Surveillance of antimicrobial resistance - what, how and whither?

Objective To express the views of a working party held to consider antibiotic resistance surveillance systems, their strengths and weaknesses, and their current and future applications. Methods The participants, all of whom were experienced in this field, discussed the development of surveillance systems in relation to the increasing prevalence of resistance to antibacterial agents and the current interest in surveillance systems shown by many official bodies, in both the human and veterinary fields. The problems inherent in surveillance systems were considered together with the applications of different systems. Results The properties of good antibiotic resistance surveillance systems were defined. Surveillance systems vary widely from those with a narrow base, focusing on few organisms in one disease area, to those covering many diseases, many organisms (including normal flora) and many compounds. Whatever their design, they should be able to detect significant differences and shifts in susceptibility to various antibacterial agents, and the information derived from them should reach as many interested parties as possible in a timely manner. In using this information to decide strategies, criteria for action need to be determined by pragmatic consensus. Funding remains a major problem, with few large studies being supported by official bodies in spite of their professed enthusiasm for surveillance. In consequence, many current systems are funded by the pharmaceutical industry and are of necessity restricted in their focus. Conclusions Antibiotic resistance surveillance studies should and can be well planned and well executed. Many current systems suffer from well-recognized but uncorrected biases. Consortium funding will be necessary for large schemes to be successful. There is no 'ideal' surveillance system.