To the Editor We agree with Dr Steinbrook1(p373) that implementation of the new policy of the European Medicines Agency (EMA) on the proactive release of clinical trial data means that “a new era of clinical trial data as a public good will begin.” However, the policy has a major flaw: it will not apply to drugs approved before the effective date.2 A previous transparency initiative, the Food and Drug Administration Amendments Act of 2007, also has the same problem.3 Thus, despite the recent EMA initiative, although comprehensive information will be available on newer drugs in the future, published information on established drugs (ie, approved before the effective date of the policy) will still remain biased, even though they will account for the lion’s share of drugs prescribed in clinical practice for years to come. This imbalance will hamper a meaningful comparison of established and newer drugs and therefore devaluate comparative effectiveness research (especially when indirect comparisons are required). In addition, open questions on established drugs will never be answered. This is particularly relevant for drugs with a large public health impact such as Tamiflu (oseltamivir phosphate; Genentech): it took independent researchers several years to obtain all of the relevant clinical trial data.