X2.1 Confirm exacerbation and categorise severity

Assessment of severity of the exacerbation includes a medi­cal history, examination, spirometry and, in severe cases (FEV1 < 40% predicted), blood gas measurements, chest x- rays and electrocardiography.

Patients should be provided with and bring a summary of their medical problems and treatment (eg, a personal health record). If available, results of previous stable lung function tests and arterial blood gas measurements are invaluable for comparison.

Spirometry: Unless confused or comatose, even the sickest of patients can perform an FEV1 manoeuvre. An FEV1 less than 1.0 L (or < 40% predicted) is usually indicative of a severe exacerbation in patients with moderate COPD. For patients with stable levels below these values (ie, severe COPD), the most important signs of a severe exacerbation will be worsening hypoxaemia, acute respiratory acidosis (carbon dioxide retention), or both.

Arterial blood gases: Arterial blood gas levels should be measured if the FEV1 is less than 1.0 L or less than 40% predicted, or if percutaneous oxygen saturation is less than 90% in the presence of adequate peripheral perfusion or cor pulmonale. Values obtained while breathing room air are the most useful for assessing ventilation–perfusion inequality. A PaO2 less than 60 mmHg (8 kPa) indicates respiratory fail­ure, while a PaCO2 greater than 45 mmHg indicates ventila­tory failure. Respiratory acidosis indicates acute respiratory failure warranting consideration for assisted ventilation.

All prospective RCTs that have demonstrated a mortality advantage with the use of NIV in exacerbations of COPD have used ABG (arterial blood gas), not VBG (venous blood gas) samples to determine need for NIV. McKeever et al examined paired ABG-VBG (venous blood gas) samples in 234 patients admitted to hospital with a doctor diagnosed exacerbation of COPD (McKeever 2016). A VBG pH ≤7.34 gave a sensitivity of 88.9% and specificity of 95.6% for an ABG pH ≤7.35. The authors argued that all patients presenting with an exacerbation of COPD should initially be assessed with a VBG and only go on to an ABG if the VBG pH ≤7.34. The primary reasons for preferring VBG samples cited by the authors were less pain and lower risk of bruising. The general applicability of these findings is limited by the fact that this cohort had relatively few patients with pH below 7.30. The authors did not propose that VBGs should replace ABGs to assess severity of respiratory failure or be used to monitor patient response to treatment/ NIV.  Caution is required due to the lesser precision with VBGs compared to ABGs.

Chest x-ray and electrocardiogram: These help to identify alternative diagnoses and complications, such as pulmonary oedema, pneumothorax, pneumonia, empyema, arrhythmias, myocardial ischaemia and others.

Studies have identified a simple clinical prediction score, the BAP-65, based on age, basal urea nitrogen, acute mental status change and pulse, which predict in-hospital mortality (Tabak 2009, Shorr 2011). In-hospital mortality in both studies increased as patient classification escalated from 1 (no risk factors, age <65 years) to 5 (3 risk factors present), the highest class being associated with an in-hospital mortality between 14.1% and >25%.

A 2012 prospective single centre study of 920 patients admitted with an exacerbation of COPD found that those with CXR confirmed pneumonia had a far higher mortality  (20.1% versus 5.8%, p<0.001).  Severity of dypsnoea in the stable state was strongly associated with both in-hospital mortality and early readmission (Steer 2012) [evidence level III-2].