The major objectives of this article are to review hemoglobin outcome studies, focusing on the utility of purely observational approaches; the design limitations of hemoglobin target randomized trials; what is known from the trials that have been performed to date; and whether confident recommendations for target ranges can be made. The commonly observed association among lower hemoglobin levels, left ventricular hypertrophy, and higher mortality also has been seen within randomized trials when assigned hemoglobin targets were ignored; critically, however, corresponding relationships were absent when intention-to-treat principles were used, strongly suggesting noncausal associations and the need for randomized designs. This being said, hemoglobin typical target trials often have undesirable features, including inadequate blinding and the use of imbalanced, nonstandardized, nonblinded co-interventions. The trials published to date, spanning hemoglobin levels of approximately 7 to 13 g/dl, suggest that higher treatment targets enhance quality of life but at the price of higher BP, thrombotic events, and reduced dialysis adequacy in hemodialysis patients. To date, there is no convincing evidence that targets that approach the physiologic range (versus intermediate targets) have an effect on left ventricular size or survival. Therefore, depending on the outcome examined, higher hemoglobin levels may have beneficial effects, harmful effects, or no effect, leading to the unsatisfactory situation of having to make opinion-based tradeoff decisions. Whereas the available evidence suggests that 11 g/dl is a reasonable lower bound for the hemoglobin target range, the upper bound remains to be defined and targets above 13 g/dl cannot be routinely recommended.
|Original language||English (US)|
|Number of pages||7|
|Journal||Clinical journal of the American Society of Nephrology : CJASN|
|State||Published - Jul 2006|