| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a Department of Thoracic and Cardiovascular Surgery, West German Heart Center Essen, University Hospital, University of Duisburg-Essen, Hufelandstrasse 55, Essen, 45122 Germany
b Institute of Medical Informatics, Biometry, and Epidemiology, University Hospital, University of Duisburg-Essen, Hufelandstrasse 55, Essen, 45122 Germany
c Department of Neurology, University Hospital, University of Duisburg-Essen, Hufelandstrasse 55, Essen, 45122 Germany
d Institute of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital, University of Duisburg-Essen, Hufelandstrasse 55, Essen, 45122 Germany
(Email: stephan.knipp{at}uk-essen.de).
In a study on 39 patients undergoing coronary bypass surgery, we reported that postoperative cognitive deficits, albeit present in some patients for as many as 3 years after surgery, were not related to perioperative cerebral ischemia as seen on diffusion-weighted magnetic resonance imaging [1]. In contrast, recently presented data on 40 valve surgery patients showed that postoperative cognitive decline was present in all patients with postoperative lesions, but in only 35% of patients without lesions [2]. In the letter to the editor [3] the authors addressed a number of issues that prompt clarification as follows:
No relationship was observed between the volume of new lesions on magnetic resonance imaging and early or late cognitive decline [1]. Also, the number of lesions and cognition dysfunction were not related to each other (odds ratio = 0.893; 95% confidence interval, 0.535 to 1.490; p = 0.664). Statistical details on the latter were omitted in the article, as we think that single or composite volumes of lesions, or both, are more accurate markers to describe the ischemic cerebral load than their numbers alone.
Ten patients had one lesion, 2 patients had two lesions, 2 patients had three lesions, and 1 patient had four, five, six, and seven lesions, respectively, adding up to a total of 42 lesions.
If the presence of lesions on postoperative magnetic resonance imaging was treated as a categorical variable, this would not change the principal result that lesions were unrelated to cognitive impairment (odds ratio = 0.700; 95% confidence interval = 0.157 to 3.130; p = 0.641). If the composite lesion volume was dichotomized (ie, 
0.2 cm3; yes/no), this does not change our results either (odds ratio = 0.374; 95% confidence interval = 0.066 to 2.106; p = 0.265).
Indeed, using unit increments in the order of 1 mm3, the confidence interval is smaller than with greater increments. The choice of the appropriate unit increase, however, is a matter of debate, and in our data (composite lesion volumes ranged from 172 to 922 mm3), the use of volumes rounded to 100 mm3 (as it was done in Barber's work [2]) would unnecessarily be inaccurate. Note that the unit increment itself does not influence the significance of the relationship.
All tests used in our study follow the recommendations of the internationally approved Consensus Statement. They are reliable instruments and well standardized, providing valid data of cognitive functioning.
Rates of postoperative neurobehavioral dysfunction are crucially influenced by the way cognitive impairment is defined. We used a strict criterion for definition where a drop of 
1 standard deviation score in at least 3 of 11 tests was required. On the contrary, in Barber and colleagues' work, a decline in only 1 of 10 tests was sufficient to define cognitive impairment. In doing so, the incidence of deficits is expectedly high, and this may well attribute for differences in the relationship between cognitive dysfunction and structural brain changes between research groups.
Studies from other groups also failed to find any association between cerebral ischemia and cognitive decline after cardiac surgery [4, 5], and by now, Barber and colleagues [2] are the only ones to show a positive relation. In addition, the range of composite lesion volume was much greater in his patients compared with our patients (<0.2 to 3.0 vs 0.172 to 0.922 cm3), yielding much greater differences in exposition. When a greater difference of lesion volume is present and a more liberal criterion of deficit is applied, this may certainly explain also why they found a positive relation while we did not.
 |
References |
|---|
 Top References |
|---|
Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
