The CO₂ GAP Project--CO₂ GAP as a prognostic tool in emergency departments.

Objective: To determine whether CO2 GAP [(a-ET) PCO2] value differs consistently in patients presenting with shortness of breath to the ED requiring ventilatory support. To determine a cut-off value of CO2 GAP, which is consistently associated with measured outcome and to compare its performance against other derived variables. Methods: This prospective observational study was conducted in ED on a convenience sample of 412 from 759 patients who underwent concurrent arterial blood gas and ETCO2 (end-tidal CO2) measurement. They were randomized to test sample of 312 patients and validation set of 100 patients. The primary outcome of interest was the need for ventilatory support and secondary outcomes were admission to high dependency unit or death during stay in ED. The randomly selected training set was used to select cut-points for the possible predictors; that is, CO2 GAP, CO2 gradient, physiologic dead space and A-a gradient. The sensitivity, specificity and predictive values of these predictors were validated in the test set of 100 patients. Results: Analysis of the receiver operating characteristic curves revealed the CO2 GAP performed significantly better than the arterial-alveolar gradient in patients requiring ventilator support (area under the curve 0.950 vs 0.726). A CO2 GAP >= 10 was associated with assisted ventilation outcomes when applied to the validation test set (100% sensitivity 70% specificity). Conclusions: The CO2 GAP [(a-ET) PCO2] differs significantly in patients requiring assisted ventilation when presenting with shortness of breath to EDs and further research addressing the prognostic value of CO2 GAP in this specific aspect is required.