This paper examines differing institutional responses to and interpretation of the same scientific and medical data, and looks at the way in which policies, ostensibly based upon these interpretations, are presented to public audiences. The case study concerns the use of AZT as a prophylactic for injured health care workers. Data was collected from personal in depth interviews in two health authorities and telephone interviews with a further 35 health authorities. Observations include the fact that widely different interpretations of scientific data by scientific and medical experts is likely to be presented to the lay audience in terms of scientific certainty, based on an institutional need for certainty and consensus. Contrary to conventional perspectives which would suggest that scientific knowledge was completed in its expert arena, then applied in different settings, this analysis suggests that the scientific knowledge is socially 'completed' or 'closed' in each of the different situations in which it is interpreted into practice. Ideal models of both the science and of organisational working practice, appear to have been used as the basis of these different constructions.