and the third half
________________
5.
The editorial which accompanied the above piece. I think this is very even-handed.
Oxygen-induced acute hypercapnia in chronic obstructive pulmonary disease: What's the problem?
Levetown, Marcia MD
"For most of the past decade, chronic obstructive pulmonary disease (COPD) has been the fourth leading cause of death in the United States, and through 1999 both the number of deaths from COPD and the age-adjusted death rate from the disease continued to increase (1). The natural history of COPD is one of progressive decline in pulmonary function, punctuated by frequent acute chest illnesses caused by infection, congestive heart failure, pneumothorax, or other acute events (2). These exacerbations are treated with bronchodilators, antibiotics, corticosteroids, and supplemental oxygen.
A notorious complication of oxygen therapy during COPD exacerbations is progressive hypercapnia, leading to stupor, profound acidemia, or cardiac dysrhythmias. Such events nearly always result in mechanical ventilation, whether noninvasively by face mask or invasively by endotracheal intubation. Hypercapnia of this magnitude leads, at a minimum, to prolonged hospitalization. Also, because the patients involved have severe obstructive ventilatory defects, they are at high risk for complications of mechanical ventilation, including barotrauma, chronic ventilator dependency, and death. This phenomenon seems to be well publicized in medical schools, although its incidence is completely unclear; students and interns from across the United States and around the world have been able to tell me about it on rounds in the intensive care unit. Almost universally in my experience, their explanation revolves around the loss of a hypercapnic drive for ventilation in chronic CO2-retaining COPD patients and removal of the hypoxic drive to ventilate by administration of supplemental oxygen. As Gomersall and colleagues (3) point out in this issue of Critical Care Medicine, that belief sometimes results in withholding of supplemental oxygen from distressed and hypoxemic COPD patients, in efforts to avoid the necessity for mechanical ventilation.
Although my students are partially correct, the explanation for acutely increasing plasma CO2 tension in this setting is not quite so simple. Working with computer models in the early 1970s, West (4) demonstrated clearly that ventilation-perfusion inequality can lead to chronic CO2 retention and objected to the term chronic hypoventilation to describe these patients, on the basis that minute ventilation is actually increased in patients with COPD. Nevertheless, hypercapnic COPD patients breathe with lower tidal volumes than nonhypercapnic COPD patients, and therefore have lower minute alveolar ventilation and increased deadspace ventilation (5, 6). In stable hypercapnic COPD patients, Paco2 increases when patients are exposed to high Fio2; however, there is no correlation of Paco2 increase with measures of ventilatory drive, as measured by P0.1(7). Dunn et al. (8) studied a different measure of ventilatory drive, the CO2 recruitment threshold, in COPD patients who had been ventilated for, on average, a week. They found a small decrease in the CO2-induced ventilatory drive when these patients were exposed to 100% oxygen.
However, patients may behave differently during acute respiratory insufficiency than they do when they are in their stable chronic state (9). Aubier and colleagues (10) demonstrated that administration of 100% oxygen to patients with acute COPD exacerbation resulted in mild decreases (14%) in minute ventilation but in increases in Paco2 out of proportion to the change in minute ventilation, suggesting an increase in deadspace ventilation induced by high Fio2. In a separate series they showed that administration of supplemental oxygen reduced P0.1 in similar COPD patients, but that P0.1 remained three times greater than that of normal controls, indicating that hypoxemia was not the only stimulus for ventilation in these patients (11). Taken together, the data suggest that a complex interaction of reduced hypoxic drive, changes in ventilation-perfusion matching, increases in deadspace ventilation, and to a small extent the added component of the Haldane effectthe unloading of CO2 by hemoglobin in hyperoxic conditionsis responsible for acute CO2 retention in these patients.
Gomersall et al. have taken a more pragmatic approach to the problem by sidestepping the issue of the mechanism of acute CO2 retention and attempting to determine whether patients whose hypoxemia is completely corrected fare worse than those patients in whom relative hypoxemia is tolerated. For this purpose, they defined relative hypoxemia as arterial Pao2 between 50 mm Hg and 70 mm Hg and completely corrected hypoxemia as Pao2 >70 mm Hg. These levels represented the targets for supplemental oxygen administration in patients who presented with room air hypoxemia. The outcomes of interest were mechanical ventilation and death; serial arterial blood gases were examined, and Glasgow Coma Score, respiratory rate, and cardiac rhythm were followed. In 34 evaluated patients, no differences were detected in the rate of mechanical ventilation or death, nor in the serial measurements of Paco2.
The principal strength of this study is its recognition that there is little in the way of evidence for any particular approach to the administration of supplemental oxygen to COPD patients in extremis. In an age when we attempt to treat patients according to empirical evidence, and given the frequency with which COPD patients suffer exacerbations of their disease, the absence of evidence in this area represents a glaring deficiency. I believe that the authors are correct in their call for a larger study to address this issue.
One would need to be careful, however, in using the results of the current study as a basis for a larger study. With only 34 patients analyzed, this study by Gomersall et al. did not have the statistical power to find any but the greatest of differences between the groups, with the majority of one group requiring mechanical ventilation and the majority of the other group not requiring it. Further, there appear to be difficulties in the mechanics of targeting the arterial Pao2 ranges that the authors chose. In actuality, the low Pao2 group in this study achieved an average Pao2 of 6368 mm Hgbased on the figures provided, since the numbers were not included in the textwhereas the high Pao2 group achieved an average Pao2 of 97115 mm Hg. Given that studies addressing the mechanism of acute CO2 retention with oxygen administration have typically induced acute CO2 retention by administering 80% to 100% oxygen, it is difficult to know whether the statistically different Pao2 levels in the current study represent a sufficient clinical difference to have engendered a problem. In this study there was, in fact, no difference in serial Paco2 measurements, although again statistical power is lacking.
What sort of clinical study or studies do we need? Given the current state of affairs, I believe that a simple observational study would be a good starting point. My exposure to physicians in numerous locales throughout the United States and my informal polling of numerous colleagues during the writing of this editorial leads me to believe that, in the absence of data, an ad hoc standard for administering supplemental oxygen to COPD patients with acute respiratory insufficiency has evolved, at least in the United States. Roughly stated, that standard appears to be that all patients in the emergency department and the intensive care unit should have continuous Spo2 monitoring, that Spo2 should be maintained in the approximate range of 90% via nasal cannula or Venturi face mask, and that it should probably not exceed 93% to 95%. Since the study of Bone et al. (12) in the mid-1970s, there have been no investigations of the incidence of acute CO2 retention during COPD exacerbation, nor of predictors for its development (7). At the time of that publication, Spo2 monitoring was not readily available, and one could not adjust the flow of supplemental oxygen quickly or continuously in response to continuous measures of oxygenation, as can be done today. What we need now is to understand the magnitude of the problem in the 21st century. I agree with Gomersall et al. that there is a paucity of data."
_______________________________________________________________
This is a controversial area. Till there is more data, I will titrate oxygen for COPD patients who require it to an Sa02 of 90%
Dont be bullied. Make up your own minds.
