The DECAF (Dyspnoea, Eosinopenia, Consolidation, Acidaemia, atrial Fibrillation) score was derived in the original derivation study of 920 adults admitted to two UK hospitals with acute exacerbations of COPD, including pneumonic and non-pneumonic cases, to predict in-hospital mortality from simple admission data.

Risk group	Score range	In-hospital mortality, derivation	In-hospital mortality, validation
Low	0-1	1.4%	1.0%
Intermediate	2	8.4%	5.4%
High	3-6	34.6%	21.4%

In the derivation cohort DECAF had an AUROC of about 0.86 for in-hospital mortality, and the validation study across four additional UK hospitals confirmed good discrimination and preserved separation between these risk groups.

Validation work confirmed the original thresholds for eosinopenia, below 0.05 x 10^9/L, and acidaemia, arterial pH below 7.30. A higher acidaemia cut-off offered no clear advantage, and patients with DECAF 0-1 and non-scoring acidaemia had very low observed mortality.

Across derivation and validation cohorts DECAF generally performed at least as well as, and often better than, other prognostic tools for acute exacerbation of COPD such as CURB-65, BAP-65, CAPS and APACHE II, including in pneumonic exacerbations where CURB-65 may underestimate short-term mortality.

Low-risk scores have been used to select patients for hospital-at-home or early supported discharge. In a randomised home-treatment trial of low-risk admissions, home treatment selected by DECAF was as safe as usual inpatient care over 90 days, reduced hospital bed-days and lowered overall costs. High DECAF scores identify patients with both a higher risk of in-hospital death and a shorter time to death, supporting early senior review, escalation decisions and timely palliative discussions.

MDCalc DECAF comparator

DECAF Score for Acute Exacerbation of COPD