It turns out that a randomized controlled trial was published early online this weekend in NEJM in which the authors concluded that a hematocrit goal of <45% was associated w/lower of risk of cardiovascular death or thrombotic events in patients avg 65yo w/polycythemia compared to a goal of 45-50%. In essence, 365 patients w/polycythemia vera were followed for an average of 31 months after randomization to either of the two stated goals.  Normal reference range for hematocrit goes up to 51% or so depending upon the lab, so it's interesting to note that normal isn't necessarily optimal!