Correspondence  |   January 2000
Anemia and Arterial Partial Pressure of Oxygen
Author Notes
  • Associate Professor
  • Department of Anesthesiology
  • The University of Alabama at Birmingham
  • Birmingham, Alabama 35249-6810
Article Information
Correspondence   |   January 2000
Anemia and Arterial Partial Pressure of Oxygen
Anesthesiology 1 2000, Vol.92, 284. doi:
Anesthesiology 1 2000, Vol.92, 284. doi:
In Reply:—
We appreciate the interest expressed by Drs. Deem and Swenson with regard to our recent article. 1 Our experimental design involved four experimental groups, three of which underwent four serial hemodilutions with concurrent measurement of several physiologic and biochemical parameters. Because one of the groups was sham-operated, it was acceptable to have used one-way analysis of variance with a rigorous post hoc  test to determine if there were any significant differences between the groups.
Of further interest, Drs. Deem and Swenson suggest that hemodilution may increase arterial partial pressure of oxygen (PaO2) by improving ventilation–perfusion distribution. They commented that there was “an apparent strong inverse correlation between PaO2and hematocrit.” When the hematocrit and PaO2data from all four groups of our study 1 are analyzed by linear regression, an inverse relationship is observed (R2= 0.14;P  < 0.0001). Although highly significant, the relationship between PaO2and hematocrit in this model is weak, with only 14% of the change in PaO2associated with the change in hematocrit. To test the hypothesis that hemodilution per se  is associated with an increase in PaO2, we analyzed the combined PaO2data from all three hemodilution groups (n = 25) of our study, using one-way repeated-measures analysis of variance with a post hoc  Tukey test. The PaO2values (mean ± SD) are as follows: 445 ± 54 mmHg before hemodilution, 472 ± 47 mmHg after the first hemodilution, 475 ± 51 mmHg after the second hemodilution, 484 ± 38 mmHg after the third hemodilution, and 519 ± 37 mmHg after the fourth hemodilution. The PaO2values observed after each hemodilution were significantly (P  < 0.05) greater than the PaO2values observed before hemodilution. Furthermore, the PaO2values observed after the fourth hemodilution (hematocrit = 5%) were significantly (P  < 0.01) greater than those observed after all preceding hemodilutions. Finally, a comparison of the sham-operated group (n = 8) with the hemodilution group with repeated-measures analysis of variance did not demonstrate a significant difference between the groups (P  = 0.3). The relative similarity of mean values between the sham-operated and hemodilution groups (at best 8% different) after most of the hemodilutions combined with the observed variability (SD of 5–10% of mean value) are the most likely reasons for the inability to demonstrate statistical significance. Although our study was not specifically designed to test the hypothesis that hemodilution is associated with an increase in PaO2, our data are, in general, similar to that of Deem et al.  2 
Nielsen VG, Baird MS, Brix AE, Matalon S: Extreme, progressive isovolemic hemodilution with 5% human albumin, PentaLyte, or Hextend does not cause hepatic ischemia or histologic injury in rabbits. A NESTHESIOLOGY 1999; 90:1428–35Nielsen, VG Baird, MS Brix, AE Matalon, S
Deem S, Hedges RG, McKinney S, Poslissar NL, Alberts MK, Swenson ER: Mechanisms of improvement in pulmonary gas exchange during isovolemic hemodilution. J Appl Physiol 1999; 87:132–41Deem, S Hedges, RG McKinney, S Poslissar, NL Alberts, MK Swenson, ER