3152 patients were included. The range was between 143 (Cantabria) and 2122 (Richmond). In 2171 patients’ respiratory rate was measured on two occasions.

The most effective reduction of respiratory rate was found in Bonn in the subgroups of severe and slight tachypnoea. In Bonn emergency physicians were able to significantly reduce respiratory rate in these subgroups by 12.8 ± 11.9 and 5.2 ± 4.0 breath/min, respectively. The physician staffed system in Cantabria check details and the paramedic systems of Richmond and Coventry achieved a reduction in these subgroups of only 6.0 ± 7.2 and 2.8 ± 2.8 breath/min, respectively. Differences between the systems were significant. In Bonn before and after treatment, 195 and 78 out of 494 patients showed slight or severe XL184 ic50 tachypnoea, respectively. In Bonn, therefore, tachypnoea could be normalized pharmacologically in 60% of the patients. In the other EMS systems 1469 patients had a respiratory rate ≥19 breath/min which could be reduced by treatment in 8.6% of the patients and the remaining 1342 continued to suffer tachypnoea. Oxygen saturation was measured in 1927

patients on two occasions, ranging from 121 in Cantabria to 1112 in Richmond. Table 4 demonstrates that in patients with hypoxemia (SpO2 < 90%) improvement of oxygen saturation was most effective in Bonn and Richmond. Oxygen saturation was increased by 13.4 ± 9.6% and 11.8 ± 10.5%, respectively. The increase in Bonn was significantly higher than in Cantabria and Coventry (8.7 ± 5.6% and 9.7 ± 5.6%,

respectively). Emergency physicians in Bonn were confronted with the highest percentage of patients with hypoxemia, but all EMS systems were able to normalize SpO2 in more than 75% of these patients. 872 patients were included, 499 in Richmond and 72 in Coventry. 818 patient records reported initial rhythm and outcome. We demonstrate that when emergency physicians treated patients with OHCA significantly more patients reached hospital alive than when treated by paramedics (Bonn: 35.6%; Cantabria: 30.1% [O.R. 1.28 (95% CI 0.75–2.2; p > 0.05]; Coventry: Etofibrate 11.9% [O.R. 4.07 (95% CI 1.81–9.61; p < 0.05] and Richmond: 9.2% [O.R. 5.47 (95% CI 3.46–8.64; p < 0.0001], respectively). In the subgroup of patients found with VF all systems were able to admit more patients with ROSC to the hospitals, but difference between the EMS systems favouring the physician staffed systems still remained (Bonn: 60%; Cantabria: 48.8% [O.R. 1.57 (95% CI 0.66–3.75; p > 0.05]; Coventry: 11.1% [O.R. 12 (95% CI 3.09–46.6; p < 0.05] and Richmond: 17.7% [O.R. 6.97 (95% CI 3.07–15.85; p < 0.05], respectively). In addition, in the subgroup of patients with PEA/Asystole as initial presenting rhythm, Bonn and Cantabria demonstrated the highest success rates (Bonn: 26.6%, Cantabria: 16.7% [O.R. 1.81 (95% CI 0.82–4.04; p > 0.