Post-marketing studies of subjective side effects; A case for strict methodological criteria and careful analysis of data