Managing Congestive Heart Failure: It Is Mostly About Water, Not Salt!

This article refers to ‘Nurse-led diuretic titration via a point-of-care urinary sodium sensor in patients with acute decompensated heart failure (EASY-HF): A single-centre, randomized, open-label study’ by E. Meekers et al., published in Eur J Heart Fail 2024;26:2129–2139. Renal retention of water and salt, also known as congestion, is a key clinical feature of worsening heart failure and an important driver of hospitalizations and deaths.1 Managing congestion is one of the most difficult tasks facing both general physicians and heart failure specialists.2 Guidelines recommend administering loop diuretics to improve signs and symptoms of congestion but there is a paucity of evidence on how they should be used. Choice of loop diuretic dose is based on opinion rather than evidence and the clinical utility of various methods of quantifying diuretics response is still disputed.3 The acuity of symptoms and the severity of congestion are major determinants of diuretic strategy. Most patients hospitalized due to worsening peripheral oedema are comfortable sitting upright at rest. However, patients admitted with pulmonary oedema and severe breathlessness at rest require urgent treatment to relieve symptoms using oxygen, a substantial bolus of intravenous loop diuretics and, with less certainty, a vasodilator for 1–2 h to reduce pulmonary capillary pressures and relieve alveolar oedema.4 The diuretic strategy should be tailored according to context: not all patients hospitalized for worsening heart failure have the same needs. Several strategies have been proposed to assess congestion and monitor diuretic response. Changes in body weight5 may be difficult to obtain, until patients are well enough to get out of bed. Weighing a patient who is able to stand is simple, but failure to provide and enforce standard procedures for measurements limits accuracy. Bioimpedance weight scales might ensure that instructions are followed with greater attention to detail, such as taking the patient's socks and shoes off. Bioimpedance technologies also enable continuous haemodynamic monitoring, simply and at low cost,6 but evidence that such measurements can be used to tailor treatment for congestion has not received widespread acceptance. Right heart catheterization carries some risks, is costly and requires considerable expertise. Changes in haematocrit, to estimate plasma volume, and natriuretic peptides have also been considered, but evidence of clinical utility in the acute setting is lacking. Imaging of heart, kidney and large veins by ultrasound can identify and quantify congestion with precision and might be useful for discharge planning in the latter part of an admission, but there is little evidence that it can guide the acute management of congestion.7, 8 Surprisingly expert position papers, as well as the European Society of Cardiology guidelines,3, 9 suggest assessing the diuretic response based on the amount of sodium in the urine 2 h following the administration of diuretics. Sodium concentrations below 50–70 mmol/L are thought to indicate substantial resistance to natriuresis and the need to intensify treatment. Compared to some other methods of quantifying diuretic response, spot urinary sodium is widely available, not too complicated nor very expensive. Two recent trials, the Efficacy of a Standardized Diuretic Protocol in Acute Heart Failure (ENACT-HF) study and the Pragmatic Urinary Sodium-based algorithm in Acute Heart Failure (PUSH-AHF) study, suggest that basing diuretic dose on spot urinary sodium results in a greater natriuresis and diuresis in the first days of a hospitalization for worsening congestive heart failure.10, 11 Is this evidence sufficient to justify these recommendations, or would simply recommending starting with higher doses of loop diuretics have similar results? Moreover, laboratory measurements of urinary sodium can present some logistical problems and inconvenience. Recently, Meekers and colleagues from Genk, Belgium, reported the feasibility of using a point-of-care urinary sodium sensor to adjust diuretic therapy for patients hospitalized with worsening congestive heart failure; those with an estimated glomerular filtration rate (eGFR) <20 ml/min/1.73 m2 or requiring intravenous inotropic or vasopressor agents were excluded.12 The authors conducted a single-centre, randomized trial (Available Urinary Sodium Analysis in Patients with Acute Decompensated Heart Failure; EASY-HF) of nurse-led care, based on a protocol guided by a sodium sensor, compared to physician-led standard of care (SoC). Nurses made urine measurements 2–4 h after each bolus of diuretic, doubling the dose every 6 h until a urinary sodium of >70 mmol/L was achieved. Physicians were blinded to sodium-sensor results and guided by clinical assessment of congestion alone. Consecutive 24-h urine collections were acquired during the first 2 days of hospitalization. The primary endpoint was natriuresis after 48 h, but safety and user friendliness of the protocol and the fidelity of the point-of-care sensor were also evaluated. Patients were old (80 ± 8 years), predominantly men (75%), with a high median N-terminal pro-B-type natriuretic peptide (4667 [2667–7709] ng/L) and impaired renal function (58% had an eGFR <50 ml/min/1.73 m2). More than one in three of those recruited had newly diagnosed heart failure, and about half were not taking loop diuretics before admission. Peripheral oedema contributed most to the overall congestion score and at least half of patients were asymptomatic at rest; dyspnoea scores at baseline were measured but, unfortunately, not reported. Natriuresis after 48 h was higher in the intervention arm versus those assigned to SoC, which was mirrored by a greater diuresis (6.0 L with SoC compared to 7.3 L with measurement of urinary sodium) and a similar trend for weight loss (4.0 kg and 4.4 kg, respectively). The median dose of loop diuretics administered during the first 48 h was 240 (200–320) mg in the SoC versus 320 (160–480) mg in the intervention group. However, the increased natriuresis and diuresis did not translate into better resolution of congestion or a shorter length of stay. Relieving congestion and achieving euvolaemia are key goals in the management of acute heart failure; diuresis, rather than natriuresis, is the way to achieve this. However, few patients become euvolaemic within 2 days and many will have residual congestion at the time of discharge, which is associated with a greater risk of rehospitalization and death compared to those who are completely decongested.13 Diuretic response (defined both as diuresis or natriuresis) is frequently considered as a surrogate of decongestion but, actually, these concepts are very different. Diuresis provides only modest incremental prognostic information compared to the amount of residual congestion at discharge. Management of congestion requires a multidimensional clinical approach, including assessment of blood pressure, serum sodium concentration and renal function; not just diuretic response. Overall, the sodium-sensor strategy was safe. No patient in either arm of the trial had a decrease in eGFR of >50%, which may reassure some, although the relevance of this endpoint is questionable.14, 15 Intensification of diuretic therapy may cause eGFR to fall, leading clinicians to reduce the dose of diuretics. However, worsening renal function is only associated with an adverse prognosis when signs of congestion persist at discharge.16 In the EASY-HF trial, the largest benefit in natriuresis and diuresis was achieved during the second day of treatment. This reflects the protocol design, which required higher doses of loop diuretics if the urinary sodium target had not been achieved. Giving lower initial doses of diuretics to gauge diuretic responsiveness is a perfectly reasonable strategy for the majority of patients who have few symptoms at rest in bed.4 Failure to intensify diuretic therapy in those with an inadequate initial response is likely to lead to delays in resolution of congestion and fitness for discharge. Maybe all that is required is to encourage higher initial doses of diuretics and further intensification for those who do not produce good volumes of urine assuming most are on bed-rest for the first day, with a urinary catheter (or penile sheath) if incontinent. Good attention to urine collection may be the best measure of diuretic effect; copious amounts of urine should be produced within 30 min of an intravenous bolus of loop diuretic (Figure 1). Meekers and colleagues report both natriuresis and diuresis. In congested patients, focusing only on sodium excretion might be misleading. Congestion is due to a large excess volume of water but only a small excess of salt (1000 g of extracellular water contains about 30 g of sodium).17 It is the size of diuresis, not natriuresis, that matters. Urinary sodium concentration might be a surrogate for diuresis but why not focus on measuring urine volume accurately? Measuring urine volume is simple, inexpensive, real-time, does not require specialist equipment and, for shorter timeframes, is accurate. When the patient is unable to get to a toilet, urine volume measurements made by staff should be accurate. Once the patient is mobile, strict attention to measurement protocols should deliver weights accurate to within about 100 g (about 100 ml) of water. Ultimately, it is patients who are the best judge of the success of treatment. Did treatment make them feel better, improving breathlessness, leg swelling and abdominal bloating? Did it get them fit for discharge sooner? Was any initial success sustained for the following weeks, months, or years? Did treatment prolong life rather than prolong dying? Many treatments for acute heart failure have been investigated, but in the last 30 years, only sodium–glucose cotransporter 2 inhibitors have been shown to improve both congestion and prognosis. They have rather modest effects on natriuresis but rather more on diuresis.18 The EASY-HF trialists should be congratulated for making us think. More effective diuretic protocols are certainly required for patients with acute heart failure. If measurements of urinary sodium provide a ‘crutch’ to help deliver such protocols that is great. However, once we learn how to ‘walk’, perhaps we can throw away the ‘crutch’? Conflict of interest: A.I. has nothing to disclose. I.E. has received speaker honoraria from Pharmacosmos, Astellas, Bristol Myers Squibb. P.P. has received consultancy honoraria and/or sponsorship support from Pharmacosmos, Vifor, and AstraZeneca in the past 3 years, not connected with this editorial and a Scotland research grant from Heart Research UK (grant number RG2676/18/21). J.G.F.C. has received financial support for research and personal honoraria from Bristol Myers Squibb, CSL-Vifor and Pharmacosmsos Ltd and has stock options for Heartfelt Technology.