PMS data may be obtained from spontaneous reporting, as with the FDA and Committee on Safety of Medicines’ data, observational studies and directly prospective investigation. Spontaneous reporting occurs against total exposure within a country and the denominator is large. However, the data are frequently incomplete and are subject to a number of inaccuracies. Nevertheless, it does provide a basis for comparison of agents. There are difficulties when the period of observation is radically different, as drugs have been much more highly scrutinized in the last 10-15 years than before. However, contemporary reports from the last decade do show that evaluation in the immediate post-marketing period can highlight the problems of some members of a class and, contrastingly, the safety of others. Data for example fluoroquinolones are shown on the slide (50). It can clearly be seen that spontaneous ADR reporting in the US rapidly identified the problems of temafloxacin and trovafloxacin (against a background of < 500,000 patients), but that the initial 3-million-patient exposure to levofloxacin revealed no real problems and a profile similar to ciprofloxacin – a trusted agent with an exposure worldwide of > 300 million patients.