LAUSR

S urgeons have always had an ambivalent relationship with innovation. On the one hand, the boldness required to traumatize the bodies of our fellow humans in order to heal them has ensured that surgeons have, throughout history, been ambitious, confident people willing to take risks. On the other hand, the traditional surgical apprenticeship model of learning encourages conservatism: as we learn, we find out why our mentors advise against some apparently attractive ideas, and develop respect for their prohibitions. We also live and die professionally by our reputation, and few things can tarnish this as fast as failed innovations that harm patients. In previous generations, when weak clinical research methodology meant than expert opinion was valued more than data, surgical innovators were acutely aware of the risks they ran if they alienated their peers. Successful innovators were usually marked by the great care they took to reassure opinion leaders by adopting an attitude of humility and caution, at least in public. Both the risks and the appeal of surgical innovation have grown rapidly and in parallel in recent decades. Technological advances have resulted in a seemingly endless rise in the number of new devices and techniques for surgical or quasi-surgical treatment. Techniques that might, in the preinternet age, have spread slowly over decades, become globally known within months, and the relentless rise in medical publishing ensures widespread dissemination of claims for new treatment. Compared with pharma, operations and devices remain very lightly regulated, and this has had an adverse effect on the quality of the evidence which supports them. It has also become increasingly clear that the complexity and variability of delivery of surgery requires different approaches to research from those that have served drug treatments well. Modern surgical practice requires valid evidence and complete transparency to ensure that efficacy and safety are demonstrated clearly. We are still in the early stages of developing the new paradigm which will ensure this, and as ideas about what constitutes ‘‘good enough’’ evidence are still in flux, it is not surprising that ideas about how innovation should be regulated are changing too. Both in North America and in Europe, regulators are demanding better initial clinical evidence and more careful surveillance for new therapeutic devices. At the institutional level, hospitals are also recognizing that changes are needed in how they ensure that innovation occurs as safely as possible, without putting up barriers that will prevent it from happening. Institutional Review Boards (IRBs) should play a major role here, but anecdotal evidence suggests that they are seen as excessively risk averse by many surgical innovators, and are relatively easily bypassed for the initial phase of introduction of a new procedure. Harm can result to patients from this kind of unreviewed semi-clandestine innovation. The irony involved here should not escape us, given the good intentions underpinning the IRB system, and the unintended consequence. The paper by Marcus et al in this issue of Annals of Surgery reports a promising new system for dealing with the challenges of surgical innovation at the institutional level in a modern setting, and restoring a sensible degree of oversight that does not result in gridlock. The evidence that their institutions’ system has allowed safe innovation is still preliminary, but the process adopted is clearly a major advance on reliance on the IRB, or the various ad hoc systems relied on elsewhere. The key features are the comprehensive nature of the process, catching all new procedures and devices, and the expertise of the panel running it, which contained both surgical experience and professional safety and quality know-how. The undeniable major improvement in the speed of decision making apparently also derived from this—the previous system at this institution was slowed by the justifiable reluctance of panels with mainly business focused expertise to make judgment calls on safety matters. Pioneers are always judged (by nonpioneers) for the shortcomings of their ranging shots, and this was by no means a perfect scientific study. The simple before-after design was likely inevitable, given the controversy that would have attended any attempt to randomize innovations to the new system or the old. No control could be imposed on the nature or number of the innovations coming through, and data collection methods were unavoidably different in the historical and prospective