Optimisation of Heart Failure Management in Nursing Homes Using Point-of-Care Ultrasonography: Harmonious Trial Rationale and Design

Screening for HF will follow the 2016 European Cardiology Society diagnostic algorithm for the diagnosis of HF of non-acute onset (10). All screening stages will take place in nursing home facilities.

The diagnostic procedure to determine HF will include: 1) the assessment of HF probability by assessing clinical history, clinical signs and ECG, 2) the NT-proBNP measurement and 3) the echocardiography with clinical judgement.

With the help of two junior researchers, the nursing home physicians will review residents’ history and perform a clinical examination to identify clinical signs of heart failure. They will have full access to residents’ medical records and will be able to assess multimorbidity. Trained nursing home staff will record ECG and collect blood samples for measurement of the NT-proBNP marker. NT-proBNP will be measured using a Roche Cobas h 232 point-of-care system on site. The ECG will be interpreted by a nursing home physician. A qualified cardiologist will perform the echocardiography exam using a GE Vivid 7 ultrasound system. While performing echocardiography, the cardiologist will be blinded for NT-proBNP value and will perform echocardiography on an additional 10% of NT-proBNP negative patients as a control measure. A predefined ultrasonography protocol will be followed and images and clips recorded.

After performing echocardiography, the cardiologist will have access to all patient data to diagnose or exclude HF. The current classification of HF will be used: heart failure with preserved (HFpEF), mid-range (HFmrEF) and reduced ejection fraction (HFrEF). For the diagnosis of HFmrEF and HFpEF, the following requirements will have to be met: presence of symptoms and/or signs of heart failure, left ventricular ejection fraction (LVEF) of ≥50% or 40–49%, NTproBNP ≥125 pg/mL, and objective evidence of other cardiac functional and structural alterations underlying heart failure.

At the inclusion point, the cardiologist performing echocardiography will recommend a therapeutic management plan for HF patients. The cardiologist and the nursing home physicians will jointly optimise medication therapy, with due consideration given to current medication guidelines and the individual patient’s characteristics. From this point on, the patient will be managed by the nursing home physician.

Patients diagnosed with HF will be randomised using block randomisation (block n=10) into intervention and control groups. The randomisation list will be performed by an independent statistics adviser. For the randomisation process, patients’ identification numbers will be used in place of a name. This is to avoid randomisation bias. The research coordinator will perform randomisation by applying the sealed envelope technique (20).

Nursing home physicians will manage HF patients for six months by assessing volume status and aiming to prevent any significant HF deterioration. The volume assessment will be performed using POCUS and clinical signs in the intervention group, and only clinically in the control group. Patients in the control group will receive standard care in line with current HF guidelines and good clinical practice by the nursing home physician. The intervention group will receive the same clinical standards plus POCUS for volume evaluation.

The follow-up of both the test and control groups will last for six months. The patients will be evaluated at regular time intervals, as presented in Table 1. Unplanned evaluations will be performed whenever HF deterioration is suspected and one week after diuretic therapy change.

In the intervention group, nursing home physicians will use Samsung SonoAce R3 point-of-care devices. POCUS volume assessment will consist of the visualisation of B-lines on eight standard positions: the mid-clavicular line in the second and fourth intercostal spaces bilaterally, and the mid-axillary line in the second and fourth intercostal spaces bilaterally (Figure 2) (21). Inferior vena cava diameter and collapsibility will also be evaluated, as presented in Table 2.

Figure 2

The inferior vena cava diameter and collapsibility will be assessed 2 cm below the junction with the right atrium (22, 23). It will be categorised as small, medium or large. The inferior vena cava will be considered collapsible if the inspiratory collapse is more than 50% of its diameter.

Nursing home physicians will monitor patients with HF for signs of volume overload. Volume overload will be assumed if any new HF-related symptoms appear (breathlessness, orthopnoea, paroxysmal nocturnal dyspnoea, reduced exercise tolerance, fatigue, tiredness, increased time to recover after exercise) that are consistent with any clinical signs of HF (pulmonary crepitation, ankle swelling, elevated jugular venous pressure, hepatojugular reflux, gallop rhythm, laterally displaced apex, weight change >2 kg/week).

Additionally in the POCUS group, volume overload will be assumed if at least three B-lines are present in at least five out of eight regions of the thorax or if the inferior vena cava is found to be large (>2.5 cm) or medium-sized, but non-collapsible.

Volume depletion will be assumed if B-lines are absent and the inferior vena cava is small and not collapsible, together with clinical signs of volume depletion.

If the nursing home physicians assess volume overload, they will double the daily dose of diuretic and re-evaluate the patient in one week. If they find the patient volume depleted one week after the increase of the dose of diuretic, they will halve the daily dose of diuretic back to the initial dose.

The use of POCUS will also be assessed from the perspective of potential additional workload for nursing home physicians. The workload will be monitored with regard to the number of non-administrative contacts, therapy modifications and referrals. Furthermore, technical and other requirements for POCUS in NH will be documented.

The primary endpoint will be HF deterioration, defined as a composite of any of the following four events: the need for an intravenous diuretic application, the need for an emergency service intervention, the need for unplanned hospitalisation for non-injury causes, or death from whatever cause.

Secondary endpoints will be:

A change in quality of life (self-evaluated)

The workload of nursing home physicians will be measured by counting the number of all and HF-related a) non-administrative contacts, b) therapy modifications and c) unplanned non-injury referrals.

The calculation of sample size was based on an estimated prevalence of HF in nursing home residents of p=0.25 and on an estimated incidence of heart failure composite events in six months of 50% (6, 17).

In the intervention part of the research, we aim to decrease the number of patients with HF deterioration from 50% to 20% in six months. With a study power of 80% and an estimated statistical error of 5%, at least 90 residents with heart failure need to be included in the randomised controlled trial. With an expected prevalence of heart failure of p=0.25, at least 360 residents should be recruited for HF screening.

The rate of response/participation in clinical studies in this population is expected to be about 30% (25). It will therefore probably be necessary to approach approx. 1,000 nursing home residents in order to yield 360 residents for HF screening.

The data will be analysed using the SPSS statistical software package. We will show frequencies, averages and the standard deviation of variables, or value ranges and median where applicable.

Differences between groups will be tested using appropriate statistical methods, such as the T-test or analysis of variance (ANOVA) for continuous variables, and the Chi-square test and multivariable logistic regression analyses for discrete variables.

As statistical significance, a p value of <0.05 will be used.