Large multicentre trials in anaesthesia: the ANZCA Clinical Trials Group.

In the search for the best available evidence, large randomized clinical trials have much to offer1-7. First, large trials (often with more than 1000 patients) can study major endpoints such as death and permanent disability. second, large trials can be powered to detect, with precise 95% confidence intervals, the modest differences in hard endpoints that are important to clinicians, patients and their families. Third, large trials can come close to reflecting real world clinical practice if the protocol is simple and the broadest possible range of patients are included from many hospitals. Fourth, subgroups from large patient samples can be analysed to generate further hypotheses and subsequent targeted clinical trials. Fifth, while observational studies can provide important data, a randomized trial will minimize biases that may taint the results, such as patient characteristics or other confounding factors differing between treatment arms. Finally, recent guidelines have further improved the information provided in trial reports7. There have been several large trials in anaesthesia and critical care medicine817. These include clear evidence of benefit for maintenance of normothermia and/or high inspired oxygen concentration in colorectal surgery8,9, intensive insulin therapy in critically ill patients10, bispectral index monitoring to prevent awareness", low-dose aspirin for hip surgery12, lowtidal volume ventilation in acute lung injury13, and recombinant human activated protein C for severe sepsis14. In contrast, there is no demonstrable benefit of low-dose dopamine or albumin in critically ill patients15'16. Successful large trials conducted by informal networks of Australasian anaesthetists11,17, and in critical care medicine, the Clinical Trials Group of the Australian and New Zealand Intensive Care Society (ANZICS CTG)15,16,18, have led to publications in high profile general medical journals such as New England Journal of Medicine and Lancet. Following on