Correspondence

Abstract

Jayasinghe et al. (1) tested the hypothesis that CBF reactivity to changes in MAP (cerebral autoregulation) or CO 2 (CO 2 reactivity) is lost in hypotensive, ventilated, preterm infants. The investigators report a difference in MAP-CBF reactivity and CO 2 -CBF reactivity among normotensive and hypotensive infants. Specifically, the authors report higher MAP-CBF reactivity, reflecting impaired cerebral autoregulation, and attenuated CO 2 reactivity in hypotensive infants compared to normotensive infants.There are some methodologic issues with this study that merit discussion. First, there is no information regarding how CO 2 reactivity was formally examined, leaving the reader to assume that aggregate and random CO 2 -CBF data were analyzed. Second, although 95% confidence intervals were used to detect differences between groups, a Type II error cannot be ruled out in the normotensive MAP-CBF group, particularly when the average reactivity was numerically equal in both groups. Third, no definitions for what constitutes impaired versus intact cerebral autoregulation was provided, making it difficult to interpret the differences between the normotensive and hypotensive groups. Fourth, although statistical analyses were used to normalize skewed data, it is not clear why a larger number of subjects was not used in each group. Given the small sample size of the study, it is not surprising that the variances were large. Finally, it is to be expected that cerebral autoregulation was impaired or abolished in hypotensive subjects since by definition, MAP in the hypotensive group is likely to be below the lower limit of cerebral autoregulation (LLA). Therefore, it is not clear why the authors would study something that is to be expected. Finally, although the authors state that the LLA in preterm infants is 30 mmHg (2), the subjects of the present study were older (Ͼ23 months). The one paper describing the LLA in infants outside the neonatal period reports LLA values comparable to that of older children (3).