Collection of data

All data were collected through Philips IntelliVue Clinical Information Portfolio (ICIP) and Philips Intelli Space Critical Care and Anesthesia (ICCA) System. SaO₂, Hb and lactic acid were measured routinely using a blood gas analyzer (Radiometer ABL 800 Flex, Radiometer, Willich, Germany).

According to the standard operation procedures all patients with impaired cardiopulmonary function were managed with a triple-lumen central venous catheter. Additionally, a transpulmonary thermodilution catheter (Pulsiocath™, Pulsion Medical Systems, Munich, Germany) was utilized in patients, when indicated by the attending physician.

The Pulse Contour Cardiac Output monitor (PiCCOplus™, Pulsion Medical Systems, Munich, Germany) was used for measuring CI and SVRI with routine calibrations around every 8h, averaging three daily DO₂I-measurements.

DO₂I was calculated using a simplified version of the standard formula:

And for calculating the SVRI we used the following formula:

In order to quantify an insufficient oxygen delivery index within 72h before the diagnosis, we calculated meanDO₂I during the stay in ICU as a surrogate for a longer lasting hypoxic status and the minDO₂I during the stay in ICU to capture shorter periods of hypoxia. Furthermore, we calculated the mean CI, mean SaO₂, mean Hb and mean SVRI during the stay in ICU.

Lactate levels were collected in the case group 72h before the diagnosis of II and in the control group we collected all lactate values over the ICU-stay. A plasma lactate concentration of 2mmol/l or less was defined as normal finding as this represents the clearing capacity for lactic acid in normal adults [25].