Early supported discharge: a valuable alternative for some stroke patients

See Articles page 501 pendent replications. Yet do same-team approaches ensure independence? Any intermingling of the process for generating and replicating the hypothesis entails the danger of somehow diluting the independence of the replication. Sample size is also essential. A recent editorial hailed the advent of “small studies with high density of data”. Well, I think there is no free lunch in good research. Microarrays need evidence and this cannot be obtained from a couple of small studies, no matter how high-tech. Small sample sizes might actually hinder the identification of truly important genes. Molecular medicine may eventually fulfil its arrays of promises. However, we should aim for many independent studies with a total of several thousand patients, a hundred-fold more than the current standard. If we truly believe that microarrays and molecular research in general are important, we should not settle for less.