Assessing comparative effectiveness research in the US

With the inclusion of $US1.1 billion earmarked for comparative effectiveness research (CER) in the recently enacted stimulus package, the US government indicated it will play an important role in informing prescribing and reimbursement decisions. However, this sizable investment does beg four important questions: what is the nature of CER data; what methods are suitable for collecting CER data; who is (should be) responsible for collecting CER data; and how will (should) CER data be used by the federal government? Using three recent high-profile cases of drugs and drug classes, we assess the current state of federal- and state-funded CER in the US. From these cases we observe that evidence is gradually emerging as a filter for certain prescribing and coverage decisions. The first case indicates evidence should not be gathered and applied in a post hoc fashion after a reimbursement decision has already been reached. Case 2 suggests limitations associated with making inferences from systematic reviews when applying the evidence to the treatment of individual patients. Case 3 points to a comprehensive, but more costly and time-consuming, way of gathering data to inform prescribing and reimbursement decisions. Despite caveats, we argue that there is room for building a more systematic and better coordinated evidence base in the US, so that all stakeholders are better equipped to understand variation in clinical outcomes while promoting appropriate prescribing patterns. Accordingly, CER could help close the gap between what we know and what we do in pharmaceutical care. For the majority of cases in which CER is carried out, we favour a pluralistic system of CER analyses with a clearing-house for systematic reviews conducted by multiple evidence-based practice centres, each uniquely suited to its constituency.