IMAGINE A PATIENT WITH ACUTE CORONARY SYNDROME ON a flight from New York to Tokyo. Given how small the world of medicine has become, it is reasonable to assume that the treatment he/she would receive would be very similar no matter if the plane returned to New York or continued to Tokyo. Now imagine that the patient had septic shock instead of heart disease—landing in New York or Tokyo would result in drastically different treatment. In New York, early goal-directed therapy (EGDT) and drotrecogin alfa are widely used while the former is less common and the latter is unavailable in Tokyo. Conversely, for more than 10 years, a patient with sepsis would likely receive therapy with an endotoxin adsorber (polymyxin B hemoperfusion) in Tokyo, a treatment unavailable in New York. Why such radical differences? Aren’t physicians treating sepsis in Japan and the United States practicing evidencebased medicine? It turns out that evidence is indeed at the heart of these international differences—but the interpretation of the same evidence seems to vary. The opponents of EGDT and drotrecogin alfa argue that there is insufficient evidence to support their use. Widespread adoption of EGDT in the United States was based on one small singlecenter randomized trial. Although approval for drotrecogin alfa was based on a large multicenter pivotal trial, subsequent negative results in low-risk patients and children, combined with higher risk of bleeding, have led to doubts in many parts of the world about the effectiveness of the drug. Similarly, polymyxin B hemoperfusion lacks rigorous clinical trial data and is therefore considered unproven by many regulatory agencies. However, if these therapies lack sufficient evidence, why are they avoided in some places and virtually “standard of care” in others? Part of the answer may be historical. For instance, achieving supraphysiologic hemodynamics for patients with sepsis through fluid resuscitation and vasoactive drugs has always been more of the “ICU culture” in the United States and Europe whereas “blood filtering” has been more readily accepted for sepsis in Asia. A more cynical interpretation might be that drotrecogin alfa was developed by a US company, whereas polymyxin B hemoperfusion was developed in Japan. Against this backdrop, in this issue of JAMA, Cruz and colleagues report findings from the EUPHAS trial, testing the effectiveness of polymyxin B hemoperfusion in several Italian centers. This preliminary study is valuable as an example of “New Yorkers” testing a Japanese intervention. This kind of cross-community validation is refreshing and necessary but unfortunately only too rare. The results, although preliminary, suggest a number of interesting hypotheses and should provoke further study. This is essential given the significant ongoing problem that sepsis represents. However, just as interesting as the results is the way the investigators and their oversight board chose to respond to results. Instead of producing the sort of solid scientific evidence that would likely eliminate the differences between the 2 communities, the investigators have, in essence, left New York and landed in Tokyo. But how did this happen and what does it say about the treatment of sepsis around the world? The EUPHAS trial was not designed as a definitive trial with a patient-centered end point. Instead it was designed to determine whether polymyxin B hemoperfusion would result in improved mean arterial pressure and less requirement for vasopressors in patients with septic shock due to presumed abdominal infections. Given the modest differences in these end points it was therefore surprising that 28-day mortality (a secondary end point) was so drastically different between groups: 11/34 (32%) with polymyxin B vs 16/30 (53%) with conventional therapy. Even more surprising was the authors’ response: “Results were discussed with the president of the ethics committee . . . who declared it unethical to deprive a potentially beneficial therapy to a group of patients that carry high mortality.” Thus, an oversight body has, on the basis of a secondary analysis of an underpowered study, changed the standard of care for a series of hospitals participating in the study and presumably affected the standard of care for other hospitals in