Dyspnoea, Eosinopenia, Consolidation, Acidaemia and atrial Fibrillation (DECAF) prognostic score

Thursday, December 1, 2016

Guest feature by Dr Stephen Bourke, Consultant Respiratory Physician at North Tyneside General Hospital, and PhD fellows John Steer and Carlos Echevarria.

Acute exacerbations of COPD (AECOPD) disrupt patients’ lives, are a common reason for hospital admission, result in over a million hospital bed days per year in the UK and are a substantial cause of morbidity and mortality. In common with many conditions, clinicians struggle to accurately predict mortality in AECOPD; prognostic assessments are typically pessimistic and are outperformed by simple prediction tools. Accurate risk stratification can be used to inform treatment, place and level of care, improving outcomes whilst ensuring optimum use of resources.

Improving prognostication in acute exacerbations of COPD

As part of our COPD research programme we have developed the Dyspnoea, Eosinopenia, Consolidation, Acidaemia and atrial Fibrillation (DECAF) prognostic score in 2,645 patients and 6 UK hospitals (ref. 1 & 2). DECAF accurately predicts in-hospital mortality in AECOPD and stratifies patients into low (DECAF 0-1= 1 - 1.4%); intermediate (DECAF 2= 5.4 – 8.4%) and high risk (3+= 21.4 – 34.7%) groups. It is simple to score at the bedside and offers excellent performance (area under the receiver operating curve = 0.82 - 0.86), surpassing alternatives. Stable state breathlessness is assessed using the Extended MRC Dyspnoea score (eMRCD); those unable to leave their home unassisted are subdivided according to whether they can wash and dress independently (ref. 3). This is essentially a combined assessment of breathlessness and frailty and is the strongest single predictor of inpatient mortality. If oxygen saturation breathing room air is > 92%, acidaemia can be assumed not to be present without arterial blood gas confirmation.

Of importance, approximately 50% of admitted patients with AECOPD are low risk, thus potentially suitable for early (supported) discharge or full hospital at home. Selection for such services by DECAF allows inclusion of many more patients than earlier models. At the Winter BTS meeting we will present the results of our RCT of hospital at home selected by DECAF (funding: Research for Patient Benefit); this model is safe (no acute deaths), effective (no increase in readmissions) and preferred by 90% of patients. Patients with high risk scores may benefit from more intensive therapy or alternative palliation.  Routine capture of DECAF is endorsed by the 2014 National COPD Audit Report and BTS COPD Specialist Advisory Group. An important aspect of our approach is that we consider it unsafe to treat pneumonia complicating an exacerbation (pAECOPD) as simple pneumonia. This ignores co-existent airway inflammation and bronchoconstriction, and excludes potentially life-saving therapy.

Pneumonia and COPD exacerbations

A minority opinion views the presence of pneumonia as an exclusion criterion for the diagnosis of an acute exacerbation of COPD (i.e. that parenchymal disease excludes airway involvement) and that such patients should be managed according to pneumonia protocols. We believe this is incorrect:

  • In community acquired pneumonia (CAP), most studies show non-invasive ventilation (NIV) does not improve outcomes and the latest NIV guidelines clearly state NIV is not indicated (see BTS/ICS NIV guideline 2016: figure 1). In an RCT of NIV in CAP complicated by acute respiratory failure, the subgroup with COPD predominantly had respiratory acidaemia and showed lower intubation rates and improved survival among those randomised to NIV (ref. 4). Across the DECAF cohorts, 176 patients with pAECOPD received NIV; 117 (66%) survived to discharge and outcomes were proportional to total DECAF score. Of importance, most had severe airflow obstruction (mean FEV1 41%) and were considered not suitable for escalation to invasive ventilation. Therefore, failure to manage the patient as pAECOPD could lead to denial of potentially life-saving NIV. Furthermore, the airway component of the exacerbation will be left untreated and the risk of use of incorrect target saturations potentially increased.
  • pAECOPD has more similarities with non-pneumonic COPD exacerbations than CAP: the spectrum of pathogens in pAECOPD is more similar to non-pneumonic exacerbations than CAP; whilst the complication rates of CAP (particularly empyema) are substantially different from pAECOPD.
  • The diagnosis of AECOPD is broad and includes non-infective and infective processes. Including airway infection in the diagnosis of AECOPD yet excluding patients where the infection has spread to the lung parenchyma does not make pathophysiological sense. When a CT scan is performed, consolidation not previously recognised on a chest x-ray is commonly present; such a distinction is also imprecise.
  • In pneumonia, management decisions are often based on application of the CURB-65 prognostic score: patients with ‘low-risk’ CURB-65 score have a reported mortality of ~1.5% and are recommended for treatment outside of hospital. Our research has shown CURB-65 to be a poor predictor of outcome in pAECOPD, whilst DECAF is reliable. In the DECAF cohorts 788 patients had pAECOPD; in-hospital mortality in those with “low risk” scores by each tool was: CURB-65= 7.2%; DECAF= 1.6%.


  1. Echevarria C, Steer J, Heslop-Marshall K, et al. Validation of the DECAF score to predict hospital mortality in acute exacerbations of COPD. Thorax 2016;71:133-40.
  2. Steer J, Gibson GJ, Bourke SC. The DECAF Score: predicting in-hospital mortality in acute exacerbations of COPD. Thorax 2012;67:970-6.
  3. Steer J, Norman E, Afolabi OA, Gibson GJ, Bourke SC. Dyspnoea severity and pneumonia as predictors of in-hospital mortality and early readmission in acute exacerbations of COPD. Thorax 2011;in press.
  4. Confalonieri M, Potena A, Carbone G, Della Porta R, Tolley EA, Meduri GU. Acute Respiratory Failure in Patients with Severe Community-acquired Pneumonia: a Prospective Randomized Evaluation of Noninvasive Ventilation. Am J Respir Crit Care Med 1999;160:1585–91.