is too much for some patients and too little for others. 3 The most appropriate way to administer intravenous opioids is to titrate them according to the pain relief observed. 4 Although the dose of 0•1 mg/kg of morphine is widely used in randomised trials in the emergency department, 5 we think that standard care should be a titration and not a fixed dose, and we recommend that future randomised trials use titration as the standard method. If not, the opioid group might not experience suffi cient pain relief and the other treatment will be considered to be better, whereas this is probably not the case when opioids are correctly used. To avoid wasted research, 6 the control group should always benefi t from the best treatment known. We think that this was not the case here.BR reports personal fees from Sanofi , Crossject, Fresenius Kabi, Thermo Fisher Scientifi c, and LFB. FA reports personal fees from Sanofi , Mylan, Mundipharma, Biocodex, Baxter, and The Medicines Company.