The author responds:
I was very interested in the comments from Drs Bruno, Berend, and Engels about our recent report.1 Dr Bruno pointed out that in our report the fact that the significantly higher mean bicarbonate value in group 3 (unstable patients) than in group 1 (stable patients) suggests the presence of a concomitant metabolic alkalosis in many of those patients, whereas the normal bicarbonate in groups 2 and 4 does not exclude the presence of a metabolic disorder. We should say that higher bicarbonate value in group 3 than in group 1 might also reflect that chronic hypercapnia was greater and ongoing for a longer period than in patients in group 3.
Furthermore, Dr Bruno proposed using a slightly modified traditional approach to detect mixed acid-base disorders. It is true that we did not use the approach suggested by Dr Bruno. The traditional approach we used in groups 1 and 3 (Table 1) was clearly unable to detect any single disturbance. We agree that this is too simplistic and a more complex approach would have been of value. Therefore, the comparison between the Stewart approach and the approach proposed by Dr Bruno should be done, since apparently it has not been done. It is likely, however, that the stepwise approach that relies on different measurements, namely biological and clinical, has first to be assessed in terms of diagnostic performance, with pre-test odds, positive and negative likelihood ratios, and post-test odds measurements. Dr Bruno also asked for additional data from our patients, which, unfortunately, were not collected in our case report form. Dr Bruno's conclusion that “I think their conclusion of better performance with the Stewart method is not justified. The traditional approach, with only minor adjustments, can provide the same practical information.” is, however, not supported by data, as outlined above.
Drs Berend and Engels pointed out 5 issues in our report.1 I do agree with their first comment, that there is no accepted standard to compare when assessing acid-base abnormalities. Second, to try to circumvent the lack of agreement between laboratories regarding the computation of the contributing variables in the Stewart approach, as suggested by Berend and Engels, we determined normal ranges locally.
I don't yet understand the third issue Berend and Engels raised. Setting the pH threshold below 7.36 to define acute respiratory failure is all but conservative. In a randomized controlled trial in COPD patients admitted for acute respiratory failure in the pneumology ward, Plant et al2 enrolled COPD patients in acute respiratory failure with pH range 7.25–7.35. The definition we used was composite and centered around dyspnea. Furthermore, we enrolled not only COPD patients but also any patient with chronic respiratory failure, defined with a priori criteria.
Fourth, Fencl et al3 separated respiratory and non-respiratory acid-base disorders. We enrolled a prospective cohort of patients who were thought by definition to have a respiratory disorder.
I also agree with Berend and Engels' fifth comment, on the hypothesis by Martinu et al,4 a study we quoted in our report.1 There is obviously the need for further pathophysiological investigations to better understand the complex acid-base derangements in the setting of chronic respiratory failure.
Footnotes
-
The author has disclosed no conflicts of interest.
- Copyright © 2011 by Daedalus Enterprises Inc.