Paco2 becomes greater than Pvco2 during apnoea testing for brain death diagnosis

Abstract

The apnoea test is performed to complete the diagnosis of brain death. The most efficient respiratory stimulus is the arterial carbon dioxide partial pressure (Paco2), hence it is used in this test [1, 2]. The patient is disconnected from the ventilator while receiving 100% oxygen via a catheter inserted into the endotracheal tube placed to the level of the carina. The goal of apnoeic oxygenation is to document the absence of breathing once Paco2 arises sufficiently to stimulate the respiratory centre. However, there is no agreement as to what should be the final Paco2 for maximal stimulation. In addition, complications may occur during the performance of the test [3]. As a result of the chance observation of the inversion of Pvco2–Paco2 gradient in a patient with brain death during the apnoea test, we performed simultaneous arterial and central venous gases during the apnoea test in 20 consecutive patients being tested for brain death. We obtained Local Research Ethics approval and relatives’ assent for this. The patients were on two different intensive care units of two university hospitals in Barcelona, between October 2005 and May 2007. They were disconnected from mechanical ventilation while receiving 100% oxygen via a catheter inserted into the tracheal tube to the level of the carina. Simultaneous arterial and central venous blood samples for Paco2 and Pvco2 measurements were obtained before and after the apnoea test. Before the test, mean (SD) Pvco2 (4.67 (0.28)) was greater than Paco2 (4.0 (0.25)) but after the onset of apnoea, this relationship was reversed with Paco2 (7.86 (0.24)) becoming greater than Pvco2 (6.66 (0.24)). This reversal of the venous-arterial CO2 gradient was observed in all the patients and in both hospitals. This finding has been previously reported in arterial and mixed venous samples by Rigg and Cruickshank [4], who investigated CO2 kinetics in two patients undergoing apnoea testing for brain death. They postulated that the reversal of the normal Paco2 and co2 (mixed venous) relationship during apnoea to be a consequence of the Haldane effect: the CO2 dissociation curves are different in reduced and oxygenated haemoglobin. As apnoea testing was performed without oxygenation, it was concluded that reduced haemoglobin was able to carry more CO2 than carbamino compounds explaining how the rise in Paco2 was greater than that in mixed venous co2. However, our patients underwent apnoeic oxygenation and we also observed that the Paco2 became greater than Pvco2, so that reversal of the arterial and venous Pco2 relationship is independent of the absence of oxygen. In two of our patients, blood gas analyses were performed before final disconnection from the ventilator once brain death was diagnosed, and in both cases Paco2 was greater than Pvco2. Fig. 5 shows the changes in one of these patients, with a normal gradient of beginning of the test and inversion from the second minute. Unlike Rigg and Cruikshank [4] we could not take mixed venous samples and our venous samples were from the central venous catheter. The repeated observation of this phenomenon in 20 patients of two hospitals over 18 months suggests that it was not a laboratory mistake. Alterations in Pvco2 (squares; grey line) and Paco2 (circles; black line) over time in a single patient. NB PCO2 given in mmHg. One possible explanation is that the lung parenchyma continues producing CO2 and consuming O2 via the bronchial arteries that empty the bronchial veins into the left auricle. As the patient is apnoeic and not eliminating CO2, an inverse CO2 gradient occurs. Fig. 5 suggests a change in the way in which CO2 is transported, as the Paco2 at 1 min is lower than the Pvco2 at 2 min. There may be an alteration in CO2 transportation from its dissolved form (corresponding to 7% of the total CO2 transported), to transportation linked to the bicarbonate ion or carbamino compounds. Finally, the goal of the apnoea test is to achieve a sufficiently high Paco2 to stimulate respiratory centres in the medulla oblongata. However, in view of a reversed Pvco2 and Paco2 gradient we may have to consider modifying our practice, assessing Pvco2 rather than Paco2 [5].