患者设备的性能
A total of 21 patients with end-stage renal disease who were established on intermittent haemodialysis for a minimum of 90 days were enrolled. This cohort was predominantly male with a mean age 60 ± 14 years (Table 1). Of the 21, 7 (33%) patients had a history of HF. The majority of people had at least one cardiovascular risk factor, most often hypertension (71.4%) followed by diabetes (28.6%) and a prior myocardial infarction (19.1%). The median volume of fluid removed by haemodialysis was 1999 ml (IQR 1600–2000). The median number of B-lines before dialysis was 58 (33–94) and after dialysis was 31 (15–67) (p-value for difference < 0.001).
共招募了 21 名处于终末期肾病的患者，这些患者已接受间歇性血液透析至少 90 天。该队列以男性为主，平均年龄为 60 ± 14 岁（表 1）。在 21 名患者中，有 7 名（33%）有心力衰竭病史。大多数人至少有一个心血管风险因素，最常见的是高血压（71.4%），其次是糖尿病（28.6%）和既往心肌梗死（19.1%）。通过血液透析去除的液体中位量为 1999 毫升（四分位数范围 1600–2000）。透析前 B 线的中位数为 58（33–94），透析后为 31（15–67）（差异的 p 值< 0.001）。
The CPM device thoracic impedance and B-lines on LUS were strongly correlated before (rsp = −0.71, p < 0.001) and after (rsp = −0.77, p < 0.001) dialysis (online supplementary Figures S8 and S9). The correlation between the change in device thoracic impedance and change in B-lines was weaker and did not reach statistical significance (rsp = −0.19, p = 0.396) (Figure 3 and online supplementary Figure S10). The correlation between the volume of fluid removed by dialysis and the change in device thoracic impedance was rsp = 0.49, p = 0.024 (Figure 3).
CPM 设备的胸部阻抗与 LUS 上的 B 线在透析前（rsp = −0.71，p < 0.001）和透析后（rsp = −0.77，p < 0.001）有很强的相关性（在线补充图 S8 和 S9）。设备胸部阻抗变化与 B 线变化之间的相关性较弱，未达到统计显著性（rsp = −0.19，p = 0.396）（图 3 和在线补充图 S10）。透析去除的液体体积与设备胸部阻抗变化之间的相关性为 rsp = 0.49，p = 0.024（图 3）。
设备测量的胸部阻抗变化与透析前（A）和透析后（B）肺部超声（LUS）中 B 线数量之间的相关性，以及胸部阻抗变化与 LUS 中 B 线变化（C）和胸部阻抗变化与透析期间去除的液体体积变化（D）。CPM，心肺监测；Rsp，斯皮尔曼相关性。
The CPM device S3 was significantly correlated with LUS B-lines before dialysis (rsp = 0.48, p = 0.028) (online supplementary Figure S8). The strength of the correlation was weaker after dialysis (rsp = 0.40, p = 0.075) (online supplementary Figure S9) but the correlation between the change in both parameters was evident but not statistically significant (rsp = 0.43, p = 0.055) (online supplementary Figure S10). Change in CPM S3 (rsp = −0.24, p = 0.283) and change in CPM tidal volume (rsp = 0.06, p = 0.810) did not correlate significantly with fluid removed by dialysis (online supplementary Figure S10).
CPM 设备 S3 在透析前与 LUS B 线显著相关（rsp = 0.48，p = 0.028）（在线补充图 S8）。透析后相关性强度减弱（rsp = 0.40，p = 0.075）（在线补充图 S9），但两个参数变化之间的相关性明显但不具有统计学意义（rsp = 0.43，p = 0.055）（在线补充图 S10）。CPM S3 的变化（rsp = −0.24，p = 0.283）和 CPM 潮气量的变化（rsp = 0.06，p = 0.810）与透析去除的液体没有显著相关性（在线补充图 S10）。
There was a correlation between spirometry tidal volume and volume of fluid removed by dialysis that did not reach statistical significance (rsp = 0.43, p = 0.052) with similar correlations in tidal volume by spirometry and CPM device tidal volume before (rsp = 0.41, p = 0.063) or after dialysis (rsp = 0.40, p = 0.073). Change in device tidal volume did not correlate significantly with change in spirometry tidal volume (rsp = 0.28, p = 0.219). Echocardiography was performed and median values (IQR) are reported for the measured echocardiographic parameters per study visit in online supplementary Table S9 and correlations between CPM device thoracic impedance, S3 and tidal volume and echocardiography measures before and after dialysis (online supplementary Tables S10, S11 and Figures S8, S9) were inconsistent.
肺活量测定的潮气量与透析去除的液体体积之间存在相关性，但未达到统计学显著性（rsp = 0.43，p = 0.052），在透析前（rsp = 0.41，p = 0.063）或透析后（rsp = 0.40，p = 0.073），肺活量测定的潮气量与 CPM 设备的潮气量之间也存在类似的相关性。设备潮气量的变化与肺活量测定的潮气量变化之间没有显著相关性（rsp = 0.28，p = 0.219）。进行了超声心动图检查，并在在线补充表 S9 中报告了每次研究访问测量的超声心动图参数的中位值（IQR），以及透析前后 CPM 设备胸部阻抗、S3、潮气量与超声心动图测量之间的相关性（在线补充表 S10、S11 和图 S8、S9）不一致。
Change in total ECCS was correlated significantly with CPM device thoracic impedance (rsp = −0.44, p = 0.046) (online supplementary Figure S11).
总 ECCS 的变化与 CPM 设备的胸部阻抗显著相关（rsp = −0.44，p = 0.046）（在线补充图 S11）。
Performance of the device in patients undergoing right heart catheterization
在接受右心导管插入的患者中设备的性能
The characteristics of the patients enrolled are shown in Table 1 and were representative of a population with advanced HF. Device-measured S3 and PCWP obtained at a single time-point were not significantly correlated (rsp = 0.30, p = 0.204) (Table 2 and online supplementary Figure S12). On analysis of additional RHC measures, pulmonary vascular resistance on RHC was significantly correlated with device thoracic impedance (rsp = −0.47, p = 0.036). Cardiac index (calculated using the Fick equation) was inversely correlated with tidal volume measured by the CPM device (rsp = −0.47, p = 0.043) (Table 2).
入组患者的特征如表 1 所示，代表了晚期心力衰竭人群。设备测量的 S3 和 PCWP 在单一时间点获得的结果没有显著相关性（rsp = 0.30，p = 0.204）（表 2 和在线补充图 S12）。在对额外的 RHC 测量进行分析时，RHC 上的肺血管阻力与设备胸部阻抗显著相关（rsp = −0.47，p = 0.036）。心脏指数（使用 Fick 方程计算）与 CPM 设备测量的潮气量呈负相关（rsp = −0.47，p = 0.043）（表 2）。
表 2. 设备衍生的第三心音、胸部阻抗与潮气量及右心导管检查的测量之间的相关性
| RHC parameter RHC 参数 | CPM device S3 CPM 设备 S3 | CPM device thoracic impedance CPM 设备胸部阻抗 |
CPM device tidal volume CPM 设备潮气量 |
|||
|---|---|---|---|---|---|---|
| rsp | p-value p 值 | rsp | p-value p 值 | rsp | p-value p 值 | |
| Cardiac output (thermodilution) 心输出量（热稀释法） |
−0.23 | 0.338 | 0.29 | 0.211 | −0.30 | 0.204 |
| Cardiac index (thermodilution) 心脏指数（热稀释法） |
−0.27 | 0.250 | 0.25 | 0.285 | −0.24 | 0.309 |
| Cardiac output (Fick) 心输出量（Fick） | −0.29 | 0.227 | 0.32 | 0.185 | −0.42 | 0.068 |
| Cardiac index (Fick) 心脏指数（Fick） | −0.40 | 0.093 | 0.36 | 0.135 | −0.47 | 0.043 |
| PA pressure – mean PA 压力 - 平均 | 0.29 | 0.208 | −0.21 | 0.364 | 0.30 | 0.195 |
| PA pressure – systolic PA 压力 - 收缩压 | 0.25 | 0.288 | −0.07 | 0.755 | 0.22 | 0.355 |
| PA pressure – diastolic PA 压力 - 舒张压 | 0.35 | 0.130 | −0.22 | 0.343 | 0.36 | 0.117 |
| PA saturation PA 饱和度 | −0.29 | 0.225 | 0.33 | 0.167 | −0.33 | 0.164 |
| RA pressure RA 压力 | 0.04 | 0.863 | 0.16 | 0.499 | −0.12 | 0.610 |
| Pulmonary vascular resistance 肺血管阻力 |
0.25 | 0.279 | −0.47 | 0.036 | 0.40 | 0.083 |
| Systemic vascular resistance 全身血管阻力 |
0.31 | 0.176 | −0.25 | 0.289 | 0.29 | 0.214 |
|
PCWP |
0.30 | 0.204 | −0.08 | 0.729 | 0.27 | 0.248 |
-
CPM, cardiopulmonary monitoring; PA, pulmonary artery; PCWP, pulmonary capillary wedge pressure; RA, right atrial; RHC, right heart catheterization; rsp, Spearman correlation; S3, third heart sound.
CPM，心肺监测；PA，肺动脉；PCWP，肺毛细血管楔压；RA，右心房；RHC，右心导管检查；rsp，斯皮尔曼相关；S3，第三心音。
The CPM device tidal volume and spirometry tidal volume were significantly correlated (rsp = 0.45, p = 0.049). B-lines on LUS were significantly correlated with CPM S3 (rsp = 0.45, p = 0.046), CPM thoracic impedance (rsp = −0.57, p = 0.010) and CPM tidal volume (rsp = 0.47, p = 0.036). There were several correlations between measures on echocardiography and CPM device measures (online supplementary Table S12, Figure S12).
CPM 设备潮气量与肺活量测定潮气量显著相关（rsp = 0.45，p = 0.049）。LUS 上的 B 线与 CPM S3 显著相关（rsp = 0.45，p = 0.046），CPM 胸部阻抗（rsp = -0.57，p = 0.010）和 CPM 潮气量（rsp = 0.47，p = 0.036）。超声心动图测量与 CPM 设备测量之间存在多个相关性（在线补充表 S12，图 S12）。
There was a trend of increasing values in CPM device S3, thoracic impedance and tidal volume across higher grades of the EVEREST dyspnoea scale that did not reach statistical significance (online supplementary Table S13). A similar general trend was observed across NYHA functional classes (online supplementary Table S13).
在 EVEREST 呼吸困难量表的较高等级中，CPM 设备 S3、胸部阻抗和潮气量的数值呈上升趋势，但未达到统计学显著性（在线补充表 S13）。在 NYHA 功能分级中也观察到了类似的一般趋势（在线补充表 S13）。
Adverse events and device deficiencies
不良事件和设备缺陷
There were no device-related adverse events in this study cohort, including no evidence of device–device interaction in any patients with an implanted cardiac device. Six device deficiencies occurred, two were due to the adhesive falling off and required replacement, three deficiencies were due to the device being unable to obtain an auscultation wave form, and one case where the device failed to connect to the mobile application.
在本研究队列中没有与设备相关的不良事件，包括在任何植入心脏设备的患者中没有设备间相互作用的证据。发生了六起设备缺陷，其中两起是由于粘合剂脱落需要更换，三起缺陷是由于设备无法获得听诊波形，以及一起设备未能连接到移动应用程序的情况。
Discussion 讨论
In this study, the change in thoracic impedance measured by the non-invasive monitoring device correlated with change in weight though not B-lines on LUS in patients hospitalized for HF as their congestion was treated, and was also correlated with patient symptoms as measured by the ECCS. In patients receiving haemodialysis change in thoracic impedance was correlated with volume of fluid removed during the dialysis session (Graphical Abstract).
在这项研究中，非侵入性监测设备测量的胸部阻抗变化与住院心力衰竭患者在治疗充血过程中体重变化相关，但与 LUS 上的 B 线无关，并且与通过 ECCS 测量的患者症状相关。在接受血液透析的患者中，胸部阻抗的变化与透析过程中去除的液体体积相关（图形摘要）。
We examined the correlation between multiple measures derived from the CPM device and clinical measures related to congestion at single time points and with changes in the measures. We found that correlations were most pronounced with changes in keeping with the purpose of the device, to detect changes in a patient over time. Correlations between CPM-derived measures and clinical measures that were not statistically significant were in a direction that was in keeping with a plausible physiological relationship. The correlation with a single measurement at RHC was low and this was in part expected. The device is designed to track changes within a patient rather than changes or differences between patients and static measures. The most robust correlations were observed in the patients undergoing decongestion during a hospitalization for HF, the target population. We were able to observe correlations between changes in fluid status over a short period of time in those patients undergoing haemodialysis. As can be seen from the results, multiple correlations exist between the measures from the CPM device and the clinical measures, and by integrating the information from each of the measurements taken, the sensitivity of the device to detect subclinical decompensation could be improved and indicate that a multiparameter approach to monitoring HF is useful.
我们检查了来自 CPM 设备的多项测量与与充血相关的临床测量之间的相关性，分别在单一时间点和测量变化时进行。我们发现，相关性在与设备目的相符的变化中最为明显，即检测患者随时间的变化。CPM 衍生测量与临床测量之间的相关性在统计上不显著，但方向与合理的生理关系一致。在 RHC 的单一测量的相关性较低，这在一定程度上是预期的。该设备旨在跟踪患者内部的变化，而不是患者之间的变化或差异以及静态测量。在接受心力衰竭住院治疗的患者中观察到最强的相关性，这是目标人群。我们能够观察到在接受血液透析的患者中，短时间内液体状态变化之间的相关性。 从结果可以看出，CPM 设备的测量值与临床测量值之间存在多重相关性，通过整合每个测量的数据信息，可以提高设备检测亚临床失代偿的敏感性，并表明多参数监测心力衰竭的方法是有用的。
We studied correlations with clinically recognized and meaningful measures of congestion. It may be surprising that in our cohort of patients hospitalized for HF, thoracic impedance did not correlate well with LUS B-lines, a sensitive marker of pulmonary fluid and predictor of rehospitalization in HF.9, 10 However, this may be because all patients were congested on admission at the start of the study, and many remained congested as indicated by the presence of B-lines at the time they were determined fit for discharge by the treating team. Therefore, the difference in B-lines may not have been large enough to detect a statistically significant correlation and perhaps with greater decongestion, the correlation would be evident. A further issue may be the phenotype of decompensated patients with HF. Patients can present with predominantly pulmonary signs but can also present with weight gain and peripheral oedema.11-13 Given that correlations between changes in CPM-derived thoracic impedance and S3 correlated with weight change, this would suggest these indices were a more sensitive measure of decongestion in this population who perhaps had more peripheral than pulmonary congestion.
我们研究了与临床认可和有意义的充血指标的相关性。在我们住院心力衰竭（HF）患者的队列中，胸部阻抗与 LUS B 线（肺部液体的敏感标志和 HF 再住院的预测因子）之间的相关性并不理想，这可能令人惊讶。然而，这可能是因为所有患者在研究开始时入院时均存在充血，许多患者在被治疗团队确定适合出院时仍然存在充血，表现在 B 线的存在上。因此，B 线的差异可能不足以检测到统计学上显著的相关性，也许随着更大的去充血，相关性会变得明显。另一个问题可能是失代偿 HF 患者的表型。患者可能主要表现为肺部症状，但也可能出现体重增加和外周水肿。考虑到 CPM 衍生的胸部阻抗变化与 S3 与体重变化之间的相关性，这表明这些指标在这个可能外周充血多于肺部充血的群体中是更敏感的去充血测量。
The device is designed to be used by one patient over time to monitor HF-related health status and warn of possible decompensation. Our findings suggest that the device can detect physiological changes associated with the worsening of HF and that it may be possible to develop an algorithm to provide early warning of decompensation and allow appropriate intervention. Clinical examination in HF is insensitive and dependent on experience, especially for specific signs such as a S3.14 The ability of the CPM device to detect S3 energy, in essence an inaudible diastolic heart sound, and its correlation with clinically meaningful changes in congestion, like change in body weight, suggest that the device could help aid decongestion by identifying otherwise difficult to detect signs and integrating these into the assessment of the patient.
该设备旨在由单个患者长期使用，以监测与心力衰竭（HF）相关的健康状态并警告可能的失代偿。我们的研究结果表明，该设备可以检测与 HF 恶化相关的生理变化，并且可能开发出一种算法，以提供失代偿的早期警告并允许适当的干预。HF 的临床检查不敏感且依赖于经验，特别是对于特定的体征，如 S3。CPM 设备检测 S3 能量的能力，实质上是一种不可听见的舒张期心音，并且与临床上有意义的充血变化（如体重变化）相关，表明该设备可以通过识别其他难以检测的体征并将其整合到患者评估中来帮助减轻充血。
Our study has several limitations. We enrolled a relatively small number of patients, although our studies were powered to make clinically meaningful correlations between the measures derived from the device and clinically relevant markers of congestion and decongestion. We did not test the device in the outpatient setting where the device was designed to detect congestion and decompensation. We did not test whether using this device could reduce hospitalizations and improve outcomes and this remains to be tested in a randomized trial. In many patients the symptoms of HF are only evident on exertion and we obtained all measurements at rest. Whether sensitivity could be improved with measurement during or shortly after exercise is to be determined in an ongoing study.
我们的研究有几个局限性。我们招募的患者数量相对较少，尽管我们的研究有足够的统计能力来建立设备测量值与临床相关的充血和去充血标志之间的临床意义相关性。我们没有在门诊环境中测试该设备，而该设备旨在检测充血和失代偿。我们没有测试使用该设备是否可以减少住院率和改善结果，这仍需在随机试验中进行测试。在许多患者中，心力衰竭的症状仅在运动时显现，而我们在静息状态下获得了所有测量值。是否可以通过在运动期间或运动后不久进行测量来提高敏感性，仍在进行中的研究中确定。
Conclusion 结论
A non-invasive monitoring device worn for less than 5 min was able to measure changes in congestion in patients undergoing decongestion in hospital for decompensated HF and fluid changes in patients requiring haemodialysis. The device, measuring multiple parameters, provides a new method of monitoring patients with HF.
一种佩戴时间少于 5 分钟的非侵入性监测设备能够测量在医院接受去充血治疗的失代偿性心力衰竭患者的充血变化，以及需要血液透析的患者的液体变化。该设备测量多个参数，为心力衰竭患者的监测提供了一种新方法。
Funding 资金
The CONGEST-HF study was funded by Analog Devices Inc. A.T., J.J.V.M., M.C.P., and P.S.J. are supported by a British Heart Foundation Centre of Research Excellence Grant RE/18/6/34217 and the Vera Melrose Heart Failure Research Fund.
CONGEST-HF 研究由模拟设备公司资助。A.T.、J.J.V.M.、M.C.P.和 P.S.J.获得了英国心脏基金会卓越研究中心的资助，资助编号为 RE/18/6/34217，以及维拉·梅尔罗斯心力衰竭研究基金的支持。
Conflict of interest: J.P.C. reports consultancy fees for FIRE-1. J.P.C.'s employer, the University of Glasgow, were paid for his time through a research grant from Analog Devices Inc. A.McI. reports his employer, the University of Glasgow, were paid for his time through a research grant from Analog Devices Inc. A.McC. reports his employer, the University of Glasgow, were paid for his time through a research grant from Analog Devices Inc. P.W. reports grant income from AstraZeneca, Boehringer Ingelheim, Novartis, and Roche Diagnostics, and speaker fees from Novo Nordisk and Raisio, all outside the submitted work. M.M.Y.L. has received research grants through his institution, the University of Glasgow, from AstraZeneca, Boehringer Ingelheim and Roche Diagnostics; is a member of a Trial Steering Committee for Cytokinetics, and Clinical Endpoints Committee for Bayer. T.A., J.S., V.G., J.H., and S.T. are employees of Analog Devices Inc. P.R.K. reports research grants from Pharmacosmos and consultancy and lectures honoraria from Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, FIRE-1, Novartis, Pfizer, Pharmacosmos, Servier and CSL Vifor. P.B.M. reports lecture and advisory boards honoraria from AstraZeneca, Boehringer Ingelheim, GSK, Pharmacosmos, Vifor, Bayer, Astellas. N.N.L. reports research grants from Roche Diagnostics, AstraZeneca and Boehringer Ingelheim, and consultancy/speaker's fees from Boehringer Ingelheim, Roche Diagnostics, Myokardia, Pharmacosmos, Akero Therapeutics, CV6 Therapeutics, Jazz Pharma and Novartis, all outside the submitted work. J.J.V.M. has received payments through Glasgow University for work on clinical trials, consulting and other activities from Alnylam, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, BMS, Cardurion, Cytokinetics, Dal-Cor, GSK, Ionis, KBP Biosciences, Novartis, Pfizer, Theracos; personal lecture fees from the Corpus, Abbott, Hikma, Sun Pharmaceuticals, Medscape/Heart.Org, Radcliffe Cardiology, Servier, Director, Global Clinical Trial Partners (GCTP). M.C.P. reports research funding from Boehringer Ingelheim, Roche, SQ Innovations, AstraZeneca, Novartis, Novo Nordisk, Medtronic, Boston Scientific, Pharmacosmos; consultancy and trial committees for Akero, Applied Therapeutics, Amgen, AnaCardio, Biosensors, Boehringer Ingelheim, Novartis, Astra, Zeneca, Novo Nordisk, Abbvie, Bayer, Horizon Therapeutics, Takeda, Cardiorentis, Pharmacosmos, Siemens, Eli Lilly, Vifor, New Amsterdam, Moderna, Teikoku, LIB Therapeutics, 3R Lifesciences. R.S.G. reports consultancy fees from Abbott, Anacardio, AstraZeneca, Boehringer Ingelheim, Boston Scientific, Novartis, Pfizer, Pharmacosmos, Roche Diagnostics, and research support from Abbott, Boston Scientific and Roche Diagnostics. P.S.J. reports speaker's fees from AstraZeneca, Novartis, Alkem Metabolics, ProAdWise Communications, Sun Pharmaceuticals; advisory board fees from AstraZeneca, Boehringer Ingelheim, Novartis; research funding from AstraZeneca, Boehringer Ingelheim, Analog Devices Inc. P.S.J.'s employer the University of Glasgow has been remunerated for clinical trial work from AstraZeneca, Bayer AG, Novartis and NovoNordisk. Director, Global Clinical Trial Partners (GCTP). All other authors have nothing to disclose.
利益冲突：J.P.C. 报告了来自 FIRE-1 的咨询费用。J.P.C. 的雇主，格拉斯哥大学，通过来自模拟设备公司的研究资助支付了他的时间。A.McI. 报告他的雇主，格拉斯哥大学，通过来自模拟设备公司的研究资助支付了他的时间。A.McC. 报告他的雇主，格拉斯哥大学，通过来自模拟设备公司的研究资助支付了他的时间。P.W. 报告了来自阿斯利康、勃林格殷格翰、诺华和罗氏诊断的资助收入，以及来自诺和诺德和 Raisio 的演讲费用，均与提交的工作无关。M.M.Y.L. 通过他的机构格拉斯哥大学获得了来自阿斯利康、勃林格殷格翰和罗氏诊断的研究资助；是细胞动力学的试验指导委员会和拜耳的临床终点委员会的成员。T.A.、J.S.、V.G.、J.H. 和 S.T. 是模拟设备公司的员工。P.R.K. 报告了来自 Pharmacosmos 的研究资助，以及来自安进、阿斯利康、拜耳、勃林格殷格翰、FIRE-1、诺华、辉瑞、Pharmacosmos、Servier 和 CSL Vifor 的咨询和讲座酬金。P.B.M. 报告了来自阿斯利康、博瑞医药、葛兰素史克、Pharmacosmos、Vifor、拜耳、安斯泰来公司的讲座和顾问委员会的酬金。N.N.L. 报告了来自罗氏诊断、阿斯利康和博瑞医药的研究资助，以及来自博瑞医药、罗氏诊断、Myokardia、Pharmacosmos、Akero Therapeutics、CV6 Therapeutics、Jazz Pharma 和诺华的咨询/演讲费用，均与提交的工作无关。J.J.V.M. 通过格拉斯哥大学获得了来自 Alnylam、安进、阿斯利康、拜耳、博瑞医药、BMS、Cardurion、Cytokinetics、Dal-Cor、葛兰素史克、Ionis、KBP Biosciences、诺华、辉瑞、Theracos 的临床试验、咨询和其他活动的付款；来自 Corpus、雅培、Hikma、Sun Pharmaceuticals、Medscape/Heart.Org、Radcliffe Cardiology、Servier 的个人讲座费用，以及全球临床试验合作伙伴（GCTP）董事的费用。M.C.P. 报告了来自勃林格殷格翰、罗氏、SQ Innovations、阿斯利康、诺华、诺和诺德、美敦力、波士顿科学、Pharmacosmos 的研究资金；为 Akero、Applied Therapeutics、安进、AnaCardio、Biosensors、勃林格殷格翰、诺华、阿斯利康、诺和诺德、Abbvie、拜耳、Horizon Therapeutics、武田、Cardiorentis、Pharmacosmos、西门子、礼来、Vifor、新阿姆斯特丹、Moderna、Teikoku、LIB Therapeutics、3R Lifesciences 的咨询和试验委员会提供服务。R.S.G.报告了来自雅培、AnaCardio、阿斯利康、勃林格殷格翰、波士顿科学、诺华、辉瑞、Pharmacosmos、罗氏诊断的咨询费，以及来自雅培、波士顿科学和罗氏诊断的研究支持。P.S.J.报告了来自阿斯利康、诺华、Alkem Metabolics、ProAdWise Communications、Sun Pharmaceuticals 的演讲费；来自阿斯利康、勃林格殷格翰、诺华的顾问委员会费用；来自阿斯利康、勃林格殷格翰、Analog Devices Inc.的研究资金。P.S.J.的雇主格拉斯哥大学因阿斯利康、拜耳、诺华和诺和诺德的临床试验工作获得报酬。全球临床试验合作伙伴（GCTP）董事。 所有其他作者没有任何披露。
