It is of utmost importance that a diagnosis should be established – and appropriate treatment initiated as soon as possible – in patients presenting with signs and symptoms of acute heart failure. In fact, it is recommended that diagnosis and identification of precipitants should occur in parallel with treatment (Shah et al., 2017). Since the symptoms and signs may be nonspecific or not sensitive, the diagnosis of acute heart failure must preferably be confirmed by the following investigations (Ponikowski et al., 2016):
Chest X-ray in the upright position (supine chest X-ray is of limited value in this setting) to confirm pulmonary venous congestion. It can also help to identify any pleural effusion.
Electrocardiogram. The ECG is seldom normal in acute heart failure, and it can aid in identifying concomitant cardiac diseases.
Echocardiography is helpful to assess the contractile function of the ventricles.
Measurement of plasma levels of NT-proBNP. The threshold level for an elevated NT-proBNP concentration is 300 pg/ml. One must note, however, that elevated NT-proBNP can also be found in many other cardiac and non-cardiac conditions.
Lung ultrasound is more accurate than auscultation or chest x-ray for the detection of pulmonary congestion in acute dyspnoea, and should be considered where available (Harjola et al., 2017).
The following laboratory tests should be carried out:
blood urea nitrogen (BUN) or urea
thyroid-stimulating hormone (TSH)
An algorithm for the diagnosis of heart failure (HF) in the non-acute setting is shown in Figure 8.
A recent pilot study explored whether a lung and cardiac ultrasound protocol could aid in the diagnosis of acute heart failure. The scans were carried out by an experienced emergency physician, and a positive result was considered to be the identification of B+ lines in both anterosuperior lung zones, plus a left ventricular ejection fraction less than 45%. Tested on 99 patients, the sensitivity was 25% (95% CI 14–41%), the specificity 100% (95% CI 94–100%). The protocol took under 2 minutes to perform on average, leading the investigators to conclude that it may represent an important screening tool for rapid assessment (Russell & Ehrman, 2017).