In 2000 the American Institute of Medicine, adviser to the federal government on policy matters relating to the health of the public, published the report To Err is Human: Building a Safer Health System, which was to become a call to arms for improving patient safety across the Western world. By re-conceiving healthcare as a system, it was argued that it was possible to transform the current culture of blame, which made individuals take defensive precautions against being assigned responsibility for error - notably by not reporting adverse events, into a culture of safety. The IOM report draws on several prominent social scientists in accomplishing this re-conceptualisation. But the analyses of these authors are not immediately relevant for health policy. It requires knowledge translation to make them so. This paper analyses the process of translation. The discussion is especially pertinent due to a certain looping effect between social science research and policy concerns. The case here presented is thus doubly illustrative: exemplifying first how social science is translated into health policy and secondly how the transformation required for this to function is taken as an analytical improvement that can in turn be redeployed in social research.